facultad de ciencias - core
TRANSCRIPT
Facu l tad de C ienc ias
C O M P L E M E N T A C T I V A T I O N I N
N O N - S M A L L C E L L L U N G C A N C E R A N D
I T S E F F E C T O N T U M O R P R O G R E S S I O N
L e t i c i a C o r r a l e s P e c i n o
Servicio de Publicaciones de la Universidad de Navarra
ISBN 84-8081-130-7
F a c u l t a d d e C i e n c i a s
Complement activation in non-small cell lung
cancer and its effect on tumor progression
Memoria presentada por Dª Leticia Corrales Pecino para aspirar al grado
de Doctor por la Universidad de Navarra
Leticia Corrales Pecino
El presente trabajo ha sido realizado bajo mi dirección en el Departamento
de Bioquímica y el Área de Oncología del CIMA y autorizo su presentación
ante el Tribunal que lo ha de juzgar.
Pamplona, 22 de febrero de 2011
Dr. Rubén Pío Osés
El presente trabajo ha sido presentado en la Facultad de Ciencias
para optar al título de Doctor por la Universidad de Navarra
Este trabajo ha sido realizado con las siguientes ayudas:
“Ayuda para la formación de personal investigador”, de la Fundación para
la Investigación Médica Aplicada.
Beca para la “Formación de Profesorado Universitario” del Ministerio de
Ciencia e Innovación.
UTE-project CIMA.
Plan Nacional de Investigación Científica, Desarrollo e Innovación
Tecnológica del Ministerio de Educación y Ciencia (SAF-2005-01302).
Programa de Promoción de la Investigación Biomédica y en Ciencias de la
Salud del Ministerio de Sanidad y Consumo (PI080923).
A mis padres
A Edu y Montse
A mi tío Jose María
Acknowledgements
Gracias en primer lugar a la Universidad de Navarra, al departamento de
Bioquímica y al área de Oncología del CIMA, por haberme dado la oportunidad de
realizar la tesis doctoral.
En segundo lugar, gracias a mi director de tesis Rubén Pío, con el que he
empezado a dar los primeros pasos en el mundo de la investigación. Gracias por todo el
tiempo dedicado y la ayuda proporcionada para que saliese adelante este trabajo.
Muchas gracias por la dirección de la tesis, por tu paciencia y por la confianza
depositada en mí.
A Luis Montuenga tengo que agradecer todos sus consejos y opiniones que han
sido de gran ayuda para realizar esta tesis.
También me gustaría dar las gracias a Daniel Ajona, una fuente inagotable de
ideas, algunas han sido indispensables en este trabajo. Gracias por tu ayuda, tus ánimos
y por tus comentarios constructivos y por supuesto, siempre “razonables”.
A todos mis compañeros del grupo de Biomarcadores con los que he pasado
muchas horas y muchos buenos momentos: Maribel, Alberto, Marta, María José, Jackie,
Ana, Elena y Uxúa. A Txiki, Cristina y Amaya quiero darles las gracias especialmente
por todo lo que pacientemente me han enseñado y ayudado. Y a los que ya se fueron:
Javi, David, Evan, Patricia, Rebeca, Iñaki y Teresa. Muchísimas gracias a todos.
No puedo olvidarme del laboratorio de la Dra. Anna Blom en Malmö. Gracias por
haberme dado la oportunidad de trabajar y aprender tanto con vosotros. Gracias también
a Marcin Okroj por toda su ayuda y apoyo durante mi estancia.
Al laboratorio de Microambiente Tumoral: Saray, Álvaro, Lili, Rafa, Xabi y los
que ya se fueron, Eva, Marta y Erik. Gracias por dejarme ser una “okupa” y luego
ascenderme a grado de hija adoptiva. Y muchas gracias Ana por tu ayuda, por darme
buenas ideas y colaborar tan activamente en este trabajo.
Quiero también agradecer a todas las personas de la planta de Oncología, a
nuestros vecinos y compañeros de cultivos 1B con los que tantas horas he pasado: y a
todos los que he ido conociendo en estos años: Alice, Leyre, Celia, Diego, Iker,
Carolina, Manu, María, Javi, Angela, Lola… ¡es imposible nombraros a todos!, vuestra
ayuda y consejos han contribuido a la realización de este trabajo. También quisiera
agradecer la ayuda que me ofrecieron otras personas del CIMA, en especial los
laboratorios del Dr. Riezu, Dr. Lasarte y Dr. Páramo.
Al Dr. Alfredo Martínez, muchas gracias por toda la ayuda y consejos que nos
ofreciste para el estudio de la angiogénesis.
Gracias a las secretarias de la planta de Oncología, Marisol y Leticia, por vuestro
trabajo que tanto facilita el nuestro.
Acknowledgements
Gracias a todas mis compañeras de piso en estos 4 años, por aguantarme y
compartir conmigo tantos buenos momentos. A Belén y Mariona, por vuestra amistad
desde que llegamos a Pamplona; a Erika, por tu alegría, generosidad y por comprender,
compartir e incluso intercambiar “frikadas”; a Gloria, Silvia y Almu, por hacerme sentir
como “en familia”; a Carmela, porque contigo el debate durante las comidas siempre
estaba asegurado; a Laura, por tu sentido del humor, tus sevillanas maneras y por tu
espíritu crítico, gracias por enseñarme a ver las cosas desde otro punto de vista; a Elena
por demostrarme que es una persona con la que siempre podré contar; y por último,
gracias a la “UNESCO”, a Mónica por tu disponibilidad y tus artes culinarias, a Susana
por tu generosidad y contagiarnos un poco tu “happy world”, y a Lili por ser tan buena
compañera de piso, de laboratorio, de planes… y tan buena amiga. No puedo olvidarme
de vuestros respectivos novios (algunos ya maridos), gracias por compartir con nosotras
muchas risas, bastantes comidas y alguna que otra excursión. En definitiva, gracias a
todos por ser mi familia en Pamplona y haber conseguido que no me encontrase sola en
ningún momento.
A Juana Mari, porque aunque fueron pocos meses de estancia en Pamplona
bastaron para empezar una amistad que estoy segura perdurará a pesar de la distancia.
Gracias por tu buen humor, tus consejos, por tus buenos detalles y en definitiva por ser
como eres, una “maravilla de bé”.
Gracias a Miriam por tu amistad, tus consejos y por decir las cosas sin rodeos, y
por ser tan sumamente servicial.
Gracias a Rudi y a Manolo por ser mucho más que mis profesores de pintura. Por
vuestra paciencia, vuestro sentido del humor, por dar siempre ánimos incluso en
momentos no muy buenos y por ser tan excelentes artistas.
Gracias a todos mis amigos de Algeciras, por vuestro apoyo y ánimos, gracias por
hacer que desconectara. Tampoco puedo olvidarme de Chico, el perro de Rubio, gracias
porque mientras viva la comunidad científica puede estar segura de que existe un animal
más feo que un ratón nude.
Gracias especialmente a Elena Domínguez, por tu optimismo y tus ánimos.
Gracias porque es tan fácil confiar en ti.
Finalmente, gracias a mi familia. A Edu y Montse, por vuestro apoyo, cariño y por
cuidar tanto y tan bien de mí. Y aunque es muy difícil resumir todo lo que os tengo que
agradecer en unas líneas, gracias papá y mamá por vuestra confianza en mí, por
escuchar los cientos de “ohu, que hoy no me ha salido el experimento” y aguantar mi
humor a veces algo agrio, por la ilusión que ponéis en todo lo que hago y por ser mi
referente a seguir. Gracias de corazón.
ABBREVIATIONS
Abbreviations
7AAD 7-amino-actinomycin D
ADCC Antibody-dependent cellular cytotoxicity
AMD Age-related macular degeneration
APC Antigen presenting cells
APH50 Alternative pathway hemolysis (50%)
ARG1 Arginase 1
ATCC American Type Culture Collection
BCR B cell antigen receptor
BEAS Bronchial epithelial cells
bFGF Basic fibroblast growth factor
BSA Bovine serum albumin
C1 INH C1 inhibitor
C3aR C3a receptor
C4BP C4-binding protein
C5aR-1 (CD88) C5a receptor-1
C5L2 C5a-like receptor 2
CCL Chemokine ligand
CCP Complement control protein
CDC Complement dependent cytotoxicity
Cdk4 Cyclin-dependent kinase 4
CFHR Complement factor H-related protein
CK2 Casein kinase 2
CR1 (CD35) Complement receptor type 1
CTLA4 (CD152) Cytotoxic T-lymphocyte antigen 4
CVF Cobra venom factor
DAF (CD55) Decay accelerating factor
DEPC Diethyl pyrocarbonate
DMSO Dimethylsulphoxide
dNTP Deoxyribonucleoside triphosphate
DTT Dithiothreitol
ECM Extracellular matrix
EDTA Ethylenediaminetetraacetic acid
EGTA Ethylene glycol tetraacetic acid
ELISA Enzyme-linked immunosorbent assay
Abbreviations
ERK Extracellular signal-regulated kinase
FACS Fluorescence activated cell sorting
FBS Fetal bovine serum
FH Factor H
FHL-1 Factor H-like protein 1
FITC Fluorescein isothiocyanate
Foxp3 Transcription factor forkhead box P3
GAPDH Glyceraldehyde 3-phosphate dehydrogenase
GPCR G-protein-coupled receptor
GPI Glycosylphosphatidylinositol
GPI-PC Glycosylphosphatidylinositol-phospholipase C
GVB Veronal buffer saline with gelatin
HBEC Human bronchial epithelial cells
HI-NHS Heat inactivated normal human serum
HI-NMS Heat inactivated normal mouse serum
HPLC High-pressure liquid chromatography
hTERT Human telomerase reverse transcriptase
HUVEC Human umbilical vein endothelial cells
IDO Indoleamine 2,3-dioxygenase
IL Interleukin
INF-γ Interferon-gamma
iNOS Inducible nitric oxide synthase
KO Knockout
LAG3 (CD223) Lymphocyte activation gene 3
LDH Lactate dehydrogenase
LLC or 3LL Lewis lung carcinoma cells
mAb Monoclonal antibody
MAC Membrane attack complex
MAPK Mitogen-activated protein kinase
MASP MBL-associated serine protease
MBL Mannose-binding lectin
MCP (CD46) Membrane cofactor protein
MCP1 Monocyte chemotactic protein-1
mCRP Membrane complement regulatory protein
Abbreviations
MDSC Myeloid-derived suppressor cells
MFI Mean fluorescence intensity
MMP Matrix metalloproteinase
NHS Normal human serum
NK-κB Nuclear factor kappa-light-chain-enhancer of
activated B cells
NMS Normal mouse serum
NO Nitric oxide
NSCLC Non-small cell lung cancer
ONOO Peroxynitrites
PBS Phosphate buffered saline
PBST Phosphate buffered saline-Tween
PCR Polymerase chain reaction
PD1 Programmed death 1
PDL1 (B7-H1) Programmed death ligand 1
PDT Photodynamic therapy
PE Phycoerythrin
PI Phosphatidylinositol
PI-PLC Phosphatidylinositol-specific phospholipase C
PKC Protein kinase C
RNS Reactive nitrogen species
ROS Reactive oxygen species
SCLC Small cell lung cancer
SCR Short consensus repeat
SDS Sodium dodecyl sulfate
SDS-PAGE SDS polyacrylamide gel electrophoresis
TAE Tris-acetate-EDTA buffer
TAM Tumor-associated macrophage
TBS Tris buffered saline
TGF-β Transforming growth factor beta
TIMP Tissue inhibitor of metalloproteinases
TNFSF4 (OX40L) Tumor necrosis factor (ligand) superfamily
member 4
Abbreviations
TNF-α Tumor necrosis factor-alpha
Tregs Regulatory T cells
VB Veronal buffer saline
VEGF Vascular endothelial growth factor
WT Wild-type
INDEX
Index
INTRODUCTION ________________________________________________ 1
1. LUNG CANCER _______________________________________________ 3
1.1. Lung cancer classification _____________________________________________4
1.2. Lung cancer staging __________________________________________________4
1.3. Lung cancer etiology _________________________________________________6
2. THE COMPLEMENT SYSTEM __________________________________ 6
2.1. Complement activation _______________________________________________8
2.1.1 The classical pathway ______________________________________________8
2.1.2 The lectin pathway ________________________________________________9
2.1.3 The alternative pathway ____________________________________________9
2.1.4 Alternative routes of complement activation ____________________________9
2.2. Complement regulation ______________________________________________10
2.2.1 Soluble regulators of the complement system___________________________11
a. C1 inhibitor _____________________________________________________11
b. C4-binding protein _______________________________________________12
c. Factor H and factor H-like protein 1 __________________________________12
d. Vitronectin______________________________________________________13
e. Clusterin _______________________________________________________13
2.2.2 Membrane complement regulatory proteins ____________________________14
a. Complement receptor type 1 ________________________________________14
b. Membrane cofactor protein _________________________________________14
c. Decay accelerating factor __________________________________________14
d. CD59 __________________________________________________________15
2.3. Biological effects of complement activation ______________________________15
2.3.1 Cell lysis by the membrane attack complex ____________________________15
2.3.2 Opsonization of the cell by C3b and its related fragments _________________15
2.3.3 Biological effects mediated by anaphylatoxins__________________________16
3. COMPLEMENT AND CANCER ________________________________ 18
3.1. Complement in the immune surveillance against tumors_____________________18
3.2. Complement regulation on cancer cells __________________________________19
3.3. Complement in the immunotherapy of tumors_____________________________21
3.4. Complement, inflammation and cancer __________________________________21
3.5. Complement and lung cancer__________________________________________23
HYPOTHESIS AND OBJECTIVES_________________________________ 27
MATERIALS AND METHODS ____________________________________ 29
1. MATERIALS _________________________________________________ 31
1.1. Animals __________________________________________________________31
Index
1.2. Clinical samples from lung cancer patients _______________________________31
1.3. Normal human serum________________________________________________31
1.4. Plasma ___________________________________________________________32
1.5. Human non-small cell lung cancer cells__________________________________32
1.6. Immortalized normal human epithelial cells ______________________________32
1.7. Endothelial cells____________________________________________________33
1.8. Murine lung cancer cells _____________________________________________33
1.9. Hybridoma ________________________________________________________33
1.10. Factor H________________________________________________________33
1.11. C5aR-1 antagonist ________________________________________________33
2. METHODS ___________________________________________________ 34
2.1. Methods for the generation of experimental material _______________________34
2.1.1 Cell culture _____________________________________________________34
a. Cell culture conditions_____________________________________________34
b. Subcultures _____________________________________________________35
c. Freezing and thawing method _______________________________________36
2.1.2 Generation of OX-23 anti-factor H antibodies __________________________36
a. Hybridoma culture________________________________________________36
b. Production of conditioned media_____________________________________36
c. Purification _____________________________________________________37
d. Determination of protein concentration________________________________37
e. Titration________________________________________________________37
f. SDS-PAGE and Coomassie staining__________________________________38
2.1.3 Anti-A549 polyclonal antibody generation_____________________________39
a. Production of antibodies in NZW rabbits ______________________________39
b. Immunoreactivity assay____________________________________________40
2.1.4 Factor H purification______________________________________________40
a. Preparation of columns ____________________________________________40
b. Factor H purification ______________________________________________41
c. Factor H integrity ________________________________________________41
d. Factor H quantification by ELISA ___________________________________42
2.1.5 Alternative pathway hemolytic assay (APH50) _________________________43
2.1.6 Total RNA extraction and retrotranscription____________________________44
2.1.7 Protein extraction ________________________________________________46
2.1.8 Genotyping of C3 deficient mice ____________________________________46
2.2. Methods for the study of complement activity on cells ______________________47
2.2.1 Complement deposition on cells after activation of the alternative or classical
pathways _______________________________________________________47
2.2.2 Complement cytotoxicity after activation of the alternative or classical pathways __48
2.2.3 C5a quantification in plasma samples and bronchoalveolar lavages by ELISA _49
Index
2.2.4 Study of complement inhibitors in NSCLC and normal epithelial cells _______49
a. Determination of mRNA levels by quantitative PCR _____________________49
b. Factor H binding to the cell membrane ________________________________50
c. Detection of membrane-bound complement inhibitors ____________________50
d. Elimination of GPI-anchored proteins from the cell membrane by PI-PLC
treatment _______________________________________________________51
e. Effect of hypoxic conditions in the expression of complement inhibitors in
NSCLC cells ____________________________________________________51
2.3. Methods to study the in vitro effects of C5a ______________________________52
2.3.1 C5aR-1 expression _______________________________________________52
a. C5aR-1 mRNA expression _________________________________________52
b. C5aR-1 protein expression _________________________________________53
2.3.2 Treatment with C5a_______________________________________________53
2.3.3 C5a signaling activation ___________________________________________53
a. Calcium influx assay ______________________________________________53
b. ERK phosphorylation _____________________________________________54
c. NF-κB translocation to the nucleus___________________________________55
2.3.4 Effect of C5a on cell proliferation____________________________________55
2.3.5 Effect of C5a on cell cycle _________________________________________56
2.3.6 Effect of C5a on anchorage independent growth ________________________57
2.3.7 Effect of C5a on cell migration______________________________________57
2.3.8 Effect of C5a on HUVEC tube formation ______________________________58
2.3.9 Effect of C5a on cell adhesion to extracellular matrix proteins _____________58
2.3.10 Effect of C5a on integrin expression_________________________________59
2.3.11 Effect of C5a on MMP activity_____________________________________60
2.4. Mice models for the study of the influence of complement activation on lung tumor
growth in vivo______________________________________________________61
2.4.1 Syngeneic animal model of lung cancer _______________________________61
2.4.2 Tissue processing ________________________________________________63
a. Cell isolation from spleens _________________________________________63
b. Total RNA extraction from tumors and quantitative PCR _________________64
c. Fixation of tumors ________________________________________________64
d. Fixation of lungs _________________________________________________64
2.4.3 Study of angiogenesis by CD31 immunohistochemistry __________________64
2.5. Statistical analysis __________________________________________________65
RESULTS______________________________________________________ 67
1. COMPLEMENT ACTIVATION AND REGULATION ARE
INCREASED IN NON-SMALL CELL LUNG CANCER CELLS______ 69
1.1. Deposition of complement proteins after activation of the alternative pathway ___69
Index
1.2. Deposition of complement proteins after activation of the classical pathway _____70
1.3. Cell cytotoxicity due to complement activation by the alternative and classical
pathways _________________________________________________________73
1.4. Expression of complement inhibitors____________________________________74
1.5. Effect of hypoxia in the expression of complement inhibitors on NSCLC cells ___78
1.6. C5a levels in patients with NSCLC _____________________________________80
2. COMPLEMENT DEFICIENCY IMPAIRS TUMOR GROWTH ______ 81
2.1. Inhibition of tumor growth by C3 deficiency______________________________81
2.2. Inhibition of tumor growth by C5aR-1 blockade ___________________________83
3. COMPLEMENT ACTIVATION REGULATES ANTITUMOR IMMUNE
RESPONSE___________________________________________________ 85
3.1. Immune activation is induced in C3-deficient animals ______________________85
3.2. C5aR-1 blockade induces immune activation _____________________________87
3.3. C3 deficiency and C5aR-1 blockade reduces the expression of immunosuppressive
cytokines within the tumor____________________________________________89
4. EFFECT OF C5a IN THE PROMOTION OF ANGIOGENESIS ______ 91
4.1. Effect of C5a on HUVEC proliferation __________________________________91
4.2. Effect of C5a on HUVEC migration ____________________________________92
4.3. Effect of C5a on HUVEC tube formation ________________________________94
4.4. Decreased tumor angiogenesis in C3 deficient animals ______________________95
4.5. C5aR-1 blockade and C3 deficiency reduces bFGF expression in the tumor
microenvironment __________________________________________________96
5. EFFECT OF C5a ON NON-SMALL CELL LUNG CANCER CELLS__ 97
5.1. Expression of C5aR-1 in lung cancer cell lines ____________________________97
5.2. Activation of the C5a signaling pathway on NSCLC cells ___________________98
5.3. Effect of C5a on NSCLC cell proliferation ______________________________100
5.4. Effect of C5a in A549 migration ______________________________________101
5.5. Effect of C5a in NSCLC cell adhesion to extracellular matrix proteins ________102
5.5.1 Mechanisms involved in the regulation of cell adhesion to ECM proteins by C5a_103
5.5.2 Regulation of metalloprotease activity _______________________________104
DISCUSSION__________________________________________________ 107
1. THE DUAL ROLE OF COMPLEMENT ACTIVATION____________ 109
2. TUMOR CELLS ACTIVATE AND EFFICIENTLY REGULATE THE
COMPLEMENT CASCADE ___________________________________ 111
2.1. NSCLC cell lines spontaneously activate the complement system ____________111
2.2. NSCLC cells regulate complement dependent cytotoxicity__________________112
Index
3. EFFECT OF COMPLEMENT ACTIVATION IN THE TUMOR
MICROENVIRONMENT______________________________________ 116
3.1. Complement deficiency impairs tumor growth ___________________________116
3.2. Complement proteins and immunosuppression ___________________________120
3.3. Complement and angiogenesis________________________________________123
4. EFFECT OF C5a ON TUMOR CELLS __________________________ 125
5. COMPLEMENT IN HEALTH AND DISEASE____________________ 128
CONCLUSIONS _______________________________________________ 131
REFERENCES ________________________________________________ 135
INTRODUCTION
Introduction
3
1. LUNG CANCER
Cancer is a major public health problem in the western countries. For men, the
most common types of cancer are prostate, lung and colon; whereas, for woman, the
most common types are breast, lung and colon1. However, lung cancer is the leading
cause of cancer death for both men and women. More people die of lung cancer than of
colon, breast, and prostate cancers combined (Figure 1). This is a global fact, even
thought some progress has been made in the treatment of lung cancer in the last years,
using adjuvant, targeted and individualized chemotherapy. Only in the United States,
the American Cancer Society estimates for 2010 about 222,520 new cases of lung
cancer and about 157,300 deaths. In Europe, there were an estimated of 386,300 new
cases in 2006 (12.1% of overall new cancers diagnosed), ranking for the third in the list,
after breast and colorectal cancer; however, lung cancer, with an estimated 334,800
deaths (19.7% of total), was the most common cause of death from cancer2.
Figure 1. Estimated new cancer cases and deaths by sex in United States in 2009 for the ten leading cancer types. From Jemal et al.1
Introduction
4
1.1. Lung cancer classification
Lung cancer classification of the World Health Organization identifies two main
groups, with differences in their biological behavior and implications for treatment and
prognosis. These groups are: small cell lung carcinoma (SCLC), which accounts for
20% of lung cancers, and non-small cell lung carcinoma (NSCLC), which is the most
prevalent type and represents 80% of lung cancers.
NSCLC is further characterized by histology as adenocarcinoma, squamous cell
carcinoma, and large cell carcinoma. Adenocarcinomas represent about 40% of
NSCLC, are typically present in the lung periphery, may involve the pleura, and are
sometimes associated with scars. Bronchioalveolar carcinoma is a subtype of
adenocarcinoma which spreads along pre-exiting alveolar septae and does not invade
lung parenchyma. Squamous cell carcinomas, accounting for about 30% of NSCLC, are
typically more centrally located and posses features of squamous differentiation such as
keratin formation and intercellular bridges. Large cell carcinomas account for about
10% of all cases of NSCLC, may be central or peripheral, and exist in several subtypes.
NSCLC originates from epithelial cells and, even thought it is less aggressive than
SCLC, the detection usually takes place once the tumor has spread and metastases at
different points already exist3.
SCLC tends to arise in the larger airways and grows rapidly, becoming quite
large. The small cells contain dense neurosecretory granules, so are known to have an
endocrine origin. While initially this type of lung cancer is more sensitive to
chemotherapy, it ultimately carries a worse prognosis and is often metastatic at
presentation4.
1.2. Lung cancer staging
NSCLC is currently staged according to the T (extent of the primary tumor), N
(extent of involvement of regional lymph nodal disease), and M (extent of spread to
distant sites) scheme. The purpose of a staging system for any malignant disease is
threefold. First, it provides an estimate of tumor burden and extent of spread; second, it
provides prognostic information that can be applied to individual patients; and third, an
effective staging system is able to direct therapy. A recent universally stage
classification for NSCLC has been accepted by the Union Internationale Contre le
Cancer (UICC) and the American Joint Committee on Cancer (AJCC)5.
Introduction
5
More than 85 percent of people who are diagnosed with lung cancer today are
diagnosed in a late stage, when the disease is more difficult, and sometimes impossible,
to cure (Table 1). Therefore, an obvious strategy to reduce mortality would be the early
detection of the disease. The International Early Lung Cancer Action Program (I-
ELCAP) has shown that 85% of lung cancers can be found in the earliest, most curable
stage when an annual CT screening program is followed. If treated promptly with
surgery, the cure rate of these early stage patients is 92%6. Moreover, any patient
diagnosed with lung cancer as a result of I-ELCAP annual CT screening has an
estimated cure rate of 80%, regardless of stage and type of treatment (www.ielcap.org).
Recently, the National Lung Screening Trial (NLST), a randomized trial, found 20%
fewer lung cancer deaths among participants screened with low-dose helical CT
(www.cancer.gov). In any case, primary prevention services are still the more crucial
strategy to reduce lung cancer mortality. Besides, more research on the biological
mechanisms involved in the disease would allow the identification of high risk
individuals, the characterization of biomarkers that may signal the presence of lung
cancer, and the development of more efficient therapies.
Table 1. Lung cancer staging, and the related prognosis, according to the TNM system.*
*Modified from Detterbeck et al.5 **Median survival time (MST) in months, and 5-year overall survival.
Stage T N M PROGNOSIS** MST (5 years)
Ia T1a,b N0 M0 60-119 50-73%
Ib T2a N0 M0 43-81 43-56%
IIa T1a,b
T2a
T2b
N1
N1
N0
M0
M0
M0
34-49 36-46%
IIb T2b
T3
N1
N0
M0
M0 18-31 25-36%
IIIa T1-3
T3
T4
N2
N1
N0,1
M0
M0
M0
14-22 19-24%
IIIb T4
T1-4
N2
N3
M0
M0 10-13 7-9%
IV Tany Nany M1a,b 6-17 2-13%
Introduction
6
1.3. Lung cancer etiology
It is estimated that about 80% of lung cancer deaths are related to tobacco
smoking7. Apart from smoking, other associations have been seem with lung cancer.
For example, individuals with a history of chronic obstructive pulmonary disease8, or
even individuals with emphysema but without airway obstruction, are associated with
an increased risk for lung cancer9. Other contributing factors may include exposure to
radon, asbestos, arsenic, chromium, nickel and air pollution10. There is also an
increasing interest in the identification of genetic variations associated with increased
risk of lung cancer. In 2008, three independent groups simultaneously identified
common inherited genetic variations in chromosome 15, containing nicotinic
acetylcholine receptors, associated with increased susceptibility to lung cancer11-13.
2. THE COMPLEMENT SYSTEM
The complement system has been recognized classically as a central part of the
innate immune response, which develops a first defense against microbes and unwanted
host molecules. Its physiological functions are: defending against pyogenic bacterial
infection, bridging innate and adaptative immunity, and disposing of immune
complexes and the products of inflammatory injury14. All these activities are mediated
by more than 30 circulating or cell-surface-bound proteins. These proteins can function
as enzymes, substrates or regulators, keeping a well-balanced activation and inhibition
of the system. A very comprehensive review has recently been published that
summarizes the function and structure of complement proteins and describes in detail
the multiple roles of complement in homeostasis15.
Classically there are three established mechanisms of complement activation,
known as classical, lectin and alternative pathways (Figure 2). Each one differs from
the others in the initial steps, the proteolytic phase, and they converge in the cleavage of
the central molecule of the pathway, C3, which generates its active fragments C3a and
C3b. Complement is also activated by additional bypass pathways that act
independently of the classical and the alternative convertases (see section 2.1.4.). The
non-proteolytic phase starts after the cleavages of C3 and C5, when the following
components ensemble on the membrane and form a multiprotein pore complex (the
membrane attack complex, MAC) which eventually leads to cell lysis. A series of
Introduction
7
complement receptors on leukocytes mediate the recognition of opsonized cells, which
results in phagocytosis and the stimulation of the adaptive immune system. Finally, the
anaphylatoxins C3a and C5a are released during complement activation and trigger a
range of chemotactic and pro-inflammatory responses, such as the recruitment of
inflammatory cells and the increase in microvasculature permeability.
MBL
C1 complex
Ab
C4
C4b C2
C4b2b
C3
C3b
C4b2b3b C3bBb3b
C3C3H2O
C3H2O& fB
fD
C3H2OfBC3
C3b & fB
C3C3bBb &
properdin
C5C5
C5b
C6C7 C8
nC9MAC
C3a
C3a
C5a
fD
LECTIN PATHWAY
CLASSICAL PATHWAY
ALTERNATIVE PATHWAY
MBL
C1 complex
Ab
C4
C4b C2
C4b2b
C3
C3b
C4b2b3b C3bBb3b
C3C3H2O
C3H2O& fB
fD
C3H2OfBC3
C3b & fB
C3C3bBb &
properdin
C5C5
C5b
C6C7 C8
nC9MAC
C3a
C3a
C5a
fD
LECTIN PATHWAY
CLASSICAL PATHWAY
ALTERNATIVE PATHWAY
Figure 2. Drawing showing the cascade of events during the activation of the complement system.
Excessive complement activation on self tissue has severe effects and can
promote the development of various diseases. For this reason, a range of soluble and
cell-surface-bound inhibitors ensures that any action of complement on host cells is
either prevented or actively inhibited.
Introduction
8
2.1. Complement activation
2.1.1 The classical pathway
The first component of the classical pathway is a complex formed by three
proteins: C1q, C1s and C1r. The activation takes place after the hexameric C1q
recognizes the antibody constant regions of µ chains (IgM) and some γ chains (IgG)
bound to the target antigen16. In addition, it has been shown more recently that the
classical pathway can also be activated in the absence of antibodies, as C1q recognizes
endogenous ligands such as dying cells17, extracellular matrix proteins18, pentraxins19,
amyloid deposits20, prions21 and DNA22. Binding of C1q activates C1s and C1r. C1s is a
serine protease which cleaves two serum proteins, C4 and C2. The cleavage of C4
generates two fragments: C4b, which keeps bound covalently to the cell surface through
a thioester bond, and C4a, a soluble anaphylatoxin that diffuses away. The third
component, complement C2, binds to C4b and becomes a target for C1s as well. The
cleavage of C2 generates also two fragments, C2b which keeps bound to C4b, and C2a.
C4bC2b forms the classical C3 convertase on the cell surface; C2b in the convertase
complex acts as a serine protease that cleaves C3 in C3b and C3a. C3b binds covalently
to the cell membrane due to its thioester bond, and joins to the C3 convertase, forming
the classical C5 convertase, C4bC2bC3b. In addition, C3b acts as an opsonin by
covering the cell surface. The soluble C3a is the second anaphylatoxin generated in the
classical pathway. The activation of C5 occurs in the same manner, after C5 binding to
its convertase, it is cleaved in C5a, the most potent anaphylatoxin, and C5b, which starts
the second phase of complement activation. This stage results in directed, non-
enzymatic assembly of the terminal pathway components. First, C6 binds to C5b, and
then C7 binds to both and inserts into the cell membrane. The subsequent binding of C8
causes a deeper penetration of the complex into the membrane. All the conformational
changes that occur during the assembly induce the polymerization of 15-19 molecules
of C9 into the cell membrane. This action generates a 100 Å pore on the membrane, as
seen by electron microscopy23. The structure formed is known as the terminal
complement complex (TCC) or the membrane attack complex (MAC), and is
responsible for the final consequence of complement activation, the osmotic lysis of the
cell24.
Introduction
9
2.1.2 The lectin pathway
The lectin pathway is analogous to the classical pathway. This pathway is
activated by homologous proteins to C1q: mannose-binding lectin (MBL) and ficolins.
These proteins recognize specific carbohydrates on pathogens’ surface, such as
mannose and N-acetyl-glucosamine (GlcNAc). After their binding to the carbohydrates,
MBL-associated serine proteases (MASP-1 and MASP-2) function in an analogous
manner to C1r and C1s25, cleaving the complement components C4 and C2 to form the
C3 convertase C4bC2b, which is common to both the lectin and classical pathway
activation routes26.
2.1.3 The alternative pathway
The alternative pathway is mechanistically distinct from the classical and lectin
pathways. It is antibody-independent and provides an initial line of innate immune
defense, as it is initiated by the low-level and spontaneous hydrolysis of C3 on
activating surfaces. The spontaneous cleavage of C3, also known as “the tickover of
C3”, forms C3(H2O), which contains an activated thioester bond. C3(H2O) can bind to
factor B, in turn allowing its cleavage by factor D into Bb and Ba, forming the initial
alternative C3 convertase, C3(H2O)Bb27. This complex starts to convert C3 into C3b
and C3a. In most cases this C3b is rapidly inactivated. However, in complement
activating surfaces, the C3b generated can bind to them and associate with factor B. In
the last complex, C3bB, factor B is again a substrate for the factor D cleavage, forming
the predominant alternative C3 convertase, C3bBb. The stability of this convertase is
enhanced by the binding of properdin, helping to amplify the alternative pathway
activation28. The fragment Bb on the C3 convertase plays an analogous role to its
homologous C2b in the classical convertase. Bb cleaves C3 and forms an amplification
loop, generating more C3b that, on one hand, can create new alternative C3 convertases,
and on the other hand, is necessary to form the C5 convertase (C3bBbC3b). After C5
cleavage into C5a and C5b, the pathway continues with the non-enzymatic assembly of
the terminal pathway components.
2.1.4 Alternative routes of complement activation
Apart from the three major activation mechanisms, there are several bypass routes
that have been shown to directly trigger complement response at various stages. The
lectin pathway can be activated directly by the binding of MBL to natural IgM bound to
Introduction
10
ischemic antigens in endothelial cells after ischemia/reperfusion injury29-30. In the
absent of C2/C4, but in the present of alternative pathway components, it is possible
that antigen-antibody complexes or oligosaccharides activate C1 or MBL,
respectively31-32. It has been also shown that anaphylatoxins can be generated due to the
action of certain proteases. C3 can be cleaved and activated by extrinsic proteases, such
as thrombin or kallikrein, pointing to a cross-talk between the complement system and
the coagulation cascade33. C5 can be cleaved by thrombin, bypassing C334. C3a and C5a
can be generated directly from C3 and C5 by proteases found in the allergenic feces
produced by dust mites35-36. And finally, C5 can be cleaved by silica37 and asbestos
fibers38, by mechanisms involving the generation of free radicals and kallikrein
activation39.
2.2. Complement regulation
Regulation of complement is of critical importance for the homeostasis of the
organism. In order to avoid tissue damage, control proteins regulate complement at
three main levels: they can inhibit the protease activity of the proteins involved in the
activation cascade, facilitate the decay and destruction of convertases, and control the
formation of MAC (Figure 3). Usually complement regulators are grouped into two
groups, soluble proteins and membrane-bound proteins (known as membrane
complement regulatory proteins or mCRP).
Regulators share a varying number of a repeating motif termed complement control
protein (CCP) repeat, also known as short consensus repeat (SCR) or sushi domain.
CCP domain is characterized by its length (each unit contains approximately 60 amino
acids), its globular structure, and a frame of conserved residues that includes four
invariant cysteines, an almost invariant tryptophan and highly conserved prolines,
glycines and hydrophobic residues40. They are thought to play a role in binding to C3,
C4 and their breakdown products.
Introduction
11
C1 INH
Factor H
FHL-1
C4BP
CD31
CD55
CD46
ClusterinVitronectin
CD59
Factor I
MBL
C1
C4b
C3bBbC3b
iC4b
iC3b
MAC
C4b2b
C1 INH
Factor H
FHL-1
C4BP
CD31
CD55
CD46
ClusterinVitronectin
CD59
Factor I
MBL
C1
C4b
C3bBbC3b
iC4b
iC3b
MAC
C4b2b
Figure 3. Drawing of the main complement inhibitors, soluble proteins and mCRPs. Red lines represent inhibitory activity (when ending in a bar) or accelerated decay activity (when ending in a square); green lines represent cofactor activity (when ending in a square) or protease activity (when ending in an arrow head).
2.2.1 Soluble regulators of the complement system
a. C1 inhibitor
C1 inhibitor (C1 INH) is a member of the serpin family of protease inhibitors,
although it presents unique features that differentiate C1 inhibitor from other serpins41.
The protease inhibition carried out by C1 inhibitor is the most important activity
of this protein. This protease inactivates C1r, C1s and MASP2, resulting in the
inhibition of classical and lectin complement pathways. In addition, C1 INH inactivates
a variety of proteases outside the complement system: intrinsic pathway proteases
(factor XII, plasma kallikrein, factor XI), thrombin and fibrinolytic proteases (plasmin,
tissue plasminogen activator)42.
Introduction
12
b. C4-binding protein
Human C4-binding protein (C4BP) is a heterogeneous oligomeric protein present
in plasma in three isoforms, with different subunit composition. The major isoform is
composed of seven identical 75 kDa alpha-chains and a unique 40 kDa beta-chain, that
are held together by disulfide bridges. Alpha and beta-chains are formed by eight and
three SCR modules, respectively.
C4BP mostly inhibits the classical complement pathway. C4BP, after binding to
C4b, inhibits complement by three different ways: prevents the assembly of the C3
convertase, accelerates the decay of the classical C3 and C5 convertases, and functions
as a cofactor in the factor I-mediated inactivation of C4b. C4BP also plays a role as a
regulator of the alternative pathway by acting as a cofactor of factor I in the cleavage of
C3b, but only in the fluid phase43.
Outside the complement system, C4BP binds to the serum amyloid P component
(SAP) and to the anticoagulant protein S, again pointing to a cross-link between
complement and coagulation systems44.
c. Factor H and factor H-like protein 1
Factor H is a 150 kDa glycoprotein composed of 20 SCRs. In addition, alternative
splicing of factor H mRNA yields a 42 kDa protein, named factor H-like protein 1
(FHL-1) or reconectin. FHL-1 is identical in sequence to the seven N-terminal SCRs of
factor H, with four additional amino acids added at the C-terminal end. FHL-1 shares
the complement inhibitory activities of factor H.
Factor H is mostly known as an inhibitor of the alternative pathway. Through its
binding to C3b, factor H competes with factor B in the formation of the C3 and C5
convertases, displaces the Bb subunit from the convertases, and acts as a cofactor for
factor I in the cleavage of C3b45.
Several recent studies have described the association of genetic variations in
complement factor H with different diseases. Mutations or polymorphisms that alter the
binding of factor H to C3b and polyanions are associated with atypical hemolytic
uremic syndrome (aHUS), whereas mutations that disrupt the plasma activities of factor
H, leading to unrestricted activation of the alternative pathway, are associated with
membranoproliferative glomerulonephritis type II (MPGN2)46. The most studied
polymorphism at the factor H locus is rs1061170, which causes a Tyr402His amino acid
substitution in SCR7. Independent studies have shown that the allele 402His confers a
Introduction
13
significantly increased risk to age-related macular degeneration (AMD) in many
different populations47-50 possibly by altering the binding of SCR to different
glycosaminoglycans and retinal pigment epithelial cells46.
In addition, factor H displays functions outside the complement system. It serves
as a chemotactic factor and a secretagogue of IL-1β for monocytes, and as an adhesion
protein for neutrophils51-52.
Five complement factor H-related proteins (CFHR) have been identified: CFHR-1
to -5. They are coded by five genes closely linked to the factor H gene, and are
composed of SCRs with different degrees of identity with SCRs in factor H53. Like
factor H, CFHRs are mostly synthesized in the liver, but their concentrations are much
lower than that of factor H, and their functional properties are not well defined. Only
FHR-5 displays a weak cofactor activity in cleavage of C3b by factor I. FHR-1 binds to
C3b and may have a possible role in complement regulation. FHR-3 binds to C3b, C3d
and heparin, and FHR-4 to C3b, which also supports the role of these proteins in the
complement pathway (reviewed by Serka and Zipfel54).
d. Vitronectin
Vitronectin (S protein, S40) exists as a 75-kDa protein in the extracellular matrix
(ECM), and as two truncated forms in plasma, which are held together by a disulfide
bond. Vitronectin inhibits the insertion of C5-7 at its membrane-binding site. The fluid
phase sC5b-7 is still able to bind C8 and C9 to form soluble non-lytic C5b-8 and C5b-9
complexes. It has also been shown that vitronectin in solution directly inhibits C9
polymerization, and hence indirectly can block pore formation55.
In addition, vitronectin is attached to the extracellular matrix via its collagen
binding- or heparin binding-domains where provides a regulatory link between cell
adhesion and physiological proteolysis (reviewed by Schavartz et al.56).
e. Clusterin
Clusterin (apolipoprotein J) is a ubiquitously expressed protein. Structurally is a
heterodimer linked by a unique five disulfide bond motif. It functions similar to
vitronectin in complement inhibition by binding to the C5b-7 complexes and preventing
C9 polymerization57. Appart from the complement inhibitory function, clusterin plays
other physiologic functions that depend on its final maturation and localization. It has
been shown that intracellular forms of the protein display direct or indirectly
Introduction
14
pro/antiapoptotic activity and act as a negative regulator of NF-κB. Clusterin can also
modulate differentiation and regulate the production of major proinflammatory
cytokines such as TNF-α and IL-6. For all these functions, it has recently been related to
inflammation and autoimmunity58.
2.2.2 Membrane complement regulatory proteins
a. Complement receptor type 1
Complement receptor type 1 (CR1, CD35) is comprised of a linear array of 30
SCR units in its extracellular region, and is found in erythrocytes, neutrophils,
eosinophils, monocytes, follicular dendritic cells, glomerular podocytes, B lymphocytes
and some T lymphocytes59. CR1 regulates complement activation acting as a cofactor
for factor I-mediated cleavage of C3b and C4b, and accelerating the decay of the
classical and alternative convertases60.
b. Membrane cofactor protein
Membrane cofactor protein (MCP, CD46) is expressed as four common isoforms,
all of them composed of four SCR units. It is expressed in most cells but erythrocytes. It
acts as a cofactor of factor I for the C3b and C4b cleavage61.
In addition to its function as complement inhibitor, CD46 has been implicated in
the regulation of T cell-induced inflammation62 and the generation of T regulatory
cells63. CD46 also serves as receptor for several human viruses and bacteria: measles
virus human herpesvirus 6, adenovirus of different serotypes, Streptococcus pyogenes
and pathogenic Neisseria (reviewed by Cattaneo64).
c. Decay accelerating factor
Decay accelerating factor (DAF, CD55) is a glycoprotein composed of four SCR
domains which is anchored to the membrane by a glycosylphosphatidylinositol (GPI)
anchor. It is present in all blood elements and most other cell types65. DAF accelerates
the decay of the classical and alternative C3 and C5 convertases66.
CD55 binds to its ligand CD97, expressed on macrophages, granulocytes, and
dendritic cells, and is rapidly up-regulated in activated T and B cells. Through this
binding, CD55 might simultaneously regulate both the innate and adaptive immune
responses67.
Introduction
15
d. CD59
CD59 (protectin) is a compact single chain of 77 amino acids with a disc-shaped
structure, and, like CD55, is attached to the cell membrane by a GPI anchor. It is widely
expressed, present on all circulating cells, vascular endothelium, epithelium and most
other cell types. CD59 binds C8 in the forming MAC and inhibits the insertion of C9
into the lipid bilayer68. Alternative roles of CD59, related to its GPI anchor signaling
properties, have also been demonstrated in cell types such as T cells, natural killer cells
and B cells69.
2.3. Biological effects of complement activation
2.3.1 Cell lysis by the membrane attack complex
Complement was first recognized because of its capacity to lyse cells, in
particular bacteria. It has been shown that a single functional MAC in the membrane of
an erythrocyte is sufficient to lyse it70. The formed pore permits entry of water into the
cell, driven by the osmotic gradient, causing the cell to swell and burst. However,
nucleated cells are more resistant to this cytotoxic effect. Not only do they need a higher
number of pores on the cell surface, but also other factors are important. One essential
factor is the calcium influx into the cell once a number of MAC is formed. An increase
of intracellular calcium concentration causes a loss of mitochondrial transmembrane
potential resulting in an energy crisis71. And then, the loss of ATP and its precursors
through the MAC leads to cell necrosis72.
Apart from causing cell necrosis, MAC can induce other responses in the cell.
MAC can activate immune cells to release inflammatory molecules, increase the
resistance of cells to further lytic attack, drive cells to proliferate, and make cells more
resistant to apoptosis73.
2.3.2 Opsonization of the cell by C3b and its related fragments
All complement pathways lead to the central and initial step in complement
opsonization, the conversion of the component C3 to C3b. The nascent C3b molecule
can lead to complement amplification through the actions of factors B and D, but can be
also inactivated by factor I and its cofactors (MCP, CR1, CR2 and factor H). Initial
inactivation of C3b is mediated by factor I, which catalyzes the proteolysis of two
peptide bonds in the α' polypeptide chain of C3b. The resulting products are the
Introduction
16
membrane bound iC3b and a small peptide, C3f, which is released from the molecule.
The third cleavage catalyzed by factor I generates the inert C3c and C3d-g, the former is
released and the later is kept on the cell membrane. Although further complement
amplification is abolished, recognition of C3b and its fragments by complement
receptors on cells promotes phagocytosis74.
Complement receptors are divided in three recognized gene families: the SCR
family members CR1 and CR2, the β2 integrin family members CR3 and CR4, and the
immunoglobulin Ig-superfamily member CRIg.
CR1 (CD35) is expressed in the majority of peripheral blood cells and recognizes
C3b, C4b, iC3b, C3dg, C1q and mannose-binding protein75-76. The main function of
CR1 is to capture immune complexes (IC) on erythrocytes for transport and clearance
by the liver77. Apart from this role in IC elimination, CR1 promotes the secretion of
proinflammatory molecules, such as interleukin (IL)-1α, IL-1β, and prostaglandins; has
a role in antigen presentation to B cells; and is an inhibitor of both classical and
alternative pathways, due to its decay-accelerating activity and cofactor activity for
C3b/C4b cleavage78-79.
CR2 (CD21) is an evolutionary homologue of CR1, but only binds to iC3b, C3d
and C3dg. On B cells, CD21 forms a co-receptor with CD19 and CD81. Binding of the
co-receptor to the B cell antigen receptor (BCR) lowers the threshold for B cell
activation80. This fact is due to the uptake of C3d-coated antigens by B cells81.
CR3 (CD11b/CD18) and CR4 (CD11c/CD18) are expressed by leukocytes and
stimulate phagocytosis when bound to iC3b. In addition, they perform functions in
leukocyte trafficking, adhesion, migration, and costimulation, which have important
consequences in the defense of the host against pathogenic invasion82.
CRIg has been identified recently as a complement receptor. It is expressed on a
restricted subset of tissue-resident macrophages. CRIg may play an important role in
phagocytosis83.
2.3.3 Biological effects mediated by anaphylatoxins
During complement activation, soluble active fragments of approximately 9 kDa
(74-77 residues) are released from C4, C3 and C5. These molecules, named C4a, C3a,
and C5a, were called anaphylatoxins as they are potent multifunctional pro-
inflammatory molecules. In addition to complement activation, C3a and C5a can be
directly generated by proteases unrelated to the complement cascade (see section
Introduction
17
2.1.4.). Anaphylatoxins are evolutionarily related to one another, sharing a relatively
high degree of homology: C5a shares with C3a 36% amino acid identity, and C4a
shares 33% and 25% homology with C5a and C3a, respectively84. They also have some
overlapping functions in the generation of the immune response.
Anaphylatoxins play their functions after the binding to their receptors, which
belong to the superfamily of G-protein-coupled receptors (GPCR). Anaphylatoxin
receptors share high sequence homology85 and are closely related to other chemotactic
receptors such as the N-formyl-methionine-leucine-phenylalanine (fMLP) receptor,
ChemR23 and the chemokine receptors CXCR1 and CXCR286. Even though their
similarity, anaphylatoxin receptors differ in ligand specificity, signal transduction
capacity and function. The anaphylatoxin receptors comprise C3aR for C3a, and C5aR
and C5L2 for C5a. To date, no specific receptor has been found for C4a, which was
originally reported to bind to the same receptor as C3a87. For C5a binding, the first
recognized receptor was C5aR-188, also known as CD88; but more recently the orphan
receptor GPR77 was identified as a second C5a receptor and was called C5a-like
receptor 2 (C5L2)89. Both are seven-transmembrane receptors that share a 35%
homology. However, C5aR is a classical G protein-coupled receptor, while C5L2 is an
enigmatic receptor deficient in G protein coupling. This fact, together with an unknown
pathway for C5L2 after C5a binding, led to suggest that C5L2 is a default receptor that
attenuates C5a biological responses by competing with C5aR-1. Nevertheless, this
definition for C5L2 seems to be incorrect, at least in part, because current contrasting
results point to anti-inflammatory functions for C5L290 or a positive regulatory activity
of C5aR91.
Initially, C3aR-1 and C5aR-1 expression was reported only on myeloid cells, but,
in the recent years, there have been numerous reports demonstrating their expression in
a wide variety of non-myeloid cell types92. C3aR is expressed in inflamed human
central nervous system93, endothelial cells94, epithelial cells, smooth muscle cells,
submucosal and parenchymal vessels of the lung from asthma patients95, and activated
human T cells96. C5aR is expressed in a broad range of non-immune cells: endothelial
cells, neurons, astrocytes and microglia, cells from kidney, lung, liver, spleen, intestine,
skin and heart97. C5L2 seems to be frequently co-expressed with C5aR in most cells or
tissues98-99. Activities of anaphylatoxins are related to the cell types which express their
receptors. Classically, they have been known to increase vascular permeability, promote
smooth muscle contraction, induce leukocyte recruitment, increase chemotaxis,
Introduction
18
migration and phagocytosis in white blood-cell, and promote the production and release
of other inflammatory mediators (e.g., histamine)100. Apart from their pro-inflammatory
properties, anaphylatoxins have also been implicated in tissue regeneration and tissue
fibrosis101, and in brain development102. In summary, they are recognized as pleiotropic
molecules that can modulate the activity of many cell types, with a broad range of
biological functions, both inside and outside of the immune system.
The potency and duration of these activities differ among the three
anaphylatoxins. C5a has been shown to be considerably the most potent, whereas C4a is
the weakest in inducing biological responses. In addition, anaphylatoxin activity, in the
circulation and in tissues, is tightly controlled by carboxypeptidases, which rapidly
cleave off a C-terminal arginine residue, generating the desArg forms103. C5a-desArg
binds with 100-fold less affinity to C5aR, but with the same affinity to C5L2. C3a-
desArg is devoid of any pro-inflammatory activity104. The clearance of C5a-desArg
from the body is achieved after its binding to C5aR on leukocytes and other cells105, or
even more effectively after binding to C5L2, as it is internalized, retained and degraded.
This function of C5L2 suggests that a major function of this receptor is to remove active
complement fragments from the extracellular environment106.
In the recent years the interest of C5a in cancer progression has increased due to
the connection between this anaphylatoxin and the immunosuppression of tumors107.
3. COMPLEMENT AND CANCER
3.1. Complement in the immune surveillance against tumors
Once a threatening body is recognized by the complement system, the activating
steps are followed, generating an inflammatory reaction, the opsonization of pathogens,
and eventually the killing of some organisms. Considering this classical role of
complement, a similar scenario could be considered for the control of tumor growth.
During carcinogenesis, cancer cells acquire several sequential genetic and epigenetic
abnormalities that result in essential alterations in the cell physiology. These alterations
dictate malignant growth and produce changes in cell morphology, generating many
tumor-associated antigens that distinguish malignant cells from their normal
counterparts; for example, a change in glycosylation phenotype is considered a hallmark
of cancer cells108. Those changes can induce the recognition of malignant cells by the
Introduction
19
immune system109. In fact, in cancer patients, complement activation, with subsequent
deposition of complement components on tumor tissue, has been demonstrated. Thus,
deposits of C3, C4 and C5b-9 were found in breast110, and papillary thyroid
carcinoma111. In lung tumors, only C3b, but not MAC deposition, was detected by
immunohistochemistry112. Elevated levels of C3a and soluble C5b-9 were present in the
intraperitoneal ascitic fluid of ovarian cancer patients113. The MBL complement
activation pathway was significantly increased in patients with colorectal cancer
compared with healthy persons114. Higher complement hemolytic activity and C3 levels
were observed in serum samples from children with neuroblastoma115 and, similarly,
elevated complement levels were reported in patients with carcinomas of the digestive
tract116 and brain tumors117. In summary, although there are studies reporting normal
total complement and C3 levels in patients with breast, gastric and colorectal
carcinomas118, most in vivo observations suggest that cancer cells activate the
autologous complement system.
Therefore, complement recognition of cancer cells may be an element of the
immune surveillance theory, which supports that the immune system surveys the body
for tumor-associated antigens, taking part of the elimination of many or most tumors119.
However, it is clinically evident that this system sometimes fails to protect the organism
from growing tumors. The principal reason is the development of active mechanisms to
escape immune recognition or resist immune attack.
3.2. Complement regulation on cancer cells
There is sufficient basis to propose that certain cancers activate, directly or
indirectly, the complement system. However, although neoplastic transformation may
be accompanied by an increased capacity to activate complement, cancer cells seem to
develop a collection of mechanisms to resist complement attack. Those mechanisms can
be divided into two categories, extracellular and intracellular mechanisms (reviewed by
Jurianz et al.120). Extracellular protection is provided by:
(1) Expression of mCRPs on the cell surface. Several studies have demonstrated
the expression of mCRPs in primary tumors and cell lines. With the exception of CR1,
most cancers, independent of their tissue origin, express at least two, if not three,
mCRPs. In addition, it has been shown that mCRPs are overexpressed on tumor cells
(reviewed by Fishelson et al.121).
Introduction
20
(2) Release of soluble complement inhibitors, C1 inhibitor, factor H, factor I, or
clusterin, and binding of such inhibitors to the cell membrane; where they regulate
complement activation and subsequent complement dependent cytotoxicity (CDC). In
this sense, our group demonstrated that NSCLC cell lines express and release to the
extracellular milieu factor H and FHL-1122; in addition, downregulation of factor H in
one of those cell lines reduces its growth in vivo123. In a different study, we
demonstrated that NSCLC cell lines express functional factor I and C4BP, which
support efficiently the cleavage of C3b and C4b in vitro124. Another similar mechanism
is the production and release of soluble forms of mCRPs, such as sCD46, sCD55 or
sCD59, by alternative splicing or enzymatic cleavage of the membrane anchoraged
form.
(3) Secretion of chondroitin sulfate proteoglycans by malignant B lymphocytes
that resemble, in structure and function, the C1q inhibitor 125.
(4) Release of ecto-proteases by tumor cells that cleave complement components,
mainly C3126.
(5) Expression of ecto-protein kinases on the cell membrane. One example is the
casein kinase 2 (CK2), an ubiquitous and constitutively expressed protein kinase. It has
been demonstrated that CK2 expressed on the surface of human leukemia cell lines
phosphorylates several exogenous substrates, including the complement C9 protein127.
Phosphorylation of C9 may protect tumor cells from complement, either by inhibiting
MAC formation or by leading to the production of an inactive or unstable MAC. It has
been shown that blocking CK2 with selective inhibitors enhances killing of
lymphoblastoid cells due to the classical pathway activation by Rituximab128, an
antibody against the protein CD20 used in the treatment of lymphomas and leukemias.
Intracellular protection is provided by sublytic doses of MAC. Surprisingly, the
MAC itself is one of the most potent mechanisms to increase resistance to complement
attack. Its insertion on the cell membrane causes a variety of biological effects, such as
entrance into cell cycle, resistance to apoptosis, expression of adhesion molecules or
augmentation of complement resistance (reviewed by Morgan129). The mechanisms
responsible for this protection are poorly understood, but may involve increase in
calcium intracellular concentrations, PKC and MAPK activation, and protein and RNA
synthesis130. It is possible that those mechanisms are orchestrated to remove MAC from
the cell surface by external vesiculation or internalization.
Introduction
21
3.3. Complement in the immunotherapy of tumors
Tumor specific and tumor associated-antigens can be used as targets for
monoclonal antibody (mAb) immunotherapy. The number of mAbs that have been used
in the clinic as therapeutic drugs has been increasing since the first mAb, Rituximab,
was used for the treatment of lymphomas131. Those mAbs target tumor-specific and
tumor-associated antigens and block important cancer activities. In addition, most of
them are of the IgG1 isotype, which effectively mediates Fc domain–based functions,
such as antibody-dependent cellular cytotoxicity (ADCC) and complement fixation. It is
more successful to use a mAb that combines different mechanisms to direct a harmful
effect on tumor cells. For that reason, researchers are motivated to try to increase the
activation of complement on tumors using mAbs that already block another function on
the cell. However, the resistance of tumor cells to complement activation has to be
overcome.
Different strategies have been developed to overcome protection by complement
regulators: use of mCRP blocking antibodies in combination with anticancer
complement-fixing antibodies121; down-regulation of mCRP expression using
cytokines132; and removal of CD55 or CD59 from the cell surface using GPI-
phospholipase C133-134. However, the mentioned mechanisms are not specific for tumor
cells, and the aim now is to achieve an inhibitory effect of complement regulators
limited to the tumor. It seems that blockade of mCRPs is the most promising way to
achieve it. Some strategies include the use of a biotin-avidin system135, to neutralize
selectively complement regulators using antibodies136, and to use therapeutic bispecific
mAbs that target a tumor antigen and simultaneously block a major complement
regulatory protein137.
3.4. Complement, inflammation and cancer
Since the beginning of the 19th century, with the Virchow’s theory, the link
between cancer and inflammation was perceived. Virchow observed that chronic
inflammation establishes an environment that promotes the initiation and growth of
malignancy138. Since then, epidemiological studies have shown the growing evidence
that chronic inflammation predisposes individuals to various types of cancer. It is
estimated that 15–20% of all deaths due to cancer worldwide are attributable to
inflammatory processes139. Different triggers of chronic inflammation are related to the
Introduction
22
increased risk of malignancy, such as microbial infections, autoimmune diseases,
chemical and physical agents, and inflammatory reactions of uncertain etiology
(summarized in Table 2). In addition, the usage of non-steroidal anti-inflammatory
drugs may protect against various tumors140.
Table 2. Some associations between inflammation and cancer risk.
*Adapted from Balkwill and Mantovani138.
An inflammatory component is also present in the microenvironment of tumors
that are not epidemiologically related to inflammation. Within the microenvironment of
tumors, there are several inflammatory elements that feed tumor cells in terms of
inducing their proliferation and survival, promoting angiogenesis and metastasis,
evading adaptive immunity, and reducing response to hormones and chemotherapeutic
agents. Recently, it has been proposed that an inflammatory microenvironment might be
summed up to the six hallmarks of cancer proposed by Hanahan and Weinberg in
2000141 (Figure 4).
Malignancy Inflammatory stimulus/condition
Bladder Schistosomiasis
Cervical Papillomavirus
Ovarian Pelvic inflammatory disease/talc/tissue remodeling
Gastric H pylori induced gastritis
MALT lymphoma H pylori
Esophageal Barrett’s metaplasia
Colorectal Inflammatory bowel disease
Hepatocellular Hepatitis virus (B and C)
Bronchial Silica, asbestos, cigarette smoke
Mesothelioma Asbestos
Kaposi’s Sarcoma human herpesvirus type 8
Introduction
23
Figure 4. Inflammation as the seventh hallmark of cancer from Colotta et al.141
The main features of cancer-related inflammation include the infiltration of white
blood cells, prominently tumor-associated macrophages (TAMs) and the presence of
polypeptide messengers of inflammation, for instance cytokines and chemokines.
Furthermore, the activation of the complement system generates some of the most
powerful proinflammatory molecules in the body, mainly the three anaphylatoxins C3a,
C4a and C5a. Complement is activated in different tumors, as many studies have found
complement deposition in the membrane of malignant cells and an elevation of
complement activity in the sera of patients with neoplastic diseases (see section 3.1.).
So, activated complement proteins within the tumor microenvironment must be added
to the inflammatory pressure that dictates malignant growth142.
3.5. Complement and lung cancer
The relationship between complement and lung cancer has been known since
some decades ago, when elevated complement levels correlating with tumor size were
found in lung cancer patients143. Afterwards, different studies have reinforced this first
observation, suggesting that, as it happens in other kind of tumors, complement is
activated by lung cancer cells.
As most cancer cells do, lung cancer cells present inhibitory mechanisms to resist
complement dependent cytotoxicity. It has been shown by immunohistochemistry that
NSCLC tumors express on the cell membrane CD59 and CD46 with variable CD55
immunoreactivity112. Several in vitro studies also revealed that lung cancer cell lines are
extremely resistant to CDC, compared to human nasal epithelium primary cell
Introduction
24
cultures134,144, although this effect could not be attributed to the inhibitory action of
mCRPs. Lung cancer is not the only cancer type in which blocking of CD46 and CD55
does not make cancer cells more sensitive to complement145, which suggests that there
are other inhibitory mechanisms for controlling the activation of complement on cancer
cells. In particular, the role of soluble factors is taking on major importance after some
interesting findings. For example, Drs Blom and Okroj, in collaboration with our group,
have demonstrated that functional C4BP and factor I, released from NSCLC cell lines,
are able to decreased C3 deposition and complement-dependent lysis124. In addition, our
group has shown that factor H is frequently expressed and secreted by NSCLC.
Released factor H is able to inhibit the activation of complement in vitro and
downregulation of factor H in A549 cells, a NSCLC cell line that expresses factor H,
reduces its growth in vivo122-123. These results suggest that factor H is protecting
efficiently those cells from complement attack.
Due to the shielding role of complement inhibitors, lung tumor cells may resist
complement attack and contribute to the generation of an inflammatory
microenvironment. As indicated in the previous section, this inflammatory
microenvironment may contribute to the malignant development of lung cells.
HYPOTHESIS AND OBJECTIVES
Hypothesis and objetives
27
HYPOTHESIS
The hypothesis of the present work is that tumor cells express on their surface tumor-
associated antigens that have the potential to activate the complement system. At the
same time, tumor cells have developed mechanisms that make them resistant to
complement cytotoxic effects. The subsequent effects of complement activation on
complement-resistant cancer cells play a pivotal role in the development of malignant
features, ending in tumor progression.
OBJECTIVES
1. To study complement activation by the alternative and classical pathways on
non-small cell lung cancer (NSCLC) cells and normal bronchial epithelial cells.
2. To study the expression, at the mRNA and protein levels, of the main
complement inhibitory proteins in NSCLC and normal bronchial epithelial cells.
3. To study the role of complement in a syngeneic mouse model of lung cancer,
with special interest in tumor-associated characteristics of the microenvironment such as
angiogenesis and immunosuppression.
4. To evaluate the tumor-promoting effects of complement on NSCLC cells.
MATERIALS AND METHODS
Materials and methods
31
1. MATERIALS
1.1. Animals
The experimental protocols were reviewed and approved by the Institutional
Animal Care and Use Committee of the University of Navarra.
Mice
C57BL wild-type and C3 knockout female mice of 8-12 weeks (The Jackson
Laboratory) were used for the syngeneic experiments with 3LL cells.
Rabbits
Polyclonal antibodies against the A549 cell line were generated in 2-3 kg female
New Zealand White (NZW) rabbits (San Bernando).
1.2. Clinical samples from lung cancer patients
Samples were collected at the Clínica Universidad de Navarra. All protocols were
approved by the Institutional Ethical Committee, and all patients gave informed
consent.
Plasma from 20 NSCLC patients and 20 healthy donors -matched by age and
smoking history- were used. Tumors included both adenocarcinomas and squamous cell
carcinomas in stages III and IV. Samples were stored at -80ºC.
Bronchoalveolar lavage samples from subjects undergoing diagnostic
bronchoscopy were also included in this study. Control samples (n=24) were obtained
from patients with non-malignant lung disease. Tumor samples (n=57) were obtained
from lung cancer patients with different histologies (both SCLC and NSCLC) in stages
III and IV. Samples were stored at -80ºC. The clinicopathological characteristics of this
series are described in a recent publication146.
1.3. Normal human serum
Normal human serum (NHS) from 12 healthy donors was used as a source of
complement factors. Blood was collected in vacuum tubes (BD Vacutainer) and allowed
to clot for 30 minutes at 37ºC. Then tubes were chilled for 15 minutes and centrifuged
at 3345 g. Serum was separated from clot and centrifuged again in order to get a cleaner
Materials and methods
32
serum. All the serum samples were pooled together, divided in aliquots and stored at -
80ºC.
To check that the NHS complement activity was within the normal values, the
alternative pathway hemolytic assay (APH50) was performed following the procedure
explained in section 2.1.5.
Heat inactivated normal human serum (HI-NHS) was obtained by heating NHS
during 30 minutes at 56ºC.
1.4. Plasma
Plasma was obtained for factor H purification. Whole blood was collected in
vacuum tubes containing EDTA (BD Vacutainer) and centrifuged for 15 minutes at 4ºC.
Plasma was separated from cells and buffy coat and kept at -20ºC.
1.5. Human non-small cell lung cancer cells
The following human non-small cell lung cancer (NSCLC) cell lines, available at
the American Type Culture Collection (ATCC), were used:
A549: bronchoalveolar carcinoma.
H157: squamous cell carcinoma.
H1299 and H460: large cell carcinoma.
H23 and HTB58: adenocarcinoma.
H727: carcinoid.
1.6. Immortalized normal human epithelial cells
Two types of immortalized normal epithelial cells were used:
- Immortalized HBECs (human bronchial epithelial cells): established by
introducing Cdk4 and hTERT into normal HBECs147. We used two HBEC lines,
from two different donors: HBEC-3KT and HBEC-10KT. Besides, we used HBEC-
3KT cells stably knocked down for p53 (HBEC-3KT p53). The three cell lines were
kindly provided by Dr JD Minna (University of Texas Southwestern Medical
Center, Dallas).
- BEAS-2B cells: established from a normal human bronchial epithelium and
transformed by an SV40 hybrid virus148-149. BEAS-2B are available at ATCC.
Materials and methods
33
1.7. Endothelial cells
Primary HUVECs (human umbilical vein endothelial cells) were isolated from
freshly collected umbilical cords (see protocol in section 2.1.1.), or obtained from
ATCC.
1.8. Murine lung cancer cells
The Lewis lung carcinoma cell line (3LL or LLC) was obtained from ATCC. This
cell line was established from the lung of a C57BL mouse bearing a tumor resulting
from an implantation of a primary Lewis lung carcinoma150.
1.9. Hybridoma
The hybridoma cell line OX-23, which produces a monoclonal antibody against
factor H, was obtained from the European Collection of Cell Cultures (ECACC).
1.10. Factor H
Factor H was purified following the method described by Sim et. al151. The final
concentration of factor H was 450 µg/ml in phosphate buffered saline (PBS: 8 mM
Na2HPO4, 2 mM NaH2PO4, 150 mM NaCl, pH 7.4).
1.11. C5aR-1 antagonist
Acetylatedphenylalanine–(ornithine-proline-(D)cyclohexylalanine-tryptophan-
arginine) or AcF(OP(D)ChaWR, a C5aR-1 antagonist, was used in in vitro and in vivo
experiments for the blockade of C5aR-1. The antagonist was used in vitro at 100 ng/ml
and in vivo at a dose of 1 mg per kg body weight. Mice received an injection every 3
days, starting 4 hours before the inoculation of tumor cells. A control peptide, the cyclic
hexapeptide acetylated phenylalanine–(ornithineproline-(D)cyclohexylalanine-alanine-
(D)arginine or AcF(OP(D)ChaA(D)R), was used at the same concentrations. Both
peptides are derived from the C terminal decapeptide sequence of C5a, and were
modified to produce either complete antagonists of the receptor or peptides with no
activity152. Both peptides were synthesized in PeptideSynthetics (Fareham, UK) with a
purity higher than 95%.
Materials and methods
34
2. METHODS
2.1. Methods for the generation of experimental material
2.1.1 Cell culture
a. Cell culture conditions
NSCLC and Lewis lung carcinoma cell lines were maintained in RPMI-1640
growth medium (Lonza) with Glutamax, 10% Fetalclone (Thermo), 100 U/ml penicillin
and 100 µg/ml streptomycin (Lonza). Henceforth, this medium will be referred to as
complete medium. For some experiments, cells were kept in RPMI medium without
Fetalclone or antibiotics; this medium will be referred to as basal RPMI.
HBEC-KT cells were maintained in keratinocyte serum free medium
supplemented with human recombinant EGF and bovine pituitary extract (Gibco).
BEAS 2B cells were maintained in DMEM:Ham’s F-12 (1:1) supplemented with
human recombinant EGF, insulin, transferrin, epinephrine, triiodothyronine, retinoic
acid, hydrocortisone, gentamicin sulfate/amphotericin-B and bovine pituitary extract
(media and supplements were from Lonza).
HUVECs were maintained in EBM-2 basal medium containing human
recombinant EGF, VEGF, FGF, R3-IGF-1, hydrocortisone, ascorbic acid, heparin,
gentamicin sulfate/amphotericin-B and 2% fetal bovine serum (all from Lonza);
henceforth, this medium with supplements will be referred to as EGM-2MV. For some
assays, HUVECs were kept overnight in EBM-2 basal medium. HUVECs were either
commercial (ATCC) or obtained from umbilical cords. Human umbilical cords were
obtained from pregnancies and processed within 24 hours. For extracting the vein
endothelial cells, cords were cleaned with 70% ethanol to wipe out the excess of blood
and both ends were cut. Two canules were inserted into the vein of each extreme and
clamped with a thread. Then 20 ml syringes where attached to them. In the first place,
PBS was pushed through the vein to eliminate the blood within it; then, 15 ml of
colagenase at 145 units/ml were pushed and retained inside the vein. The cord was
incubated for 15 minutes at 37ºC for the enzyme to digest the tissue and liberate
endothelial cells. After incubation, the colagenase solution was collected, the remains
were washed with PBS, and gathered together with the first solution. The suspension of
Materials and methods
35
cells was spun down and the pellet resuspended in 199 medium (Lonza) with 10% fetal
bovine serum (FBS), 100 U/ml penicillin and 100 µg/ml streptomycin. Cell suspension
was placed in 25 mm3 flasks previously coated with 0.1% gelatin (Sigma), and
incubated overnight at 37°C in a 5% CO2 atmosphere. One day later, medium was
removed and replaced with EGM-2MV.
The mouse hybridoma cell line OX-23 was grown in complete RPMI medium.
For antibody production, hybridoma cells were kept for 10 days in 175 mm2 flasks
containing 25 ml of protein free medium (CD Hybridoma Medium, Invitrogen)
supplemented with 8 mM L-glutamine.
In all cases, cells were grown at 37ºC in a 5% CO2 atmosphere.
Every month and before cell freezing, a biochemical test for mycoplasma
detection (MycoAlert Mycoplasma Detection Kit, Lonza) was performed. The test
consists in the detection of certain mycoplasma enzymes in the medium. Viable
mycoplasma are lysed and the enzymes react with the kit’s substrate catalyzing the
conversion of ADP to ATP, which is measured by luminescence. To perform the test,
conditioned media from 72-96 hour cell cultures were collected and centrifuged at 500 g
for 5 minutes. Afterwards, 100 µl of medium were mixed with 100 µl of MycoAlert
Reagent and 100 µl of MycoAlert Substrate. Luminescence was measured before and
after adding the substrate. If the ratio between the second and the first measure was
higher than one, the cell culture was considered contaminated and discarded.
b. Subcultures
All the cell lines used in this study were adherent to culture flasks, except 3LL
cells which were semi-adherent. Adherent cells were subcultured by removing growth
medium and rinsing them with PBS without Ca2+ or Mg2+. Then, cells were detached
from the flasks by incubation with trypsin/EDTA (Invitrogen) for approximately 5
minutes at 37ºC. Afterwards, trypsin was neutralized with fresh complete medium, cells
were centrifuged at 470 g for 5 minutes, and the pellet was resuspended in fresh
medium. One half or one fourth of the total volume (depending on the cell line) was
mixed with fresh medium and transferred to the culture flask again. The procedure for
the 3LL cells was the same, except that the medium where the cells were grown was
used to neutralize the trypsin. In that way, the non-adherent cells were subcultured
together with the adherent cells.
Materials and methods
36
c. Freezing and thawing method
Cells were frozen in order to preserve and maintain a stock of all the cell lines.
For freezing, cells were grown until they reached an 80% confluence. After
trypsinization and neutralization of the trypsin with fresh medium, cells were
resuspended at a concentration of approx. 5 x 106 cells/ml in freezing medium, which
contained 20% dimethylsulphoxide (DMSO, Sigma) in Fetalclone. Samples were
aliquoted in cryotubes (Nalgene) and placed into an isopropanol container, which was
kept at -80ºC for 24 hours. After this time, the cryotubes were transferred to liquid
nitrogen tanks for long-term storage.
For thawing cells, cryotubes were placed for a few minutes in a 37ºC bath, until
the freezing medium got liquid. Then, cells were transferred rapidly to 15 ml tubes with
fresh medium and centrifuged at 470 g for 5 minutes, in order to get rid of the DMSO.
The pellets of cells were seeded in 25 mm3 flasks filled with fresh medium and
incubated at 37ºC in the cell incubator until the confluence was 80-90%. Then, cells
were trypsinized and transferred to 100 mm2 flasks.
2.1.2 Generation of OX-23 anti-factor H antibodies
a. Hybridoma culture
After thawing, the hybridoma cell line OX-23 was grown and subcultured in
complete RPMI medium until confirmation of antibody production by an immunoassay
(see section 2.1.2.e). Once confirmed, cells were grown in protein free medium (CD
Hybridoma Medium, Invitrogen) supplemented with 8 mM glutamine. Conditioned
medium from cells incubated for three days was checked again in order to confirm the
antibody production in the new medium.
b. Production of conditioned media
In order to produce mouse IgG1 monoclonal antibodies against factor H, 20
million OX-23 hybridoma cells were seeded in 175 mm2 flasks with 25 ml of protein
free hybridoma medium supplemented with 8 mM L-glutamine. Cells were grown in
this medium for 10 days, without any change, to increase antibody concentration.
Afterwards, conditioned medium was collected in 50 ml tubes and centrifuged for 5
minutes at 470 g. Cells were discarded, and supernatants from different flasks were put
Materials and methods
37
together and kept at -20ºC until antibody purification. An aliquot was saved apart to test
its immunoreactivity (see section 2.1.2.e).
c. Purification
For antibody purification, 100 ml of supernatant was used. First, proteins were
precipitated by mixing the supernatant with 76.1 g of ammonium sulfate (Sigma), added
carefully to obtain a saturated dilution. The solution was maintained with slight shaking
during 16 hours at 4ºC. Then, the solution was filtrated using a Whatman paper and a
Buchner funnel connected to a vacuum pump. The precipitate was recovered from the
paper by placing it in a 50 ml tube with 10 ml of PBS and shaking it slightly for five
minutes. Then, the paper was discarded and the solution was centrifuged for 5 minutes
at 3345 g at 4ºC. The paper-free solution was introduced into a dialysis membrane and
placed on a 500 ml beaker filled with PBS. The dialysis was performed at 4ºC for 2
days. The buffer was changed twice. After dialysis, the immunoreactivity was evaluated
in an aliquot of the solution (see section 2.1.2.e). Monoclonal antibodies within the
solution were purified by chromatography (AKTApurifier, Amersham) using a HiTrap
Protein G column (Amersham). The elution buffer was 100 mM glycine (Sigma) at pH
2.5. The elution buffer was neutralized by adding 50 µl of 1 M Trizma-base (Sigma) at
pH 9. Finally, the antibodies were dialyzed against PBS as mentioned above.
d. Determination of protein concentration
Protein concentration was measured with the Bradford reagent (BioRad). A BSA
(bovine serum albumin) standard curve was used. Samples or standards (10 µl) were
added to the plates in duplicate; then, 190 µl of Bradford reagent diluted 1:5 in distilled
water was added to the wells. After 10 minutes of incubation, absorbance at 595 nm and
450 nm was measured in a Sunrise microplate reader with the Magellan software
(Tecan).
e. Titration
An immunoreactivity assay was performed in order to evaluate the presence of
OX-23 antibodies in four different samples: conditioned medium previous to the
purification procedure, solutions obtained previous and after HPLC purification, and
final dialyzed solution.
Materials and methods
38
For the assay, 50 ng of purified factor H (Sigma) in PBS was added per well in
one half of a polystyrene 96 well plate (flat bottom, non-sterile, Costar). The other half
was filled with 50 µl of PBS per well, and was used to calculate the background for
each antibody dilution. Factor H was allowed to attach to the plate for 1 hour at room
temperature. Then, wells were washed three times with assay buffer (PBS with 1% BSA
and 0.1% Tween-20; all reagents from Sigma) and blocked with 200 µl of the same
buffer overnight at 4ºC. Afterwards, 50 µl of serial dilutions of the samples in assay
buffer (from 1:10 to 1:106) was added in triplicate onto wells with and without factor H.
As a positive control, a previous OX-23 batch was used in the same conditions. Plates
were incubated at room temperature for 2 hours; then, wells were washed twice with
assay buffer, and 50 µl of peroxidase-conjugated rabbit anti-mouse IgG antibody
(Amersham) diluted 1:1000 in assay buffer was added, and incubated at room
temperature for 30 minutes. After two washes, the assay was developed with 200 µl of
Sigmafast OPD (o-phenylenediamine dihydrochloride), previously prepared according
to the manufacturer’s instructions. Absorbance was measured at 450 nm in a microplate
spectrophotometer (Tecan).
f. SDS-PAGE and Coomassie staining
To check antibody purity, an electrophoresis was performed in a polyacrylamide
gel with sodium dodecyl sulfate (SDS) under reducing and non-reducing conditions,
followed by Coomassie staining (Figure 5).
Precast 4-12% Bis-tris gradient gels (NuPAGE, Invitrogen) were used. Ten
microliters of purified antibody (5 µg of proteins, approximately) were mixed with
commercial loading buffer NuPAGE LDS Sample Buffer (Invitrogen). For reducing
conditions, 10% β-mercaptoethanol was also added. The mix was boiled for 5 minutes,
snap-cooled to avoid protein refolding, and loaded into the gel.
Electrophoresis was carried out during 1 hour at 120 V in running buffer (50 mM
MOPS, 50 mM Trizma-base, 0.1% SDS and 1 mM EDTA, pH 7.7; all from Sigma).
After electrophoresis, gels were carefully placed in cuvettes, and rinse with ultrapure
water. Then, cuvettes were filled with a Coomassie blue solution (Pierce) and gels were
incubated with slight shaking for one hour. To remove excess of colorant, the gels were
washed three times with ultrapure water and photographed in a digital gel
documentation system (Gel Doc, BioRad).
Materials and methods
39
150 kDa
45 kDa
30 kDa
150 kDa
45 kDa
30
OX 24 OX 23
RNRNR R
150 kDa
45 kDa
30 kDa
150 kDa
45 kDa
30
OX 24 OX 23
RNRNR R
Figure 5. Coomassie staining obtained after electrophoresis in reducing (R) and non reducing conditions (NR) of purified OX23 antibody. OX24 antibody, previously purified in our laboratory, was used as positive control. Staining shows three bands: in non reducing conditions the 150 kDa band corresponds to the complete antibody molecule; and in reducing conditions the 45 kDa and 30 kDa bands correspond to the heavy and the light chains of the antibody, respectively.
2.1.3 Anti-A549 polyclonal antibody generation
a. Production of antibodies in NZW rabbits
Before the procedure, A549 cells were grown and tested for mycoplasma. Cells
were cultured in 175 mm2 flasks until 80% confluence, and then were harvested. Three
aliquots with 107 cells in 500 µl of PBS were prepared and kept at -80ºC until
inoculation. Female NZW rabbits were also prepared before the procedure as follows.
The hair from the dorsal area was removed using hair-removing cream (Veet), and
periphery blood was collected by venepuncture of the marginal ear vein in order to
obtain preimmune serum. For this purpose, blood was allowed to clot at 4ºC for 16
hours and then centrifuged at 677 g on a benchtop centrifuge for 15 minutes. Serum was
transferred to new tubes and centrifuged again. Finally, serum was collected, heated for
30 minutes at 56ºC to inactivate complement, and kept at -20ºC until further use.
Immunization was performed by three injections of 107 cells at 15-day intervals.
The first injection was prepared by mixing cells with 500 µl of Complete Freud
Adjuvant (Difco Laboratories) using the Sonifier 150 (Branson) for 30 seconds. The
emulsion formed was transferred to a 1 ml syringe (BD Plastipack) and maintained on
ice till the moment of injection. Then, it was injected intradermically in six different
spots on the dorsal area of the rabbit, using a 25G needle (BD Microlance). Fifteen days
later, another emulsion was prepared with 107 cells, but in this case with 500 µl of
Incomplete Freud Adjuvant (Difco Laboratories). The injection was done this time
Materials and methods
40
subcutaneally in two different spots of the dorsal area of the rabbit. Another fifteen days
later, the third and last emulsion was prepared in the same way as the second one, but
was injected intramuscularly in one spot on a footpad of the rabbit.
Fifteen days later from the last injection, whole blood from rabbits was collected.
For the procedure, animals were anesthetized with 35 mg/kg ketamine (Imalgene 1000,
Merial), 5 mg/ml xylazine (Rompun 2%, Bayer) and 1 mg/ml acepromazine (Calmo
Neosan; Tpizer). Blood was collected directly from the heart using 10 ml syringes with
25G needles. After transferring it to 50 ml tubes, blood was allowed to clot and then
centrifuged. The antiserum was finally prepared by heating the supernatant for 30
minutes at 56ºC. Then, it was aliquoted and maintained at -20ºC for further
experiments.
b. Immunoreactivity assay
Immunoreactivity against A549 cells was confirmed by flow cytometry. A549
cells (105) were incubated with 50 µl of serial dilutions of the antiserum in binding
buffer (PBS with 1% BSA and 0.1% sodium azide; all reagents from Sigma), ranging
from 1:25 to 1:200; or preimmune serum diluted 1:25 as a negative control. After 30
minutes at 4ºC, cells were washed twice with binding buffer and resuspended in 50 µl of
FITC conjugated anti-rabbit IgG antibody (Molecular Probes) diluted 1:100 in binding
buffer. Cells were again incubated for 30 minutes at 4ºC and washed twice in binding
buffer. Finally, cells were resuspended in 100 µl of FACS Flow (BD Bioscience) with
7-amino-actinomycin D (7AAD; Sigma) at 50 µg/ml and incubated for 10 minutes at
4ºC protected from light. Afterwards, cells were analyzed in a FACSCalibur flow
cytometer with the CellQuest Pro software (BD Biosciences).
The same assay was later performed in all NSCLC and normal epithelium cell
lines, to check the ability of the antibody to bind to the different cell types.
2.1.4 Factor H purification
To purify factor H from plasma, we followed a method already described151.
a. Preparation of columns
Purification of factor H and FHL1 implicates passing of plasma through three
different columns previously prepared according to manufacturer’s instructions.
Materials and methods
41
The first one, a lysine-Sepharose column (Lysine Sepharose TM 4B; Amersham),
removes the plasmin/plasminogen. The second one, the preadsorption column, is a
column of IgG-conjugated Sepharose (AminoLink Plus Immobilization Kit), that
removes proteins that bind to immunoglobulins or to modified agarose (e.g. C1q,
fibronectin, rheumatoid factors). This column was prepared in three phases, which
implies: protein immobilization, blockade of remaining active sites, and column
washing; according to the manufacturer’s instructions. The third column contains the
monoclonal antibody anti-factor H OX-24 immobilized on Sepharose, and retains factor
H and FHL-1.
b. Factor H purification
For factor H purification, columns were equilibrated in running buffer (PBS with
15 mM EDTA, pH 7.2). Plasma was prepared by adding EDTA (Sigma) and Pefabloc
[4-(2-aminoethyl) benzenesulfonyl fluoride hydrochloride, AEBSF, Fluka], both to a
final concentration of 0.5 mM. Plasma (5 ml) was added to the first column, where the
plasmin/plasminogen was retained, and eluted with running buffer. Pefabloc was added
again to the elute to made a final concentration of 0.5 mM. The sample was added to the
second column, and eluted with running buffer. Finally, the elute was passed through
the third column, where factor H was retained. The column was then washed with TBS
(25 mM Tris-HCl, 140 mM NaCl, 0.5 mM EDTA, pH 7.4) and factor H and FHL-1
were eluted with 3M MgCl2.
The elute was then dialyzed twice; first, against TBS for 16 hours at 4ºC, and
second, against PBS for 24 hours. The final solution was concentrated by centrifugation
at 2000 g and 4ºC using Amicon centricons with a 10 kDa cutoff (Millipore).
c. Factor H integrity
To check factor H purity and integrity, an electrophoresis (SDS-PAGE) and
Coomassie staining were done as explained above (section 2.1.2.f). Under reducing
conditions, the gel showed two bands of 150 and 42 kDa, which correspond to FH and
FHL-1, respectively (Figure 6).
Materials and methods
42
150 KDa
c FH p FH
NR R
c FH p FH
42 KDa
150 KDa
c FH p FH
NR R
c FH p FH
42 KDa
Figure 6. Factor H and FHL-1 purification. Coomassie staining obtained after electrophoresis in reducing (R) and non reducing conditions (NR) of purified factor H and FHL-1 (p FH). Comercial factor H (c FH) was used as positive control. Staining showed two bands corresponding to factor H and FHL-1.
d. Factor H quantification by ELISA
An ELISA assay was performed in order to calculate factor H concentration at
different stages throughout the procedure. For this purpose, several aliquots were
collected from these solutions: starting plasma, elutes from each column, elute in the
concentration step, and final solution.
To measure factor H/FHL-1 concentration, a polystyrene 96-well plate (flat
bottom, non sterile, Costar) was coated with 50 ng/well of anti-factor H monoclonal
antibody OX-23 in 50 µl of 50 mM sodium bicarbonate (pH 8), and incubated for 1
hour at room temperature. After three washings with assay buffer (PBS with 1% BSA
and 0.1% Tween 20; all from Sigma), plates were blocked overnight at 4ºC with the
same buffer. Afterwards, 50 µl of each sample collected throughout the purification
procedure, or serial dilutions of commercial factor H (Sigma) ranging from 1.5 to 200
ng/ml, were added in triplicate, and the plate was incubated for two hours at room
temperature. Wells were washed three times with assay buffer and 50 µl of goat anti-
factor H antibody (Quidel) diluted 1:1000 in assay buffer was added. After 30 minutes
of incubation at room temperature, wells were washed three times and incubated for 30
minutes with peroxidase-conjugated polyclonal rabbit anti-goat IgG antibody (Dako)
diluted 1:4000 in assay buffer. Finally, the assay was developed with OPD (Sigma),
according to manufacturer’s instructions. The plate was read at 450 nm in a microplate
reader (Tecan). The final concentration of factor H was 450 µg/ml.
Materials and methods
43
2.1.5 Alternative pathway hemolytic assay (APH50)
APH50 was performed as previously described153. This method is used to
determine the amount of complement required to lyse 50% of red cells in 1 ml of a
solution that contains 108 erythrocytes/ml. It is expressed in APH50 units and in humans
its value is comprised between 38 and 93 U/ml154.
Rabbit erythrocytes were used for the procedure, as they activate the alternative
pathway of complement spontaneously. For this purpose, blood from female NZW
rabbits was collected directly from hearts as explained in section 2.1.3. After allowing
blood to clot for 2 hours, the pellet of cells was transferred to a 50 ml tube and washed
three times with ten times the pellet’s volume in Alsever’s solution (113.7 mM
dextrose, 27.2 mM sodium citrate, 2.6 mM citric acid, 72 mM sodium chloride, pH 6.2;
all reagents from Sigma) at 677 g during 10 minutes, and resuspended in this solution
for maintenance at 4ºC.
For the procedure, a 108 erythrocytes/ml dilution is needed. To obtain this
concentration of cells, 1 ml of cells kept on Alsever’s solution was centrifuged, and the
pellet resuspended in 15 ml of GVB+ (3.1 mM barbituric acid, 1.8 mM barbital, 145
mM NaCl, 0.1 % gelatin, 0.1 M EGTA, 0.1 mM MgCl2, pH 7.4). Erythrocytes in 100 µl
of this solution were lysed with 2900 µl of distilled water. Then, serial dilutions were
prepared (1:2, 1:4 and 1:8) and their absorbances measured at 412 nm. Beer-Lambert
equation was used to calculate each dilution’s concentration knowing that 0.294 is the
absorbance of a 108 erythrocytes/ml dilution. Afterwards, a solution with 2 x 108
erythrocytes/ml was prepared in GVB+ using the stock solution maintained in Alsever’s.
Hemolytic assays were performed in U bottom 96-well plates (Costar). Different
volumes: 50, 40, 30, 20, 10 or 5 µl, of 1:2 diluted serum in GVB+ were added in
triplicate. A final volume of 50 µl was obtained adding GVB+. For the 0% and 100%
lysis, 50 µl of GVB+ or distilled water were added, respectively. Then, 50 µl of 2 x 108
erythrocyte/ml solution was added to each well and the plate incubated for 30 minutes
at 37ºC with slight agitation. Afterwards, 150 µl of chilled 0.9% NaCl (Sigma) was
added to each well and the plate centrifuged at 1000 g for 5 minutes at 4ºC.
Supernatants (120 µl) were transferred to flat bottom 96-well plates, and absorbance
was read at 412 nm in a microplate reader (Tecan).
Materials and methods
44
To determine APH50, first, the percentage of lysis (y) for each serum dilution was
calculated as:
Absorbance serum dilution x – Absobance 0% lysis
Absobance 100% lyisis – Absorbance 0% lysis
y =Absorbance serum dilution x – Absobance 0% lysis
Absobance 100% lyisis – Absorbance 0% lysis
Absorbance serum dilution x – Absobance 0% lysis
Absobance 100% lyisis – Absorbance 0% lysis
y = x 100Absorbance serum dilution x – Absobance 0% lysis
Absobance 100% lyisis – Absorbance 0% lysis
y =Absorbance serum dilution x – Absobance 0% lysis
Absobance 100% lyisis – Absorbance 0% lysis
Absorbance serum dilution x – Absobance 0% lysis
Absobance 100% lyisis – Absorbance 0% lysis
y =Absorbance serum dilution x – Absobance 0% lysis
Absobance 100% lyisis – Absorbance 0% lysis
Absorbance serum dilution x – Absobance 0% lysis
Absobance 100% lyisis – Absorbance 0% lysis
y =Absorbance serum dilution x – Absobance 0% lysis
Absobance 100% lyisis – Absorbance 0% lysis
Absorbance serum dilution x – Absobance 0% lysis
Absobance 100% lyisis – Absorbance 0% lysis
y = x 100
Then, the von Krog equation for the sigmoid curve of hemolysis was used:
x = K (y/(1-y))1/n
In this case, x is the volume of serum added to the tube (µl) and 1/n is 0.2 ± 10%,
depending on conditions used. K is a constant. At the point of 50% hemolysis:
(y/(1-y))1/n = 1, hence x = K
To calculate the value of K, first, the value of (y/(1-y)) for each serum dilution was
calculated, and then, log x was plotted against log (y/(1-y)) (Figure 7). The intercept on
the x-axis was K´ (as log 1= 0). K was the K´ antilog.
The number of APH50 U/ml of the original serum was obtained as:
APH50 = (1000/K) x 2 (which is the dilution of serum in our particular case).
-20
0
20
40
60
80
100
120
5 10 20 30 40 50-1,2
-0,8
-0,4
0
0,4
0,8
1,0 1,3 1,5
Lysis
(%
)
Lo
g (
y/(
1-y
))
Serum volume (µl) Serum volume (Log)
A B
-20
0
20
40
60
80
100
120
5 10 20 30 40 50-1,2
-0,8
-0,4
0
0,4
0,8
1,0 1,3 1,5
Lysis
(%
)
Lo
g (
y/(
1-y
))
Serum volume (µl) Serum volume (Log)
A B
Figure 7. Calculating APH50. (A) The graph shows the percentage of lysis vs. volume of diluted serum for a test sample. (B) The graph shows a log-log plot of (y/(1-y)) vs. volume of diluted serum. The interception with X-axis corresponds to K’ value. In this particular example K value was 22.5 µl of serum, which corresponded to 88.8 U/ml of APH50.
In our experiments we used pools of sera with APH50 values within the normal
values in humans, 38 and 93 U/ml154.
2.1.6 Total RNA extraction and retrotranscription
To extract total RNA from cell cultures, cells were seeded in 100 mm2 plates until
reaching 80% confluence. Then, cell medium was removed and 1 ml of Ultraspec
(Biotecx) was added. Cell homogenates were detached from plates using a scraper,
mixed with 200 µl of chloroform and vortexed. Mixtures were laid for 5 minutes on ice
Materials and methods
45
and then centrifuged at 12,000 g at 4ºC for 15 minutes. The aqueous phase was
carefully transferred to a clean tube and the same volume of isopropanol was added.
Then, tubes were slightly mixed and kept for 10 minutes at 4ºC or frozen at -80ºC, as
cold helps precipitation. After the precipitation step, samples were centrifuged for 10
minutes at 12,000 g and 4ºC. The white pellets were washed twice with 50 µl of 75%
ethanol in DEPC water (Sigma) and dried at 50ºC. Finally, extracted RNA was
resuspended in 20 µl DEPC water.
From frozen tumors, RNA was extracted using the RNeasy Micro kit (QIAGEN).
Before starting the protocol, samples were homogenized by mixing the tissues in a hand
held homogenizer with three-four crystal beads, a little spoon of quartz (Merk) and
700 µl of β-mercaptoethanol diluted 1:100 in RLT buffer (provided in the kit). After
three cycles of shaking, the mixtures were centrifuged at high speed for 3 minutes and
supernatants were transferred into QIAshredder spin columns (QIAGEN). The columns
were centrifuged at high speed for 3 minutes in a benchtop centrifuge and the elutes
were used for RNA extraction following the RNeasy Micro kit protocol.
RNA samples extracted from cultured cells or tumor tissues were analyzed for
integrity by agarose gel electrophoresis, and their concentrations measured using a
Nanodrop 1000 spectrophotometer (Thermo Scientific). RNA was stored at -80ºC until
further use.
Complementary DNA (cDNA) was obtained by standard retrotranscription, using
the SuperScript III kit (Invitrogen).
For the procedure, the following mixture was prepared:
Total RNA (extracted from cells or tissues) 1 µg
Random deoxynucleotide hexamers (60 ng/µl) 1 µl
dNTPs (2.5 mM, each) 1 µl
DEPC water (Sigma) up to 11 µl
The mixture was incubated for 5 minutes at 65ºC in a PTC-100 thermal cycler
(BioRad). After this first step of denaturalization, samples were snap-cooled on ice, to
avoid formation of secondary structures within the RNA template, while the following
mixture was prepared:
Materials and methods
46
First strand reaction buffer (5X) 4 µl
DTT (0.1 M) 1 µl
RNase OUT (40 U/µl) 1 µl
SuperScript III 1 µl
This mixture was added to samples and the retrotranscription was performed in
the thermal cycler using two temperature steps: 50 minutes at 50ºC for the enzyme
reaction, and 15 minutes at 70ºC for the enzyme denaturalization. Then, samples were
placed on ice and 1 µl of RNase H (5 U/µl, Invitrogen) was added in order to remove
RNA templates. Tubes were warmed for 20 minutes at 37ºC and, finally, cDNA was
stored at -20ºC.
2.1.7 Protein extraction
For protein extraction, cells were washed twice with PBS and then lysed during
15 minutes at 4ºC in RIPA buffer (10 mM Tris, 150 mM NaCl, 1% Triton X-100, 1%
sodium deoxycholate and 0.1% sodium dodecyl sulfate, pH 8; all from Sigma) with one
tablet per 10 ml of protease inhibitor cocktail (Roche). Samples were centrifuged at high
speed on a benchtop centrifuge and supernatants containing dissolved proteins were
transferred to clean tubes.
Protein concentration was measured with the Bradford reagent (BioRad) as
explained in section 2.1.2.d.
2.1.8 Genotyping of C3 deficient mice
Genotyping of C3 deficient mice was performed every six months as follows. Ear
biopsies from C3 knockout mice and wild type controls were digested at 55°C overnight
in lysis buffer (1 M Tris, 0.5 M EDTA, 10% SDS, 5 M NaCl and 0.1 mg/ml proteinase
K, pH 8.5). After centrifugation at high speed for 5 minutes to remove debris,
supernatants were transferred to new tubes and the same volume of 100% isopropanol
(Panreac) was added to them, in order to precipitate genomic DNA. After precipitation,
the pellet was washed in 70% ethanol, dried at room temperature for 10 minutes, and
resuspended in ultrapure water. Finally, DNA concentration was measured using the
Nanodrop 1000 spectrophotometer (Thermo Scientific). Genotyping was performed by
conventional PCR following the protocol and primers from The Jackson Laboratory
(Table 3).
Materials and methods
47
Table 3. Primers used in the genotyping of wild-type and C3 knockout mice by conventional PCR.
2.2. Methods for the study of complement activity on cells
2.2.1 Complement deposition on cells after activation of the alternative or
classical pathways
For the activation of the alternative pathway of complement, cells were detached
from culture dishes with trypsin/EDTA (Lonza), washed twice with PBS, and
resuspended in veronal buffer with magnesium (1.8 mM barbital, 3.1 mM barbituric
acid, 141 mM NaCl, 0.5 mM MgCl2, pH 7.4; hereafter referred as VB+). Then, cells (2 x
105) were incubated with 10% NHS or HI-NHS for 30 minutes at 37ºC. After two
washes in PBS, deposition of C3, iC3b, C5b or C5b-9 was assessed by incubating cells
with the specific primary antibodies diluted in binding buffer (PBS with 1% BSA and
0.1% sodium azide; all from Sigma) as follows: 1:100 FITC-conjugated polyclonal goat
anti-human C3 (ICN Biomedicals), 1:200 monoclonal mouse anti-human iC3b
neoantigen (Quidel), 1:100 monoclonal mouse anti-human C5b-9 (clone aE11; Dako),
or 1:100 polyclonal goat anti-human C5 (Quidel). After 30 minutes at 4ºC, cells were
washed twice in binding buffer and incubated in 1:100 Alexa Fluor-conjugated goat
anti-mouse IgG (Invitrogen) or FITC-conjugated rabbit anti-goat IgG (Invitrogen) for
30 minutes at 4ºC. Afterwards, cells were washed twice in binding buffer and
resuspended in 100 µl of FACS flow (BD Biosciences) with 7-AAD at a final
concentration of 50 µg/ml. After ten minutes of incubation with the dye, flow
cytometric analyses were performed.
For the classical pathway activation, cells were treated as explained above but
resuspended in veronal buffer with magnesium and calcium (1.8 mM barbital, 3.1 mM
barbituric acid, 141 mM NaCl, 0.5 mM MgCl2, 0.15 mM CaCl2, pH 7.4; hereafter
referred as VB2+). Cells (2 x 105) were incubated with the rabbit antiserum against A549
cells diluted 1:50 in a total volume of 100 µl. After two washes with VB2+, cells were
resuspended in 10% NHS or HI-NHS in VB2+. From this step forward, cells were
treated as explained for the alternative pathway.
Gene Sense Antisense
C3 WT 5´ATCTTGAGTGCACCAAGCC3´ 5´GGTTGCAGCAGTCTATGAAGG3´
C3 MUT 5´CTTGGGTGGAGAGGCTATTC3´ 5´AGGTGAGATGACAGGAGATC3´
GAPDH 5´ACTTTGTCAAGCTCATTTCC3´ 5´TGCAGCGAACTTTATTGATG3´
Materials and methods
48
2.2.2 Complement cytotoxicity after activation of the alternative or classical
pathways
The lactate dehydrogenase (LDH) assay, which measures membrane integrity as a
function of the amount of cytoplasmic LDH released into the medium, was used to
determine cytotoxicity due to complement activation on the cell membrane. LDH
activity is determined in a coupled enzymatic reaction: in the first step, NAD+ is
reduced to NADH/H+ by the LDH-catalyzed conversion of lactate to pyruvate. In the
second step, the catalyst (diaphorase) transfers H/H+ from NADH/H+ to the
iodotetrazolium chloride (tetrazolium salt INT) which is reduced to formazan, that is
detected by spectrophotometry.
LDH assay (Cytotoxicity detection kitPLUS, Roche) was performed according to
the manufacturer’s instructions, using the supernatants from cells treated with
complement. In order to obtain the supernatants, 10,000 cells were prepared in GVB2+
(1.8 mM barbital, 3.1 mM barbituric acid, 141 mM NaCl, 0.5 mM MgCl2, 0.15 mM
CaCl2, and 0.1% gelatin, pH 7.4), or in GVB+ (GVB2+ without CaCl2) for the activation
of the classical and alternative pathways, respectively. In the case of the classical
pathway, cells were preincubated for 30 minutes at 4ºC with the antiserum against A549
cells diluted 1:50 in GVB+, and washed twice in the same buffer before incubation with
the human serum. Cells were incubated in GVB+ or GVB2+ containing 10% NHS, 10%
HI-NHS (background lysis control) or 0.5% Triton X-100 (100% lysis control) for 30
minutes at 37ºC; all conditions were prepared in triplicate. After complement activation,
cells were pelleted in a benchtop centrifuge at 500 g for 5 minutes at 4ºC. Supernatants
were recollected for the LDH test.
For the assay, 50 µl of supernatants diluted 1:5 in PBS were placed in triplicate in
a 96-well plate, that made 9 repetitions per condition. In order to subtract the
background absorbance of the serum and Triton X-100, 50 µl of 2% NHS, 2% HI-NHS
and 1% Triton X-100 were added in triplicate. Then, 50 µl of freshly prepared reaction
mixture (one part of Catalyst, that contains diaphorase/NAD+ mixture, per 45 parts of
Dye solution, which contains INT and lactate) were added in each well and incubated
25 minutes at room temperature. The reaction was stopped by adding 25 µl of Stop
solution, that contains Triton X-100 and hydrochloric acid. Finally, absorbance was
measured at 492 nm with a reference wavelength of 690 nm in a microplate reader
(Tecan).
Materials and methods
49
Percentage of lysis was determined using the mean value of triplicates, by the
following formula:
(AbsorbanceNHS – Absobancebackground NHS)-(AbsorbanceHI-NHS – Absobancebackground HI-NHS)
(Absobance100% lyisis - Absorbancebackground 100% lysis)Lysis (%) = x 100
(AbsorbanceNHS – Absobancebackground NHS)-(AbsorbanceHI-NHS – Absobancebackground HI-NHS)
(Absobance100% lyisis - Absorbancebackground 100% lysis)
(AbsorbanceNHS – Absobancebackground NHS)-(AbsorbanceHI-NHS – Absobancebackground HI-NHS)
(Absobance100% lyisis - Absorbancebackground 100% lysis)Lysis (%) = x 100
2.2.3 C5a quantification in plasma samples and bronchoalveolar lavages by
ELISA
C5a concentration in plasma samples from patients with NSCLC or healty donors,
and in BALs from patients with lung cancer (SCLC or NSCLC) or with non-malignant
lung disease, were quantified using the commercial kit Human Complement Component
C5a (R&D), following the manufacturer’s instructions. Prior to the analysis, samples
were spun down at 300 g for 10 min, and the supernatants were collected.
2.2.4 Study of complement inhibitors in NSCLC and normal epithelial cells
a. Determination of mRNA levels by quantitative PCR
Real time PCR was performed to analyze the expression of the main complement
inhibitors on NSCLC and normal bronchial epithelial cells. Total RNA was isolated
from cell cultures and reverse transcribed as described in section 2.1.6.
For the PCR reaction, a mix containing the following components was prepared
per well, all samples were done in triplicate:
cDNA 0.2 µl
Sybr Green PCR Master Mix (Applied Biosystems) 12.5 µl
Primers S and AS (10 µM each), listed in Table 4 0.75 µl each
Ultrapure water up to 25 µl
Reactions were performed using the 7300 Real Time PCR System (Applied
Biosystems) according to the following scheme:
50°C for 2 minutes
95°C for 10 minutes
95°C for 15 seconds
60°C for 1 minute
And an extra dissociation stage, which consists on a continuous increase of
temperature while measuring fluorescence. This permits the detection of the
40 cycles
Materials and methods
50
temperature at which drastic decrease of fluorescence takes place, due to double strand
uncoupling and Sybr Green dye detaching from DNA. This temperature is the amplicon
Tm.
Relative levels of expression were assessed by the threshold cycle (Ct) values and
expressed as a percentage relative to the glyceraldehyde 3-phosphate dehydrogenase
(GAPDH) mRNA.
Table 4. List of primers used for the analysis of complement inhibitors by quantitative PCR.
*FH primers were designed to recognize the 17th exon; therefore, they recognize factor H but not FHL1; ** FH/FHL-1 primers were designed to recognize the second exon of the gen; therefore, they recognize both factor H and FHL-1.
b. Factor H binding to the cell membrane
Binding of factor H/FHL-1 was assessed in NSCLC and normal epithelial cells by
incubating 105 cells in binding buffer (PBS with 1% BSA and 0.1% sodium azide; all
reagents from Sigma) with purified factor H (450 µg/ml in PBS; see section 2.1.4.)
diluted 1:4 in binding buffer or 25% HI-NHS. After 30 minutes of incubation at 37ºC,
cells were washed twice in binding buffer and incubated with goat anti-factor H primary
antibody (Quidel) diluted 1:100 in binding buffer for 30 minutes at 4ºC. Afterwards,
cells were washed in binding buffer and incubated with FITC-conjugated rabbit anti-
goat IgG (Molecular Probes), diluted 1:100 in binding buffer. Finally, cells were
washed twice and incubated with 50 µg/ml of 7-AAD for 10 minutes in the dark, then,
analyzed by flow cytometry in a FACScalibur with CellQuest Pro software (BD
Biosciences).
c. Detection of membrane-bound complement inhibitors
Expression of CD55, CD59 and the protein-kinase casein kinase 2 (CK2) was
assessed in both NSCLC and normal bronchial epithelial cells. Cells were detached
Gene Sense Antisense
FH* 5´GATGTGTATAAGGCGGGTGAG3´ 5´GGAGGTGTCTCTGCATGTTG3´
FH/FHL-1** 5´GAAGGCACCCAGGCTATCTA3´ 5´ATCTCCAGGATGTCCACAGG3´
FI 5´GAGGAAAGCGAGCACAACTG3´ 5´GTCGGGGTGTATCCAGTCTACTA3´
CD46 5´GTGCTGCTCCAGAGTGTAAAGTG3´ 5´CGCTGCCATCGAGGTAAA3´
CD55 5´CCACAAAAACCACCACACC3´ 5´GCCCAGATAGAAGACGGGTAGTA3´
CD59 5´GGAATCCAAGGAGGGTCTGT3´ 5´CAGTCAGCAGTTGGGTTAGGA3´
GAPDH 5´GAAGGTGAAGGTCGGAGTC3´ 5´GAAGATGGTGATGGGATTTC3´
Materials and methods
51
from the culture dishes with trypsin/EDTA (Lonza), washed twice with PBS, and
resuspended in binding buffer. Cells (2 x 105) were washed twice with binding buffer
and incubated for 30 minutes at 4ºC with mouse anti-human CD55 (BRIC 216,
IBGRL), rat anti-human CD59 (YTH51, Serotec), both diluted 1:25 in binding buffer,
or mouse anti-human CK2 (1AD9, Sigma) diluted 1:100 in binding buffer. After
incubation, cells were washed twice in binding buffer and incubated for 30 minutes in
the dark with Alexa Fluor-conjugated goat anti-mouse IgG (Invitrogen) or FITC-
conjugated rabbit anti-rat IgG (Serotec), both at 1:100 dilution in binding buffer.
Finally, inhibitor expression was analyzed by flow cytometry in a FACScalibur with the
CellQuest Pro software (BD Biosciences).
d. Elimination of GPI-anchored proteins from the cell membrane by PI-PLC
treatment
In some experiments, CD55 and CD59 were removed from NSCLC and normal
epithelial cells by a treatment with the enzyme phosphatidylinositol-specific
phospholipase C (PI-PLC) from Bacillus cereus (Molecular Probes). PI-PLC cleaves the
phospholipid phosphatidylinositol (PI), which is part of the molecule
glycosylphosphatidylinositol (GPI) that anchors hundreds of proteins, including CD55
and CD59, to the cell membrane.
To remove these complement regulators from the cell membrane, cells (2 x 105)
were incubated one hour at 37ºC in 100 µl of basal RPMI with 0.5% BSA and one unit
of PI-PLC. Afterwards, cells were washed twice in PBS. These cells were used to
determine C5b-9 deposition and cell cytotoxicity due to the activation of the alternative
pathway of complement.
e. Effect of hypoxic conditions in the expression of complement inhibitors in
NSCLC cells
In order to mimic hypoxic conditions, cells were incubated at 37°C in a
humidified hypoxia chamber (Hypoxia Workstation 400; Ruskinn Technology),
connected to a Ruskinn gas mixer module supplying 94% N2, 5% CO2 and 1% O2. Cells
were either kept in the hypoxic chamber for 48 hours (referred to as hypoxia) or for 24
hours followed by 24 h at normal O2 tension (referred to as hypoxia/reoxygenation).
Control cells (normoxia) were cultured for 48 hours under normal O2 tension. Cells
Materials and methods
52
were then used to analyze the expression of complement inhibitors at both the mRNA
and protein levels.
2.3. Methods to study the in vitro effects of C5a
2.3.1 C5aR-1 expression
a. C5aR-1 mRNA expression
Conventional PCR was used for the screening of C5aR-1 mRNA expression.
Total RNA was extracted from cells and retrotranscribed to cDNA. The PCR reaction
mixture was as follows:
Master Mix buffer (Promega) 12.5 µl
Primer S and AS (10 µM each), listed in Table 5 1 µl
cDNA 2 µl
Ultrapure water up to 25 µl
The reactions were performed according to the following conditions:
94ºC 5 minutes
94ºC 45 seconds
56º 45 seconds
72ºC 1 minute
72ºC 5 minutes
To identify PCR products, electrophoresis in agarose gels was performed. Gels
were casted by mixing agarose (Pronadisa) 1% w/v with 1x Tris-acetate-EDTA buffer
(TAE), then, the mixture was boiled in order to disolve the agarose. Ethydium bromide
(Invitrogen) was added at a final concentration of 0.5 µg/ml. Gels were left to solidify
into gel casters (BioRad). Prior to electrophoresis, samples were mixed with loading
buffer (0.25% bromophenol blue, 0.25% xylenecyanol and 50% glycerol; all from
Sigma) in a 1:5 proportion. Samples were loaded and at least one well per gel was
preserved for a 100 bp molecular weight marker. Electrophoresis was performed at
100 V during 45 minutes. Afterwards, imaging and analysis of the gels was done in a
digital gel scanner (Gel Doc, BioRad).
30 cycles
Materials and methods
53
Table 5. List of primers used in the analysis of C5aR-1 by conventional PCR.
b. C5aR-1 protein expression
Cells (2x105) were incubated for 30 minutes at 4ºC with a mouse anti-human
C5aR-1 (S5/1, Serotec) or an IgG2b isotype control (R&D Systems) diluted 1:50 or
1:100, respectively, in binding buffer (PBS with 1% BSA and 0.1% sodium azide; all
reagents from Sigma). After two washes, cells were incubated in Alexa Fluor-
conjugated goat anti-mouse IgG (Invitrogen) diluted 1:100 in binding buffer for 30
minutes at 4ºC. Finally, cells were washed twice and analyzed by flow cytometry on a
FACSCalibur using CellQuest Pro software (BD Biosciences).
2.3.2 Treatment with C5a
The following treatment with C5a was performed for most experiments; however,
C5a concentrations and incubation times may differ, as it will be indicated in the
corresponding section.
Treatment of cells with C5a (Calbiochem) was performed in 10 cm2 plates. Cells
(5 x 105) were seeded in complete medium and left to attach for 24 hours at 37ºC in the
cell incubator. Afterwards, cells were washed three times with 10 ml of PBS and kept in
basal RPMI overnight. Next day, the medium was removed and 10 ml of fresh basal
medium (negative control) or basal medium containing C5a was added. In some
experiments, the C5aR-1 antagonist was used to study the specificity of C5a; in those
cases an additional plate was treated with C5a (10 ng/ml) and C5aR-1 antagonist (100
ng/ml). Cells were incubated for different periods of time. In some cases, conditioned
media were also collected at the end of the experiments and concentrated using Amicon
centricons with a 10 kDa cutoff (Millipore).
2.3.3 C5a signaling activation
a. Calcium influx assay
Granulocytes were used as positive controls. They were obtained from fresh
heparinized blood by centrifugation at 750 g during 10 minutes. Then, the pellet was
resuspended in lysis buffer (55 mM NH4Cl, 10 mM KHCO3, pH 7.6) and incubated for
Gene Sense Antisense
C5aR-1 5´ATGAACTCCTTCAATTATACC3´ 5´TGGTGGAAAGTACTCCTCCCG3´
GAPDH 5´ACTTTGTCAAGCTCATTTCC3´ 5´CACAGGGTACTTTATTGATG3´
Materials and methods
54
5 minutes at room temperature for erythrocytes elimination. After incubation, the
dispersion of cells was centrifuged and the pellet washed with basal RPMI. Cells were
counted and prepared at 8x106 cells in 500 µl of basal RPMI. Tumor cells were left 24
hours in basal medium before the experiment, and then trypsinized and resuspended
(8x106) in 500 µl of basal RPMI. To each sample, 1 µl of Fluo-4 (Millipore) was added,
and incubated for 30 minutes at 37ºC in the dark. Cells were then washed twice and
resuspended in 500 µl of basal RPMI. The increase in fluorescence due to changes in
the intracellular calcium content was measured on the FACScalibur flow cytometer
after gating the proper population, granulocytes (in the case of whole blood) or tumor
cells. Baseline fluorescence was established by measuring Fluo-4 labeled cells during 2
minutes. After that, either 10 µl of C5a at 100 µg/ml or PBS were added, and changes in
calcium mobilization were measured for an additional 3 minutes. Data were analyzed
using the FlowJo software (Tree Star).
b. ERK phosphorylation
A549 cells were treated for 45 minutes with or without C5a, as explained in
section 2.3.2. Proteins extracted in RIPA buffer (30 µg) were mixed with commercial
loading sample buffer (NuPAGE LDS Sample Buffer; Invitrogen) and 1 µl of β-
mercaptoethanol. The mixture was boiled during 10 minutes at 70ºC. Afterwards,
samples were snap-cooled to avoid protein refolding, and loaded into the gel. To
estimate the molecular weight of proteins under study, a standard commercial molecular
weight marker, MultiMark (Invitrogen), was used. Samples were loaded into
commercial denaturing gels of 4-12% polyacrilamide (NuPAGE Bis-Tris; Invitrogen)
and the electrophoresis was carried out in an Xcell SureLock system (Novex) in running
buffer (50 mM MOPS, 50 mM Tris, 0.1% sodium dodecyl sulfate and 1 mM EDTA, pH
7.7; all from Sigma), during 45 minutes at 180 V. After electrophoresis, proteins were
transferred to 0.45 µm nitrocellulose membranes (BioRad) in an Xcell II Blot Module
system (Invitrogen) during 90 minutes at 30 V. Afterwards, membranes were rinsed
twice with PBS-0.1%Tween (PBST), and blocked overnight with blocking solution
(PBST with 10% skim milk) at 4ºC. Next day, blots were incubated for one hour with a
rabbit anti-human phospho p44/p42 MAPK (Thr202/tyr204) and a rabbit anti-human
p44/p42 MAPK (both from Cell Signaling) diluted 1:2000 in blocking solution. After
three washes with PBST (10 minutes each), blots were incubated for 30 minutes at room
temperature with a horseradish peroxidase-conjugated anti-rabbit IgG secondary
Materials and methods
55
antibody (Amersham) diluted 1:2000 in blocking solution. Afterwards, membranes were
washed three times with PBST and immunoreactive bands were visualized by
chemoluminiscence using the Lumi-LightPLUS kit (Roche) following the manufacturer’s
instructions. Blots were slightly dried and placed into radiography cassettes together
with ECL Amersham Hyperfilm sheets during variable amounts of time, and
radiography sheets were developed automatically in a Curix 60 device (AGFA).
c. NF-κκκκB translocation to the nucleus
NF-κB translocation to the nucleus was tested in A549 cells treated with C5a.
Cells (10,000) were cultured in 8-well glass chambers (BD Biosciences) in complete
RPMI and incubated for 48 hours at 37ºC and 5% CO2. Cells were then washed twice
and left in basal medium overnight. Next day, cells were treated in basal medium with
10 ng/ml C5a, with or without 100 ng/ml C5aR-1 antagonist or control peptide. Basal
medium was used also as negative control. After 45-minute incubation, cells were
washed twice with PBS and fixed in 4% paraformaldehyde for 15 minutes at room
temperature. After two additional washes with PBS, cells were permeabilized with 0.5%
Tween in PBS for 10 minutes, blocked with 10% goat serum in PBS and incubated
overnight at 4ºC with mouse 1:200 anti-NF-κB antibody (Santa Cruz Biotech) in PBS
containing 1% BSA. Cells were washed with PBS and incubated with 1:200
phycoerythrin (PE) anti-rabbit IgG secondary antibody (Invitrogen) for 1 hour at room
temperature. Finally, cells were stained with 10 µl of DAPI and visualized by using an
Axio Imager ZI with Axio Vision software (Zeiss).
2.3.4 Effect of C5a on cell proliferation
A colorimetric MTT assay, which is a means of measuring the activity of living
cells by assessing the activity of mitochondrial dehydrogenases, was used to study the
effect of C5a on cell proliferation. Mitochondrial dehydrogenase activity on viable cells
is measured by a reduction reaction, where yellow MTT [3-(4,5-dimethylthiazol-2-yl)-
2,5-diphenyltetrazolium bromide] is reduced to formazan crystals, giving a purple color.
The building up of crystals inside the cells causes their lysis and release of cellular
content to the exterior. Then formazan crystals, which are insoluble in aqueous
solutions, are dissolved in a solubilization solution that contains DMSO.
The assay was used to study the proliferation of HUVEC and NSCLC cell lines.
For endothelial cells, 1500 HUVECs were seeded in 50 µl of assay medium (EGM-
Materials and methods
56
2MV medium without rhVEGF and rhFGF) in a 96-well sterile plate (Costar). After 24
hours of incubation, 50 µl of C5a (Calbiochem) or bFGF (Sigma) were added in assay
medium two times more concentrated than the desired final concentrations. bFGF at a
final concentration of 10 ng/ml was used as positive control, and cells with only assay
medium were used to assess 100% proliferation. All treatments were performed in six
wells. In order to subtract the background absorbance of the medium, 100 µl of medium
was added in six empty wells. Cells were placed in the cell incubator for 96 hours.
Afterwards, 10 µl of 5 mg/ml MTT solution (Sigma) was added to each well. Plates
were incubated for an additional 4 hours at 37ºC to allow MTT reduction by viable
enzymes. Afterwards, 100 µl of solubilization buffer (10% SDS, 50% N-N-
dimethylformamide; both from Sigma) was added and plates were left on the cell
incubator overnight. Next day, absorbance was measured at 540 and 690 nm
(corresponding to formazan and reference wavelengths, respectively) in a Sunrise
microplate reader (Tecan). In some experiments, bFGF was also added to the assay
medium at a final concentration of 0.31 ng/ml, in order to induce cell proliferation.
For NSCLC cell lines, the procedure was similar to that explained for HUVEC. In
this case, 1500 NSCLC cells (A549, H157 and H661 cells) were seeded in 50 µl of
RPMI complemented with 1% Fetalclone and allowed to attach overnight. Next day,
50 µl of C5a two times more concentrated than the final concentration was added in
basal RPMI (this made a medium with 0.1% Fetalclone). In some experiments, medium
was removed before the addition of C5a and cells were washed twice in basal medium.
Then, C5a dilutions in basal medium were added to cells; so no serum was present.
Cells were incubated for 96 hours at 37ºC. The MTT reaction and development was
done as indicated above for HUVEC.
To measure the percentage of proliferation, the mean value of each condition was
calculated and used in the formula:
Absorbance treatment concentration – Absobance background
Absobance – Absorbance background
Absorbance treatment concentration – Absobance background
Absobance 100% proliferation – Absorbance background
Proliferation (%) = x 100Absorbance treatment concentration – Absobance background
Absobance – Absorbance background
Absorbance treatment concentration – Absobance background
Absobance 100% proliferation – Absorbance background
Proliferation (%) = x 100
2.3.5 Effect of C5a on cell cycle
A549 cells were treated with C5a as indicated in section 2.3.2. Afterwards, cells
were harvested with trypsin/EDTA and washed twice in PBS. For fixation, cells (1x106)
were resuspended in 5 ml of 70% cold ethanol and left at 4ºC for at least one hour (this
step can be extended for one week). After fixation, cells were washed twice in PBS and
Materials and methods
57
the pellets resuspended in 500 µl of RNase (Invitrogen) at 200 µM in PBS. Then, cells
were incubated for one hour at 37ºC and, finally, 20 µl of propidium iodide at 1 mg/ml
was added. Cells were analyzed in the FACScalibur flow cytometer.
2.3.6 Effect of C5a on anchorage independent growth
A549 cells were treated with C5a as indicated in section 2.3.2. Anchorage-
independent assays were performed by covering 6-well plates with 1.5 ml of 0.6%
sterile noble agar (BD Biosciences) dissolved in 1% medium (RPMI with 1%
Fetalclone, 100 U/ml penicillin and 100 µg/ml streptomicin). Cells (1000 cells/well)
were mixed with 1 ml of 0.3% agar dissolved in 1% medium with or without 10 ng/ml
C5a. Plates were incubated for 8, 10 and 12 days at 37ºC. After incubation, 500 µl of 10
mg/ml MTT was added to each well and plates were incubated again at 37ºC for
additional 4 hours. Finally, excess of MTT was removed carefully with a pipete, 500 µl
of DMSO was added to wells, and plates were incubated at 37ºC overnight. For
analyzing the number of colonies in each well, plates were digitally scanned and
colonies were counted.
2.3.7 Effect of C5a on cell migration
The effect of C5a on cell migration was evaluated in HUVEC and A549 cells
using a transwell migration assay. This assay consists in seeding cells in specially
designed inserts with microscopic pores in 24-well plates (Corning-Costar). Then, a
chemoattractant solution is added where the inserts are laid. If cells are attracted to the
chemoattractant, they will migrate through the pores of the transwells and keep attached
to the lower side of them (in case they are attaching cells).
HUVEC (75,000) and A549 (10,000) cells were seeded on the upper chamber of
transwells in 150 µl of EBM-2 or RPMI basal medium, respectively. In both cases,
chambers with 8 µm pore size filters were used. C5a dilutions ranging from 500 to
0.25 ng/ml in EBM-2 or RPMI basal medium were added in the 24-well plates, in a
final volume of 400 µl. EBM-2 or RPMI basal medium was used as negative control,
and 10% FBS in EBM-2 or RPMI basal medium as positive control. Cells were allowed
to migrate for 48 hours at 37ºC with 5% CO2. After incubation, cells from the upper
chambers were removed by swiping with cotton swabs, and transwells were washed
twice in PBS. Cells adhered to the lower side of transwells were fixated with 4%
formaldehyde (Panreac) for one hour at room temperature. Afterwards, cells were
Materials and methods
58
stained with 0.5% crystal-violet (Sigma) on PBS for 15 minutes at room temperature.
Excess of staining was removed by washing twice with PBS. Migrated cells on the
lower surface of the filter were visualized using an inverted microscope (Leyca DM
ILM), and at least five fields at 10x magnification were photographed (Leica EC3
digital camera). Migrated cells were then counted using ImageJ software (NIH).
2.3.8 Effect of C5a on HUVEC tube formation
C5a ability to promote angiogenesis in vitro was tested with the tube formation
assay using HUVECs. For the assay, HUVECs were used between the 5th and 7th
passages and were cultured in 10 cm2 plates coated with 0.1% gelatin (Sigma). Cells
were cultured in EGM-2MV until they got an optimal confluence (between 80-90%);
then, the medium was changed to EGM-2MV/EBM-2 basal medium (1:1) and cells
were incubated for 8 hours. The reason to use this transition medium (with two-fold
diluted supplements) is to avoid a sudden change from the normal conditions to a total
lack of supplements. Afterwards, cells were washed and kept in EBM-2 basal medium
overnight. Next day, 48-well plates were coated with 150 µl of growth factor reduced
Matrigel (BD Biosciences) and the matrix was allowed to gel for 2 hours at 37ºC.
HUVECs (25,000) were seeded in the gelled Matrigel in 150 µl of EBM-2 basal
medium, and 150 µl of 20 ng/ml C5a in EBM-2 basal medium, or only 150 µl of EBM-
2 basal medium (negative control), was added to cells. At least five repetitions were
done per condition and plates were incubated at 37ºC for 5 hours. Afterwards, the
medium was carefully removed by turning the plates upside down, and the cells were
fixed and stained with Diff Quik (Panoptico Rapido; QCA) following the
manufacturer’s instructions, and handling the plate with care, as tubes are very fragile
structures. Photographs of three random optical fields at 10x magnifications in the
central area of wells were taken with a digital camera coupled to a phase microscope
(Leica DM IL LED). Total length of tubes in each well was calculated using the ImageJ
software (NIH).
2.3.9 Effect of C5a on cell adhesion to extracellular matrix proteins
This assay was performed in order to study the effect of C5a on the adhesion of
NSCLC cells to several intracellular matrix components, as previously described by
Irigoyen et al.155
Materials and methods
59
NSCLC cells, A549, H157 and H661, were treated as indicated in section 2.3.2.
For the adhesion assay, 96-well plates with flat bottom (Costar) were cover with 35 µl
of different matrix dilutions: 5 µg/cm2 rat tail collagen type I (BD Bioscience) in
20 mM acetic acid, 5 µg/cm2 collagen type IV (Sigma) in 50 mM acetic acid, 1 µg/cm2
fibronectin (Sigma) in PBS, 0.1 µg/cm2 vitronectin (Sigma) in ultrapure water, 1 µg/cm2
laminin (Sigma) in PBS, or 0.1 µg/cm2 gelatin (Sigma) in ultrapure water; 3% BSA in
PBS was used as a negative control. All matrix dilutions were added in sextuplicate.
Plates were incubated for 2 h at 37°C. After incubation, they were washed twice with
PBS, and blocked with 200 µl of 1% BSA in PBS for 1 h at 37°C. At the same time,
cells were trypsinized, washed twice in PBS, and incubated in 300 µl of basal RPMI
containing 10 µM calcein-AM (Fluka) for 20 min at 37°C, in a water bath. After
labeling, cells were washed three times in basal RPMI to eliminate excess of calcein-
AM, and pellets were resuspended in a concentration of 2 × 105 cells/ml in adhesion-
medium (RPMI with 0.5% BSA and 20 mM HEPES). Before adding cells to the plates,
the blocking solution was removed from wells. Then, 100 µl of each cell dilution
(control cells, cells treated with C5a, and cells treated with C5a and C5aR1 antagonist)
was added in the different matrixes. The remaining cells were maintained for the last
step (see below). Plates were incubated for 30 minutes at 37ºC in order to allow cells to
attach to the matrixes. After the incubation time, non-adherent cells were removed by
gentle washing with adhesion-medium; the remains of medium were wiped carefully
with a multi-channel pipette. Finally, 100 µl of basal RPMI were added to the attached
cells and, to establish the 100% adhesion, 100 µl of each cell dilution (that were kept
previously for this step) were added in six empty wells. Fluorescent adherent cells were
measured on a Polar Star Galaxy plate reader (BMG Labtechnologies), with an
excitation wavelength of 485 nm and an emission filter of 520 nm.
To calculate the percentages of adhesion to the different matrixes in each
condition, fluorescence mean of replicates was calculated, and the following formula
was used:
Fluorescence cells adhered to matrix – Fluorescence cells adhered to 3% BSA
Fluorescence 100% adhesion – Fluorescence cells adhered to 3% BSA
Adhesion (%) = x 100 Fluorescence cells adhered to matrix – Fluorescence cells adhered to 3% BSA
Fluorescence 100% adhesion – Fluorescence cells adhered to 3% BSA
Fluorescence cells adhered to matrix – Fluorescence cells adhered to 3% BSA
Fluorescence 100% adhesion – Fluorescence cells adhered to 3% BSA
Adhesion (%) = x 100
2.3.10 Effect of C5a on integrin expression
Integrin expression on the cell membrane was analyzed by flow cytometry. Cells
were treated as explained in section 2.3.2. for 48 hours with 10 ng/ml C5a. Cells (2x105)
Materials and methods
60
were incubated with primary monoclonal antibodies against integrins α1, α4, αv, β1,
β3, α5β1, αvβ3, and vβ4. All antibodies came from the Integrin Classics Kit
(Chemicon International) and were diluted 1:100 in binding buffer (PBS with 1% BSA
and 0.1% sodium azide; all reagents from Sigma), except for the antibody anti-αvβ4,
which was from BD Pharmigen and was diluted 1:40. An isotype control anti-IgG2b
(R&D Systems) diluted 1:10 was also used. After 30 minutes of incubation, cells were
washed twice in binding buffer and incubated for 30 minutes in Alexa fluor goat anti-
mouse IgG (Invitrogen) diluted 1:100 in binding buffer. Then, after two washes in
binding buffer, cells were analyzed by flow cytometry in a FACScalibur with the
CellQuest software (BD Biosciences).
2.3.11 Effect of C5a on MMP activity
For measuring the matrix metalloproteinase (MMP) activity on the conditioned
media, fluorogenic peptide substrates were used, those peptides incorporate a
fluorophore and a quencher moiety positioned on either side of the enzyme cleavage
site. When an active protease cleaves the substrate, the fluorescence is increased
(Figure 8). The increase in fluorescence correlates with the amount of protease activity.
Donor
QuencherRecognition
site
Enzyme
Generation of
fluorescent signal
Donor
QuencherRecognition
site
Enzyme
Generation of
fluorescent signal Figure 8. Schematic representation of the fluorogenic peptide recognition by specific proteases and subsequent cleveage and fluorescence generation. (Adapted from LABTECH).
For the assay, conditioned media from cells treated as explained in section 2.3.2.
with 10 ng/ml C5a, with or without 100 ng/ml C5aR-1 antagonist, were collected after
24 or 48 hours and concentrated using Amicon centricons with a 10 kDa cutoff
(Millipore). Concentrated media were incubated in triplicate with the fluorogenic
peptides Mca-R-P-K-P-V-E-Nval-W-R-K(Dnp)-NH2 and Mca-PLGL-Dpa-AR-NH2
(R&D Systems), at a final concentration of 20 µM, in TCNB buffer (50 mM Tris, 150
Materials and methods
61
mM NaCl, 10 mM CaCl2 and 0.5% Brij, pH 7.5). As a positive control, MMP-10 at 20
nM was used; and as an inhibition control, 100 µM GM 6001 (Chemicon International)
was added to the samples with the highest metalloproteinase activity. GM 6001 is a
wide range hydroxamate MMP inhibitor. Plates were incubated at 37ºC for 2 hours
while the increase of the fluorescence intensity during substrate cleavage was monitored
in a fluorescence spectrofluorimeter (SpectraMAX GeminiXS, Molecular Devices). The
measurements were carried out at the excitation λex=320 nm and at the emission
λem=405 nm. The increase in fluorescence was plotted against the incubation time and
the slope (rate of increase in fluorescence) was calculated by linear regression.
2.4. Mice models for the study of the influence of complement activation on lung
tumor growth in vivo
2.4.1 Syngeneic animal model of lung cancer
The murine Lewis Lung Cancer cell line (also called LLC or 3LL) implanted in
the flanks of female C57BL/6J mice, a model of syngeneic lung cancer, was used to
analyze the effect of complement in tumor progression, angiogenesis and immune
response against tumors.
For the experiments, 3LL cells were trypsinized and neutralized with the
conditioned medium, as this cell line is a semi-adherent cell culture, and then washed
twice with PBS. Mice were anesthetized with a mixture containing 35 mg/kg ketamine
(Imalgene 1000, Merial) and 5 mg/ml xylazine (Rompun 2%, Bayer). Two different
inoculation sites were used: subcutaneous in the mouse back, and intravenous in the tail
vein. For subcutaneous tumor growth, 2.5x104 cells were resuspended in 100 µl of PBS
and mixed with 100 µl of growth factor reduced Matrigel (BD Biosciences). Two
hundred microliters of the cell/Matrigel mixture was injected subcutaneously into the
flanks of mice (wild type or C3-deficient mice). Diameters of the tumors were measured
every three days with calipers, and the volume was calculated by the formula:
(L × W2)/2, where L is the length and W is the width. Around the 30th day after cell
injection, animals were sacrificed and spleen and tumors were removed, weighed, and
processed for flow cytometry analysis or used for total RNA extraction, as explained in
section 2.4.2. For intravenous injection, mice received an injection into the tail vein of
1x105 cells in 100 µl of PBS. After 3 weeks, mice were anesthetized as above and
Materials and methods
62
sacrificed by exsanguination. Lungs were harvested and fixed as explained in section
2.4.2.d.
In some experiments, C5aR-1 was blocked pharmacologically in wild type
C57BL mice to study specifically the effect of C5a through its main receptor. Mice
received an injection every 3 days of C5aR-1 antagonist [AcF(OP(D)ChaWR)] or
control peptide [AcF(OP(D)ChaA(D)R)], at a dose of 1 mg per kg body weight in
100 µl of PBS, starting 4 hours before the inoculation of tumor cells.
In some experiments, C3 deficient animals were treated with cobra venom factor
(CVF, Aczon) in order to eliminate any possible effect of C5 in those animals. CVF
forms a stable C3/C5 convertase with factor B; this characteristic allows it to cleave
both complement components C3 and C5, resulting in complement depletion. In a pilot
experiment, CVF was injected in wild type mice in order to set the administration
timing by monitoring C3 levels in serum. From the pilot experiment a dose of 5 µg per
gram body weight in 100 µl of PBS injected intraperitoneally was settled, with the
following regimen of injections: three injections at 28, 24 and 4 hours before cell
inoculation (in order to deplete complement activity before cell implantation), and
injections every 3 days after inoculation (to avoid complement recovery during the
experiment). Complement depletion was monitored by ELISA quantification of C3 in
mouse sera. This assay was also used to confirm C3 deficiency in C3 knockout mice.
C3 ELISA was carried out in 96-well plates coated with goat anti-mouse C3 (Cappel
Laboratories), diluted 1:1000 in 50 mM sodium bicarbonate, pH 8.3. After 1 hour at
room temperature, plates were washed in the assay buffer (PBS with 1% BSA and 0.1%
Tween, all reagents from Sigma) and blocked with the same buffer. Afterwards, 1:1000,
1:2000 and 1:4000 dilutions of samples in assay buffer were added in triplicate. When
the ELISA was used to confirm C3 deficiency in C3 knockout mice, dilutions of a pool
of sera from 8 normal wild type mice ranging from 1:500 to 1:256,000 were also added
in triplicate. Plates were placed at room temperature for 2 hours and then washed twice
with assay buffer. Finally, the assay was developed using a horseradish peroxidase-
conjugated goat anti-mouse C3 antibody (Cappel) diluted 1:5000 in assay buffer, and
OPD (Sigma), according to manufacturer’s instructions. Plates were read at 450 nm in a
Sunrise microplate reader (Tecan). Percentage of C3 depletion in mice treated with CVF
was calculated by setting the C3 levels before CVF injection as 100%. C5a levels were
also measured in wild type and C3 knockout mice. For this purpose the mouse
Materials and methods
63
complement component C5a DuoSet (R&D Systems) was used following the
manufacturer’s instructions.
2.4.2 Tissue processing
a. Cell isolation from spleens
Spleens were weight and cut in two parts. To obtain single-cell suspensions, one
portion was mechanically disaggregated in 5 ml of PBS using cell strainers with 70 µm
pore size (BD Biosciences). Cells were then centrifuged at 500 g during 5 minutes and
resuspended in 500 µl of PBS. Single-cell suspensions were used to calculate total cell
numbers in spleens, using a Neubauer chamber, and to analyze, by flow cytometry, the
main populations within the spleens.
Single-cell suspensions from spleens (approx. 5 x 105 splenocytes) were
centrifuged and incubated with primary antibodies. Antibodies were diluted in binding
buffer (PBS with 1% BSA and 0.1% sodium azide; all reagents from Sigma) following
the conditions summarized in Table 6. In the case of Treg, they were fixated,
permeabilizated and stained according to the manufacturer’s instructions. After 30
minutes of incubation in each antibody dilution, splenocytes were washed twice and
analyzed by flow cytometry on the FACSCalibur with CellQuest Pro software (BD
Biosciences).
Table 6. Antibodies used for staining splenocytes*
*All antibodies, except those for Tregs, were from BD Pharmingen. Antibodies for Tregs were obtained on the kit for mouse regulatory T cells from eBioscience. MDSC: myeloid-derived suppressor cells.
Cell population Fluorochrome-conjugated monoclonal antibody Dilution
T cells CD4+ FITC-conjugated rat anti mouse CD4 (RM4-5) 1:1000
T cells CD8+ PE-conjugated rat anti mouse CD8a (53-6.7) 1:1000
B cells CD19+ FITC-conjugated rat anti mouse CD19 (ID3) 1:400
MDSC FITC-conjugated rat anti mouse CD11b (M1/70) 1:500
APC-conjugated rat anti mouse Ly6C/Gr1 (AL-2) 1:100
Tregs FITC-conjugated rat anti mouse CD4 (RM4-5) 0.125 µg/test (1:40)
APC-conjugated rat anti mouse CD25 (PC61.5) 0.06 µg/test (1:30)
PE-conjugated rat anti mouse Foxp3 (FJK-16S) 0.2 µg/test (1:20)
Materials and methods
64
b. Total RNA extraction from tumors and quantitative PCR
Total RNA extracted from tumors was analyzed for the expression of different
molecules related to the immune response and angiogenesis. Portions from the edge of
tumors with a volume of approx. 0.1 cm3 (a sphere with 0.3 cm radio) were cut, frozen
in dry ice, and maintained at -80ºC until the extraction procedure. Total RNA was
extracted and retrotranscribed to cDNA as explained in section 2.1.6. Quantitative PCR
was performed by Drs Lasarte and Riezu, from the Division of Gene Therapy &
Hepatology at CIMA. RNA expression levels of the different genes in the study were
related to β-actin expression levels.
c. Fixation of tumors
Once the portion for RNA extraction was obtained, tumors were fixed in 4%
formaldehyde for 24 hours.
d. Fixation of lungs
Lungs harvested from mice receiving an injection into the tail vein of 3LL cells
were fixed by intratracheal instillation of 4% formaldehyde at a constant pressure of 20
cm H20156. One day later, lungs were transferred to 70% ethanol and paraffin blocks
were made from each lobe in the Morphology Core Facility at CIMA. One section per
block was cut and stained with hematoxylin and eosin, resulting on a total of five slides
per mouse. Lung nodules were counted in a microscope. Consecutive slices to those
with nodules were used for CD31 immunohistochemistry.
2.4.3 Study of angiogenesis by CD31 immunohistochemistry
Slices of fixed lungs from the intravenous injection model of lung cancer were
used. Fixed lung sections were kept at 60ºC during 20 minutes to allow paraffin
melting. Immediately, they were placed in xylol (Panreac) for 15 minutes to remove
paraffin leftovers. Then, sections were hydrated with decreasing alcohol stages for 5
minutes: 96%, 80%, 70% ethanol, and finally rinsed in distilled water for another 5
minutes. To block endogenous peroxidases, sections were incubated with 3% hydrogen
peroxide (Merck) for 10 minutes. Afterwards, they were rinsed and subjected to antigen
unmasking. This step consisted in boiling the sections in citrate buffer (10 mM citrate,
pH 6.0) in a standard microwave oven for a maximum of 10 minutes in two consecutive
times. After that, slides were kept at room temperature for 5 minutes to cool down, and
Materials and methods
65
washed in TBS (50 mM Tris, 150 mM NaCl, pH 7.4). An incubation of one hour with
1:20 anti-mouse CD31 (LabVision) in REAL antibody diluent (Dako) was done. Slides
were washed in TBS for five minutes and incubated with 1:50 rabbit anti-rat IgG
(Dako) during 30 minutes. After another washing step, samples were incubated in
Envision anti-rabbit (Dako) for 30 minutes. Samples were revealed by incubation with
diaminobenzidine for 8 minutes. The reaction was stopped with water. Finally, samples
were stained with hematoxylin and dehydrated with increasing alcohol solutions to be
finally driven to xylol and mounted with DPX (VWR). Images were capture using a
microscope with a digital camera and ACT-2U software (Nikon). Quantification of the
microvascular density on tumor sections was carried out by measuring the fraction of
area immunostained using the AnalySIS FIVE software (Olympus BioSystems).
2.5. Statistical analysis
The effect of hypoxic conditions in the expression of mCRPs and soluble factors
was evaluated by ANOVA and the Tukey post-test. The Mann-Whitney test was used
for evaluation of the expression of different citokines within the tumors, and for
evaluation of the differences of factor H in bronchoalveolar lavage supernatants. The
other comparisons were carried out with the Student’s t test. Relationship between
variables was quantified using Spearman’s rank correlation. Box plots were used to
present the results of factor H and C5a quantification on clinical samples. In these plots,
the bottom and top of the boxes represent the lower and upper quartiles, the band inside
the box is the median, and the whiskers the minimum and maximum of all the data.
Receiver operating characteristic (ROC) curves were used to define the best cut-off
point and evaluate the performance of the factor H test. Area under the ROC curve, as
well as the 95% confidence interval, was also calculated. Data were analyzed using
SPSS software or GraphPad Prism (GradPad Software). For all analyses, p values of
0.05 or less were considered statistically significant.
RESULTS
Results
69
1. COMPLEMENT ACTIVATION AND REGULATION ARE
INCREASED IN NON-SMALL CELL LUNG CANCER CELLS
Initially, complement activation and its regulation were compared in vitro between
non-small cell lung cancer (NSCLC) cells and normal bronchial epithelial cells. In
terms of activation, both classical and alternative pathways were studied, and deposition
of complement proteins on the cell membrane and cell cytotoxicity were analyzed. With
respect to regulation, expression of the main complement inhibitors was analyzed, as
well as the effect of hypoxia on the expression of these inhibitors.
1.1. Deposition of complement proteins after activation of the alternative
pathway
Seven human NSCLC cells lines were chosen for these experiments, representing
the principal subgroups of the disease. The activation of the alternative complement
pathway in these cell lines was compared with that in four immortalized human normal
bronchial epithelial cells.
Activation of the alternative pathway of complement was analyzed by measuring
the increase of C3 deposition on the cell membrane in cells treated with 10% NHS
versus cells treated with 10% HI-NHS (Figure 9A). A significant increase in C3
deposition was exhibited in all NSCLC cell lines, except HTB58, compared to BEAS-
2B normal bronchial epithelial cells (pA549=0.002; pH23<0.001; pH157<0.001;
pH460=0.030; p H727=0.002; pH460=0.011). No significant differences were observed
between the different normal bronchial epithelial cells analyzed. This result indicates
that there is an increase in spontaneous complement activation on malignant lung cells
when compared to normal lung epithelial cells. In order to determine if there was any
difference in C3b inactivation among cells, detection of iC3b on the surface of cells was
carried out (Figure 9B). In each cell line, the amount of iC3b on the membrane was
comparable with the amount of C3b, and the significant differences with normal
epithelial cells were maintained (pA549= 0.007; pH23=0.017; pH157=0.010; pH460=0.038;
pH727<0.001; pH1299=0.029; pHTB58=0.003). Finally, we analyzed the deposition of C5b
on the cell membrane (Figure 9C). Compared to C3b and iC3b deposition, this factor
Results
70
appeared at lower levels on the surface of all the cell lines tested, but the differences
between groups were still significant (pA549=0.005; pH23=0.044; pH157=0.025;
pH727=0.006; pH1299=0.040; pHTB58<0.001).
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10K
T
BEA
S
0
10
20
30
C3 d
epositio
n (M
FI ra
tio)
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10K
T
BEA
S
0
20
40
60
80
iC3b d
epositio
n (M
FI ra
tio)
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC
3KT
BEA
S
0
1
2
3
4
C5 d
epositio
n (M
FI ra
tio)
∗∗
∗∗∗
∗∗∗
∗∗
∗
∗
∗∗∗
∗∗
∗∗∗
∗
∗∗∗
∗∗
∗
∗∗∗∗
∗∗∗
A B
C
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10K
T
BEA
S
0
10
20
30
C3 d
epositio
n (M
FI ra
tio)
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10K
T
BEA
S
0
20
40
60
80
iC3b d
epositio
n (M
FI ra
tio)
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC
3KT
BEA
S
0
1
2
3
4
C5 d
epositio
n (M
FI ra
tio)
∗∗
∗∗∗
∗∗∗
∗∗
∗
∗
∗∗∗
∗∗
∗∗∗
∗
∗∗∗
∗∗
∗
∗∗∗∗
∗∗∗
A B
C
Figure 9. Deposition of complement proteins on NSCLC (shaded bars) and normal bronchial epithelial cells (white bars) after activation of the alternative complement pathway. Graphs show flow cytometry data of cells incubated in VB+ with 10% NHS vs. HI-NHS (as a negative control), and stained for C3b (A), iC3b (B) and C5b (C). At least three independent experiments for each cell line were performed. Results are expressed as the ratio of the mean fluorescence intensity (MFI) between NHS and HI-NHS, and shown as mean ± SEM. Statistical differences were analyzed using BEAS cells as reference. *p<0.05; **p<0.01; ***p<0.001.
1.2. Deposition of complement proteins after activation of the classical pathway
In order to activate the classical pathway, two antibodies were evaluated, an
antiserum obtained from rabbits immunized against whole cell extracts of A549 cells and
an antiserum against β2-microglobulin. Both polyclonal antibodies showed high and
similar levels of binding to NSCLC and normal bronchial epithelial cell lines, however
the antiserum against β2-microglobulin failed to trigger the classical complement
pathway (data not shown). Therefore, we used the antiserum against A549 cells to
activate the classical complement pathway on all the cell lines studied (Figure 10).
Results
71
Fluorescence intensity
HTB58H1299 HBEC 3KT
HBEC 10KT BEAS
H727
Cell
nu
mb
er
HBEC 3KTp53
H23 H157A549 H460
Fluorescence intensity
HTB58HTB58H1299H1299 HBEC 3KTHBEC 3KT
HBEC 10KTHBEC 10KT BEASBEAS
H727H727H727
Cell
nu
mb
er
HBEC 3KTp53HBEC 3KTp53
H23H23H23 H157H157H157A549 A549 A549 H460
Figure 10. Binding of the polyclonal antibody against A549 cells to NSCLC and normal bronchial epithelial cells. Cells were incubated with serum obtained from rabbits immunized against A549 cell extracts (white peak) or with pre-immune serum (shaded peak), with a FITC-conjugated anti-rabbit IgG, and analyzed by flow cytometry. Binding of the polyclonal antibody to the membrane was measured as the increase in intensity in the green channel.
Complement activation by the classical pathway was analyzed in the same manner
as that used for the alternative pathway. First, C3 deposition was compared between
NSCLC and normal epithelial cells. This pathway is more powerful than the alternative
pathway, so all cell lines showed much higher C3 levels on their cell membrane (Figure
11A). However, in this case, NSCLC cells did not show higher C3 deposition when
compared to normal cells (using BEAS-2B cells as reference), except for H157 cells
(p=0.050). In addition, it was surprising that two cell lines, H460 and HTB58, showed a
significant decreased in C3 deposition compared to normal cells (pH460=0.005;
pHTB58=0.004). In view of these results, C3b inactivation was analyzed by iC3b
detection (Figure 11B). Analogous to the alternative pathway, cells showed a similar
pattern in iC3b deposition to that observed in C3 deposition. Finally, the study of C5b
deposition showed no significant differences among the two groups (Figure 11C); only
the H727 cell line showed a higher C5b deposition compared to BEAS-2B cells
(p=0.001).
Results
72
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC 3KT
HBEC 3KTp
53
HBEC
10K
T
BEA
S
0
50
100
150
200
C3 d
epositio
n (M
FI ra
tio)
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC 3KT
HBEC 3KTp
53
HBEC
10K
T
BEA
S
0
50
100
150
200
iC3b d
epositio
n (M
FI ra
tio)
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC
3KT
BEA
S
0
1
2
3
4
5
C5 d
epositio
n (M
FI ra
tio)
∗
∗∗ ∗∗
∗∗
∗
∗
∗∗
A B
C
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC 3KT
HBEC 3KTp
53
HBEC
10K
T
BEA
S
0
50
100
150
200
C3 d
epositio
n (M
FI ra
tio)
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC 3KT
HBEC 3KTp
53
HBEC
10K
T
BEA
S
0
50
100
150
200
iC3b d
epositio
n (M
FI ra
tio)
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC
3KT
BEA
S
0
1
2
3
4
5
C5 d
epositio
n (M
FI ra
tio)
∗
∗∗ ∗∗
∗∗
∗
∗
∗∗
A B
C
Figure 11. Deposition of proteins of the complement system on NSCLC (shaded bars) and normal bronchial epithelial cells (white bars) after activation of the classical complement pathway. Graphs show flow cytometry data of cells incubated in VB2+ with 10% NHS or 10% HI-NHS, as control, and stained for C3b (A), iC3b (B) and C5b (C). At least three independent experiments for each cell line were performed. Results are expressed as the ratio of the mean fluorescence intensity (MFI) between NHS and HI-NHS, and shown as mean ± SEM. Statistical differences were analyzed using BEAS cells as reference. *p<0.05; **p<0.01.
We next compared C3 deposition by the classical and the alternative pathways in
each cell line (Figure 12). The ratio of C3 deposition between both pathways was
significantly lower on NSCLC cell lines than on normal bronchial epithelial cells, using
BEAS-2B cells as reference (pA549<0.001; pH23=0.005; pH157=0.005; pH727=0.003;
pH460=0.003; pH1299=0.004 pHTB58=0.002). The reduced increase in C3 deposition on
NSCLC cell lines strongly suggests that these cells have powerful regulatory
mechanisms to inhibit complement activation on their surface.
Results
73
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10K
T
BEA
S
0
5
10
15
20
CP v
s. AP (M
FI ra
tio)
∗∗∗ ∗∗ ∗∗∗∗ ∗∗ ∗∗
∗∗
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10K
T
BEA
S
0
5
10
15
20
CP v
s. AP (M
FI ra
tio)
∗∗∗ ∗∗ ∗∗∗∗ ∗∗ ∗∗
∗∗
Figure 12. Relationship between C3 deposition after activation of the classical and alternative complement pathways (CP and AP, respectively). Alternative and classical pathway activation was performed simultaneously in each cell line and C3 deposition was analyzed. The ratios between C3 deposition in the classical and the alternative pathways were calculated. At least three independent experiments for each cell line were performed, and results are shown as mean ± SEM. Statistical differences were analyzed using BEAS cells as reference. **p<0.01; ***p<0.001.
1.3. Cell cytotoxicity due to complement activation by the alternative and
classical pathways
Complement activation leads to the assembly of the membrane attack complex
(MAC), a pore-forming agent that can penetrate into the membrane and eventually
cause cell lysis. We questioned whether NSCLC cell lines were more or less sensible
than normal bronchial epithelial cells to the lytic effect of the complement system. To
study the cytotoxicity mediated by complement activation we measured the release of
LDH to the medium after triggering complement by both pathways. A meaningless
LDH release was observed in all cells (both malignant and non-malignant) after
activation of the alternative pathway (Figure 13A). This indicates that NSCLC and
normal epithelial cell lines are resistant to the lysis induced by the spontaneous
activation of the complement system, even though NSCLC cell lines present significant
levels of C3b and C5b deposition. Regarding the classical pathway, LDH release
increased in all cell lines (Figure 13B). The cytotoxic effect due to the classical
pathway in NSCLC cell lines was comparable with that observed in BEAS-2B cells;
however, normal HBEC-KT cells showed higher sensitivity to lysis (pHBEC3KT=0.029;
pHBEC3KTp53=0.026; pHBEC10KT=0.017). This result confirms that, although lung cancer
cells activate complement, they are well-protected against its damaging effects.
Results
74
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC 3KT
HBEC 3KTp
53
HBEC 10K
T
BEAS
0
10
20
30
40
50
LDH rele
ase (%
)
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC 3KT
HBEC 3KTp
53
HBEC 10K
T
BEAS
0
10
20
30
40
50
LDH rele
ase (%
)
A B∗
∗∗
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC 3KT
HBEC 3KTp
53
HBEC 10K
T
BEAS
0
10
20
30
40
50
LDH rele
ase (%
)
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC 3KT
HBEC 3KTp
53
HBEC 10K
T
BEAS
0
10
20
30
40
50
LDH rele
ase (%
)
A B∗
∗∗
Figure 13. Complement-mediated cytotoxicity after stimulation of the alternative and classical pathways in NSCLC and normal epithelial cell lines. Cells were incubated with 10% NHS or HI-NHS, and cytotoxicity after activation of the alternative (A) or the classical (B) pathway was measured using an LDH release assay. The classical pathway was induced in all cell lines with an A549 antiserum. The percentage of complement-mediated lysis was calculated subtracting the percentage of lysis obtained after incubation with HI-NHS. Graphs show a representative experiment for the alternative pathway (A), and mean ± SEM of three independent experiments for the classical pathway (B). Statistical differences were analyzed using BEAS cells as reference. *p<0.05.
1.4. Expression of complement inhibitors
The fact that NSCLC cell lines present similar cytotoxicity to normal bronchial
epithelial cells, although they activate more efficiently complement, suggests
differences in the expression of complement regulators. Therefore, the expression of the
main soluble and membrane-bound complement regulators (factor H, FHL-1, factor I,
CD46, CD55 and CD59) was analyzed at the mRNA level by real time PCR (Figure
14). The expression of the inhibitors was heterogeneous throughout the cell lines, with
no apparent patterns of differential expression among groups. Another fact worthy of
mention is that the soluble regulators factor H and FHL-1 showed high levels of
expression in some NSCLC cell lines (e.i. A549, H23, H460, and HTB58). This is in
agreement with previous results from our group demonstrating that some NSCLC
tumors express and release factor H and FHL-1122, affecting tumor growth in vivo123.
We assessed the binding of factor H to the cell surface of NSCLC and normal bronchial
epithelial cells, using purified factor H or 25% HI-NHS (as a source of factor H). No
differences were seen among the two groups of cells (data not shown).
Results
75
A B
C D
E F
A54
9H23
H15
7
H72
7
H46
0
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10K
T
BEA
S
0.00
0.05
0.10
0.15
FH
(G
APD
H %
)
A54
9H23
H15
7
H72
7
H46
0
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10K
T
BEA
S
0.0
0.1
0.2
0.3
FH
and F
HL-1
(G
APD
H %
)
A54
9H23
H15
7
H72
7
H46
0
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10K
T
BEA
S
0.00
0.01
0.02
0.03
0.04
FI (G
APD
H %
)
A54
9H23
H15
7
H72
7
H46
0
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10K
T
BEA
S
0.0
0.5
1.0
1.5
2.0
2.5
CD
46 (G
APD
H %
)
A54
9H23
H15
7
H72
7
H46
0
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10K
T
BEA
S
0
2
4
6
8
CD
55 (G
APD
H %
)
A54
9H23
H15
7
H72
7
H46
0
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10K
T
BEA
S
0
5
10
15
CD
59 (G
APD
H %
)
A B
C D
E F
A54
9H23
H15
7
H72
7
H46
0
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10K
T
BEA
S
0.00
0.05
0.10
0.15
FH
(G
APD
H %
)
A54
9H23
H15
7
H72
7
H46
0
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10K
T
BEA
S
0.0
0.1
0.2
0.3
FH
and F
HL-1
(G
APD
H %
)
A54
9H23
H15
7
H72
7
H46
0
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10K
T
BEA
S
0.00
0.01
0.02
0.03
0.04
FI (G
APD
H %
)
A54
9H23
H15
7
H72
7
H46
0
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10K
T
BEA
S
0.0
0.5
1.0
1.5
2.0
2.5
CD
46 (G
APD
H %
)
A54
9H23
H15
7
H72
7
H46
0
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10K
T
BEA
S
0
2
4
6
8
CD
55 (G
APD
H %
)
A54
9H23
H15
7
H72
7
H46
0
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10K
T
BEA
S
0
5
10
15
CD
59 (G
APD
H %
)
Figure 14. Expression of complement inhibitors by NSCLC and normal epithelial cells. mRNA semi-quantification of complement inhibitors factor H (A), FHL-1 (B), factor I (C), CD46 (D), CD55 (E) or CD59 (F) is presented as percentage relative to the GAPDH mRNA. Representative graphs for each inhibitor analysis are shown.
We next compared the concentration of factor H in bronchoalveolar lavage
samples from patients with non-malignant lung disease and from patients with lung
malignancies (both SCLC and NSCLC) (Figure 15A). Cancer patients showed
significantly higher levels of factor H in bronchoalveolar lavage samples than non-
cancer patients (p=0.002). A cytologic confirmation of the presence of malignant cells
in the clinical samples was also obtained. A ROC curve was generated for factor H. The
area under the curve was 0.79 (95% confidence interval, 0.63-0.94). A cut-off point of
Results
76
1 µg/ml was established in order to provide a sensitivity of at least 80%. Using this cut-
off, the sensitivity and specificity of the factor H test was 82% and 77%, respectively
(Figure 15B). These results were validated in an independent set of bronchoalveolar
lavage samples and in a series of sputum supernatants, with nearly identical results (data
not shown). Therefore, it can be concluded that factor H is elevated in biological fluids
from lung cancer patients and its quantification may be used as an adjunct to cytology in
the diagnosis of malignant pulmonary diseases146.
A
∗∗∗∗∗∗∗∗
Non
-mal
igna
nt
lung
dis
ease
0.0
2.5
5.0
7.5
Fa
cto
r H
(µµ µµ
g/m
l)
Lung
can
cer
Se
nsit
ivit
y
1-Specificity
BA
∗∗∗∗∗∗∗∗
Non
-mal
igna
nt
lung
dis
ease
0.0
2.5
5.0
7.5
Fa
cto
r H
(µµ µµ
g/m
l)
Lung
can
cer
Se
nsit
ivit
y
1-Specificity
B
Se
nsit
ivit
y
1-Specificity
B
Figure 15. Quantification of complement factor H in bronchoalveolar lavage supernatants A) Box plot for the quantification of factor H in bronchoalveolar lavage fluids from patients with a benign respiratory disease and patients with lung cancer. The levels of factor H were significantly higher in the lung cancer group than in the control group (**p<0.01). B) A ROC curve was generated for factor H. Area under the curve was 0.79 (95% confidence interval, 0.63-0.94; p=0.003).
Following with the characterization of the membrane-bound regulators in
NSCLC, the role of regulators of the MAC formation was more deeply studied. We first
analyzed the expression of the main inhibitor at this level, CD59. Our intention was to
study the inhibitory effect of CD59 on the cell membrane before and after the treatment
with phosphatidylinositol-specific phospholipase C (PI-PLC). CD59, as well as CD55,
is anchored in the plasma membrane via a conserved glycan-phosphatidylinositol (GPI)
anchor. The anchor consists of a phosphatidylinositol group and a carbohydrate-
containing linker that binds to the C-terminal amino acid of a mature protein. This type
of binding is shared by a heterogeneous group of proteins and can be released from cell
membranes by purified bacterial PI-PLCs. The first observation we made was that
CD59 protein expression on the cell membrane was irregular among NSCLC and
normal epithelial cells (Figure 16A), as expected from our previous results in mRNA
Results
77
expression. We next analyzed the remaining CD59 on the cell membrane after PI-PLC
treatment. The reason to perform this analysis was that, in some cases, palmitate, one of
the two fatty acids within the hydrophobic phosphatidylinositol group, is esterified to
the 2-OH group of myo-inositol; this feature prevents the release of GPI-proteins after
PI-PLC action and distinguishes two types of proteins, PI-PLC resistant and PI-PLC
sensible157. We noticed that both the sensitivity and resistance to PI-PLC treatment were
commonly present in the two cell groups (Figure 16B). When the remaining CD59 in
the cells was compared to that on BEAS-2B cells, four NSCLC cell lines showed to be
more resistant to PI-PLC action (pA549=0.028; pH23=0.021; pH460=0.014; pH1299=0.023),
although HBEC10KT cells also showed this resistance to the treatment
(pHBEC10KT=0.019). After activation of the alternative complement pathway, C5b-9
deposition on the cell membrane increased in cells treated with PI-PLC compared to
untreated cells; however, no correlation with the levels of CD59 that remained attached
to the cell was observed (data not shown). These results prompted us to look for
alternative regulatory mechanisms. Casein kinase 2 (CK2), an ubiquitous and
constitutively protein kinase that is expressed on the cell membrane, regulates MAC
assembly. We analyzed the expression of CK2 on the cell membrane of tumor and
normal epithelial cells. However, no difference in CK2 expression on the cell surface
was uncovered (Figure 16C).
Results
78
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10 KT
BEA
S
0
500
1000
1500
2000
2500
CD
59 e
xpre
ssio
n (M
FI units)
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10 KT
BEA
S
0
20
40
60
80
100
CD
59 a
fter
PI-PLC
(%
)
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10 KT
BEA
S
0
5
10
15
20
CK
2 e
xpre
ssio
n (M
FI units)
A B
C
∗
∗∗
∗ ∗∗
∗∗
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10 KT
BEA
S
0
500
1000
1500
2000
2500
CD
59 e
xpre
ssio
n (M
FI units)
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10 KT
BEA
S
0
20
40
60
80
100
CD
59 a
fter
PI-PLC
(%
)
A54
9H23
H15
7
H46
0
H72
7
H12
99
HTB
58
HBEC
3KT
HBEC
3KTp
53
HBEC
10 KT
BEA
S
0
5
10
15
20
CK
2 e
xpre
ssio
n (M
FI units)
A B
C
∗
∗∗
∗ ∗∗
∗∗
Figure 16. Expression of CD59 and ecto-phosphokinase CK2 on NSCLC and normal epithelial cells. CD59 expression (A), remaining CD59 on the cell membrane after the treatment with PI-PLC (B), and CK2 expression (C) were measured by flow cytometry. Data represent mean fluorescence intensity (MFI) from three independent experiments and are expressed as mean ± SEM. *p<0.05; **p<0.01.
1.5. Effect of hypoxia in the expression of complement inhibitors on NSCLC
cells
In a previous work, Drs Anna Blom and Marcin Okroj from the University of
Lund (Malmö, Sweeden), in collaboration with our group, had shown that two NSCLC
cell lines, H2087 and H358, express not only the whole set of soluble complement
inhibitors (factor H, C4BP and factor I), but also the main membrane-bound
complement inhibitors (CD46, CD55 and CD59)124. This work prompted us to study the
effect of different physiological conditions in complement regulation. In particular, we
examined how hypoxic conditions influence the expression of complement
regulators158. For the study, H358 and H2087 cells were incubated in hypoxia (48
hours), in hypoxia followed by normoxia (24 hours each) or in normoxia (48 hours).
Expression of complement inhibitors, at mRNA and protein levels, was analyzed. Real-
time PCR showed a significant reduction in the expression of soluble and membrane
inhibitors in both cell lines, except for CD59 (Figure 17A). CD59 levels on the cell
membrane, analyzed by flow cytometry, were not affected by hypoxic conditions. CD55
Results
79
exhibited a significant decrease in its expression after 24 h of hypoxia followed by
reoxygenation in H358 cells. CD46 presented a substantial loss in both cell lines after
24 h of hypoxia/reoxygenation, which was intensified after 48 h of hypoxia (Figure
17B). Factor I and factor H levels were measured by ELISA. Conditioned media
collected from both cell lines showed a strong decrease in the levels of the soluble
factors in hypoxia/reoxygenation and hypoxia, in comparison to normoxic conditions
(Figure 17C). These results imply that under hypoxic conditions NSCLC cells give up
some of their defense mechanisms and become more prone to complement attack.
CFH CFI CD46 CD55 CD590
50
100
150
mR
NA e
xpessio
nvs. norm
oxic
cells
(%
)
CFH CFI CD46 CD55 CD590
50
100
150
mR
NA e
xpessio
nvs. norm
oxic
cells
(%
)
CD46 CD55 CD590
50
100
150
Pro
tein
expre
ssio
nvs. norm
oxic
cells
(%
)
CD46 CD55 CD590
50
100
150
Pro
tein
expre
ssio
nvs. norm
oxic
cells
(%
)
FH FI0
50
100
150
Pro
tein
rele
ased
vs. norm
oxic
cells
(%
)
FH FI0
50
100
150
Pro
tein
rele
ased
vs. norm
oxic
cells
(%
)
H358 cells H2087 cells
A
B
C
∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗
∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗
Normoxia 24h Hypoxia/Reoxygenation 48h Hypoxia
∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗
CFH CFI CD46 CD55 CD590
50
100
150
mR
NA e
xpessio
nvs. norm
oxic
cells
(%
)
CFH CFI CD46 CD55 CD590
50
100
150
mR
NA e
xpessio
nvs. norm
oxic
cells
(%
)
CD46 CD55 CD590
50
100
150
Pro
tein
expre
ssio
nvs. norm
oxic
cells
(%
)
CD46 CD55 CD590
50
100
150
Pro
tein
expre
ssio
nvs. norm
oxic
cells
(%
)
FH FI0
50
100
150
Pro
tein
rele
ased
vs. norm
oxic
cells
(%
)
FH FI0
50
100
150
Pro
tein
rele
ased
vs. norm
oxic
cells
(%
)
H358 cells H2087 cells
A
B
C
∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗
∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗
Normoxia 24h Hypoxia/Reoxygenation 48h HypoxiaNormoxia 24h Hypoxia/Reoxygenation 48h Hypoxia
∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗
Figure 17. Expression of complement inhibitors by NSCLC cells in hypoxic conditions. Expression in normoxic cells was set as 100% (indicated by a dotted line). Data from at least three experiments are summarized in the graphs and expressed as mean ± SEM. A) mRNA expression of each complement inhibitor was studied by real-time PCR and calculated as percentage relative to the GAPDH mRNA. B) Presence of membrane-bound complement inhibitors, CD46, CD55 and CD59, on NSCLC cells exposed to hypoxia were analyzed by flow cytometry, and expressed as percent of mean fluorescence intensity (MFI) versus normoxic cells. C) Concentrations of soluble factors were measured by ELISA. *p<0.05; **p<0.01; ***p<0.001.
Results
80
1.6. C5a levels in patients with NSCLC
Our in vitro data suggest that NSCLC cells are able to activate complement
spontaneously, leading to the deposition of complement proteins on the cell membrane
and, likely, to the release of complement components, such as anaphylatoxins, to the
extracellular medium. To assess whether this may be the case in vivo, we evaluated the
concentration of C5a, the most potent anaphylatoxin, in clinical samples from patients
diagnosed with lung cancer. We compared it with the concentration of this protein in
samples from control patients. Quantification of C5a was first performed in plasma
from NSCLC (both adenocarcinoma and squamous cell carcinoma) patients and healthy
donors matched by age and smoking history (Figure 18A). Cancer patients showed
significantly higher levels of C5a in plasma than healthy donors (9.62 ± 2.15 ng/ml vs.
5.13 ± 1.15 ng/ml; p<0.001). No histological differences were found. C5a was also
quantified in bronchoalveolar lavage supernatants from patients with non-malignant
lung disease and patients with lung malignancies (both SCLC and NSCLC) (Figure
18B). C5a levels in cancer patients were significantly higher than in patients with non-
malignant respiratory diseases (40.16 ± 37.08 ng/ml vs. 25.69 ± 25.79 ng/ml; p=0.049).
Again, no statistical differences were observed between histological subtypes.
∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
Con
trol
NSC
LC
0
10
20
30
40
C5a (ng/m
l)
Con
trol
Lung
can
cer
0
50
100
150
200
C5a (ng/m
l)
A B
∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
Con
trol
NSC
LC
0
10
20
30
40
C5a (ng/m
l)
Con
trol
Lung
can
cer
0
50
100
150
200
C5a (ng/m
l)
A B
Figure 18. C5a concentration in biological fluids from patients with lung cancer and controls. A) Box plot for the quantification of C5a in plasma samples from patients with NSCLC or from healthy donors. B) Box plot for the quantification of C5a in bronchoalveolar lavage supernatants from patients with lung cancer (NSCLC and SCLC) and patient with non-malignant respiratory diseases. *p<0.05; ***p<0.001.
Results
81
2. COMPLEMENT DEFICIENCY IMPAIRS TUMOR GROWTH
Our in vitro experiments have shown that lung tumor cells activate complement
spontaneously; however this activation is not able to damage them. Most probably, this
is due to the acquisition of complement regulatory mechanisms which inhibit the
progression of complement activation before it causes harm to the cells. At this point,
we speculated with the possibility that this partial complement activation may have
other functional consequences to the cell or its microenvironment. In particular, the
anaphylatoxin C5a, overexpressed in biological fluids from lung cancer patients, have
been shown to have promoting effects on tumor cells, and on cells present within the
tumor microenvironment. Bearing this fact in mind, the next step was to evaluate the in
vivo consequences of complement deprivation, to check if complement activation could
be beneficial for lung tumorigenesis. We used C3 deficient mice and the
pharmacological blockade of C5aR-1 (the main C5a receptor).
A syngeneic model of lung cancer was used to perform the in vivo experiments.
The model consisted in the subcutaneous injection of Lewis lung carcinoma (LLC or
3LL) cells into the flanks of C57BL/6J mice. We first confirmed that 3LL cells were
able to activate the alternative pathway of complement in vitro. Cells incubated with
10% normal mouse serum (NMS) presented higher C3 deposition than cells incubated
with 10% heat-inactivated NMS (Figure 19A).
2.1. Inhibition of tumor growth by C3 deficiency
The formation of the C3 convertase is a main step in complement activation. At
this point, the three central pathways converge. Consequently, C3 deficiency is a good
manner to prevent the generation of complement effectors. Therefore, using C3
knockout mice we were able to study the effect of complement, in a wide and general
way, in the promotion of tumors.
Before starting the experiments, mice were genotyped and C3 expression was
checked. The presence of C3 gene was evaluated by PCR and the concentration of
serum C3 by ELISA. The assays confirmed that C3-deficient mice lacked the C3 gene
and had undetectable concentrations of C3 in their serum (data not shown).
We next assessed tumor growth in C3-deficient and wild-type mice after
subcutaneous inoculation of 3LL cells. Tumor growth was significantly impaired in the
Results
82
absence of C3 (Figure 19B). Tumor volumes measured every three days, starting about
one week after subcutaneous inoculation of tumor cells, were significantly lower in the
C3-deficient mice than in the wild-type controls from the 15th day after cell injection
(p=0.002) until the end of the experiment (p=0.009). At the end of the experiment, we
observed no increase in C3 levels in wild-type tumor-bearing mice when compared with
a pool of serum from ten tumor-free wild-type mice (data not shown). This implies that
these tumor cells are not producing detectible C3 to reverse its deficiency in C3
knockout mice. This result indicates that complement activation promotes lung tumor
growth.
It has been reported that C5 can be proteolyzed into C5a and C5b in the genetic
absence of C334. If that was the case in our in vivo model, this would imply that C5a and
C5b may be present in C3 deficient mice and would not be responsible for the tumor
promoting effect of complement. Moreover, an ELISA for murine C5a showed similar
levels of the anaphylatoxin in C3-deficient and wild-type animals (data not shown). To
rule out any possible effect of C3-independent C5 convertases in our model, we
repeated the last experiment using C3 deficient animals injected with cobra venom
factor (CVF). Injection of CVF has been widely used to deplete complement in several
animal models. CVF resembles functionally and structurally C3b; this characteristic
allows it to bind factor B and to form a stable C3/C5 convertase that cleaves both
complement components C3 and C5. The uncontrollable activity of the convertase
induces consumption of complement proteins C3 and C5, resulting in complement
depletion159. Therefore, the injection of CVF in C3 deficient animals guaranteed a
complete depletion of C5 from circulation.
To settle the timing of CVF injection, we monitored C3 levels in wild-type mice
after a dose of 5 µg CVF. C3 levels were undetectable during three days after the
injection and started to recover after the forth day. After one week of injection, C3
levels became normal again (Figure 19C). Based on these results, we decided to inject
CVF in C3 deficient animals at a dose of 5 µg every three days. Using this regiment of
administration, we again assessed in vivo tumor growth in C3-deficient mice after
subcutaneous inoculation of 3LL tumor cells. As previously described, cell growth was
significantly reduced in C3-deficient animals. This difference was maintained in C3-
deficient mice injected with CVF (Figure 19D).
Results
83
0
1000
2000
3000
4000
9 12 15 18 21 24
WT
C3 KO
0
1000
2000
3000
4000
9 12 15 18 21 24
WT
C3 KO
C3 KO + CVF
∗∗
∗
∗∗
A B
C D
Tu
mo
r vo
lum
e (
mm
3)
Tu
mo
r vo
lum
e (
mm
3)
100 101 102 103 104
Fluorescence intensity
025
0C
ell
nu
mb
er
Basal 1 2 3 4 5 6 7 8
50
100
150
Basal 1 2 3 4 5 6 7 8 Days after CVF injection
Basal 1 2 3 4 5 6 7 8
50
100
150
Basal 1 2 3 4 5 6 7 8 Days after injection
C3 v
sb
asal
valu
e (
%)
∗∗
n.s.
Days after injection
0
1000
2000
3000
4000
9 12 15 18 21 24
WT
C3 KO
0
1000
2000
3000
4000
9 12 15 18 21 24
WT
C3 KO
C3 KO + CVF
∗∗
∗
∗∗
A B
C D
Tu
mo
r vo
lum
e (
mm
3)
Tu
mo
r vo
lum
e (
mm
3)
100 101 102 103 104
Fluorescence intensity
025
0C
ell
nu
mb
er
100 101 102 103 104
Fluorescence intensity
025
0C
ell
nu
mb
er
Basal 1 2 3 4 5 6 7 8
50
100
150
Basal 1 2 3 4 5 6 7 8 Days after CVF injection
Basal 1 2 3 4 5 6 7 8
50
100
150
Basal 1 2 3 4 5 6 7 8 Days after injection
C3 v
sb
asal
valu
e (
%)
∗∗
n.s.
Days after injection
Figure 19. Tumor growth of Lewis lung carcinoma (3LL) cells injected in C57BL mice with normal or deficient complement activity. A) Activation of the alternative pathway of complement in 3LL cells measured by flow cytometry. Cells were incubated with normal mouse serum (white peak) or heat-inactivated normal mouse serum (shaded peak). B) Tumor growth of the 3LL cell line injected s.c. in C3-deficient mice (C3 KO) and wild-type mice (WT). Significant differences were observed from day 15th after injection. Data are representative of three independent experiments with at least 10 mice per group. Tumor volume is expressed as mean ± SEM. C) Consumption of serum C3 in C57BL mice after injection of three i.p. doses of 5 µg CVF; doses were injected 28, 24 and 4 hours before the collection of the first blood sample till the 8th day. C3 concentration in serum was determined by ELISA and expressed as percentage of basal C3 in mice without CVF treatment. D) Tumor growth of the 3LL cell line injected in wild-type and C3-deficient mice treated with or without CVF. Data correspond to one experiment with at least 10 mice per group. Tumor volume is expressed as mean ± SEM. *p<0.05; **p<0.01.
2.2. Inhibition of tumor growth by C5aR-1 blockade
We postulated that the C5a anaphylatoxin may be a good candidate for the tumor
promoting effect of complement activation. C5a causes a number of biological effects,
which are essential in the primary innate immune response160. However, it has been
demonstrated that the functions of C5a are not limited to the immune response. C5a also
elicits several cellular responses related to tumor growth and progression. In addition,
we observed increased levels of C5a in plasma and bronchioalveolar lavage fluid from
patients with NSCLC. These data, together with our results in C3-deficient mice,
directed us towards the study of the implications of C5a on tumor growth in our
syngeneic model of lung cancer. To achieve this purpose, C5aR-1 was blocked with the
Results
84
cyclic hexapeptide AcF(OP(D)ChaWR). Tumor growth was compared between wild-
type animals treated every three days with the C5aR antagonist and wild-type animals
treated with the control cyclic hexapeptide AcF(OP(D)ChaA(D)R). The blockade of
C5aR-1 decreased tumor growth, relative to the effect of the control peptide (Figure
20A). This reduction reached statistical significance from the 24th day of tumor cell
injection (p=0.035) to the end of the experiment (p=0.005). The pharmacological
blockade of C5aR-1 was also used in animals with C3-deficiency. We questioned
whether blocking the C5a receptor in animals deficient in C3 had an additive effect in
tumor growth. For this purpose, C3-deficient mice were treated with the C5aR-1
antagonist or with the control peptide after injection of 3LL cells. Tumor growth was
compared with wild-type mice treated with the control peptide. Tumor growth was
reduced in C3 knockout mice independently of the C5aR-1 blockade (Figure 20B). As
with the CVF studies, these results suggest that no C3-independent C5 convertases are
generated in the C3 knockout model which may have affected the interpretation of the
results. More importantly, our results demonstrate the participation of C5aR-1 in lung
tumor growth in vivo.
0
500
1000
1500
2000
2500
12 15 18 21 24 27 30 32
WT + control peptide
WT + C5aR antagonist
0
500
1000
1500
2000
2500
12 15 18 21 24 27 30 32
WT + control peptide
C3 KO + control peptide
C3 KO + C5aR-1 antagonist∗∗
n.s.
Tu
mo
r vo
lum
e (
mm
3)
Days after injection
Tu
mo
r vo
lum
e (
mm
3)
Days after injection
∗
∗
∗∗
A B
0
500
1000
1500
2000
2500
12 15 18 21 24 27 30 32
WT + control peptide
WT + C5aR antagonist
0
500
1000
1500
2000
2500
12 15 18 21 24 27 30 32
WT + control peptide
C3 KO + control peptide
C3 KO + C5aR-1 antagonist∗∗
n.s.
Tu
mo
r vo
lum
e (
mm
3)
Days after injection
Tu
mo
r vo
lum
e (
mm
3)
Days after injection
∗
∗
∗∗
A B
Figure 20. Pharmacological blockade of C5aR-1 decreases tumor growth. A) Tumor volumes of 3LL cells in wild-type mice treated with the C5aR-1 antagonist or the control peptide. B) Tumor volumes in wild-type or C3-deficient animals treated with the C5aR-1 antagonist or the control peptide. In each case, data are representative of two independent experiments, each one with at least 10 mice per group. Tumor volume is expressed as mean ± SEM. *p<0.05; **p<0.01.
Results
85
3. COMPLEMENT ACTIVATION REGULATES ANTITUMOR
IMMUNE RESPONSE
Once we had demonstrated that the activation of C5aR-1 is important for tumor
growth in vivo, we were interested in evaluating the mechanisms by which this effect
takes place. Immune system activation is a crucial mechanism that contributes to the
regulation of tumor growth, and C5a is a modulator of the immune response. Therefore,
it was likely that C5a was mediating some sort of control in the immune response that
would explain its role in tumor growth. Using our syngeneic animal model of lung
cancer, we analyzed the effect of C3 deficiency and C5aR-1 blockade in the immune
response to the tumor.
3.1. Immune activation is induced in C3-deficient animals
Immune cells within the spleens of animals from the experiment described in
section 2.1. (in which C3 deficient and wild-type mice were inoculated subcutaneously
with 3LL cells) were analyzed by flow cytometry. We observed that complement
deficiency quantitatively altered the populations of cells present in the spleen. As shown
in Figure 21, animals deficient in C3 (which presented smaller tumors than wild-type
ones) exhibited important changes in some immune populations. There was an increase
of CD19+ (p=0.004) and CD8+ (p=0.017) populations. There was also a tendency to the
increase of CD4+ cells, although the difference did not reach statistical significance
(p=0.099). In addition, myeloid-derived suppressor cells (MDSCs; CD11b+Ly6c+)
decreased in C3 knockout mice (p=0.022). MDSCs consist of two major
subpopulations, granulocytic and monocytic, both differ in the expression of two
epitopes, Ly6C and Ly6G, in the GR1 protein. The Ly6G molecule is expressed
primarily on granulocytes161, whereas Ly6C is typically expressed on monocytes162.
Using an antibody against Ly6C we differentiated granulocytic MDSCs, with Ly6Clow,
from monocytic MDSCs, with Ly6Chigh. We observed that both subpopulations were
reduced in the spleens of C3 knockout mice compared to those in the spleens of wild-
type ones (pLy6C-low= 0.036; pLy6C-high= 0.007). We also analyzed the population of Treg
cells (CD4+CD25+Foxp3+) within the spleen. The percentage of Treg cells tended to
Results
86
decrease in C3 knockout mice, but the changes were not statistically significant.
However, the ratio between Treg and CD4 cells (Treg %/CD4 %) was clearly different
between groups (p<0.001). These results indicate that complement sufficient animals
present an immune suppressive microenvironment that may tolerate tumor progression.
WT
C3 KO
0
10
20
30
40
50
CD
19 (
%)
WT
C3 KO
0
5
10
15
CD
4 (%
)
WT
C3 KO
0
5
10
15
CD
8 (%
)
WT
C3 KO
0
1
2
3
Tre
gs (
%)
WT
C3 KO
0
5
10
15
20
Tre
gs (
%)
/ C
D4 (%
)
WT
C3 KO
0
2
4
6
8
10
MD
SC
(Ly6c low
) (%
)
WT
C3 KO
0.0
0.5
1.0
1.5
2.0
MD
SC
(Ly6c h
igh) (%
)
WT
C3 KO
0
5
10
15
Tota
l M
DS
C (%
)
∗∗ns
∗
∗∗ ∗∗
∗∗∗ns
WT
C3 KO
0
10
20
30
40
50
CD
19 (
%)
WT
C3 KO
0
5
10
15
CD
4 (%
)
WT
C3 KO
0
5
10
15
CD
8 (%
)
WT
C3 KO
0
1
2
3
Tre
gs (
%)
WT
C3 KO
0
5
10
15
20
Tre
gs (
%)
/ C
D4 (%
)
WT
C3 KO
0
2
4
6
8
10
MD
SC
(Ly6c low
) (%
)
WT
C3 KO
0.0
0.5
1.0
1.5
2.0
MD
SC
(Ly6c h
igh) (%
)
WT
C3 KO
0
5
10
15
Tota
l M
DS
C (%
)
∗∗∗∗nsns
∗∗
∗∗∗∗ ∗∗∗∗
∗∗∗∗∗∗nsns
Figure 21. C3 deficiency affects populations of splenocytes in tumor-bearing mice. Graphs represent data from the flow cytometry analysis of splenocytes obtained from wild-type (WT) or C3 knockout mice (C3 KO). Splenocytes were stained with specific antibodies to CD19+, CD4+, CD8+, CD11b+Ly6c+ (MDSC), and CD4+CD25+Foxp3+ (Tregs). Individual results per animal and mean per group are shown. *p<0.05; **p<0.01; ***p<0.001.
Results
87
3.2. C5aR-1 blockade induces immune activation
Due to the last result, and considering that C5a has been proved to be a key factor
in the immune regulation of tumors, mainly by the recruitment of MDSCs107; we
analyzed the splenocytes from tumor-bearing animals treated with the C5aR-1
antagonist or the control peptide. Similarly to the results obtained in C3 knockout mice,
splenocytes from animals treated with the control peptide showed an
immunosuppressive microenvironment compared to the animals treated with the C5aR-
1 antagonist (Figure 22). In this case, the increased in CD19+, CD4+ and CD8+
populations with C5aR-1 blockade did not reach significance, although a tendency was
observed. Total MDSCs, as well as MDSC subpopulations, were significantly reduced
when C5aR-1 was blocked (pTotal= 0.015; pLy6C-low=0.018; pLy6C-high= 0.045). Finally,
there were no differences among groups in the percentage of Treg cells or the ratio Treg
%/CD4 %.
Results
88
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
0
10
20
30
CD
19 (%
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
0
2
4
6
8
10
CD
4 (%
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
0
5
10
15
CD
8 (%
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
0.0
0.5
1.0
1.5
2.0
Tre
gs (%
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
0
5
10
15
20
Tre
gs (%
) / C
D4 (%
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
0
5
10
15
20M
DS
C (Ly6c low
) %
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
0
1
2
3
4
MD
SC
(Ly6c h
igh) %
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
0
5
10
15
20
MD
SC
%
ns
∗
nsns
∗ ∗
nsns
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
0
10
20
30
CD
19 (%
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
0
2
4
6
8
10
CD
4 (%
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
0
5
10
15
CD
8 (%
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
0.0
0.5
1.0
1.5
2.0
Tre
gs (%
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
0
5
10
15
20
Tre
gs (%
) / C
D4 (%
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
0
5
10
15
20M
DS
C (Ly6c low
) %
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
0
1
2
3
4
MD
SC
(Ly6c h
igh) %
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
0
5
10
15
20
MD
SC
%
nsns
∗∗
nsnsnsns
∗∗ ∗∗
nsnsnsns
Figure 22. C5aR-1 blockade affects populations of splenocytes in tumor-bearing mice. Graphs represent data from the flow cytometry analysis of splenocytes obtained from mice treated with the control peptide or the C5aR-1 antagonist. Splenocytes were stained with specific antibodies to CD19+, CD4+, CD8+, CD11b+Ly6c+ (MDSC), and CD4+CD25+Foxp3+ (Tregs). Individual results per animal and mean per group are shown. *p<0.05.
Results
89
3.3. C3 deficiency and C5aR-1 blockade reduces the expression of
immunosuppressive cytokines within the tumor
Analysis of splenocytes is a valuable information about the immune response
generated against the tumor, but direct evaluation of the tumor microenvironment is also
needed. For this purpose, total RNA was extracted from explanted tumors of wild-type
animals treated with the control peptide or the C5aR-1 antagonist and from C3 deficient
mice treated with the control peptide. In collaboration with Dr. Lasarte and Dr. Riezu
(Division of Gene Therapy and Hepathology, CIMA), we performed real time PCR for
twenty four molecules related to the immune response. The study showed that
expression of PDL1, CTLA4, LAG3 and ARG1 significantly decreased in tumors from
animals treated with the C5aR-1 antagonist and in tumors from C3-deficient animals
(Figure 23). IL10, IL6 and IL12p35 expression decreased in tumors from animals
treated with the antagonist; and CD70 in tumors from C3-deficient mice. The other
molecules studied (IDO, TNF-α, IL-2, IL-4, Foxp3, PD1, CCL17, CCL22, GITR, TGF-
β, INF-γ, OX40L, MCP1, granzyme A, granzyme B and perforin) showed similar
expression levels between the different groups (data not shown).
Results
90
∗
ns∗∗
∗
∗∗
∗
∗∗
∗∗∗ ∗
∗
∗
ns
∗
ns ∗
ns
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
C3 KO + con
trol pep
tide
0.000
0.005
0.010
0.015
0.020IL
10 m
RN
A (
ββ ββ-a
ctin %
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
C3 KO + con
trol pep
tide
0.0
0.2
0.4
0.6
0.8
PD
L1 m
RN
A (
ββ ββ-a
ctin %
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
C3 KO + con
trol pep
tide
0.00
0.02
0.04
0.06
0.08
0.10
CTLA4 m
RN
A (
ββ ββ-a
ctin %
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 ant
agon
ist
C3 KO + con
trol pep
tide
0
2
4
6
LAG
3 m
RN
A (
ββ ββ-a
ctin %
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
C3 KO + con
trol pep
tide
0
5
10
15
20
AR
G1 m
RN
A (
ββ ββ-a
ctin %
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
C3 KO + con
trol pep
tide
0.0000
0.0005
0.0010
0.0015
IL12p35 m
RN
A (
ββ ββ-a
ctin %
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
C3 KO + con
trol pep
tide
0.00
0.01
0.02
0.03
0.04
IL6 m
RN
A (
ββ ββ-a
ctin %
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 ant
agon
ist
C3 KO + con
trol pep
tide
0.0000
0.0002
0.0004
0.0006
0.0008
0.0010
CD
70 m
RN
A (
ββ ββ-a
ctin %
)
∗∗
nsns∗∗∗∗
∗∗
∗∗∗∗
∗∗
∗∗∗∗
∗∗∗∗∗∗ ∗∗
∗∗
∗∗
nsns
∗∗
nsns ∗∗
nsns
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
C3 KO + con
trol pep
tide
0.000
0.005
0.010
0.015
0.020IL
10 m
RN
A (
ββ ββ-a
ctin %
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
C3 KO + con
trol pep
tide
0.0
0.2
0.4
0.6
0.8
PD
L1 m
RN
A (
ββ ββ-a
ctin %
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
C3 KO + con
trol pep
tide
0.00
0.02
0.04
0.06
0.08
0.10
CTLA4 m
RN
A (
ββ ββ-a
ctin %
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 ant
agon
ist
C3 KO + con
trol pep
tide
0
2
4
6
LAG
3 m
RN
A (
ββ ββ-a
ctin %
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
C3 KO + con
trol pep
tide
0
5
10
15
20
AR
G1 m
RN
A (
ββ ββ-a
ctin %
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
C3 KO + con
trol pep
tide
0.0000
0.0005
0.0010
0.0015
IL12p35 m
RN
A (
ββ ββ-a
ctin %
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
C3 KO + con
trol pep
tide
0.00
0.01
0.02
0.03
0.04
IL6 m
RN
A (
ββ ββ-a
ctin %
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 ant
agon
ist
C3 KO + con
trol pep
tide
0.0000
0.0002
0.0004
0.0006
0.0008
0.0010
CD
70 m
RN
A (
ββ ββ-a
ctin %
)
Figure 23. RNA expression in 3LL tumors of molecules related to the immune response. Total mRNA was extracted from tumors of wild-type mice (WT) treated with the control peptide or the C5aR-1 antagonist, and from tumors of C3 knockout mice (C3 KO) treated with the control peptide. mRNA levels of the different immune mediators were calculated by real time PCR and expressed as percentage relative to the β-actin mRNA in each tumor. Individual results per animal and mean per group are shown. *p<0.05; **p<0.01; *** p<0.001.
Results
91
4. EFFECT OF C5a IN THE PROMOTION OF ANGIOGENESIS
At this point of the study, we can conclude that lung cancer cells activate
complement and that this activation is important for tumor growth, because released
factors such as C5a generate an immunosupressive microenvironment. Nevertheless, the
relevance of C5a may not be limited to the regulation of immune response. In this sense,
it has been suggested that C5a may affect the endothelium during inflammation.
Therefore, our next step in the study was to evaluate the effect of C5a in human
umbilical vein endothelial cell (HUVEC) proliferation and migration, as well as the
capacity of C5a to induce angiogenesis.
4.1. Effect of C5a on HUVEC proliferation
First, the expression of C5aR-1 on HUVECs was demonstrated by flow
cytometric analysis (Figure 24A). Cell proliferation assays were performed in HUVECs
stimulated with different concentrations of C5a. At first, a wide range of C5a
concentrations was used, and proliferation was measured after 96 hours. In these
conditions, C5a did not affect the proliferation of HUVECs (Figure 24B). Next, we
tested the capacity of C5a to synergize the proliferative activity of bFGF, since this
molecule is one of the supplements needed for HUVEC growth. To select the
appropriate bFGF concentration, we performed a dose-response assay where HUVECs
were treated with bFGF at different concentrations, ranging from 0.02 to 10 ng/ml
(Figure 24C). We selected 0.31 ng/ml of bFGF because this was one of the lowest
concentrations in the linear slope of the dose-response curve sufficient to support cell
proliferation. MTT assays were performed adding bFGF at 0.31 ng/ml in two different
conditions: C5a at different concentrations ranging from 12.5 to 100 ng/ml, or C5a at a
constant concentration of 100 ng/ml, with the other two anaphylatoxins, C3a and C4a,
at 25, 50 or 100 ng/ml. None of these attempts had any effect in HUVEC proliferation
(Figure 24D).
Results
92
contro
l
FGF (10 ng
/ml)
0.5 5 25 50 25
050
010
00
0.0
0.1
0.2
0.3
0.4
0.5
C5a (ng/ml)
Absorb
ance (540 n
m)
0.0
0.1
0.2
0.3
Absorb
ance (540 n
m)
00.02
0.04
0.08
0.15
0.31
0.62
1.25
2.5 5 10
0.0
0.1
0.2
0.3
0.4
bFGF (ng/ml)
Absorb
ance (540 n
m)
100 101 102 103 104
20
00
bFGF (ng/ml)
C5a (ng/ml)
C3a and C4a (ng/ml)
- 10 0.31 0.31 0.31 0.31 0.31 0.31 0.31 0.31
- - - 12.5 25 50 100 - - -
- - - - - - - 25 50 100
A B
C D
Fluorescence intensity
Cell
nu
mb
er
contro
l
FGF (10 ng
/ml)
0.5 5 25 50 25
050
010
00
0.0
0.1
0.2
0.3
0.4
0.5
C5a (ng/ml)
Absorb
ance (540 n
m)
0.0
0.1
0.2
0.3
Absorb
ance (540 n
m)
00.02
0.04
0.08
0.15
0.31
0.62
1.25
2.5 5 10
0.0
0.1
0.2
0.3
0.4
bFGF (ng/ml)
Absorb
ance (540 n
m)
100 101 102 103 104
20
00
100 101 102 103 104
20
00
bFGF (ng/ml)
C5a (ng/ml)
C3a and C4a (ng/ml)
- 10 0.31 0.31 0.31 0.31 0.31 0.31 0.31 0.31
- - - 12.5 25 50 100 - - -
- - - - - - - 25 50 100
A B
C D
Fluorescence intensity
Cell
nu
mb
er
Figure 24. Expression of C5aR-1 in human umbilical vein endothelial cells (HUVECs) and effect of C5a on HUVEC proliferation. A) Detection of C5aR-1 on the cell membrane of HUVECs was assessed by flow cytometry. Cells were incubated with an anti-C5aR-1 antibody (white peak) or with an isotype control (shaded peak). B) Proliferation assay with cells incubated with C5a at different concentrations in basal medium for 96 hours. C) Proliferation assay with bFGF at different concentrations in basal medium. D) Proliferation assay with combinations prepared with 0.31 ng/ml bFGF and C5a at different concentrations (black bars), or 0.31 ng/ml bFGF with 100 ng/ml C5a and the other two anaphylatoxins, C3a and C4a, at different concentrations (grey bars). Negative and positive controls in B and D were done with basal medium and with 10 ng/ml bFGF, respectively. In all graphs, one representative experiment is shown, and results are expressed as the mean optical density of the MTT assay ± SEM.
4.2. Effect of C5a on HUVEC migration
We next investigated the capacity of C5a to induce HUVEC migration using the
transwell chamber assay. HUVECs were plated on transwells and were stimulated with
different concentrations of C5a, ranging from 0.03 to 500 ng/ml, during 48 hours.
HUVECs showed a bell shaped dose-response (Figure 25A and 25B). A significant
increase in migration, compared to the untreated cells, was observed within 0.125 and
5 ng/ml of C5a (p0.125=0.015; p0.25<0.001; p0.5=0.012; p5<0.001).
Results
93
cont
rol
0.03
0.06
0.12
50.25 0.
5 5 50 500
0
20
40
60
80
100
C5a (ng/ml)
Cell
num
ber
/ field
∗
∗∗∗
∗∗∗
∗
Control
C5a 0.125 ng/ml C5a 0.25 ng/ml
C5a 0.5 ng/ml C5a 5 ng/ml
A B
cont
rol
0.03
0.06
0.12
50.25 0.
5 5 50 500
0
20
40
60
80
100
C5a (ng/ml)
Cell
num
ber
/ field
∗
∗∗∗
∗∗∗
∗
cont
rol
0.03
0.06
0.12
50.25 0.
5 5 50 500
0
20
40
60
80
100
C5a (ng/ml)
Cell
num
ber
/ field
∗
∗∗∗
∗∗∗
∗
cont
rol
0.03
0.06
0.12
50.25 0.
5 5 50 500
0
20
40
60
80
100
C5a (ng/ml)
Cell
num
ber
/ field
∗
∗∗∗
∗∗∗
∗
Control
C5a 0.125 ng/ml C5a 0.25 ng/ml
C5a 0.5 ng/ml C5a 5 ng/ml
A B
Figure 25. Effect of C5a on HUVEC migration in the transwell assay. C5a was added at different concentrations in basal medium on the lower side of the transwells. HUVECs were plated in the upper chamber in basal medium. At 48 hours after plating, cells that had migrated to the underside of the transwells were fixed and stained with crystal violet. A) Cell migration in each transwell was analyzed by counting cells in eight different fields of the transwell and calculating the mean value. Data are collected from three independent experiments and shown as mean ± SEM. *p<0.05; ***p<0.001. B) Representative images of HUVEC migration at different concentrations are shown at x10 magnification.
Results
94
4.3. Effect of C5a on HUVEC tube formation
In the presence of angiogenic compounds, HUVECs can form tube-like vessels on
Matrigel, a commercial basement membrane matrix derived from the Engelbreth-Holm-
Swarm mouse sacoma163. With this assay, we tested the potential of C5a to induce
angiogenesis in vitro. HUVECs were plated on gelled Matrigel with 10 ng/ml C5a in
EBM-2 medium. After 3 hours of incubation, tube-like structures were starting to form,
and firm tubes appeared from 4 to 6 hours after seeding the cells; over that time,
proteases released by cells started to disintegrate the existing tubes. So, in our
experiments, HUVECs were fixed and stained five hours after plating them in the
matrix. HUVECs treated with C5a showed more numerous tube-like structures than
cells in basal medium (Figure 26A). Analysis of the photographs taken in the central
area of wells demonstrated that total tube length was superior in cells treated with C5a
than in control cells (p=0.002) (Figure 26B).
Con
trol
C5a
0
1000
2000
3000
Tube length
(µµ µµm
/ field
)
**
A
B
Control C5a
Con
trol
C5a
0
1000
2000
3000
Tube length
(µµ µµm
/ field
)
**
A
B
Control C5a
Figure 26. C5a promotes tube formation by endothelial cells. HUVECs were plated on gelled Matrigel in the absence or presence of C5a at 10 ng/ml. After 5 hours, cells were fixed and stained. Three high-power fields per culture condition were examined using the 10X objective. A) Representative pictures are shown. B) For quantification of tube formation, the total length of each picture was measured using ImageJ software (NIH). Mean ± SEM is represented. Results are representative of three independent experiments, each one with at least five wells per condition. **p<0.01.
Results
95
4.4. Decreased tumor angiogenesis in C3 deficient animals
Due to our in vitro results demonstrating a role of C5a in the promotion of
angiogenesis, together with the in vivo impairment of tumor growth in animals deficient
in complement, we decided to analyze the effect of complement deficiency on the
promotion of angiogenesis. For this attempt, LLC cells were injected i.v. in the tail vein
of animals with normal (wild-type) or deficient complement (C3 knockout). Their lungs
were removed three weeks later and embedded in paraffin. The number of macroscopic
and microscopic (in H/E preparations) metastases were countered. No significant
differences between groups were found (data not shown). However the aspect of tumors
was qualitatively different between animals with normal complement and animals with
C3-deficiency. In the former, tumors seemed more hemorrhagic (data not shown).
CD31 immunostaining in lung slides showed a clear difference in microvessel density.
Tumors from C3-deficient mice presented a lower number of vessels than those from
wild-type mice (Figure 27A). The percentage of CD31 stained area was significantly
higher in tumors from wild-type mice than in those from C3-deficient animals
(p<0.001) (Figure 27B).
B
WT
C3 KO
0
2
4
6
8
Imm
unosta
ined a
rea (%
)
∗∗∗
AWT C3 KO
B
WT
C3 KO
0
2
4
6
8
Imm
unosta
ined a
rea (%
)
∗∗∗
WT
C3 KO
0
2
4
6
8
Imm
unosta
ined a
rea (%
)
∗∗∗
AWT C3 KO
Figure 27. Microvascular density, determined by CD31 immunohistochemistry, in lung tumors from wild type (WT) and C3 knockout (C3 KO) mice. A) A representative preparation from a tumor of each group is shown. B) Quantification of the microvascular density was carried out by measuring the fraction of immunostained area on tumor sections. Data show mean ± SEM. ***p < 0.001.
Results
96
4.5. C5aR-1 blockade and C3 deficiency reduces bFGF expression in the tumor
microenvironment
The expression of angiogenic factors was also studied in lung tumors from wild-
type animals treated with the C5aR-1 antagonist, the control peptide, or C3-deficient
animals treated with the control peptide. For this purpose, total RNA was extracted from
tumors and the expression of the angiogenic factors bFGF and VEGF was analyzed by
real time PCR. In the case of bFGF, tumors from wild-type animals treated with the
C5aR-1 antagonist showed a significantly lower expression of the factor (p.=0.015), and
this difference was even more pronounced in tumors from C3-deficient animals
(p=0.004) (Figure 28A). In the case of VEGF, the expression within the tumor
microenvironment was similar in all groups (Figure 28B).
∗∗
∗
ns
ns
A B
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
C3 KO + con
trol pep
tide
0.00
0.02
0.04
0.06
bFG
F m
RN
A (
ββ ββ-a
ctin %
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 ant
agon
ist
C3 KO + con
trol pep
tide
0.00
0.01
0.02
0.03
VEG
F m
RN
A (
ββ ββ-a
ctin %
)
∗∗
∗
∗∗
∗
ns
ns
A B
WT + co
ntro
l pep
tide
WT + C5a
R-1 antag
onist
C3 KO + con
trol pep
tide
0.00
0.02
0.04
0.06
bFG
F m
RN
A (
ββ ββ-a
ctin %
)
WT + co
ntro
l pep
tide
WT + C5a
R-1 ant
agon
ist
C3 KO + con
trol pep
tide
0.00
0.01
0.02
0.03
VEG
F m
RN
A (
ββ ββ-a
ctin %
)
Figure 28. Expression of the angiogenic factors bFGF and VEGF within the tumor microenvironment. Total mRNA was extracted from tumors of wild-type (WT) mice treated with the control peptide, wild-type mice treated with the C5aR-1 antagonist, and C3 knockout (KO) mice treated with the control peptide. Quantification of bFGF (A) and VEGF (B) was carried out by real-time PCR and the expression of each gene was calculated as percentage relative to the β-actin mRNA in each tumor. Individual results per animal and mean per group are shown. *p<0.05; **p<0.01.
Results
97
5. EFFECT OF C5a ON NON-SMALL CELL LUNG CANCER
CELLS
Once we had demonstrated the action of C5a on the regulation of immune
response and angiogenesis, our last question was whether C5a has a direct effect on
cancer cells.
5.1. Expression of C5aR-1 in lung cancer cell lines
First, we analyzed whether lung cancer cells express C5aR-1. By conventional
PCR, we demonstrated that many lung cancer cells express C5aR-1 (Figure 29). In
particular, the expression in small-cell lung cancer (SCLC) cells was very frequent, with
only one negative cell line (H446) out of eight cell lines analyzed. Among NSCLC
cells, most of them also expressed the receptor, although the levels were apparently
lower than in SCLC cells. Carcinoid cells also showed a high expression of the receptor.
Since our current study focused on the effects of complement on NSCLC cells, we
selected three cell lines with this histology for further studies. We chose A549 and
H157 cells as C5aR-1 expressing cells. H661, a cell line with undetected levels of the
protein, was selected as control. Expression of C5aR-1 on these NSCLC cell lines was
confirmed by flow cytometry (Figure 29). Protein levels on the cell membrane
correlated with mRNA levels.
Results
98
Fluorescence intensity
Ce
lln
um
ber
100 101 102 103 104
H661
20
00
100 101 102 103 104
H15720
00
100 101 102 103 104
A549
20
00
A
B
H 6
9
H 8
2
H 1
87
H 2
09
H 3
45
H 4
17
H 4
46
H 5
10
H 2
3
H 4
41
H 2
087
Cal
u3
H 1
648
H C
C82
7
H 3
58
A54
9
H 3
22
SCLC ADC BAC
H 6
9
H 8
2
H 1
87
H 2
09
H 3
45
H 4
17
H 4
46
H 5
10
H 2
3
H 4
41
H 2
087
Cal
u3
H 1
648
H C
C82
7
H 3
58
A54
9
H 3
22
SCLC ADC BAC
H52
0
H 1
703
2170
H 1
57
H 2
26
H T
B58
H 1
299
H 4
60
97TM
1
103H
H 6
61
H 7
27
H 7
20
SQ LC Carc.
H52
0
H 1
703
2170
H 1
57
H 2
26
H T
B58
H 1
299
H 4
60
97TM
1
103H
H 6
61
H 7
27
H 7
20
SQ LC Carc.
C5aR-1
GAPDH
C5aR-1
GAPDH
Fluorescence intensity
Ce
lln
um
ber
100 101 102 103 104
H661
20
00
100 101 102 103 104
H15720
00
100 101 102 103 104
A549
20
00
Fluorescence intensity
Ce
lln
um
ber
100 101 102 103 104
H661
20
00
100 101 102 103 104
H15720
00
100 101 102 103 104
A549
20
00
100 101 102 103 104
H661
20
00
100 101 102 103 104
H661
20
00
100 101 102 103 104
H15720
00
100 101 102 103 104
H157
100 101 102 103 104
H15720
00
100 101 102 103 104
A549
20
00
100 101 102 103 104
A549
20
00
A
B
H 6
9
H 8
2
H 1
87
H 2
09
H 3
45
H 4
17
H 4
46
H 5
10
H 2
3
H 4
41
H 2
087
Cal
u3
H 1
648
H C
C82
7
H 3
58
A54
9
H 3
22
SCLC ADC BAC
H 6
9
H 8
2
H 1
87
H 2
09
H 3
45
H 4
17
H 4
46
H 5
10
H 2
3
H 4
41
H 2
087
Cal
u3
H 1
648
H C
C82
7
H 3
58
A54
9
H 3
22
SCLC ADC BAC
H52
0
H 1
703
2170
H 1
57
H 2
26
H T
B58
H 1
299
H 4
60
97TM
1
103H
H 6
61
H 7
27
H 7
20
SQ LC Carc.
H52
0
H 1
703
2170
H 1
57
H 2
26
H T
B58
H 1
299
H 4
60
97TM
1
103H
H 6
61
H 7
27
H 7
20
SQ LC Carc.
C5aR-1
GAPDH
C5aR-1
GAPDH
Figure 29. Expression of C5aR-1 in lung cancer cells. A) Conventional PCR analysis of C5aR-1 in lung cancer cell lines. Detection of GAPDH mRNA was used to ensure equal loading and RNA integrity. SCLC: small cell lung cancer; ADC: adenocarcinoma; BAC: bronchioloalveolar carcinoma; SQ: squamous cell carcinoma; LC: large cell carcinoma; Carc: carcinoid. B) Detection of C5aR-1 by flow cytometry in some NSCLC cell lines. The expression of C5aR-1 was measured as the increase in intensity in the green channel (white peak). Incubation of cells with a control isotype antibody was used as a negative control (shaded peak).
5.2. Activation of the C5a signaling pathway on NSCLC cells
We next assessed whether C5a was able to activate C5aR-1 on the surface of
NSCLC cells. C5aR-1 is a member of the seven transmembrane G protein-coupled
receptor superfamily. Upon interaction with its ligand, C5aR-1 is activated and induces
calcium mobilization. Therefore, intracellular calcium after incubation with C5a was
analyzed in A549 and H157 cells. Granulocytes were used as a positive controls.
Treatment with high concentrations of C5a (2 µg/ml) increased calcium influx in
granulocytes, but did not produce any change in the calcium content in A549 or H157
cells. Of note, these cell lines showed very low basal levels of calcium compared to
granulocytes. After evaluating calcium basal levels in other NSCLC cells, we found that
H1299 cells, which also express C5aR-1, have higher basal levels of calcium than A549
and H157. In this cell line, high concentrations of C5a (2 µg/ml) enhanced calcium
influx (Figure 30A).
It has been reported that in some cell types, ERK1/2 is activated by C5a
stimulation, leading to NF-κB activation and translocation to the nucleus160. ERK
phosphorylation was analyzed in A549 cells by Western blotting after 30-minute
Results
99
incubation with 10 ng/ml C5a. An increased in ERK phosphorilation was detected
compared to the phosporylation level in cells treated with PBS (Figure 30B). The
translocation of NF-κB to the nucleus after C5aR-1 activation in A549 cells was also
studied. After 45 minutes of incubation with 10 ng/ml C5a, a clear translocation of NF-
κB to the nucleus was observed. This effect was blocked with the C5aR-1 antagonist,
but not with the control peptide (Figure 30C).
PBS C5a
- Phospho-p44 MAPK- Phospho-p42 MAPK
- p44 MAPK- p42 MAPK
B C Control C5a
C5a + C5aR-1 antagonist C5a + control peptide
A
0 50 100 150 200 250 300 0 50 100 150 200 250 300
0 50 100 150 200 250 300 0 50 100 150 200 250 300
0 50 100 150 200 250 300 0 50 100 150 200 250 300
0 50 100 150 200 250 300 0 50 100 150 200 250 300
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
250
200
150
100
50
500
400
300
200
100
0 50 100 150 200 250 300 0 50 100 150 200 250 300
0 50 100 150 200 250 300 0 50 100 150 200 250 300
0 50 100 150 200 250 300 0 50 100 150 200 250 300
0 50 100 150 200 250 300 0 50 100 150 200 250 300
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
250
200
150
100
50
500
400
300
200
100
0 50 100 150 200 250 300 0 50 100 150 200 250 300
0 50 100 150 200 250 300 0 50 100 150 200 250 300
0 50 100 150 200 250 300 0 50 100 150 200 250 300
0 50 100 150 200 250 300 0 50 100 150 200 250 300
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
250
200
150
100
50
250
200
150
100
50
500
400
300
200
100
PBS C5a
Granulocytes
A549
H157
H1299
Time (seconds)
Flu
ore
sce
nc
e in
ten
sit
y
PBS C5aPBS C5a
- Phospho-p44 MAPK- Phospho-p42 MAPK
- p44 MAPK- p42 MAPK
B C Control C5a
C5a + C5aR-1 antagonist C5a + control peptide
A
0 50 100 150 200 250 300 0 50 100 150 200 250 300
0 50 100 150 200 250 300 0 50 100 150 200 250 300
0 50 100 150 200 250 300 0 50 100 150 200 250 300
0 50 100 150 200 250 300 0 50 100 150 200 250 300
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
250
200
150
100
50
500
400
300
200
100
0 50 100 150 200 250 300 0 50 100 150 200 250 300
0 50 100 150 200 250 300 0 50 100 150 200 250 300
0 50 100 150 200 250 300 0 50 100 150 200 250 300
0 50 100 150 200 250 300 0 50 100 150 200 250 300
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
250
200
150
100
50
500
400
300
200
100
0 50 100 150 200 250 300 0 50 100 150 200 250 300
0 50 100 150 200 250 300 0 50 100 150 200 250 300
0 50 100 150 200 250 300 0 50 100 150 200 250 300
0 50 100 150 200 250 300 0 50 100 150 200 250 300
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
80
60
40
20
250
200
150
100
50
250
200
150
100
50
500
400
300
200
100
PBS C5a
Granulocytes
A549
H157
H1299
Time (seconds)
Flu
ore
sce
nc
e in
ten
sit
y
Figure 30. Activation of the C5a signaling pathway in lung cancer cells after C5a treatment. A) Calcium flux in granulocytes and NSCLC cell lines after stimulation with C5a. Fluo-4-loaded cells were initially analyzed by flow cytometry to establish calcium basal levels. After two minutes, either PBS (negative control) or C5a were added and increase in the intracellular calcium flux was recorded in real time over 150 s. B) Western blot analysis for phospho-p44/p42 MAPK and total p44/p42 MAPK was carried out in extracts of A549 cells after 30 minutes of treatment with 10 ng/ml C5a. C) Confocal images showing NF-κB translocation to the nucleus in cells treated with 10 ng/ml C5a, with or without the C5aR-1 antagonist or the control peptide. NF-κB staining is shown in red. DAPI (blue) was used to stain nuclei.
Results
100
5.3. Effect of C5a on NSCLC cell proliferation
The next aspect to be analyzed was the effect of C5a on cell proliferation using
the MTT assay. The two positive cell lines for C5aR-1, A549 and H157, and the
negative one, H661, were tested. We established two conditions for the assay: C5a at
different concentrations, ranging from 0.03 to 500 ng/ml, in RPMI basal medium
(Figure 31A), or in RPMI medium with 0.5% FBS (Figure 31B). The last condition
was tested considering that cells may need a minimal quantity of serum, which would
not disturb the effect of C5a but benefit cell viability during the assay. In none of the
situations C5a affected cell proliferation.
contro
l
FBS 10
%0.03
0,06
0,12
50,25 0,
5 5 50 500
0.0
0.5
1.0
1.5
C5a (ng/ml)
Absorb
ance (540 n
m)
contro
l
FBS 10
%0.03
0,06
0,12
50,25 0,
5 5 50 500
0.0
0.1
0.2
0.3
0.4
0.5
C5a (ng/ml)
Absorb
ance (540 n
m)
contro
l
FBS 10
%0.03
0,06
0,12
50,25 0,
5 5 50 500
0.0
0.1
0.2
0.3
0.4
0.5
C5a (ng/ml)Absorb
ance (540 n
m)
contro
l
FBS 10
%0.03
0.06
0.12
50.25 0.
5 5 50 500
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0.5% FBS + C5a (ng/ml)
Absorb
ance (540 n
m)
contro
l
FBS 10
%0.03
0,06
0,12
50,25 0,
5 5 50 500
0.0
0.2
0.4
0.6
0.8
1.0
0.5% FBS + C5a (ng/ml)
Absorb
ance (540 n
m)
cont
rol
FBS 10
%0.03
0,06
0,12
50,25 0,
5 5 50 500
0.0
0.2
0.4
0.6
0.8
1.0
0.5% FBS + C5a (ng/ml)
Absorb
ance (540 n
m)
A549 H157 H661A
B
contro
l
FBS 10
%0.03
0,06
0,12
50,25 0,
5 5 50 500
0.0
0.5
1.0
1.5
C5a (ng/ml)
Absorb
ance (540 n
m)
contro
l
FBS 10
%0.03
0,06
0,12
50,25 0,
5 5 50 500
0.0
0.1
0.2
0.3
0.4
0.5
C5a (ng/ml)
Absorb
ance (540 n
m)
contro
l
FBS 10
%0.03
0,06
0,12
50,25 0,
5 5 50 500
0.0
0.1
0.2
0.3
0.4
0.5
C5a (ng/ml)Absorb
ance (540 n
m)
contro
l
FBS 10
%0.03
0.06
0.12
50.25 0.
5 5 50 500
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0.5% FBS + C5a (ng/ml)
Absorb
ance (540 n
m)
contro
l
FBS 10
%0.03
0,06
0,12
50,25 0,
5 5 50 500
0.0
0.2
0.4
0.6
0.8
1.0
0.5% FBS + C5a (ng/ml)
Absorb
ance (540 n
m)
cont
rol
FBS 10
%0.03
0,06
0,12
50,25 0,
5 5 50 500
0.0
0.2
0.4
0.6
0.8
1.0
0.5% FBS + C5a (ng/ml)
Absorb
ance (540 n
m)
A549 H157 H661A
B
Figure 31. Effect of C5a on NSCLC cell proliferation. Cells were incubated with C5a at different concentrations ranging from 0.03 to 500 ng/ml, in the absence (A) or presence (B) of 0.5% FBS. Positive controls were performed with 10% FBS. The proliferative response was evaluated using MTT assay. One representative experiment is shown in each case, and results are expressed as the mean absorbance ± SEM.
The absence of any effect on cell proliferation was confirmed by the analysis of
the cell cycle in cells treated with C5a. For this purpose, A549 cells were incubated with
10 ng/ml C5a or RPMI basal medium for 48 hours, and the cell cycle distribution was
assessed. No differences were seen between the two treatments (Figure 32A).
Finally, anchorage-independent growth of A549 cells was studied after treatment
with C5a. As a common malignant cell line, A549 cells have the ability to proliferate
Results
101
without attachment to a substratum. C5a did not modify this malignant feature of A549
cells (Figure 32B).
0
100
200
300
control
C5a
Co
lon
ies
/ w
ell
8 10 12
Days of incubation
Cell
nu
mb
er
Fluorescence intensity
200 400 600 800 1000
SubG0
SG0/G1
G2/M
0
100
2
00
3
00
40
0
500
A B
nsns ns
control C5aSub G0/G1 0.7% 0.63%
G0/G1 70.75% 67.92%
S 15.89% 19.52% G2/M 7.20% 6.36%
0
100
200
300
control
C5a
control
C5a
Co
lon
ies
/ w
ell
8 10 12
Days of incubation
Cell
nu
mb
er
Fluorescence intensity
200 400 600 800 1000
Cell
nu
mb
er
Fluorescence intensity
200 400 600 800 1000
SubG0
SG0/G1
G2/M
0
100
2
00
3
00
40
0
500
A B
nsnsnsns nsns
control C5aSub G0/G1 0.7% 0.63%
G0/G1 70.75% 67.92%
S 15.89% 19.52% G2/M 7.20% 6.36%
Figure 32. Cell cycle distribution and anchorage-independent growth of A549 cells incubated with C5a. Cells were treated 48 hours with or without 10 ng/ml C5a in basal medium. A) Cell cycle profile of control cells is represented in grey, and cell cycle profile of cells treated with C5a is depicted with a black line. B) After treatment with C5a, one thousand A549 cells were plated in soft agar with C5a at 10 ng/ml. Control cells were incubated in the absence of C5a. Experiments were performed in triplicate. The graph shows the number of colonies formed after 8, 10 and 12 days of incubation (mean ± SEM).
5.4. Effect of C5a in A549 migration
C5a is a potent chemoattractant that can stimulate monocytes/macrophages to
migrate to sites of wound healing and inflammation164. We studied the effect of C5a on
NSCLC cell migration.
The Boyden chamber assay was used to study the migration of A549 cells through
a microporous filter toward different concentrations of C5a placed under the filter. Cells
were allowed to migrate for 48 h to the other side. Afterwards, cells were fixed, stained,
and counted. C5a stimulated the migration of A549 cells when added within 0.5 and
50 ng/ml (p0.5=0.009; p5=0.004; p50=0.001) (Figure 33). It is interesting to note that the
addition of C5a at a high concentration (500 ng/ml) had the inverse effect (p<0.001).
Results
102
cont
rol
0.25 0.
5 5 50 500
0
2
4
6
8
10
C5a (ng/ml)
Cell
num
ber
/ field
Control 5 ng/ml C5a
A B
∗∗
∗∗ ∗∗
∗∗∗
cont
rol
0.25 0.
5 5 50 500
0
2
4
6
8
10
C5a (ng/ml)
Cell
num
ber
/ field
Control 5 ng/ml C5aControl 5 ng/ml C5a
A B
∗∗
∗∗ ∗∗
∗∗∗
Figure 33. Effect of C5a on A549 migration in the transwell assay. C5a was added at different concentrations in basal medium (0.25, 0.5, 5, 50, and 500 ng/ml) on the lower side of the transwell, while A549 cells were plated in the upper chamber. After 48 hours of incubation, cells that had migrated to the underside of the transwell were fixed and stained with crystal violet. A) Cell migration in each transwell was analyzed by counting the cells in eight different fields of the transwell (at 10x magnification) and calculating the mean value. Data are collected from three independent experiments and shown as mean ± SEM. **p<0.01; ***p<0.001. B) Representative images at x4 magnification of control cells and cells treated with C5a added at 5 ng/ml.
5.5. Effect of C5a in NSCLC cell adhesion to extracellular matrix proteins
To better characterize the mechanism by which NSCLC cells increase its
migration ability in the presence of C5a, we examined the role of C5a in the capacity of
A549, H157 and H661 cells to adhere to different extracellular matrix (ECM) proteins.
A549 cells showed high adhesion to collagens I and IV, intermediate adhesion to
fibronectin and laminin, and low adhesion to vitronectin and gelatin (Figure 34A).
Adhesion to collagen I was reduced when cells were treated with C5a at 10 ng/ml for 48
hours (p=0.007). This effect was inhibited in the presence of the C5aR-1 antagonist. A
more accurate analysis was done by giving to the basal adhesion a 100% value (Figure
34B). With this analysis, the differences between treatments became more significant. A
clear decrease in the adhesion to collagens and fibronectin was observed (pcol.I=0.001;
pcol.IV=0.002; pfib.<0.001). In the case of collagens, the C5aR-1 antagonist blocked the
effect, but not in the case of fibronectin (p=0.002).
The H157 cell line showed a similar behavior in the adhesion to ECM proteins,
but these cells presented higher adhesion to fibronectin than to collagens (Figure 34A).
Adhesion to collagen I and fibronectin was reduced in cells treated with C5a
(pcol.I<0.001; pfib.=0.043). C5aR-1 antagonist blocked the effect in the case of
fibronectin, but not in the case of collagen I (p=0.002). Using the basal adhesion level
as reference (100%), we observed a significant decrease in cell adhesion to collagen I
(p<0.001), which was not inhibited with the antagonist (p=0.014) (Figure 34B). There
Results
103
was also a decrease in the adhesion to fibronectin and to vitronectin (pfib.=0.006;
pvit.=0.013), which was blocked by the presence of the antagonist.
Finally, the negative cell line, H661, showed a different pattern of adhesion to
ECM proteins. These cells presented low adhesion to collagens and laminin, but high
adhesion to gelatin and fibronectin, and intermediate adhesion to vitronectin. As
expected, adhesion to those ECM proteins was not affected by C5a treatment (Figure
34a and 34B).
Collage
n I
Collage
n IV
Fibr
onec
tin
Vitron
ectin
Laminin
Gelatin
0
20
40
60
80
Adhesio
n to E
CM
pro
tein
s (%
)
Collage
n I
Collage
n IV
Fibr
onec
tin
Vitron
ectin
Laminin
Gelatin
0
20
40
60
80Adhesio
n to E
CM
pro
tein
s (%
)
Collage
n I
Collage
n IV
Fibr
onec
tin
Vitron
ectin
Laminin
Gelatin-10
20
50
80
Adhesio
n to E
CM
pro
tein
s (%
)
Collage
n I
Collage
n IV
Fibr
onec
tin
0
50
100
150
Adhesio
n to E
CM
pro
tein
s v
s. contr
ol cells
(%
)
Collage
n I
Collage
n IV
Fibr
onec
tin
Vitron
ectin
0
50
100
150
Adhesio
n to E
CM
pro
tein
s v
s. contr
ol cells
(%
)
Fibr
onec
tin
Gelatin
0
50
100
150
200
250
Adhesio
n to E
CM
pro
tein
s v
s. contr
ol cells
(%
)
Control C5a C5a + C5aR-1 antagnist
A549 H157 H661A
B
∗∗
∗∗∗
∗∗
∗
∗∗
∗∗
∗∗∗∗∗
∗∗∗
∗∗ ∗∗
Collage
n I
Collage
n IV
Fibr
onec
tin
Vitron
ectin
Laminin
Gelatin
0
20
40
60
80
Adhesio
n to E
CM
pro
tein
s (%
)
Collage
n I
Collage
n IV
Fibr
onec
tin
Vitron
ectin
Laminin
Gelatin
0
20
40
60
80Adhesio
n to E
CM
pro
tein
s (%
)
Collage
n I
Collage
n IV
Fibr
onec
tin
Vitron
ectin
Laminin
Gelatin-10
20
50
80
Adhesio
n to E
CM
pro
tein
s (%
)
Collage
n I
Collage
n IV
Fibr
onec
tin
0
50
100
150
Adhesio
n to E
CM
pro
tein
s v
s. contr
ol cells
(%
)
Collage
n I
Collage
n IV
Fibr
onec
tin
Vitron
ectin
0
50
100
150
Adhesio
n to E
CM
pro
tein
s v
s. contr
ol cells
(%
)
Fibr
onec
tin
Gelatin
0
50
100
150
200
250
Adhesio
n to E
CM
pro
tein
s v
s. contr
ol cells
(%
)
Control C5a C5a + C5aR-1 antagnist
A549 H157 H661A
B
∗∗
∗∗∗
∗∗
∗
∗∗
∗∗
∗∗∗∗∗
∗∗∗
∗∗ ∗∗
Figure 34. Effect of C5a on NSCLC adhesion to different ECM proteins. NSCLC cell lines A549, H157 and H661 were incubated with basal medium alone (control), 10 ng/ml C5a, or C5a and the inhibitor peptide for C5aR1 at 100 ng/ml. After 48 hours, cells were collected, loaded with calcein, and plated in 96 well plates, where the different ECM proteins were set: collagen I, collagen IV, fibronectin, vitronectin, laminin and gelatin. After 30 minutes, non-adherent cells were removed and fluorescence was measured. The percentage of cells adhered to each ECM protein was calculated using the fluorescence of the wells containing the total number of cell loaded into them (A). The percentage of adhesion respect to the basal conditions is also shown (B). Data are presented as mean ± SEM. *p<0.05; **p<0.01; ***p<0.001.
5.5.1 Mechanisms involved in the regulation of cell adhesion to ECM proteins
by C5a
5.4.1. Regulation of integrin expression.
Due to the fact that integrins are heterodimeric transmembrane glycoproteins that
mediate attachment between a cell and the ECM, we studied the possible effect of C5a
Results
104
on integrin expression on the cell membrane. The expression of several well-
characterized integrins, α and β subunits (α1, α4, αv, β1 and β3), and three
hetorodimers (α4β1, α5β1, αvβ3), was apparently unaffected by C5a treatment (Figure
35).
100 101 102 103 104
β1
100 101 102 103 104
αv
100 101 102 103 104
α4β1 (VLA4)
100 101 102 103 104
αVβ3
100 101 102 103 104
α5β1
100 101 102 103 104100 101 102 103 104
α1 α4
100 101 102 103 104
β3
Fluorescence intensity
Cell
nu
mb
er
20
00
20
00
20
00
20
00
20
00
20
00
20
00
20
00
100 101 102 103 104
β1
100 101 102 103 104
β1
100 101 102 103 104
αv
100 101 102 103 104
αv
100 101 102 103 104
α4β1 (VLA4)
100 101 102 103 104
α4β1 (VLA4)
100 101 102 103 104
αVβ3
100 101 102 103 104
αVβ3
100 101 102 103 104
α5β1
100 101 102 103 104
α5β1
100 101 102 103 104100 101 102 103 104
α1 α4
100 101 102 103 104100 101 102 103 104
α1 α4
100 101 102 103 104
β3
100 101 102 103 104
β3
Fluorescence intensity
Cell
nu
mb
er
20
00
20
00
20
00
20
00
20
00
20
00
20
00
20
00
Figure 35. Expression of different integrins in A549 cells after incubation with C5a. Cells were incubated 48 hours with or without 10 ng/ml C5a in basal medium. Expression of integrins in treated (dotted line) or untreated cells (black line) was assessed by flow cytometry. An isotype control antibody was used as a negative control (shaded peak).
5.5.2 Regulation of metalloprotease activity
The lower attachment of C5a treated cells to the extracellular matrix proteins may
be explained by the production of proteases that degrade those proteins. Therefore, we
analyze the metalloproteinase activity within the conditioned medium of cells treated
with 10 ng/ml C5a for 48 hours. We measured the kinetics of hydrolysis of two
synthetic fluorogenic peptides which contain specific amino acid sequences for different
metalloproteases. In particular, one of the peptides, named here as Substrate 1, is a
substrate for MMP-3 (stromelysin 1) and MMP-10 (stromelysin 2). The other
fluorogenic peptide, named here as Substrate 2, is recognized by MMP-1 (collagenase
1), MMP-2 (gelatinase A), MMP-7 (matrilysin), MMP-8 (collagenase 2), MMP-9
(gelatinase B), MMP-12 (macrophage elastase), MMP-13 (collagenase 3), MMP-14
(MT1-MMP), MMP-15 (MT2-MMP) and MMP-16 (MT3-MMP).
We observed that the conditioned medium from C5a-treated A549 cells presented
a higher protease activity against Substrate 1 than the conditioned medium from control
cells or from cells treated with C5a and the C5aR-1 antagonist (Figure 36A). The
increase in the slope in the linear phase of the fluorescence curve was used to quantify
the difference in the protease activity. The medium from A549 cells treated with C5a
showed 3.43 ± 1.5 times more activity than the control medium. In addition, in the case
of cells treated with C5a and the C5aR-1 antagonist, the ratio between slopes was 0.66 ±
Results
105
0.64, confirming that the increase in protease activity could be blocked with the
antagonist.
When Substrate 2 was used, the conditioned medium of C5a-treated cells showed
also a higher protease activity than the control medium, with a ratio of slopes of 2.59 ±
1.46 (Figure 36B). The medium from cells treated with C5a and the C5aR-1 antagonist
showed a similar protease activity than the control one. Of note, the linear phase of the
fluorescence curve was clearly delayed compared to that obtained with Substrate 1.
In the case of H157 cells, the conditioned medium from C5a treated-cells showed
more protease activity for Substrate 1 than the control medium. However, no differences
were found in the activity against Substrate 2 (data not shown).
Time (seconds)
Flu
ore
scen
ce in
ten
sit
y
BA
60
100
140
80
120
160
1000 2000 3000 4000 5000 6000 7000
C-
Ant.C5aR-1
170
250
330
410
490
210
290
370
450
1000 2000 3000 4000 5000 6000 7000
Ant.C5aR-1
C-
C5a C5a
Time (seconds)
Flu
ore
scen
ce in
ten
sit
y
BA
60
100
140
80
120
160
1000 2000 3000 4000 5000 6000 7000
C-
Ant.C5aR-1
60
100
140
80
120
160
1000 2000 3000 4000 5000 6000 7000
C-
Ant.C5aR-1
170
250
330
410
490
210
290
370
450
1000 2000 3000 4000 5000 6000 7000
Ant.C5aR-1
C-
C5a
170
250
330
410
490
210
290
370
450
1000 2000 3000 4000 5000 6000 7000
Ant.C5aR-1
C-
C5a C5a
Figure 36. Kinetic analysis of metalloproteinase-mediated hydrolysis of fluorogenic peptide substrates. Kinetic data for cleavage of Substrate 1, a fluorogenic peptide recognized by MMP-3 and -10 (A), or Substrate 2, recognized by MMP-1,-2,-7,-8,-9-12,-13,-14,-15 and -16 (B). Conditioned media from A549 cells incubated with basal medium (C-), 10 ng/ml C5a or C5a with 100 ng/ml C5aR-1 antagonist (Ant.C5aR-1) were used to perform the experiments. The reactions were allowed to proceed for 120 minutes, and fluorescence data were evaluated at 1-minute intervals. Fluorescence was quantified as arbitrary units, and progression curves were plotted for each reaction as a function of time. Data show representative experiments for each substrate.
DISCUSSION
Discussion
109
1. THE DUAL ROLE OF COMPLEMENT ACTIVATION
In this work, we have demonstrated that lung cancer cells are able to activate
complement spontaneously, leading to the deposition of complement proteins on the
cell membrane and to the release of complement components, such as C5a, to the
extracellular medium. The release of this anaphylatoxin is a crucial event that
contributes to tumor progression by the induction of inmunosuppression, angiogenesis
and tumor cell migration.
In the cancer setting, researchers have traditionally been focused on the
conventional role of complement for tagging, clearance and lysis of tumor cells; and,
consequently, have been working in strategies such as the development of antibody
mediated responses through complement dependent cytotoxicity (CDC)165. But our
results in the present work, in line with other recent findings, challenge the conventional
view of complement. It seems that elements of the complement system perversely assist
tumor growth; in particular, C5a anaphylatoxin, generated within the tumor tissue
through the activation of complement, seems to have a major role. During the course of
the present study, some other works have been published supporting the role of C5a on
tumor progression. C5a activity has been shown to be connected to the recruitment and
activation of myeloid-derived suppressor cells (MDSCs)107. MDSCs shield cancer cells
from T-cell-mediated immune attack166. C5a released on the tumor environment may
play a key role as a chemoattractant for a subpopulation of MDSCs morphologically
related to neutrophils (PMN-MDSCs) and as an activator for the production of reactive
oxygen species (ROS) and reactive nitrogen species (RNS) in the monocyte-like
subpopulation (MO-MDSCs). This effect on MDSCs is critical for their suppressive
capabilities on T-cell function167. In addition, inhibition of the C5aR-1 was found to be
as efficient in reducing tumor growth as the treatment with the well-recognized anti-
cancer chemotherapeutic paclitaxel107. Nevertheless, MDSCs may not be the only type
of cells involved in the immunosuppressive response to tumors. The present work
suggests that complement activation and C5a may influence the presence of different
populations of immune cells within the tumors, tilting the immune response to
immunosuppression.
Although the finding that complement elements can act as tumor promoters is
novel, the idea is entirely consistent with the growing recognition of the homeostatic
Discussion
110
function of the complement system. The uncovering of this new function of
complement connects with two prominent aspects of cancer research: first, the role of
the tumor microenvironment in the initiation, development and progression of
malignant lesions; second, the impact of the inflammatory response in cancer. The
debate about inflammation and cancer came up more than a century ago, with the
Virchow’s and Coley’s theories. While the former supported that chronic inflammation
can be a factor that increases the risk of cancer, the latter defended that acute
inflammation can result in the regression of malignant tumors. Both theories have been
sustained by a huge number of clinical examples. There is a growing body of evidence
that many malignancies are initiated by infections and the subsequent continuous
inflammatory reaction168. In the contrary, the use of Coley’s toxins was the only known
systemic treatment for cancer until 1940s, and represents a pioneering method to exploit
the host response against tumors169. Although both theories would seem contradictory,
the recent progress in tumor immunology put both of them in the same context. While
Virchow defined a long lasting inflammation, that has been demonstrated to encourage
the genesis and spread of tumors, Coley observed a brisk inflammation that could fight
cancer with some grade of success. The paradigm of how complement fits in this puzzle
is now easily understood considering the kind of inflammatory response that is
involved. As a component of the innate immunity, the activation of complement
generates some of the most powerful proinflammatory molecules in the body, including
most notably the anaphylatoxins C3a and C5a. Therefore, the tumor-promoting
properties of the complement system are consistent with the currently accepted view of
inflammation as exacerbating the growth of various cancers and the recognition of C5a
as a key regulator of the inflammatory response160. Conversely, an anticancer function
for complement is well recognized, for example in its contribution to the clinical
efficacy of monoclonal antibodies for the treatment of neoplasias170-171.
Discussion
111
2. TUMOR CELLS ACTIVATE AND EFFICIENTLY REGULATE
THE COMPLEMENT CASCADE
2.1. NSCLC cell lines spontaneously activate the complement system
Tumorigenesis is a multistep process in which normal cells acquire progressively
genetic and epigenetic alterations. This progression is analogous to Darwinian
evolution, in which the succession of genetic changes, each conferring one or another
type of essential alterations in cell physiology, leads to the progressive conversion of
normal human cells into cancer cells172. During the evolution of a cancer, the sequential
alterations produce changes in cell morphology and result in the expression of many
neoantigens that can be represented by a change in protein glycosylation or in the
expression and structure of membrane-bound proteins. These new elements that
distinguish cancer cells from their normal counterparts may well be recognized by the
immune system109. The complement system, as part of the innate immunity, can
recognize those changes and be activated in the suitable surfaces that bear the tumor-
associated antigens. According to this theory, clinical studies have suggested that the
complement system is activated in patients with cancer in response to the expression of
tumor-associated antigens173. Deposited complement proteins have been observed on
the surface of many human tumors, and elevations in complement levels or activity in
the sera of patients with neoplastic diseases have been described. Therefore, these
observations indicate that complement is activated in the tumor tissue or in its vicinity.
In the present work we have performed for the first time a comparative study of
the in vitro activation of the complement system in tumor and normal cells. In
particular, we have compared non-small cell lung cancer (NSCLC) cells and normal
bronchial epithelial cells. The spontaneous activation of complement was higher on
tumor cells than on normal cells. This result is in agreement with the immune
surveillance theory, which supports that the immune system surveys the body for tumor-
associated antigens for the elimination of tumors119. This motivating result opens more
questions than answers; for example, which are those tumor antigens that trigger the
spontaneous activation of the complement system? In the case of the alternative
pathway, it has been shown that C3b attaches preferably sugars174 and amino acid
hydroxyl groups (Thr, Ser and Tyr)175-176. Then, the activation is more efficient on some
surfaces depending on their polysaccharide and protein composition. Another
Discussion
112
mechanism that triggers the alternative pathway, uncovered in the recent years, is the
direct ability of properdin to initiate complement activation in specific surfaces
associated with infection or injury, acting as a pattern recognition molecule177-178. This
suggests that properdin may become localized on the altered surface of tumor cells.
There is evidence for the binding of properdin to mouse tumors previously treated with
photodynamic therapy (PDT). In those mice, the injection of properdin increased the
cure rates of PDT-treated tumors, suggesting that this protein can localize on the cell
surface and initiate autologous complement activation through the alternative
pathway179. However, till the date, there is not any study that demonstrates a selective
deposit of properdin on untreated tumor tissues.
Regarding our results, another possibility is the activation of the classical or the
lectin pathways by molecules on the surface of tumor cells which are able to bind and
activate C1q or MBL. To elucidate the possible activation of these two pathways, new
experiments using deficient serum in different complement factors, C1q, C2, C4, MBL
or factor B, should be performed. Nevertheless, the potential role of the classical or the
lectin pathways would not reject the importance of the alternative pathway, since this
pathway acts as a loop after the initial deposition of C3b, amplifying the activation of
the other two pathways. It has recently been shown that the alternative pathway might
account for up to 80–90% of total complement activation, even when this pathway is
not the one initially triggered180.
2.2. NSCLC cells regulate complement dependent cytotoxicity
In our experiments, the efficient and spontaneous activation of complement on
NSCLC cells did not lead to a noticeable effect in cell lysis. Although at first glance this
result appeared unexpected, it is well understood considering that the efficiency of
complement-mediated tumor cytotoxicity is hampered by various protective
mechanisms120. Most of the mechanisms to resist complement attack are also exploited
by normal cells to avoid accidental activation or bystander effects from a local
activation of complement. However, it is reasonable to think that cancer cells develop
additional mechanisms to avoid the harm caused by complement activation. In vitro
studies have shown that lung cancer cells are extremely resistant to CDC, and this
resistance is much higher than that observed in normal cells, such as human nasal
epithelium primary cell cultures134,144. One of the best understood protective mechanism
used by cancer cells is the over-expression of membrane-associated and/or soluble
Discussion
113
complement regulatory proteins. Many current hypotheses propose that expression of
these proteins on the neoplastic cell membrane protects tumors from complement
activation.
Several studies have analyzed the expression of complement regulators in primary
tumors and cancer cell lines. Membrane-bound complement regulatory proteins
(mCRPs) appear to be overexpressed on many tumor cell types. Most cancers express at
least two mCRPs, with the exception of CR1. However, it is surprising the huge
variation in mCRP expression that can be observed among tumor types and even among
different specimens of the same tumor type121,181. The evident consequence for a high
expression of one or more mCRPs by a tumor cell would be a higher resistance to
complement cytotoxicity. The expression of CD59 correlated with the resistance of
different types of malignant cell lines133,182-184. Regarding the role of CD55 and CD46 in
complement activation, an increased expression of those molecules has been reported in
many tumors185-191. The secretion of soluble forms of mCRPs can also regulate
complement activation in cancer patients112-113,120,184,192-196, although this mechanism is
not restricted to malignant diseases and can be also found under normal conditions65,197.
Based on our results, we can propose that NSCLC cells present efficient
mechanisms to resist complement activation due to two fundamental signs. First, the
insignificant cytotoxicity produced by the activation of the alternative pathway, even
when we observed a higher C3 deposition on tumor cells than on normal cells. Second,
the low increase in C3 deposition by the classical pathway compared to the alternative
one. These facts imply indirectly that tumor cells are equipped with potent mechanisms
to challenge strong complement activation on their surface. However, in the present
study no clear differences were detected among lung tumor and normal cells in terms of
the expression of inhibitors. It is possible that there is not a predominant mechanism for
the control of complement activation by lung cancer cells. In fact, although there are a
high number of studies suggesting that tumor cells increase the expression of inhibitors,
some of them also show no differences or a reduction in the expression of particular
complement inhibitory proteins. The loss of one complement inhibitory protein may be
compensated by an increased expression of one of the others194. Although not uniformly
observed in all cell lines, in some lung cancer cell lines, we detected high expression
levels of the soluble regulator complement factor H. The contribution of soluble
complement regulatory proteins in the control of complement activation on cancer cells
had been underestimated until some works highlighted that these factors have a key role
Discussion
114
in the resistance of tumor cells to complement activation and CDC. This is the case of
factor H and its splicing product factor H-like protein 1 (FHL-1), whose expressions
have been described in primary tumors and cell lines from different origins198-202. In
fact, a clinically approved immunoassay for the detection of bladder cancer in urine is
based in the quantification of factor H203. We have found that factor H is elevated in
bronchoalveolar lavage and sputum from lung cancer patients. This finding is doubly
interesting. On the one hand, reasserts factor H as an active inhibitory mechanism in
complement activation in lung cancer. On the other hand, it introduces factor H as a
molecular biomarkers that may enhance the sensitivity of cytologic examination in the
detection of lung cancer146. Moreover, complement factor H antibodies in serum have
recently been found in lung cancer patients and its use as a molecular marker for early-
stage disease has been proposed204.
The expression of factor H and FHL-1, and their subsequent binding to the cell
membrane, may have consequences in the resistance of cancer cells against
complement. H2 glioblastoma cells showed an exceptional resistance to CDC even after
the combined neutralization of CD46, CD55, and CD59, but the use of anti-factor H
monoclonal antibodies enhanced cell death, confirming that factor H (or FHL-1) was
involved in the complement resistance of this cancer cell line205. Similarly, an anti-
factor H antibody also enhanced complement-mediated killing of Raji cells, a cell line
obtained from a Burkitt's lymphoma206. In lung cancer cells, the neutralization of CD46
and CD59 using specific antibodies was not effective in increasing the susceptibility to
CDC134, whereas the same antibodies successfully facilitated CDC on normal
respiratory epithelial cells144. These findings suggested that lung cancer cells have other
mechanisms to resist complement activation in addition to CD46 and CD59. In this
sense, our group demonstrated that NSCLC commonly expressed and secreted factor H
and FHL-1 to the extracellular milieu, where they were able to bind to lung tumor cell
surfaces and inhibit efficiently the activation of complement122. In addition, factor H
expression was knockdown in A549 lung cell line, reducing its growth significantly when
injected into athymic mice123.
The inhibitor C4b-binding protein (C4BP), the functional analogue of factor H in the
classical pathway, may be also playing a key role in complement regulation on cancer cells.
In a collaborative study with Professor Anna Blom our laboratory reported that NSCLC
cells produce the soluble complement inhibitors factor I and C4BP, and these cells are able
Discussion
115
to bind C4BP to their surface and support the cleavage of C3b and C4b124. This issue may
enable NSCLC cells to turn off the complement cascade.
In a different scenario, using hypoxic conditions to mimic in vitro the
circumstances of growing tumors, we observed that H358 and H2087 cell lines, two
NSCLC cell lines that express factor H, factor I and mCRPs, presented substantial C3
and C9 deposition after 24 hours of hypoxia and 24 hours of reoxygenation, and even
more after 48 hours of hypoxia. Nevertheless, the considerable complement activation
on those cells was not sufficient to cause CDC. The existence of hypoxic areas is a
common feature that occurs within tumors. Its appearance can be explained by the rapid
consumption of oxygen due to the sudden expanding of tumor cells and the bad quality
of the new vessels generated. Tumor vessels are often blind ended, have incomplete
endothelial lining and basement membrane and have a tendency to collapse207.
Furthermore, sudden restoration of oxygen (reoxygenation) worsens hypoxic injury
through increased ROS production208. The activation of complement on cells
undergoing hypoxic stress or hypoxia/reoxygenation has been widely observed in
several pathologies 209, although the molecules that trigger the different pathways are
unknown. One possibility is the generation of molecular changes in proteins and lipid
membranes after reacting with ROS208, or the production of new carbohydrate patterns
by the activation of certain genes that control sugar metabolism210. Those antigens may
potentially activate the three complement pathways211. This is in agreement with our
results, because NSCLC cell lines exposed to hypoxia or hypoxia/reoxygenation
showed an increase in C3 and C9 deposition when the experimental conditions
promoted exclusively the activation of the alternative pathway (Veronal buffer with
Mg2+ and EGTA) or the classical and lectin pathways (factor B-depleted serum in buffer
with Mg2+ and Ca2+). The binding of immunoglobulins, present in serum, C1q or MBL
to the surface of hypoxic cells sustained that classical and lectin pathways were also
activated. By contrast, those tumor cells, that efficiently activated complement,
appeared to be really resistant to CDC, as no increase in cell lysis was observed. This
again suggests that inhibitory mechanisms stopped complement cascade to elicit its
dreadful effects. Another remarkable finding of this study was that NSCLC cell lines
downregulated the expression of factor H, factor I, CD46 and, in a lower extent, CD55,
both in hypoxia/reoxygenation and even more in hypoxia. This implies that under
hypoxic conditions NSCLC cells reduce their mechanisms to face complement attack
while keep free from CDC.
Discussion
116
3. EFFECT OF COMPLEMENT ACTIVATION IN THE TUMOR
MICROENVIRONMENT
3.1. Complement deficiency impairs tumor growth
In the last years, complement activation has been exploited in the monoclonal
antibody-based immunotherapy against tumors, due to the improvement in efficacy
when different cytotoxic mechanisms are combined. There are several examples of
chimerized and humanized mouse monoclonal antibodies containing the human IgG1
Fc-region that binds the heterooligomeric complex C1q and activates the classical
complement pathway170. However, little attention has been paid to the promoting effect
of complement in malignant tumors. Using a syngeneic mouse model of lung cancer, we
have demonstrated that deficiency of complement impairs normal tumor growth. In
particular, we have shown that the deficiency in C3 and the blockade of C5aR-1 are
associated with tumor growth delay. This result indicates that complement is involved
in mechanisms that benefit the development of malignant tumors. Similar results were
obtained in a syngeneic mouse model of cervical cancer107, whereas in a transgenic
mouse model of epidermal neoplasia the authors showed that mechanisms related to
inflammation and tumor progression were independent of C3 deficiency212. These
contradictory findings may be due to the different situations where the complement
contribution to malignancy was studied. We and Markiewski’s group studied
complement effects in cells from advanced invasive tumors, whereas in the transgenic
model of epidermia neoplasia only premalignant skin lesions were considered.
Complement activation generates pro-inflammatory polypeptides named
anaphylatoxins, C3a, C4a and C5a. Within the group, C5a and C3a are both well known
for their immunoregulatory functions, whilst C4a function as an anaphylatoxin has been
recently questioned because of the lack of its own receptor and specific biological
functions100. The inflammatory potential of anaphylatoxins C3a and C5a must link
complement activation with the progression of tumors, in the same way it has been
implicated in the pathogenesis of many inflammatory and immunological diseases such
as sepsis, acute respiratory distress syndrome, rheumatoid arthritis, glomerulonephritis,
multiple sclerosis, ischemia-reperfusion injury, and asthma160. The harmful roles of
complement may be mainly due to the release of these activated products. Although
anaphylatoxins are not necessarily the initiating factors in the inflammatory disorders,
Discussion
117
they appear to be responsible for promoting and perpetuating inflammatory reactions.
The cleavage of C3 precedes the generation of complement inflammatory mediators,
although a local C5a generation in the absence of C3 in some pathological conditions
has been demonstrated34,36. In C3 deficient mice, it is difficult to ascertain whether the
pro-tumorigenic effects of complement activation are due to C3 or its downstream
products. However, the pharmacological blockade of C5aR-1 uncovered the role of C5a
in tumor progression. In any case, the effect of tumor regression was superior using C3
deficient mice than C5aR-1 antagonist. There may be two explanations for this fact:
first, the use of two different ways to achieve complement disruption, genetic deficiency
versus pharmacological blockade; and second, the assumption that C3a might also have
a tumor-promoting effect. Regarding to the second explanation, both C3a and C5a have
been shown to exert key regulatory functions in and out the immune system, through
their binding to three cognate receptors, i.e. C3aR-1, C5aR-1 and the C5a receptor-like
2, C5L2. However, C5aR-1 signaling has been the most characterized. The study of
C5aR-1 in a wide variety of cell types and the pleitropic biologic effects originated from
C5a binding (summarized in Table 7) made this anaphylatoxin an attractive starting
point for the study of complement effects in the tumor microenvironment. Besides, we
observed that C5a levels were increased in plasma and bronchoalveolar lavage
specimens from patients with lung cancer; suggesting that C5a is generated extensively
within tumors and diffuses systemically. In the tumor microenvironment, C5a may play
a key role in promoting tumor-associated effects. But it is important to keep in mind
that the role of C5a does not exclude the effects of C3a in tumor progression, an aspect
that warrants future studies.
Table 7. C5aR-1 functions on myeloid and non-myeloid cells.
Loci Cell type Action of C5a Reference
Myeloid cells Neutrophils Expression of adhesion molecules Superoxide release Release of granule enzymes Release of cytokines Reduction of apoptosis
Guo et al.160
Eosinophils Chemotaxis Release of granule enzymes
Guo et al.160
Basolphils Release of granule enzymes Kurimoto et al.213 Monocytes Release of cytokines
Chemotaxis Guo et al.160
Macrophages Enhancement of phagocytosis Release of cytokines
Guo et al.160
Mast cells Chemotaxis Release of histamine Change to prothombotic phenotype
Guo et al.160
Dendritic cells Activation and allostimulation of T cells Peng et al.214
Lymphocytes T cells Co-stimulation and suppression of apoptosis Strainic et al.92
Tonsil B cells Chemotaxis Ottonello et al.215
Thymocytes Apoptosis Riedemann216
Circulatory system HUVEC Induction of tissue factor activity Regulation of cytokine expression Upregulation of adhesion molecules in response to inflammation
Ikeda et al.217 Monsinjon et al.94 Foreman et al.218
Human immortalized dermal microvascular endothelial cells
Chemotaxis Schraufstatter et al.219
Septic cardiomiocites Reduction of contractibility Niederbichler220
Cerebral microvascular endothelial cells
Alteration of iNOS, ROS, and actin reorganization Jacob et al.221
Mouse microvascular endothelial cells
C5a synergistically enhanced chemokine release from MDMEC following exposure of endothelial cells to IL-6, IFN-γ, or LPS
Laudes et al.222
Loci Cell type Action of C5a Reference
CNS Neurons Possible protection during local inflammatory processes Apoptosis; neurodegeneration
O'Barr et al.223 Farkas et al.224
Microglia, reactive astrocytes Possible role in activation and recruitment of astrocites and microglia (gliosis) Gasque et al.225
Cultured neural stem cells Possible function in mammalian cerebrogenesis and basal adult neurogenesis Rahpeymai et al.226
Oligodendrocytes
Possible role in oligodendrocytes functions: migration, proliferation, regeneration, elimination of oligodendrocytes
Nataf et al.227
Connective tissue Synoviocytes Production of TNF and possible role in joint inflammation process Yuan et al.228
Synovial mast cells
Release of histamine in mast cells from rheumatoid arthritis Production of the proinflammatory cytokine IL-17A, which is considered a crucial player in rheumatoid arthritis
Kiener et al.229 Hueber et al.230
Human articular chondrocytes Possible effect of the cartilage pathology in rheumatoid arthritis Onuma et al.231
Eye Retinal pigment epithelial cells
Increase in the secretion of VEGF from ARPE-19 cells (role in AMD) Expression of VEGF, leukocyte recruitment and CNV formation in a mouse model of laser injury Induction of IL-8 mRNA (role in inflammation)
Cortright et al.232 Nozaki et al.233 Fukuoka et al.234
Kidney Cultured human renal glomerular mesangial cells
Promotion cellular proliferation and production of cytokines (MCP-1) and growth factors (PDGF-AB)
Braun et al.235
Human renal proximal tubular cells Unknown function in normal kidney Fayyazi et al.236
Mouse tubular epithelial cells Overexpression of both PDGF-B and -D and TGF-beta-1, acting as a profibrotic factor
Boor et al.237
Liver Human hepatoma-derived cell line HepG2
Production of alpha 1-antitrypsin Regulation of the expression of acute phase proteins
Buchner et al.238 McCoy et al.239
Stimulated hepatocytes
Enhancement of glucose output Regeneration of rat hepatocytes
Koleva et al.240 Daveau et al.241
Hepatic stellate and Kupfer cells Upregulation of extracellular matrix proteins Schlaf et al.242
Lung Human and mouse bronchial epithelial and smooth muscle cells, alveolar epithelial cells, and pulmonary smooth muscle endothelial cells
Participation in the pathology of diseases such as sepsis and asthma
Drouin et al.243
Rat alveolar epithelial cells Release of inflammatory mediators Riedemann et al.244
Skin Inflamed keratinocytes
Possible role in the pathogenesis of skin disorders beyond its well-defined function as a chemoattractant and activator of leukocytes
Fayyazi et al.245
92,94,160,213-245
Discussion
120
3.2. Complement proteins and immunosuppression
In this work we have also demonstrated that both C3 deficiency and the
pharmacological blockade of C5aR-1 quantitatively alters the populations of
splenocytes in the syngeneic model of lung cancer. CD19+ and CD8+ populations were
increased in C3 deficient mice while myeloid-derived suppressor cells (MDSC) and
regulatory T cells (Tregs) were decreased in those animals as compared to wild-type
mice. The same tendency was observed after blocking C5aR-1, although significant
differences were only seen in total MDSC, and the two characterized subpopulations of
these cells. MDSCs are a heterogeneous population of myeloid progenitor cells and
immature myeloid cells. Their immunosuppressive function consists in reducing
activated T-cells and disrupting their effects by multiple mechanisms, such as depletion
of L-arginine by arginase-1 (ARG-1) and production of nitric oxide by inducible nitric
oxide synthase (iNOS), which leads to an increased production of reactive oxygen
species (ROS), including peroxynitrites (ONOO) and hydrogen peroxide (H2O2).
MDSCs lack the expression of cell-surface markers that are specifically expressed by
monocytes, macrophages or dendritic cells. In mice, MDSCs are characterized by the
co-expression of the myeloid-cell lineage differentiation antigen GR1 and CD11b. The
differential expression of two epitopes in GR1, Ly6G and Ly6C (encoded by different
genes), has revealed a morphological heterogeneity within MDSCs. The use of two
epitope-specific antibodies has led to the identification of two MDSC subsets:
granulocytic MDSCs have a CD11b+Ly6G+Ly6Clow phenotype, whereas monocytic
MDSCs are CD11b+Ly6G-Ly6Chigh. In summary, the Ly6G molecule is expressed
primarily on granulocytes161, whereas Ly6C is typically highly expressed on
monocytes162. In our experiments, we distinguished the two subpopulations using the
antibodies anti CD11b and anti-Ly6C, and assumed that within the CD11b+ cells those
with low and high expression of Ly6C corresponded to granulocytic and monocytic
subpopulations of MDSCs, respectively. The recognition of these two subsets is
important as each may have different functions in cancer and infectious and auto
immune diseases246. Both subsets of MDSC have different mechanisms of action to
suppress antigen-specific T-cell proliferation. In brief, granulocytic MDSCs express
high levels of ROS, and monocytic MDSCs express high levels of NO; both subsets
express ARG-1247. However it is not clear if both subsets differ in the suppression of T-
Discussion
121
cells. Apart from this, MDSCs have been implicated in a whole array of non-
immunological functions, such as the promotion of angiogenesis and metastasis248-249.
Besides cancer, an increased production of MDSCs has also been reported in a
number of pathological conditions including traumatic stress250, bacterial and parasitic
infections251, sepsis252 and transplantation253. In cancer, this population of cells
increases up to ten fold in the spleen of patients with different tumors; and in mice,
expansion of MDSC has been detected in practically all tumor models. However, the
extent of expansion varies between different tumor models (from 2 to 20 times), which
can be explained by the fact that MDSCs are produced in response to a variety of tumor-
derived factors247. We observed that the percentage of MDSCs increases approximately
three fold in the spleen of animals bearing tumors with normal complement activity,
which is in agreement with other studies that used the same syngeneic model of lung
cancer247. Then, C3 deficiency or C5aR-1 blockade disrupted this expansion, halving
the percentage of MDSCs. MDSCs, together with regulatory T cells (Tregs), also
decreased in the spleen of C3 deficient mice, promoting an immunosuppressive
environment in tumor-bearing hosts. Classically Tregs have been characterized by the
cellular markers CD4+CD25+FoxP3+, although there are other phenotypically distinct
regulatory T-cell populations that have been suggested, including CD8+ subsets254.
CD25, the IL-2 receptor, is essential for Tregs survival. Foxp3, the transcription factor
forkhead box P3, controls the genes encoding proteins capable of mediating Treg
suppressive function255. Tregs suppress both T cells and antigen presenting cells (APC)
by several proposed mechanisms that include production of IL-10, which can drastically
suppress APC and T cell function, production of TGF-β, which may suppress natural
killer cell function, competitive consumption of IL-2, which is a survival factor for
activated T cells, perforin- and granzyme-dependent killing of T cells and APCs,
CTLA-4 induction of indolamine 2,3-dioxygenase (IDO)-expressing APCs, which
suppress T cell activation and promote tolerance, and also, induction of PDL-1
expression on APCs, which renders them immunosuppressive256.
Foxp3+CD25+CD4+ Treg cells are produced in the thymus as a functionally
mature T-cell subpopulation, and constitute 5-10% of peripheral CD4+ cells in healthy
mice and humans257. But they accumulate in tumors due to the recruitment through the
chemokine ligand 22 (CCL22) produced by tumor cells and macrophages258, and self-
antigens that Tregs recognize on tumors259. Differentiation or expansion within the
tumor stroma and conversion from normal T cells have also be proposed260.
Discussion
122
Interestingly, we observed that the ratio between Tregs and total CD4+ was significantly
reduced in C3 deficient mice. Finally, CD8+ and CD19+ populations were increased
quantitatively in spleens of C3 deficient animals. This may indicate that complement
activation disrupts a cytotoxic response raised against tumors, as well as an immune
response mediated by B cells.
We also analyzed the immunosuppressive microenvironment within tumors.
Unfortunately, due to the wide heterogeneity of cell populations that made difficult to
select the desired immune cells, we were not able to draw clear conclusions from these
data. In order to check whether the analysis of splenocytes represented the tumor
microenvironment, the expression of different molecules was analyzed. Interestingly,
the expression of immunosuppressive molecules was significantly decreased in C3
deficient mice and mice treated with the C5aR-1 antagonist. Those molecules included:
IL-10, PDL-1, CTLA-4, LAG-3, ARG-1 and IL12p35. IL-10 is an important
immunomodulatory cytokine well known because of its anti-inflammatory properties261.
PDL-1 (programmed death ligand 1, also known as B7-H1) is constitutively expressed
in cells of the macrophage lineage and regulates cellular and humoral immune responses
through the PD-1 receptor on activated T and B cells. An aberrant PDL-1 expression,
described in a number of human malignancies, impairs both T cell function and
survival262. CTLA-4 (cytotoxic T-lymphocyte antigen 4, also known as CD152) inhibits
T-cell activation when binds to CD80 and CD86 on APCs263. Based on the results of a
phase III clinical trial, ipilimumab, a human monoclonal antibody against CTLA-4, will
likely be approved soon by the FDA for metastatic melanoma. According to a research
presented in December 2010 at the Chicago Multidisciplinary Symposium in Thoracic
Oncology, ipilimumab also showed benefit in advanced non-small cell lung cancer.
LAG-3 (lymphocyte activation gene 3, also known as CD223) is a CD4-related
activation-induced cell surface molecule that binds to MHC class II molecules with high
affinity and negatively regulates T cell expansion and homeostasis264. CTLA-4 and
LAG-3 expression by Tregs are involved in their immune-suppressive function265-266.
ARG-1, together with iNOS, takes part in the catabolism of L-arginine, a non-essential
amino acid necessary for T-cell proliferation. These two enzymes are expressed in
MDSCs and their action depletes L-arginine from the microenvironment, inhibiting T-
cell proliferation267. IL-12p35 is one of the subunits of the heterodimeric cytokine IL-12
and the more recently described IL-35; while IL-12 stimulates the growth and function
of T cells, IL-35 may be specifically produced by Tregs and is required for their
Discussion
123
suppressive activity268. The expression of IL-6 was also decreased in animals treated
with the C5aR-1 antagonist. IL-6 has been linked to the promotion of some types of
cancers due to its proinflammatory nature269. Regarding to the complement activation,
deficiency in C3a, C5a or blockade of C5aR-1 impairs the induction of IL-6 in
hepatocytes, which is necessary to prevent cell death270. Similar mechanisms may occur
in the tumor setting. Finally, CD70, a member of the TNF receptor superfamily,
decreased in tumors from C3-deficient mice. Upon binding to its ligand, CD27 sends
costimulatory signals for the expansion and maturation of antigen-specific T effector
cells, thus governing the magnitude and maintenance of the cytotoxic response. Its
normal expression is highly restricted, but is aberrantly expressed in hematologic
malignancies and solid tumors271.
Globally, considering the different populations of immune cells in spleens and the
cytokines within tumors in C3 deficient mice and mice treated with the C5aR-1
antagonist peptide, we can conclude that normal complement activation contributes to
an immunosuppressive microenvironment necessary for tumor growth.
3.3. Complement and angiogenesis
Angiogenesis is another key mechanism that contributes to tumor growth. At
present, the role of complement in angiogenesis is a matter of debate. There are studies
that support the pro-angiogenic potential of complement activation, whereas there are
others that sustain the anti-angiogenic properties of complement. Recently, the anti-
angiogenic effect of C3 and C5 was shown in a model of retinopathy of prematurity,
where C5a indirectly produced an anti-angiogenic effect by stimulating macrophages
toward an angiogenesis-inhibitory phenotype272. On the contrary, the role of
complement in the activation of angiogenesis has been demonstrated in age-related
macular degeneration (AMD), a disease caused by choroidal neovascularization (CNV).
Both C3a and C5a are present in the lipoproteinaceous deposits that appear between the
coroid and retinal pigmented epithelium (RPE), also called drusen, in patients with
AMD or animals with laser-induced CNV. Both anaphylatoxins seem to be involved in
the induction of VEGF expression in vitro and in vivo in RPE cells, thus promoting the
generation of new vessels233. In addition, MAC and C3 are deposited in the eyes of
animals submitted to laser-induced CNV, concomitantly with an increase in the
expression of the angiogenic factors VEGF, TGF-β2 and bFGF273.
Discussion
124
We investigated the extent of angiogenesis in tumor slices from wild-type and C3
deficient mice or from mice treated with either the C5aR antagonist or the control
peptide. The immunohistochemistry of CD31 showed only minimal differences among
groups that did not reach statistical significance. The same finding had been previously
observed in a mouse model of cervical cancer107. Nevertheless, using extracts from
those tumors, we observed that the expression of bFGF, a potent mitogen and
chemotactic factor for endothelial cells274, was significantly reduced in animals treated
with the C5aR-1 antagonist or in C3 deficient mice. This result points to a role of
complement activation in the angiogenesis of tumors, although the question of why
there were not differences in vascular density remains unclear. Two possible
explanations could justify this finding. First, angiogenesis is orchestrated by a balance
between a variety of activators and inhibitors275, so a deeper analysis of the different
anti- and pro-angiogenic factors within the tumors would be needed to clarify their
angiogenic potential. Second, complement activation may be important in the
promotion of angiogenesis in the early steps of tumor formation. To evaluate this
hypothesis we injected syngeneic lung cancer cells into the vein of wild-type and C3
deficient animals, and analyzed the vessels formed in the disseminated nodules three
weeks after injection. Interestingly, C3 deficient mice presented a significantly lower
vessel density within the lung micrometastases than wild-type animals, supporting the
role of complement in angiogenesis. In vitro studies, using human umbilical vein
endothelial cells (HUVECs), support this conclusion. The activation of C5aR-1 in
endothelial cells causes cell retraction, increased paracellular permeability and
eosinophil transmigration276. It has also been demonstrated that C5a induces genes that
promote the effects of VEGF277. With this in mind, we studied the effects of C5a on
HUVECs. Of note, we detected considerable levels of C5aR-1 by flow cytometry in
HUVECs. However, the mRNA levels, detected by real-time PCR, seemed very low
(data not shown). In fact, the expression of C5aR-1 in HUVECs is controversial278. In
any case, it would be incorrect to conclude that the activation of C5aR-1 in HUVECs is
functionally irrelevant based on its low levels of expression, because engagement of a
few hundred G protein-coupled receptors per cell is sufficient for maximal cell
activation279. Bearing this in mind, we investigated the effect of C5a in HUVEC
proliferation, migration and formation of tube-like structures in a gelled matrix. The
formation of tubes is a method widely utilized to assess compounds that either inhibit or
stimulate angiogenesis280. We did not observe any effect in proliferation after C5a
Discussion
125
treatment, although this effect was previously observed by other groups in HUVECs
and immortalized human microvascular endothelial cells281-282. On the other hand, C5a
induced migration of HUVECs and formation of tubes in gelled Matrigel. The
chemotactic property of C5a in HUVECs was also seen in previous studies219. However,
to our knowledge, this is the first work in which the capacity of C5a to induce the
formation of tube-like structures by endothelial cells is demonstrated. Taken together,
these facts strongly suggest that C5a has angiogenic properties.
4. EFFECT OF C5a ON TUMOR CELLS
In the present study we have analyzed for the first time the expression of C5aR-1
in different lung cancer cell lines. The expression of C5aR-1 in human lung epithelial
cells had previously been demonstrated283. Of note, small-cell lung cancer (SCLC) and
carcinoid cells presented higher levels of expression than NSCLC cells. The present
work focused on NSCLC cells, but the high expression in lung cancer cells of
neuroendocrine origin merits further studies.
C5a signaling through C5aR-1, a seven transmembrane G protein-coupled receptor
that signals through Gαi and G16, has been well characterized in both immune and non-
immune cells. ERK1/2, Akt, and JNK are activated in monocytes and neutrophils,
whereas macrophages activate MAPK p38 pathway in response to C5a. In endothelial
cells, C5a/C5aR-1 induces transactivation of the epidermal growth factor receptor,
enhancing their migration284. We observed that C5aR activation in A549 cells produced
ERK phosphorylation and NF-κB translocation to the nucleus, as it was previously
reported for monocytes160. In undifferentiated human neuroblastoma cells, NF-κB
activation after C5aR-1 stimulation produced neoplastic proliferation223. However, we
were unable to appreciate any difference in lung cancer cell proliferation after treatment
with C5a or the three anaphylatoxins combined. Although calcium influx has been
described for both C5a and C3a on several cell types285, preceding NF-κB translocation
to the nucleus286, we did not observe a calcium influx after treating A549 or H157 cells
with high doses of C5a. In H1299 cells, which have high levels of intracellular calcium,
C5a treatment increased calcium influx, suggesting that the basal levels of calcium are a
key factor for this effect.
Cancer cell migration and adhesion to extracellular matrix (ECM) proteins are
processes highly related to invasion and metastasis. The migration and establishment of
Discussion
126
neoplastic colonies, distant from their origin, is the culprit of around 90% of cancer
deaths287. Several types of proteins are involved in the process, but of great importance
are those which reorganize their surrounding microenvironment and permit cell
migration and invasion through the extracellular matrix and into distant tissues. Those
proteins include cell-to-cell adhesion proteins and proteases. C5a is well recognized as a
chemoattractant for leukocytes and endothelial cells, but we present here for the first
time the chemotactic effect of C5a on cancer cells. In addition, the significant reduction
in ECM adhesion after C5a treatment supports this capacity. In our attempt to uncover
the mechanism beyond this process, we investigated the effect of C5a in the expression
of several integrins that mediate the adhesion to different ECM proteins. Integrins are a
large family of heterodimeric transmembrane glycoproteins. They are implicated in cell
proliferation, migration and survival288. In humans, 18 α-subunits and 8 β-subunits can
form a total of 24 integrin heterodimers. For our analysis, we selected those involved
preferentially in the adhesion to collagen I, collagen IV and fibronectin, which were the
ECM proteins for which we observed significant differences. However, no changes in
the expression of these molecules were observed, suggesting that other mechanisms are
responsible for the changes in adhesion.
Alternative mechanisms that would explain the relationship between complement
and degradation of ECM proteins have previously been proposed. The degradation can
be achieved directly by complement proteins289 or by the activation of matrix
metalloproteinases (MMPs). For example, C1s was found to activate the zymogen of
MMP-9, a protease with established roles in the metastasis of cancers290. In general,
MMPs are a family of zinc-dependent endopeptidases categorized by their architectural
features. In humans, twenty three MMPs have been indentified. All of them share a
general structural blueprint with three common domains: the pro-peptide, the catalytic
domain, and the hemopexin-like C-terminal domain291. The interaction of a cysteine
residue of the pro-domain with the zinc ion of the catalytic site keeps MMP in an
enzymatically inactive state. To become proteolytically active, the disruption of this
interaction has to occur by chemical modification of the cysteine residue or proteolytic
removal of the pro-domain, catalyzed extracellularly by convertases such as plasmin or
other MMPs, or intracellularly by furin. Once activated, MMPs realize their functions.
The most known function is the degradation of the ECM proteins, but they are also
involved in complex mechanisms that modulate angiogenesis, signaling pathways and
the inflammatory response292. The function of MMPs in vivo depends on the local
Discussion
127
balance between them and their physiological inhibitors. The most important inhibitors
are the four tissue inhibitors of metalloproteinases (TIMPs)291.
Using specific fluorogenic peptides for MMPs, we observed that the supernatants
of cells treated with C5a held a higher proteolytic activity than those from cells treated
without the anaphylatoxin or with C5a and the antagonist of C5aR-1. The use of two
peptides with different specificity for MMPs allowed us to discern the possible MMPs
involved in the ECM cleavage. The cleavage of the specific peptide for MMP-3 and
MMP-10 (stromelysin 1 and 2, respectively) was more efficient than the cleavage of the
peptide recognized by MMP-1, -2, -7, -8, -9, -13, -14, -15, and -16. In our experiments,
we also used positive controls: recombinant MMP-10 and human serum. As expected,
MMP-10 efficiently cleaved the first peptide. However, MMP-10 also showed some
activity on the second peptide. This indicates that one specific protease can cleave its
substrate efficiently, but may also have some activity over other different substrates. By
extrapolating this fact to our results, we conclude that stromelysins may be the
metalloproteinases involved in the proteolysis of the ECM proteins once cells have been
treated with C5a. Further work is needed to confirm this point and to identify the
specific MMP involved in the process. In preliminar studies, the increase in MMP
activity seemed not to be regulated by gene expression, as no changes in the expression
of specific MMPs at the RNA level were observed (data not shown). Since MMP
function is regulated by the shift from the inactive to the active form and by the
inactivation by MMP inhibitors, the increase in MMP activation or the decrease in
MMP inhibition should be evaluated.
Discussion
128
5. COMPLEMENT IN HEALTH AND DISEASE
The present work offers additional support to the transition from the classical
perception of the human complement system to a wider perspective. That is to say, from
its role as a microbial defense system to a complex system that takes part on cell
homeostasis, immune surveillance and the sustention of disease when the balance
between activation-regulation becomes awry.
The immunosurveillance functions of complement and their biological results
have recently been reviewed by Ricklin et al.15, and are illustrated in Figure 37. The
spontaneous C3 activation (tick-over) on normal cell surfaces is well checked by the
normal presence of mCRPs and the recruitment of soluble inhibitors from their
surroundings; hence the silent controlled complement activation does not upset the
homeostasis in healthy tissues (Figure 37A). Nevertheless, there are different situations
where the complement activation and regulation becomes relevant. In the case of
microbial or foreign cells, complement activation triggered by different means (such as
patron recognition molecules (PRM) or antigen-antibody complexes), in the absence of
complement inhibitors, permits amplification of the cascade. As a result, complement
plays its classical role of eliminating intruders by opsonization, CDC and stimulation of
downstream immune responses triggered by complement degradation products (Figure
37B). In the case of apoptotic cells, the balance between complement activation and
regulation is extremely delicate. Complement is activated by PRMs bore on those
altered cells and, together with the simultaneous shedding of CD46 and CD59,
apoptotic cells become opsonized by C3b and C4b, and phagocyted. The presence of
residual mCRPs on their surface and the binding of soluble inhibitors prevent the
release of inflammatory effectors and the harmful result of complement mediated lysis
(Figure 37C). In the case of some inflammatory, autoimmune, neurodegenerative and
infectious diseases, the balance between complement activation and regulation is
disrupted. Globally, the cause of this imbalance is an inappropriate initiation of the
cascade or deficiencies in specific complement inhibitors. Several diseases have been
related to both situations. Systemic lupus erythematosus, rheumatoid arthritis,
isquemia/reperfusion injury and Alzheimer’s disease are few examples of an excessive
activation. Age-related macula degeneration, atypical hemolytic uremic syndrome,
hereditary angioedema and paroxistical nocturnal hemoglobinuria, are caused by
inadequate complement inhibition293. In such pathological conditions, complement
Discussion
129
sustains inflammation and perpetuates tissue damage. (Figure 37D). In recent years,
several reports have demonstrated that complement dysregulation in those diseases is
associated with mutations or polymorphisms in the components and regulators of the
alternative pathway C3 convertase294.
tick-over
sufficient
regulation
strong
activation
opsonization
amplification
danger
signaling
lysis
activation
sufficient
regulation
attenuated
signalingattenuated
lysis
clearance
activation
insufficient
regulation
lysis
proinflammatory
signaling
A B
C D
tick-over
sufficient
regulation
strong
activation
opsonization
amplification
danger
signaling
lysis
activation
sufficient
regulation
attenuated
signalingattenuated
lysis
clearance
activation
insufficient
regulation
lysis
proinflammatory
signaling
A B
C D
Figure 37. Drawing showing the immune surveillance functions of complement in different settings: healthy cells (A), foreign or microbial cells (B), apoptotic cells (C) and host cells in disease conditions (D). Based on Ricklin et al.15
At this point of the work, based on our results and the most recent literature, we
can propose a fifth biological situation for lung cancer cells, and most probably for
many other cancer types (Figure 38):
Discussion
130
- Cancer cells are recognized by complement due to the presence of tumor-
associated antigens on their surface, causing a substantial activation of the system in the
tumor microenvironment.
- Cancer cells regulate complement activation, preventing their elimination.
- Complement activated proteins, such as C5a, orchestrate several biological
processes that promote immunosuppression, angiogenesis and invasion.
activation
sufficient
regulation
proinflammatory
signaling
deficient
lysis
Invasion
Angiogenesis
Immunosuppression
activation
sufficient
regulation
proinflammatory
signaling
deficient
lysis
Invasion
Angiogenesis
Immunosuppression
Figure 38. Potential oncogenic roles for complement proteins in NSCLC cells.
CONCLUSIONS
Conclusions
133
1. Non-small cell lung cancer (NSCLC) cells activate complement spontaneously.
This activation is significantly higher than in normal bronchial epithelial cells, and may
be responsible for the high levels of C5a anaphylatoxin found in biological fluids from
NSCLC patients. However, acquisition of different inhibitory mechanisms by cancer
cells, such as factor H overexpression, makes them resistant to the harmful effects of
complement.
2. In a syngeneic mouse lung cancer model, complement activation favors tumor
progression. This tumor promoting-effect of complement is mediated at least in part by
C5aR-1.
3. Three major mechanisms may be involved in the tumor promoting-effects
mediated by C5aR-1:
3a. C5a, a potent chemoattractant for different immune cells, increases splenic
suppressor cell populations and immunosuppressive cytokines within the tumor.
3b. C5a promotes migration of endothelial cells and facilities angiogenesis.
3c. C5a reduces adhesion of NSCLC cells to extracellular matrix proteins and
enhances cell migration, possibly by the release of active metalloproteinases.
4. Overall, these results suggest that the complement system can perversely assist
lung cancer growth. This finding may be relevant in the development of complement-
mediated therapies against lung cancer.
REFERENCES
References
137
1 Jemal, A. et al. Cancer statistics, 2009. CA Cancer J Clin 59, 225-249 (2009).
2 Ferlay, J. et al. Estimates of the cancer incidence and mortality in Europe in
2006. Ann Oncol 18, 581-592 (2007).
3 Ginsberg, R. Lung Cancer. (Hamilton: BC Decker, 2002).
4 Zakowski, M. F. Pathology of small cell carcinoma of the lung. Semin Oncol 30,
3-8 (2003).
5 Detterbeck, F. C., Boffa, D. J. & Tanoue, L. T. The new lung cancer staging
system. Chest 136, 260-271 (2009).
6 Henschke, C. I. et al. Survival of patients with stage I lung cancer detected on
CT screening. N Engl J Med 355, 1763-1771 (2006).
7 Schrump, D. S., Altorki, N.K., Henschke, C.L., Carter, D., Turrisi, A.T.,&
Gutierrez, M.E. Cancer: Principles and practice of oncology 7th edn, 753–776
(Lippincott Williams and Wilkins, 2005).
8 Tockman, M. S., Anthonisen, N. R., Wright, E. C. & Donithan, M. G. Airways
obstruction and the risk for lung cancer. Ann Intern Med 106, 512-518 (1987).
9 de Torres, J. P. et al. Assessing the relationship between lung cancer risk and
emphysema detected on low-dose CT of the chest. Chest 132, 1932-1938
(2007).
10 Alberg, A. J., Ford, J. G. & Samet, J. M. Epidemiology of lung cancer: ACCP
evidence-based clinical practice guidelines (2nd edition). Chest 132, 29S-55S
(2007).
References
138
11 Amos, C. I. et al. Genome-wide association scan of tag SNPs identifies a
susceptibility locus for lung cancer at 15q25.1. Nat Genet 40, 616-622 (2008).
12 Thorgeirsson, T. E. et al. A variant associated with nicotine dependence, lung
cancer and peripheral arterial disease. Nature 452, 638-642 (2008).
13 McKay, J. D. et al. Lung cancer susceptibility locus at 5p15.33. Nat Genet 40,
1404-1406 (2008).
14 Walport, M. J. Complement. First of two parts. N Engl J Med 344, 1058-1066
(2001).
15 Ricklin, D., Hajishengallis, G., Yang, K. & Lambris, J. D. Complement: a key
system for immune surveillance and homeostasis. Nat Immunol 11, 785-797
(2010).
16 Law, S. K. A., & Reid, K.B.M. Complement. 2nd edn, (Oxford University
Press, 1995).
17 Nauta, A. J. et al. Direct binding of C1q to apoptotic cells and cell blebs induces
complement activation. Eur J Immunol 32, 1726-1736 (2002).
18 Sjoberg, A. P. et al. Short leucine-rich glycoproteins of the extracellular matrix
display diverse patterns of complement interaction and activation. Mol Immunol
46, 830-839 (2009).
19 McGrath, F. D. et al. Evidence that complement protein C1q interacts with C-
reactive protein through its globular head region. J Immunol 176, 2950-2957
(2006).
20 Ying, S. C., Gewurz, A. T., Jiang, H. & Gewurz, H. Human serum amyloid P
component oligomers bind and activate the classical complement pathway via
References
139
residues 14-26 and 76-92 of the A chain collagen-like region of C1q. J Immunol
150, 169-176 (1993).
21 Mitchell, D. A., Kirby, L., Paulin, S. M., Villiers, C. L. & Sim, R. B. Prion
protein activates and fixes complement directly via the classical pathway:
implications for the mechanism of scrapie agent propagation in lymphoid tissue.
Mol Immunol 44, 2997-3004 (2007).
22 Jiang, H., Cooper, B., Robey, F. A. & Gewurz, H. DNA binds and activates
complement via residues 14-26 of the human C1q A chain. J Biol Chem 267,
25597-25601 (1992).
23 Mayer, M. M. Mechanism of cytolysis by complement. Proc Natl Acad Sci U S
A 69, 2954-2958 (1972).
24 Muller-Eberhard, H. J. The membrane attack complex of complement. Annu Rev
Immunol 4, 503-528 (1986).
25 Thiel, S. et al. Interaction of C1q and mannan-binding lectin (MBL) with C1r,
C1s, MBL-associated serine proteases 1 and 2, and the MBL-associated protein
MAp19. J Immunol 165, 878-887 (2000).
26 Matsushita, M., Thiel, S., Jensenius, J. C., Terai, I. & Fujita, T. Proteolytic
activities of two types of mannose-binding lectin-associated serine protease. J
Immunol 165, 2637-2642 (2000).
27 Bexborn, F., Andersson, P. O., Chen, H., Nilsson, B. & Ekdahl, K. N. The tick-
over theory revisited: formation and regulation of the soluble alternative
complement C3 convertase (C3(H2O)Bb). Mol Immunol 45, 2370-2379 (2008).
28 Hourcade, D. E. Properdin and complement activation: a fresh perspective. Curr
Drug Targets 9, 158-164 (2008).
References
140
29 McMullen, M. E. et al. Mannose-binding lectin binds IgM to activate the lectin
complement pathway in vitro and in vivo. Immunobiology 211, 759-766 (2006).
30 Zhang, M. et al. Activation of the lectin pathway by natural IgM in a model of
ischemia/reperfusion injury. J Immunol 177, 4727-4734 (2006).
31 Wagner, E., Platt, J. L., Howell, D. N., Marsh, H. C., Jr. & Frank, M. M. IgG
and complement-mediated tissue damage in the absence of C2: evidence of a
functionally active C2-bypass pathway in a guinea pig model. J Immunol 163,
3549-3558 (1999).
32 Selander, B. et al. Mannan-binding lectin activates C3 and the alternative
complement pathway without involvement of C2. J Clin Invest 116, 1425-1434
(2006).
33 Markiewski, M. M., Nilsson, B., Ekdahl, K. N., Mollnes, T. E. & Lambris, J. D.
Complement and coagulation: strangers or partners in crime? Trends Immunol
28, 184-192 (2007).
34 Huber-Lang, M. et al. Generation of C5a in the absence of C3: a new
complement activation pathway. Nat Med 12, 682-687 (2006).
35 Maruo, K., Akaike, T., Ono, T., Okamoto, T. & Maeda, H. Generation of
anaphylatoxins through proteolytic processing of C3 and C5 by house dust mite
protease. J Allergy Clin Immunol 100, 253-260 (1997).
36 Huber-Lang, M. et al. Generation of C5a by phagocytic cells. Am J Pathol 161,
1849-1859 (2002).
37 Governa, M. et al. Variability of biological effects of silicas: different degrees of
activation of the fifth component of complement by amorphous silicas. Toxicol
Appl Pharmacol 208, 68-77 (2005).
References
141
38 Governa, M. et al. In vitro cleavage by asbestos fibers of the fifth component of
human complement through free-radical generation and kallikrein activation. J
Toxicol Environ Health A 59, 539-552 (2000).
39 Governa, M. et al. Cleavage of the fifth component of human complement and
release of a split product with C5a-like activity by crystalline silica through free
radical generation and kallikrein activation. Toxicol Appl Pharmacol 179, 129-
136 (2002).
40 Kirkitadze, M. D. & Barlow, P. N. Structure and flexibility of the multiple
domain proteins that regulate complement activation. Immunol Rev 180, 146-
161 (2001).
41 Beinrohr, L. et al. C1 inhibitor serpin domain structure reveals the likely
mechanism of heparin potentiation and conformational disease. J Biol Chem
282, 21100-21109 (2007).
42 Davis, A. E., 3rd, Mejia, P. & Lu, F. Biological activities of C1 inhibitor. Mol
Immunol 45, 4057-4063 (2008).
43 Blom, A. M., Villoutreix, B. O. & Dahlback, B. Complement inhibitor C4b-
binding protein-friend or foe in the innate immune system? Mol Immunol 40,
1333-1346 (2004).
44 Garcia de Frutos, P., Hardig, Y. & Dahlback, B. Serum amyloid P component
binding to C4b-binding protein. J Biol Chem 270, 26950-26955 (1995).
45 Zipfel, P. F., Jokiranta, T. S., Hellwage, J., Koistinen, V. & Meri, S. The factor
H protein family. Immunopharmacology 42, 53-60 (1999).
References
142
46 de Cordoba, S. R. & de Jorge, E. G. Translational mini-review series on
complement factor H: genetics and disease associations of human complement
factor H. Clin Exp Immunol 151, 1-13 (2008).
47 Edwards, A. O. et al. Complement factor H polymorphism and age-related
macular degeneration. Science 308, 421-424 (2005).
48 Hageman, G. S. et al. A common haplotype in the complement regulatory gene
factor H (HF1/CFH) predisposes individuals to age-related macular
degeneration. Proc Natl Acad Sci U S A 102, 7227-7232 (2005).
49 Haines, J. L. et al. Complement factor H variant increases the risk of age-related
macular degeneration. Science 308, 419-421 (2005).
50 Klein, R. J. et al. Complement factor H polymorphism in age-related macular
degeneration. Science 308, 385-389 (2005).
51 Nabil, K. et al. Identification of human complement factor H as a chemotactic
protein for monocytes. Biochem J 326 ( Pt 2), 377-383 (1997).
52 Iferroudjene, D., Schouft, M. T., Lemercier, C., Gilbert, D. & Fontaine, M.
Evidence for an active hydrophobic form of factor H that is able to induce
secretion of interleukin 1-beta or by human monocytes. Eur J Immunol 21, 967-
972 (1991).
53 Zipfel, P. F. & Skerka, C. Complement factor H and related proteins: an
expanding family of complement-regulatory proteins? Immunol Today 15, 121-
126 (1994).
54 Skerka, C. & Zipfel, P. F. Complement factor H related proteins in immune
diseases. Vaccine 26 Suppl 8, I9-14 (2008).
References
143
55 Podack, E. R. & Muller-Eberhard, H. J. SC5b-9 complex of complement:
formation of the dimeric membrane attack complex by removal of S-protein. J
Immunol 124, 1779-1783 (1980).
56 Schvartz, I., Seger, D. & Shaltiel, S. Vitronectin. Int J Biochem Cell Biol 31,
539-544 (1999).
57 Jenne, D. E. & Tschopp, J. Molecular structure and functional characterization
of a human complement cytolysis inhibitor found in blood and seminal plasma:
identity to sulfated glycoprotein 2, a constituent of rat testis fluid. Proc Natl
Acad Sci U S A 86, 7123-7127 (1989).
58 Falgarone, G. & Chiocchia, G. Chapter 8: Clusterin: A multifacet protein at the
crossroad of inflammation and autoimmunity. Adv Cancer Res 104, 139-170
(2009).
59 Fischer, E., Appay, M. D., Cook, J. & Kazatchkine, M. D. Characterization of
the human glomerular C3 receptor as the C3b/C4b complement type one (CR1)
receptor. J Immunol 136, 1373-1377 (1986).
60 Fearon, D. T. Regulation of the amplification C3 convertase of human
complement by an inhibitory protein isolated from human erythrocyte
membrane. Proc Natl Acad Sci U S A 76, 5867-5871 (1979).
61 Seya, T. & Atkinson, J. P. Functional properties of membrane cofactor protein
of complement. Biochem J 264, 581-588 (1989).
62 Marie, J. C. et al. Linking innate and acquired immunity: divergent role of CD46
cytoplasmic domains in T cell induced inflammation. Nat Immunol 3, 659-666
(2002).
References
144
63 Kemper, C. et al. Activation of human CD4+ cells with CD3 and CD46 induces
a T-regulatory cell 1 phenotype. Nature 421, 388-392 (2003).
64 Cattaneo, R. Four viruses, two bacteria, and one receptor: membrane cofactor
protein (CD46) as pathogens' magnet. J Virol 78, 4385-4388 (2004).
65 Medof, M. E., Walter, E. I., Rutgers, J. L., Knowles, D. M. & Nussenzweig, V.
Identification of the complement decay-accelerating factor (DAF) on epithelium
and glandular cells and in body fluids. J Exp Med 165, 848-864 (1987).
66 Medof, M. E., Kinoshita, T. & Nussenzweig, V. Inhibition of complement
activation on the surface of cells after incorporation of decay-accelerating factor
(DAF) into their membranes. J Exp Med 160, 1558-1578 (1984).
67 Abbott, R. J. et al. Structural and functional characterization of a novel T cell
receptor co-regulatory protein complex, CD97-CD55. J Biol Chem 282, 22023-
22032 (2007).
68 Meri, S. et al. Human protectin (CD59), an 18,000-20,000 MW complement
lysis restricting factor, inhibits C5b-8 catalysed insertion of C9 into lipid
bilayers. Immunology 71, 1-9 (1990).
69 Kimberley, F. C., Sivasankar, B. & Paul Morgan, B. Alternative roles for CD59.
Mol Immunol 44, 73-81 (2007).
70 Koski, C. L., Ramm, L. E., Hammer, C. H., Mayer, M. M. & Shin, M. L.
Cytolysis of nucleated cells by complement: cell death displays multi-hit
characteristics. Proc Natl Acad Sci U S A 80, 3816-3820 (1983).
71 Campbell, A. K., Daw, R. A., Hallett, M. B. & Luzio, J. P. Direct measurement
of the increase in intracellular free calcium ion concentration in response to the
action of complement. Biochem J 194, 551-560 (1981).
References
145
72 Morgan, B. P. & Walport, M. J. Complement deficiency and disease. Immunol
Today 12, 301-306 (1991).
73 Cole, D. S. & Morgan, B. P. Beyond lysis: how complement influences cell fate.
Clin Sci (Lond) 104, 455-466 (2003).
74 Davis, A. E., 3rd, Harrison, R. A. & Lachmann, P. J. Physiologic inactivation of
fluid phase C3b: isolation and structural analysis of C3c, C3d,g (alpha 2D), and
C3g. J Immunol 132, 1960-1966 (1984).
75 Klickstein, L. B., Barbashov, S. F., Liu, T., Jack, R. M. & Nicholson-Weller, A.
Complement receptor type 1 (CR1, CD35) is a receptor for C1q. Immunity 7,
345-355 (1997).
76 Ghiran, I. et al. Complement receptor 1/CD35 is a receptor for mannan-binding
lectin. J Exp Med 192, 1797-1808 (2000).
77 Taylor, R. P. et al. Immune complexes bound to the primate erythrocyte
complement receptor (CR1) via anti-CR1 mAbs are cleared simultaneously with
loss of CR1 in a concerted reaction in a rhesus monkey model. Clin Immunol
Immunopathol 82, 49-59 (1997).
78 Bacle, F., Haeffner-Cavaillon, N., Laude, M., Couturier, C. & Kazatchkine, M.
D. Induction of IL-1 release through stimulation of the C3b/C4b complement
receptor type one (CR1, CD35) on human monocytes. J Immunol 144, 147-152
(1990).
79 Krych-Goldberg, M. & Atkinson, J. P. Structure-function relationships of
complement receptor type 1. Immunol Rev 180, 112-122 (2001).
References
146
80 Fearon, D. T. & Carter, R. H. The CD19/CR2/TAPA-1 complex of B
lymphocytes: linking natural to acquired immunity. Annu Rev Immunol 13, 127-
149 (1995).
81 Dempsey, P. W., Allison, M. E., Akkaraju, S., Goodnow, C. C. & Fearon, D. T.
C3d of complement as a molecular adjuvant: bridging innate and acquired
immunity. Science 271, 348-350 (1996).
82 van Lookeren Campagne, M., Wiesmann, C. & Brown, E. J. Macrophage
complement receptors and pathogen clearance. Cell Microbiol 9, 2095-2102
(2007).
83 Helmy, K. Y. et al. CRIg: a macrophage complement receptor required for
phagocytosis of circulating pathogens. Cell 124, 915-927 (2006).
84 Gorski, J. P., Hugli, T. E. & Muller-Eberhard, H. J. Characterization of human
C4a anaphylatoxin. J Biol Chem 256, 2707-2711 (1981).
85 Lee, D. K. et al. Identification of four novel human G protein-coupled receptors
expressed in the brain. Brain Res Mol Brain Res 86, 13-22 (2001).
86 Samson, M. et al. ChemR23, a putative chemoattractant receptor, is expressed in
monocyte-derived dendritic cells and macrophages and is a coreceptor for SIV
and some primary HIV-1 strains. Eur J Immunol 28, 1689-1700 (1998).
87 Ames, R. S., Tornetta, M. A., Foley, J. J., Hugli, T. E. & Sarau, H. M. Evidence
that the receptor for C4a is distinct from the C3a receptor.
Immunopharmacology 38, 87-92 (1997).
88 Boulay, F., Mery, L., Tardif, M., Brouchon, L. & Vignais, P. Expression cloning
of a receptor for C5a anaphylatoxin on differentiated HL-60 cells. Biochemistry
30, 2993-2999 (1991).
References
147
89 Ohno, M. et al. A putative chemoattractant receptor, C5L2, is expressed in
granulocyte and immature dendritic cells, but not in mature dendritic cells. Mol
Immunol 37, 407-412 (2000).
90 Gerard, N. P. et al. An anti-inflammatory function for the complement
anaphylatoxin C5a-binding protein, C5L2. J Biol Chem 280, 39677-39680
(2005).
91 Chen, N. J. et al. C5L2 is critical for the biological activities of the
anaphylatoxins C5a and C3a. Nature 446, 203-207 (2007).
92 Strainic, M. G. et al. Locally produced complement fragments C5a and C3a
provide both costimulatory and survival signals to naive CD4+ T cells. Immunity
28, 425-435 (2008).
93 Gasque, P. et al. The receptor for complement anaphylatoxin C3a is expressed
by myeloid cells and nonmyeloid cells in inflamed human central nervous
system: analysis in multiple sclerosis and bacterial meningitis. J Immunol 160,
3543-3554 (1998).
94 Monsinjon, T. et al. Regulation by complement C3a and C5a anaphylatoxins of
cytokine production in human umbilical vein endothelial cells. FASEB J 17,
1003-1014 (2003).
95 Fregonese, L. et al. Expression of the anaphylatoxin receptors C3aR and C5aR
is increased in fatal asthma. J Allergy Clin Immunol 115, 1148-1154 (2005).
96 Werfel, T. et al. Activated human T lymphocytes express a functional C3a
receptor. J Immunol 165, 6599-6605 (2000).
References
148
97 Monk, P. N., Scola, A. M., Madala, P. & Fairlie, D. P. Function, structure and
therapeutic potential of complement C5a receptors. Br J Pharmacol 152, 429-
448 (2007).
98 Okinaga, S. et al. C5L2, a nonsignaling C5A binding protein. Biochemistry 42,
9406-9415 (2003).
99 Gao, H. et al. Evidence for a functional role of the second C5a receptor C5L2.
FASEB J 19, 1003-1005 (2005).
100 Haas, P. J. & van Strijp, J. Anaphylatoxins: their role in bacterial infection and
inflammation. Immunol Res 37, 161-175 (2007).
101 Strey, C. W. et al. The proinflammatory mediators C3a and C5a are essential for
liver regeneration. J Exp Med 198, 913-923 (2003).
102 Benard, M. et al. Characterization of C3a and C5a receptors in rat cerebellar
granule neurons during maturation. Neuroprotective effect of C5a against
apoptotic cell death. J Biol Chem 279, 43487-43496 (2004).
103 Matthews, K. W., Mueller-Ortiz, S. L. & Wetsel, R. A. Carboxypeptidase N: a
pleiotropic regulator of inflammation. Mol Immunol 40, 785-793 (2004).
104 Ember, J. A., Jagels, M.A. & Hugli, T.E. Characterization of complement
anaphylatoxins and their biological responses. 241–284 (Marcel Dekker,
1998).
105 Oppermann, M. & Gotze, O. Plasma clearance of the human C5a anaphylatoxin
by binding to leucocyte C5a receptors. Immunology 82, 516-521 (1994).
References
149
106 Scola, A. M., Johswich, K. O., Morgan, B. P., Klos, A. & Monk, P. N. The
human complement fragment receptor, C5L2, is a recycling decoy receptor. Mol
Immunol 46, 1149-1162 (2009).
107 Markiewski, M. M. et al. Modulation of the antitumor immune response by
complement. Nat Immunol 9, 1225-1235 (2008).
108 Hakomori, S. Tumor malignancy defined by aberrant glycosylation and
sphingo(glyco)lipid metabolism. Cancer Res 56, 5309-5318 (1996).
109 Pardoll, D. Does the immune system see tumors as foreign or self? Annu Rev
Immunol 21, 807-839 (2003).
110 Niculescu, F., Rus, H. G., Retegan, M. & Vlaicu, R. Persistent complement
activation on tumor cells in breast cancer. Am J Pathol 140, 1039-1043 (1992).
111 Yamakawa, M. et al. Protection of thyroid cancer cells by complement-
regulatory factors. Cancer 73, 2808-2817 (1994).
112 Niehans, G. A., Cherwitz, D. L., Staley, N. A., Knapp, D. J. & Dalmasso, A. P.
Human carcinomas variably express the complement inhibitory proteins CD46
(membrane cofactor protein), CD55 (decay-accelerating factor), and CD59
(protectin). Am J Pathol 149, 129-142 (1996).
113 Bjorge, L. et al. Ascitic complement system in ovarian cancer. Br J Cancer 92,
895-905 (2005).
114 Ytting, H., Jensenius, J. C., Christensen, I. J., Thiel, S. & Nielsen, H. J.
Increased activity of the mannan-binding lectin complement activation pathway
in patients with colorectal cancer. Scand J Gastroenterol 39, 674-679 (2004).
References
150
115 Carli, M. et al. Fluctuation of serum complement levels in children with
neuroblastoma. Cancer 43, 2399-2404 (1979).
116 Maness, P. F. & Orengo, A. Serum complement levels in patients with digestive
tract carcinomas and other neoplastic diseases. Oncology 34, 87-89 (1977).
117 Matsutani, M. et al. Cellular immunity and complement levels in hosts with
brain tumours. Neurosurg Rev 7, 29-35 (1984).
118 Mangano, A., Messina, L., Birgillito, S., Stivala, F. & Bernardini, A.
Complement and its fractions (C3-C4) pattern in subjects with neoplasia. J
Immunopharmacol 6, 147-162 (1984).
119 Smyth, M. J., Godfrey, D. I. & Trapani, J. A. A fresh look at tumor
immunosurveillance and immunotherapy. Nat Immunol 2, 293-299 (2001).
120 Jurianz, K. et al. Complement resistance of tumor cells: basal and induced
mechanisms. Mol Immunol 36, 929-939 (1999).
121 Fishelson, Z., Donin, N., Zell, S., Schultz, S. & Kirschfink, M. Obstacles to
cancer immunotherapy: expression of membrane complement regulatory
proteins (mCRPs) in tumors. Mol Immunol 40, 109-123 (2003).
122 Ajona, D. et al. Expression of complement factor H by lung cancer cells: effects
on the activation of the alternative pathway of complement. Cancer Res 64,
6310-6318 (2004).
123 Ajona, D., Hsu, Y. F., Corrales, L., Montuenga, L. M. & Pio, R. Down-
regulation of human complement factor H sensitizes non-small cell lung cancer
cells to complement attack and reduces in vivo tumor growth. J Immunol 178,
5991-5998 (2007).
References
151
124 Okroj, M., Hsu, Y. F., Ajona, D., Pio, R. & Blom, A. M. Non-small cell lung
cancer cells produce a functional set of complement factor I and its soluble
cofactors. Mol Immunol 45, 169-179 (2008).
125 Kirschfink, M., Blase, L., Engelmann, S. & Schwartz-Albiez, R. Secreted
chondroitin sulfate proteoglycan of human B cell lines binds to the complement
protein C1q and inhibits complex formation of C1. J Immunol 158, 1324-1331
(1997).
126 Ollert, M. W. et al. C3-cleaving membrane proteinase. A new complement
regulatory protein of human melanoma cells. J Immunol 144, 3862-3867 (1990).
127 Paas, Y., Bohana-Kashtan, O. & Fishelson, Z. Phosphorylation of the
complement component, C9, by an ecto-protein kinase of human leukemic cells.
Immunopharmacology 42, 175-185 (1999).
128 Bohana-Kashtan, O., Pinna, L. A. & Fishelson, Z. Extracellular phosphorylation
of C9 by protein kinase CK2 regulates complement-mediated lysis. Eur J
Immunol 35, 1939-1948 (2005).
129 Morgan, B. P. Complement membrane attack on nucleated cells: resistance,
recovery and non-lethal effects. Biochem J 264, 1-14 (1989).
130 Reiter, Y., Ciobotariu, A. & Fishelson, Z. Sublytic complement attack protects
tumor cells from lytic doses of antibody and complement. Eur J Immunol 22,
1207-1213 (1992).
131 Schrama, D., Reisfeld, R. A. & Becker, J. C. Antibody targeted drugs as cancer
therapeutics. Nat Rev Drug Discov 5, 147-159 (2006).
References
152
132 Blok, V. T., Gelderman, K. A., Tijsma, O. H., Daha, M. R. & Gorter, A.
Cytokines affect resistance of human renal tumour cells to complement-
mediated injury. Scand J Immunol 57, 591-599 (2003).
133 Brasoveanu, L. I. et al. Levels of cell membrane CD59 regulate the extent of
complement-mediated lysis of human melanoma cells. Lab Invest 74, 33-42
(1996).
134 Varsano, S., Rashkovsky, L., Shapiro, H., Ophir, D. & Mark-Bentankur, T.
Human lung cancer cell lines express cell membrane complement inhibitory
proteins and are extremely resistant to complement-mediated lysis; a comparison
with normal human respiratory epithelium in vitro, and an insight into
mechanism(s) of resistance. Clin Exp Immunol 113, 173-182 (1998).
135 Paganelli, G. et al. Three-step monoclonal antibody tumor targeting in
carcinoembryonic antigen-positive patients. Cancer Res 51, 5960-5966 (1991).
136 Macor, P. et al. Complement activated by chimeric anti-folate receptor
antibodies is an efficient effector system to control ovarian carcinoma. Cancer
Res 66, 3876-3883 (2006).
137 Gelderman, K. A., Lam, S. & Gorter, A. Inhibiting complement regulators in
cancer immunotherapy with bispecific mAbs. Expert Opin Biol Ther 5, 1593-
1601 (2005).
138 Balkwill, F. & Mantovani, A. Inflammation and cancer: back to Virchow?
Lancet 357, 539-545 (2001).
139 Mantovani, A., Allavena, P., Sica, A. & Balkwill, F. Cancer-related
inflammation. Nature 454, 436-444 (2008).
References
153
140 Langman, M. J., Cheng, K. K., Gilman, E. A. & Lancashire, R. J. Effect of anti-
inflammatory drugs on overall risk of common cancer: case-control study in
general practice research database. BMJ 320, 1642-1646 (2000).
141 Colotta, F., Allavena, P., Sica, A., Garlanda, C. & Mantovani, A. Cancer-related
inflammation, the seventh hallmark of cancer: links to genetic instability.
Carcinogenesis 30, 1073-1081 (2009).
142 Laufer, J., Katz, Y. & Passwell, J. H. Extrahepatic synthesis of complement
proteins in inflammation. Mol Immunol 38, 221-229 (2001).
143 Nishioka, K., Kawamura, K., Hirayama, T., Kawashima, T. & Shimada, K. The
complement system in tumor immunity: significance of elevated levels of
complement in tumor bearing hosts. Ann N Y Acad Sci 276, 303-315 (1976).
144 Varsano, S., Frolkis, I., Shapiro, H. & Ophir, D. Human nasal epithelium
adsorbs complement C3-related fragments and expresses cell membrane
complement regulatory proteins. Laryngoscope 106, 599-604 (1996).
145 Gorter, A. et al. Expression of CD46, CD55, and CD59 on renal tumor cell lines
and their role in preventing complement-mediated tumor cell lysis. Lab Invest
74, 1039-1049 (1996).
146 Pio, R. et al. Complement factor H is elevated in bronchoalveolar lavage fluid
and sputum from patients with lung cancer. Cancer Epidemiol Biomarkers Prev
19, 2665-2672 (2010).
147 Sato, M. et al. Multiple oncogenic changes (K-RAS(V12), p53 knockdown,
mutant EGFRs, p16 bypass, telomerase) are not sufficient to confer a full
malignant phenotype on human bronchial epithelial cells. Cancer Res 66, 2116-
2128 (2006).
References
154
148 Reddel, R. R. et al. Development of tumorigenicity in simian virus 40-
immortalized human bronchial epithelial cell lines. Cancer Res 53, 985-991
(1993).
149 Ramirez, R. D. et al. Immortalization of human bronchial epithelial cells in the
absence of viral oncoproteins. Cancer Res 64, 9027-9034 (2004).
150 Bertram, J. S. & Janik, P. Establishment of a cloned line of Lewis Lung
Carcinoma cells adapted to cell culture. Cancer Lett 11, 63-73 (1980).
151 Sim, R. B., Day, A. J., Moffatt, B. E. & Fontaine, M. Complement factor I and
cofactors in control of complement system convertase enzymes. Methods
Enzymol 223, 13-35 (1993).
152 Finch, A. M. et al. Low-molecular-weight peptidic and cyclic antagonists of the
receptor for the complement factor C5a. J Med Chem 42, 1965-1974 (1999).
153 Morgan, B. Complement Methods and Protocols. Vol. 150 (Humana Press,
2000).
154 Servais, G., Walmagh, J. & Duchateau, J. Simple quantitative haemolytic
microassay for determination of complement alternative pathway activation
(AP50). J Immunol Methods 140, 93-100 (1991).
155 Irigoyen, M. et al. Hypoxia alters the adhesive properties of lymphatic
endothelial cells. A transcriptional and functional study. Biochim Biophys Acta
1773, 880-890 (2007).
156 Fehrenbach, H. Animal models of pulmonary emphysema: a stereologist’s
perspective. European Respiratory Review 15 (2006).
References
155
157 Treumann, A., Lifely, M. R., Schneider, P. & Ferguson, M. A. Primary structure
of CD52. J Biol Chem 270, 6088-6099 (1995).
158 Okroj, M., Corrales, L., Stokowska, A., Pio, R. & Blom, A. M. Hypoxia
increases susceptibility of non-small cell lung cancer cells to complement attack.
Cancer Immunol Immunother 58, 1771-1780 (2009).
159 Kock, M. A., Hew, B. E., Bammert, H., Fritzinger, D. C. & Vogel, C. W.
Structure and function of recombinant cobra venom factor. J Biol Chem 279,
30836-30843 (2004).
160 Guo, R. F. & Ward, P. A. Role of C5a in inflammatory responses. Annu Rev
Immunol 23, 821-852 (2005).
161 Fleming, T. J., Fleming, M. L. & Malek, T. R. Selective expression of Ly-6G on
myeloid lineage cells in mouse bone marrow. RB6-8C5 mAb to granulocyte-
differentiation antigen (Gr-1) detects members of the Ly-6 family. J Immunol
151, 2399-2408 (1993).
162 Sunderkotter, C. et al. Subpopulations of mouse blood monocytes differ in
maturation stage and inflammatory response. J Immunol 172, 4410-4417 (2004).
163 Kubota, Y., Kleinman, H. K., Martin, G. R. & Lawley, T. J. Role of laminin and
basement membrane in the morphological differentiation of human endothelial
cells into capillary-like structures. J Cell Biol 107, 1589-1598 (1988).
164 Snyderman, R., Pike, M. C., McCarley, D. & Lang, L. Quantification of mouse
macrophage chemotaxis in vitro: role of C5 for the production of chemotactic
activity. Infect Immun 11, 488-492 (1975).
References
156
165 Clynes, R. A., Towers, T. L., Presta, L. G. & Ravetch, J. V. Inhibitory Fc
receptors modulate in vivo cytoxicity against tumor targets. Nat Med 6, 443-446
(2000).
166 Nagaraj, S. & Gabrilovich, D. I. Tumor escape mechanism governed by
myeloid-derived suppressor cells. Cancer Res 68, 2561-2563 (2008).
167 Gabrilovich, D. I. & Nagaraj, S. Myeloid-derived suppressor cells as regulators
of the immune system. Nat Rev Immunol 9, 162-174 (2009).
168 Coussens, L. M. & Werb, Z. Inflammation and cancer. Nature 420, 860-867
(2002).
169 Davis, I. D., Jefford, M., Parente, P. & Cebon, J. Rational approaches to human
cancer immunotherapy. J Leukoc Biol 73, 3-29 (2003).
170 Gelderman, K. A., Tomlinson, S., Ross, G. D. & Gorter, A. Complement
function in mAb-mediated cancer immunotherapy. Trends Immunol 25, 158-164
(2004).
171 Hsu, Y. F. et al. Complement activation mediates cetuximab inhibition of non-
small cell lung cancer tumor growth in vivo. Mol Cancer 9, 139 (2010).
172 Hanahan, D. & Weinberg, R. A. The hallmarks of cancer. Cell 100, 57-70
(2000).
173 Guidi, L. et al. Immune complexes in solid tumours precipitable by 3.5%
polyethylene glycol: analysis of some nonspecific components. Diagn Clin
Immunol 5, 284-288 (1988).
References
157
174 Sahu, A., Kozel, T. R. & Pangburn, M. K. Specificity of the thioester-containing
reactive site of human C3 and its significance to complement activation.
Biochem J 302 ( Pt 2), 429-436 (1994).
175 Sahu, A. & Pangburn, M. K. Covalent attachment of human complement C3 to
IgG. Identification of the amino acid residue involved in ester linkage formation.
J Biol Chem 269, 28997-29002 (1994).
176 Sahu, A. & Pangburn, M. K. Tyrosine is a potential site for covalent attachment
of activated complement component C3. Mol Immunol 32, 711-716 (1995).
177 Spitzer, D., Mitchell, L. M., Atkinson, J. P. & Hourcade, D. E. Properdin can
initiate complement activation by binding specific target surfaces and providing
a platform for de novo convertase assembly. J Immunol 179, 2600-2608 (2007).
178 Kemper, C., Atkinson, J. P. & Hourcade, D. E. Properdin: emerging roles of a
pattern-recognition molecule. Annu Rev Immunol 28, 131-155 (2010).
179 Korbelik, M. & Cecic, I. Complement activation cascade and its regulation:
relevance for the response of solid tumors to photodynamic therapy. J
Photochem Photobiol B 93, 53-59 (2008).
180 Harboe, M. & Mollnes, T. E. The alternative complement pathway revisited. J
Cell Mol Med 12, 1074-1084 (2008).
181 Weichenthal, M., Siemann, U., Neuber, K. & Breitbart, E. W. Expression of
complement regulator proteins in primary and metastatic malignant melanoma. J
Cutan Pathol 26, 217-221 (1999).
182 Jarvis, G. A. et al. Expression and function of the complement membrane attack
complex inhibitor protectin (CD59) in human prostate cancer. Int J Cancer 71,
1049-1055 (1997).
References
158
183 Bjorge, L., Vedeler, C. A., Ulvestad, E. & Matre, R. Expression and function of
CD59 on colonic adenocarcinoma cells. Eur J Immunol 24, 1597-1603 (1994).
184 Donin, N. et al. Complement resistance of human carcinoma cells depends on
membrane regulatory proteins, protein kinases and sialic acid. Clin Exp Immunol
131, 254-263 (2003).
185 Loberg, R. D., Wojno, K. J., Day, L. L. & Pienta, K. J. Analysis of membrane-
bound complement regulatory proteins in prostate cancer. Urology 66, 1321-
1326 (2005).
186 Mustafa, T. et al. Expression of CD97 and CD55 in human medullary thyroid
carcinomas. Int J Oncol 24, 285-294 (2004).
187 Loberg, R. D., Day, L. L., Dunn, R., Kalikin, L. M. & Pienta, K. J. Inhibition of
decay-accelerating factor (CD55) attenuates prostate cancer growth and survival
in vivo. Neoplasia 8, 69-78 (2006).
188 Durrant, L. G. et al. Enhanced expression of the complement regulatory protein
CD55 predicts a poor prognosis in colorectal cancer patients. Cancer Immunol
Immunother 52, 638-642 (2003).
189 Madjd, Z. et al. Loss of CD55 is associated with aggressive breast tumors. Clin
Cancer Res 10, 2797-2803 (2004).
190 Madjd, Z. et al. Do poor-prognosis breast tumours express membrane cofactor
proteins (CD46)? Cancer Immunol Immunother 54, 149-156 (2005).
191 Rushmere, N. K. et al. Analysis of the level of mRNA expression of the
membrane regulators of complement, CD59, CD55 and CD46, in breast cancer.
Int J Cancer 108, 930-936 (2004).
References
159
192 Sadallah, S. et al. Soluble complement receptor 1 is increased in patients with
leukemia and after administration of granulocyte colony-stimulating factor. J
Leukoc Biol 65, 94-101 (1999).
193 Seya, T. et al. Purification and functional properties of soluble forms of
membrane cofactor protein (CD46) of complement: identification of forms
increased in cancer patients' sera. Int Immunol 7, 727-736 (1995).
194 Li, L., Spendlove, I., Morgan, J. & Durrant, L. G. CD55 is over-expressed in the
tumour environment. Br J Cancer 84, 80-86 (2001).
195 Morgan, J., Spendlove, I. & Durrant, L. G. The role of CD55 in protecting the
tumour environment from complement attack. Tissue Antigens 60, 213-223
(2002).
196 Hakulinen, J., Junnikkala, S., Sorsa, T. & Meri, S. Complement inhibitor
membrane cofactor protein (MCP; CD46) is constitutively shed from cancer cell
membranes in vesicles and converted by a metalloproteinase to a functionally
active soluble form. Eur J Immunol 34, 2620-2629 (2004).
197 Davies, A. et al. CD59, an LY-6-like protein expressed in human lymphoid
cells, regulates the action of the complement membrane attack complex on
homologous cells. J Exp Med 170, 637-654 (1989).
198 Gasque, P. et al. Expression of complement components of the alternative
pathway by glioma cell lines. J Immunol 149, 1381-1387 (1992).
199 Gasque, P., Thomas, A., Fontaine, M. & Morgan, B. P. Complement activation
on human neuroblastoma cell lines in vitro: route of activation and expression of
functional complement regulatory proteins. J Neuroimmunol 66, 29-40 (1996).
References
160
200 Junnikkala, S. et al. Secretion of soluble complement inhibitors factor H and
factor H-like protein (FHL-1) by ovarian tumour cells. Br J Cancer 87, 1119-
1127 (2002).
201 Katz, Y., Guterman, M. & Lahat, E. Regulation of synthesis of complement
proteins in HEp2 cells. Clin Immunol Immunopathol 67, 117-123 (1993).
202 Legoedec, J., Gasque, P., Jeanne, J. F. & Fontaine, M. Expression of the
complement alternative pathway by human myoblasts in vitro: biosynthesis of
C3, factor B, factor H and factor I. Eur J Immunol 25, 3460-3466 (1995).
203 Cheng, Z. Z. et al. Complement factor H as a marker for detection of bladder
cancer. Clin Chem 51, 856-863 (2005).
204 Amornsiripanitch, N. et al. Complement factor H autoantibodies are associated
with early stage NSCLC. Clin Cancer Res 16, 3226-3231 (2010).
205 Junnikkala, S. et al. Exceptional resistance of human H2 glioblastoma cells to
complement-mediated killing by expression and utilization of factor H and
factor H-like protein 1. J Immunol 164, 6075-6081 (2000).
206 Corey, M. J., Kinders, R. J., Brown, L. G. & Vessella, R. L. A very sensitive
coupled luminescent assay for cytotoxicity and complement-mediated lysis. J
Immunol Methods 207, 43-51 (1997).
207 Brown, J. M. & Giaccia, A. J. The unique physiology of solid tumors:
opportunities (and problems) for cancer therapy. Cancer Res 58, 1408-1416
(1998).
208 Li, C. & Jackson, R. M. Reactive species mechanisms of cellular hypoxia-
reoxygenation injury. Am J Physiol Cell Physiol 282, C227-241 (2002).
References
161
209 Arumugam, T. V., Shiels, I. A., Woodruff, T. M., Granger, D. N. & Taylor, S.
M. The role of the complement system in ischemia-reperfusion injury. Shock 21,
401-409 (2004).
210 Kannagi, R. Carbohydrate antigen sialyl Lewis a--its pathophysiological
significance and induction mechanism in cancer progression. Chang Gung Med
J 30, 189-209 (2007).
211 Carroll, M. C. & Holers, V. M. Innate autoimmunity. Adv Immunol 86, 137-157
(2005).
212 de Visser, K. E., Korets, L. V. & Coussens, L. M. Early neoplastic progression
is complement independent. Neoplasia 6, 768-776 (2004).
213 Kurimoto, Y., de Weck, A. L. & Dahinden, C. A. Interleukin 3-dependent
mediator release in basophils triggered by C5a. J Exp Med 170, 467-479 (1989).
214 Peng, Q. et al. Dendritic cell function in allostimulation is modulated by C5aR
signaling. J Immunol 183, 6058-6068 (2009).
215 Ottonello, L. et al. rC5a directs the in vitro migration of human memory and
naive tonsillar B lymphocytes: implications for B cell trafficking in secondary
lymphoid tissues. J Immunol 162, 6510-6517 (1999).
216 Riedemann, N. C. et al. C5a receptor and thymocyte apoptosis in sepsis. FASEB
J 16, 887-888 (2002).
217 Ikeda, K. et al. C5a induces tissue factor activity on endothelial cells. Thromb
Haemost 77, 394-398 (1997).
218 Foreman, K. E. et al. C5a-induced expression of P-selectin in endothelial cells. J
Clin Invest 94, 1147-1155 (1994).
References
162
219 Schraufstatter, I. U., Trieu, K., Sikora, L., Sriramarao, P. & DiScipio, R.
Complement c3a and c5a induce different signal transduction cascades in
endothelial cells. J Immunol 169, 2102-2110 (2002).
220 Niederbichler, A. D. et al. An essential role for complement C5a in the
pathogenesis of septic cardiac dysfunction. J Exp Med 203, 53-61 (2006).
221 Jacob, A. et al. C5a alters blood-brain barrier integrity in experimental lupus.
FASEB J 24, 1682-1688 (2010).
222 Laudes, I. J. et al. Expression and function of C5a receptor in mouse
microvascular endothelial cells. J Immunol 169, 5962-5970 (2002).
223 O'Barr, S. A. et al. Neuronal expression of a functional receptor for the C5a
complement activation fragment. J Immunol 166, 4154-4162 (2001).
224 Farkas, I. et al. Complement C5a receptor-mediated signaling may be involved
in neurodegeneration in Alzheimer's disease. J Immunol 170, 5764-5771 (2003).
225 Gasque, P., Singhrao, S. K., Neal, J. W., Gotze, O. & Morgan, B. P. Expression
of the receptor for complement C5a (CD88) is up-regulated on reactive
astrocytes, microglia, and endothelial cells in the inflamed human central
nervous system. Am J Pathol 150, 31-41 (1997).
226 Rahpeymai, Y. et al. Complement: a novel factor in basal and ischemia-induced
neurogenesis. EMBO J 25, 1364-1374 (2006).
227 Nataf, S., Levison, S. W. & Barnum, S. R. Expression of the anaphylatoxin C5a
receptor in the oligodendrocyte lineage. Brain Res 894, 321-326 (2001).
References
163
228 Yuan, G. et al. Expression of C5aR (CD88) of synoviocytes isolated from
patients with rheumatoid arthritis and osteoarthritis. Chin Med J (Engl) 116,
1408-1412 (2003).
229 Kiener, H. P. et al. Expression of the C5a receptor (CD88) on synovial mast
cells in patients with rheumatoid arthritis. Arthritis Rheum 41, 233-245 (1998).
230 Hueber, A. J. et al. Mast cells express IL-17A in rheumatoid arthritis synovium.
J Immunol 184, 3336-3340 (2010).
231 Onuma, H. et al. Expression of the anaphylatoxin receptor C5aR (CD88) by
human articular chondrocytes. Rheumatol Int 22, 52-55 (2002).
232 Cortright, D. N., Meade, R., Waters, S. M., Chenard, B. L. & Krause, J. E. C5a,
but not C3a, increases VEGF secretion in ARPE-19 human retinal pigment
epithelial cells. Curr Eye Res 34, 57-61 (2009).
233 Nozaki, M. et al. Drusen complement components C3a and C5a promote
choroidal neovascularization. Proc Natl Acad Sci U S A 103, 2328-2333 (2006).
234 Fukuoka, Y. & Medof, E. M. C5a receptor-mediated production of IL-8 by the
human retinal pigment epithelial cell line, ARPE-19. Curr Eye Res 23, 320-325
(2001).
235 Braun, M. & Davis, A. E., 3rd. Cultured human glomerular mesangial cells
express the C5a receptor. Kidney Int 54, 1542-1549 (1998).
236 Fayyazi, A. et al. The C5a receptor is expressed in normal renal proximal
tubular but not in normal pulmonary or hepatic epithelial cells. Immunology 99,
38-45 (2000).
References
164
237 Boor, P. et al. Complement C5 mediates experimental tubulointerstitial fibrosis.
J Am Soc Nephrol 18, 1508-1515 (2007).
238 Buchner, R. R., Hugli, T. E., Ember, J. A. & Morgan, E. L. Expression of
functional receptors for human C5a anaphylatoxin (CD88) on the human
hepatocellular carcinoma cell line HepG2. Stimulation of acute-phase protein-
specific mRNA and protein synthesis by human C5a anaphylatoxin. J Immunol
155, 308-315 (1995).
239 McCoy, R. et al. N-formylpeptide and complement C5a receptors are expressed
in liver cells and mediate hepatic acute phase gene regulation. J Exp Med 182,
207-217 (1995).
240 Koleva, M. et al. Induction of anaphylatoxin C5a receptors in rat hepatocytes by
lipopolysaccharide in vivo: mediation by interleukin-6 from Kupffer cells.
Gastroenterology 122, 697-708 (2002).
241 Daveau, M. et al. Expression of a functional C5a receptor in regenerating
hepatocytes and its involvement in a proliferative signaling pathway in rat. J
Immunol 173, 3418-3424 (2004).
242 Schlaf, G. et al. Upregulation of fibronectin but not of entactin, collagen IV and
smooth muscle actin by anaphylatoxin C5a in rat hepatic stellate cells. Histol
Histopathol 19, 1165-1174 (2004).
243 Drouin, S. M. et al. Expression of the complement anaphylatoxin C3a and C5a
receptors on bronchial epithelial and smooth muscle cells in models of sepsis
and asthma. J Immunol 166, 2025-2032 (2001).
244 Riedemann, N. C. et al. Expression and function of the C5a receptor in rat
alveolar epithelial cells. J Immunol 168, 1919-1925 (2002).
References
165
245 Fayyazi, A. et al. C5a receptor and interleukin-6 are expressed in tissue
macrophages and stimulated keratinocytes but not in pulmonary and intestinal
epithelial cells. Am J Pathol 154, 495-501 (1999).
246 Youn, J. I. & Gabrilovich, D. I. The biology of myeloid-derived suppressor
cells: the blessing and the curse of morphological and functional heterogeneity.
Eur J Immunol 40, 2969-2975 (2010).
247 Youn, J. I., Nagaraj, S., Collazo, M. & Gabrilovich, D. I. Subsets of myeloid-
derived suppressor cells in tumor-bearing mice. J Immunol 181, 5791-5802
(2008).
248 Priceman, S. J. et al. Targeting distinct tumor-infiltrating myeloid cells by
inhibiting CSF-1 receptor: combating tumor evasion of antiangiogenic therapy.
Blood 115, 1461-1471 (2010).
249 Yang, L. et al. Abrogation of TGF beta signaling in mammary carcinomas
recruits Gr-1+CD11b+ myeloid cells that promote metastasis. Cancer Cell 13,
23-35 (2008).
250 Makarenkova, V. P., Bansal, V., Matta, B. M., Perez, L. A. & Ochoa, J. B.
CD11b+/Gr-1+ myeloid suppressor cells cause T cell dysfunction after traumatic
stress. J Immunol 176, 2085-2094 (2006).
251 Voisin, M. B., Buzoni-Gatel, D., Bout, D. & Velge-Roussel, F. Both expansion
of regulatory GR1+ CD11b+ myeloid cells and anergy of T lymphocytes
participate in hyporesponsiveness of the lung-associated immune system during
acute toxoplasmosis. Infect Immun 72, 5487-5492 (2004).
252 Delano, M. J. et al. MyD88-dependent expansion of an immature GR-
1(+)CD11b(+) population induces T cell suppression and Th2 polarization in
sepsis. J Exp Med 204, 1463-1474 (2007).
References
166
253 Suzuki, J., Ricordi, C. & Chen, Z. Immune tolerance induction by integrating
innate and adaptive immune regulators. Cell Transplant 19, 253-268 (2010).
254 Cosmi, L. et al. Human CD8+CD25+ thymocytes share phenotypic and
functional features with CD4+CD25+ regulatory thymocytes. Blood 102, 4107-
4114 (2003).
255 Gavin, M. A. et al. Foxp3-dependent programme of regulatory T-cell
differentiation. Nature 445, 771-775 (2007).
256 Cao, X. Regulatory T cells and immune tolerance to tumors. Immunol Res 46,
79-93 (2010).
257 Sakaguchi, S. Naturally arising Foxp3-expressing CD25+CD4+ regulatory T
cells in immunological tolerance to self and non-self. Nat Immunol 6, 345-352
(2005).
258 Curiel, T. J. et al. Specific recruitment of regulatory T cells in ovarian
carcinoma fosters immune privilege and predicts reduced survival. Nat Med 10,
942-949 (2004).
259 Nishikawa, H. et al. Definition of target antigens for naturally occurring CD4(+)
CD25(+) regulatory T cells. J Exp Med 201, 681-686 (2005).
260 Wilke, C. M., Wu, K., Zhao, E., Wang, G. & Zou, W. Prognostic significance of
regulatory T cells in tumor. Int J Cancer 127, 748-758 (2010).
261 Grutz, G. New insights into the molecular mechanism of interleukin-10-
mediated immunosuppression. J Leukoc Biol 77, 3-15 (2005).
262 Dong, H. et al. Tumor-associated B7-H1 promotes T-cell apoptosis: a potential
mechanism of immune evasion. Nat Med 8, 793-800 (2002).
References
167
263 Scalapino, K. J. & Daikh, D. I. CTLA-4: a key regulatory point in the control of
autoimmune disease. Immunol Rev 223, 143-155 (2008).
264 Workman, C. J., Dugger, K. J. & Vignali, D. A. Cutting edge: molecular
analysis of the negative regulatory function of lymphocyte activation gene-3. J
Immunol 169, 5392-5395 (2002).
265 Read, S., Malmstrom, V. & Powrie, F. Cytotoxic T lymphocyte-associated
antigen 4 plays an essential role in the function of CD25(+)CD4(+) regulatory
cells that control intestinal inflammation. J Exp Med 192, 295-302 (2000).
266 Huang, C. T. et al. Role of LAG-3 in regulatory T cells. Immunity 21, 503-513
(2004).
267 Bronte, V. & Zanovello, P. Regulation of immune responses by L-arginine
metabolism. Nat Rev Immunol 5, 641-654 (2005).
268 Collison, L. W. et al. The inhibitory cytokine IL-35 contributes to regulatory T-
cell function. Nature 450, 566-569 (2007).
269 Danese, S. & Mantovani, A. Inflammatory bowel disease and intestinal cancer: a
paradigm of the Yin-Yang interplay between inflammation and cancer.
Oncogene 29, 3313-3323 (2010).
270 Markiewski, M. M. et al. The regulation of liver cell survival by complement. J
Immunol 182, 5412-5418 (2009).
271 McEarchern, J. A. et al. Engineered anti-CD70 antibody with multiple effector
functions exhibits in vitro and in vivo antitumor activities. Blood 109, 1185-
1192 (2007).
References
168
272 Langer, H. F. et al. Complement-mediated inhibition of neovascularization
reveals a point of convergence between innate immunity and angiogenesis.
Blood 116, 4395-4403 (2010).
273 Bora, P. S. et al. Role of complement and complement membrane attack
complex in laser-induced choroidal neovascularization. J Immunol 174, 491-497
(2005).
274 Cross, M. J. & Claesson-Welsh, L. FGF and VEGF function in angiogenesis:
signalling pathways, biological responses and therapeutic inhibition. Trends
Pharmacol Sci 22, 201-207 (2001).
275 Bergers, G. & Benjamin, L. E. Tumorigenesis and the angiogenic switch. Nat
Rev Cancer 3, 401-410 (2003).
276 Jagels, M. A., Daffern, P. J. & Hugli, T. E. C3a and C5a enhance granulocyte
adhesion to endothelial and epithelial cell monolayers: epithelial and endothelial
priming is required for C3a-induced eosinophil adhesion. Immunopharmacology
46, 209-222 (2000).
277 Albrecht, E. A. et al. C5a-induced gene expression in human umbilical vein
endothelial cells. Am J Pathol 164, 849-859 (2004).
278 Zwirner, J., Fayyazi, A. & Gotze, O. Expression of the anaphylatoxin C5a
receptor in non-myeloid cells. Mol Immunol 36, 877-884 (1999).
279 Sklar, L. A. et al. Signal transduction and ligand-receptor dynamics in the
human neutrophil. Transient responses and occupancy-response relations at the
formyl peptide receptor. J Biol Chem 260, 11461-11467 (1985).
280 Ponce, M. L. Methods in Molecular Medicine. Vol. 46 (Humana Press Inc.,
2001).
References
169
281 Langeggen, H., Johnson, E. & Hetland, G. Effects of C5a and FMLP on
interleukin-8 production and proliferation of human umbilical vein endothelial
cells. Inflammation 25, 83-89 (2001).
282 Kurihara, R. et al. C5a promotes migration, proliferation, and vessel formation
in endothelial cells. Inflamm Res 59, 659-666 (2010).
283 Floreani, A. A. et al. Expression of receptors for C5a anaphylatoxin (CD88) on
human bronchial epithelial cells: enhancement of C5a-mediated release of IL-8
upon exposure to cigarette smoke. J Immunol 160, 5073-5081 (1998).
284 Rutkowski, M. J., Sughrue, M. E., Kane, A. J., Mills, S. A. & Parsa, A. T.
Cancer and the complement cascade. Mol Cancer Res 8, 1453-1465 (2010).
285 Monk, P. N., Barker, M. D. & Partridge, L. J. Multiple signalling pathways in
the C5a-induced expression of adhesion receptor Mac-1. Biochim Biophys Acta
1221, 323-329 (1994).
286 Crane, J. W. et al. The C5a anaphylatoxin receptor CD88 is expressed in
presynaptic terminals of hippocampal mossy fibres. J Neuroinflammation 6, 34
(2009).
287 Christofori, G. Changing neighbours, changing behaviour: cell adhesion
molecule-mediated signalling during tumour progression. EMBO J 22, 2318-
2323 (2003).
288 Desgrosellier, J. S. & Cheresh, D. A. Integrins in cancer: biological implications
and therapeutic opportunities. Nat Rev Cancer 10, 9-22 (2010).
289 Andrades, J. A. et al. Complement proteins are present in developing
endochondral bone and may mediate cartilage cell death and vascularization.
Exp Cell Res 227, 208-213 (1996).
References
170
290 Sakiyama, H. et al. Immunolocalization of complement C1s and matrix
metalloproteinase 9 (92kDa gelatinase/type IV collagenase) in the primary
ossification center of the human femur. Cell Tissue Res 277, 239-245 (1994).
291 Kessenbrock, K., Plaks, V. & Werb, Z. Matrix metalloproteinases: regulators of
the tumor microenvironment. Cell 141, 52-67 (2010).
292 Egeblad, M. & Werb, Z. New functions for the matrix metalloproteinases in
cancer progression. Nat Rev Cancer 2, 161-174 (2002).
293 Ricklin, D. & Lambris, J. D. Complement-targeted therapeutics. Nat Biotechnol
25, 1265-1275 (2007).
294 Rodriguez de Cordoba, S., Harris, C. L., Morgan, B. P. & Llorca, O. Lessons
from functional and structural analyses of disease-associated genetic variants in
the complement alternative pathway. Biochim Biophys Acta 1812, 12-22 (2011).