Report

Impaired neutrophil functions in patients withleukocytoclastic vasculitisMarcelo H. Grunwald, MD, Ofra Shriker, BSC, Sima Halevy, MD, Michael Alkan, MD,and Rachel Levy, PhD

From the Clinical Biochemistry Unit,Infectious Diseases Institute, andDepartment of Dermatology, Facultyof Health Sciences, Soroka MedicalCenter of Kupat Holim, Ben-GurionUniversity of the Negev, Beer Sheva,israel

CorrespondenceRachel Levy, phDInfectious Diseases LaboratoryClinical Biochemistry Unit r.v,.--Faculty of Health SciencesBen-Gurion University of the NegevBeer Shevaisrael 84105

AbstractBackground Leukocytoclastic vasculitis (LV) is characterized by segmental inflammation

of small blood vessels, resulting in ischemic damage to the surrounding tissue. It is

considered to be related to a type III hypersensitivity reaction, although the exact etiologic

mechanism is not clear.

Objective The purpose of this study was to evaluate neutrophil functions in patients with

LV in order to understand their role in the pathogenesis of the disease.

Methods Neutrophil functions were examined in 25 LV patients. The patients were "*'

divided into two groups: Group A consisted of 14 patients with drug-induced LV and

Group B consisted of 11 patients where LV was induced by other factors.

Results Both groups of patients showed significantly reduced chemotaxis and

phagocytosis. Superoxide generation was significantly lower (P < 0.001) only in "

neutrophils from patients in Group A: 5.8 ± 0.5 nmoles 02/10^ cells/min compared to

9.08 ± 0.8 nmoles 02/10^ cells/min in the controls. Preincubation on normal neutrophils

with the patients' sera caused an increase in their superoxide generation in accordance

with the high IL-8 levels in these sera.

Conclusions Neutrophil functions were significantly impaired in patients with LV. It is

likely that factors present in LV plasma may chronically activate neutrophils, so that they

become refracfory to further stimulation. Our study showed that neutrophil superoxide

generation is low only in drug-induced LV; this test may assist in distinguishing such

patients from those with LV induced by other causes.

Neutrophils play a critical role in host defense against avariety of microbial pathogens.^ These rapidly moving,avidly phagocytic cells are essentially the first line of defenseagainst invasion of hody tissues. Engulfment of microorgan-istns induces phagocytic cells to undergo an ahrupt consump-tion of oxygen, leading to the production of microbicidaloxidants.^ The biochemical basis for this respiratory burst isthe activation of a superoxide-generating NADPH oxidase;'however, during altered physiologic states, neutrophils mayreact in a manner that leads to tissue injury,'* whereby theydischarge their toxic substances and induce endothelial celldestruction.

Leukocytoclastic vasculitis (LV) is a disease characterizedby segmental inflammation of blood vessels. It is consideredto be triggered by a type III hypersenstivitiy reaction, andhas been related to several causes such as drug reaction,infections, autoimmune disorders, malignancies, and other

disorders.5 It is characterized by vessel wall necrosis, inducedby a sequence of factors including activation of the comple-ment system by circulating immune complexes, recruitmentof neutrophils which release superoxide and lysosomalenzymes that induce auto-oxidative tissue injury at the site ofinflammation causing further damage to perivascular tissue.*Conflicting results of neutrophil functions have beenreported in diseases related to vasculitis, like enhancementof chemotaxis, phagocytosis and generation of oxygen inter-mediate in Behcet's disease,^ impairrnent of granulocytechemotaxis and phagocytosis in Kawasaky disease,^ increaseof superoxide production by systemic lupus erythematosus,'defective phagocytosis in rheumatoid arthritis,'" andBehcet's syndrome.''

The aim of the present investigation was to study neutro-phil functions in two etiologic settings of LV; patients with 509

© 1997 Blackwell Science Ltd International Journal of Dermatology 1997, 36, 509-513

510 Report Impaired neutrophil functions in patients with leukooytoclastiG vasculitis Grunwaid el a\.

drug-induced vasculitis and patients with vasculitis resultingfrom other causes.

Patients and methods

Patients

During the years 1989-1993, 25 patients were diagnosed as

suffering from ieukocytociastic vasculitis. The diagnosis was

based on ciinicai, histoiogic, and direct immunofluorescence

findings.

These patients were divided into two groups. Group A

consisted of 14 patients with drug-induced LV. The drugs most

commonly implicated were two types of penicillins (amoxycilliri

trihydrate and cloxacillin sodium) and paracetamol.

Hypersensitivity of these patients to the drugs was based on

clinical evidence which included the time relationship between

drug intake and the appearance of vasculitis and the

disappearance of skin lesions once the drug was discontinued.

Group B consisted of 11 patients: in three of the cases LV was

caused by streptococcal infection, in another three by

cryoglobulinemia, and in five cases no etiologic cause was

identified. Neutrophil functions were studied in 25 age and

sex-matched healthy controls.

Separation of neutrophils

Granulocytes (95% purity) were separated by Ficoll/Hypaque

centrifugation, dextran sedimentation, and hypotonic lysis of

erythrocytes.^^

Superoxide anion measurements

The production of superoxide anion (02~) by intact cells was

measured as the superoxide dismutase inhibltable reduction of

ferricytochrome c by a microtiter plate technique, as described

earlier.•'2 Cells were suspended (2 x 10^ cells/well) in 100 ^L

in Hanks' Balanced Salts Solution (HBSS) containing 150 ̂ M

ferricytochrome c and stimulated by the addition of 50 ng/mL

phorbol myristate acetate (PMA), 1 mg/mL opsonlzed zymosan

(OZ) or 10"'' M formylmethionylphenylalanine (FMLP) and the

reduction of acetyl ferricytochrome c was followed by a change

of absorbance at 550 nm every two minutes on a Thermomax

Microplate Reader (Menio Park, CA). The maximal rates of

superoxide generation were determined and expressed

as nmoles Og^/IO^ cells/10 min using Extinction Coefficient

E550 = 21 mM~̂ cm~^ Opsonlzed zymosan was prepared as

follows: 20 mg of zymosan (Sigma) was incubated with 1 mL of

human pooled serum for 1 h at 37°C and washed three times

with HBSS.

Leukocyte chemotaxis assay

The assay was performed as described earlier.''^ Agarose was

dissolved in sterile, distilled water by heating in a boiling water

bath for 10 min. After cooling to 48°C in a water bath, the

agarose was mixed with an equal volume of prewarmed 2 x

MEM (Bet-Haemek, Israel) supplemented to achieve a final ,«

concentration of 10% heat-inactivated fetal calf serum and *

7.5% sodium bicarbonate. Five milliliters of the agarose > S

medium were delivered to each 60 x i5-mm tissue culture dish

(3002, Falcon, Oxnard, CA) and allowed to harden. Six series

of three wells, 2.4 mm in diameter and spaced 2.4 mm apart,

were marked. The agarose plugs were plucked out using a ;

hypodermic needle. Neutrophils were suspended in MEM

(minimal essential medium). The center well of each three-well

series received a lO-pL volume of the cell suspension

containing 2.5 x 10^ of purified neutrophils. The outer well

received 10 mL of fMLP at 10'^ M concentration. The inner well

received control medium, MEM. The completed dishes were ^,

incubated at 37°C in a humidified atmosphere containing 5% c

CO2 in air. After incubation for 2 h, the plates were fixed by •

addition of 3 mL methanol at 4°C overnight or 30 min at room

temperature. After pouring off the methanol, the plates were

incubated with 2.5% glutaraldehyde for 30 min at room

temperature. The gel was removed intact after fixation and the

plates were stained with Wright's stain and air dried.

Quantitation of migration was done by measurement of the "'

linear distance the cells had moved from the margin of the well

toward the chemotactic factor (distance A\ chemotaxis) and the

linear distance the cells had moved from the margin of the well

toward the control medium (distance S: spontaneous si

migration). Chemotaxis was defined as A/B. ,̂ ,,

Phagocytosis

The assay was performed as described earlier:''^ 5 x 1 0 ^

neutrophil suspensions in MEM containing 10% inactivated

FCS (fetal calf serum) were incubated at 37°C for 10 min with

1 mg/mL of zymosan opsonized with pooled human serum.

Phagocytosis was determined under the microscope as the

percentage of neutrophils which phagocytized more than two

particles of OZ.

interieui<in-8 (iL-8) ievels in serum .-,,,,..';.iThe measurements of IL-8 levels in the serum were performedwith the Biokine Test kits (T Cell Sciences, Inc., Cambridge,MA).

•J

Statistics

The results are expressed as the mean ± SE. Comparisons of

data sets were performed using a one way ANOVA. The

Schoffe test was used to compare the means a posteriori when

the ANOVA test was statistically significant (P < 0.05).

Results . '

Figure 1 presents chemotaxis values of neutrophils frompatients with drug-induced LV (Group A) and patients withLV due to other causes (Group B). Neutrophils from thetwo groups of patients showed lower rates of chemotaxis

International Journal of Dermatology 1997, 36, 509-513 © 1997 Biackweii Science Ltd

Grunwald el al. Impaired neutrophii functions in patients with ieukocytociastic vasculitis Report 511

. * .

Figure 1 Chemotaxis of neutrophils from patients withvasculitis and matched controls. A, patients with druginduced vasculitis; B, patients with vasculitis from othercauses. Symbols: D, patients; • , controls. The differencebetween the values of chemotaxis of neutrophils for patientsfrom Group A and their matched controls was significant(P < 0.003) as was that between Group B patients and theircontrols (P < 0.005)

I!b" 8

Figure 3 The maximal rates of PMA stimulated superoxidegeneration in neutrophils from patients with vasculitis. Theconcentration of PMA was 50 ng/mL. The lines representthe mean ± SE. Symbols: D, patients; • , controls. Thedifference between the rates in neutrophils of patients fromGroup A and the matched controls was significant(P < o.ooi). The difference between the rates in neutrophilsof patients from Group B and the matched controls was notsignificant (P = 0.4)

^

Figure 2 Phagocytosis of OZ by neutrophils from patientswith vasculitis and matched controls. Symbols: D, patients;• , controls. The difference between the phagocytosis ofneutrophils of patients from the two groups and thematched controls was significant (P < o.ooi) for Group Aand (P < 0.001) for Group B

compared with the matched controls. The values weresimilar in both groups: 1.13 ± 0.2 (Group A) and 1.2 ± 0.3(Group B) and significantly lower in the matched controls:2.0 ± 0.2 and 2.0 ± 0.3, respectively (F < 0.003 '̂̂ dP < 0.005). Similarly, phagocytosis of OZ particles wassignificantly lower in neutrophils from patients with LV,51 ± 5% with P < O.OOI (Group A) and 59 ± 4% withP < 0.003 (Group B), compared with 67 ± 3% and70 ± 4% in the controls (Fig. 2).

Figure 3 presents the rates of neutrophii superoxideproduction in the two groups of patients. The generationof superoxide stimulated by 50 ng/ml of PMA wassignificantly lower (P < o.ooi) in neutrophils fromGroup A patients than that of the controls (5.8 ± 0.5and 9.08 ± 0.8 limoles OJio^ cells/min). In contrast,superoxide production in patients from Group B wassimilar to that of the control (8.1 ± i.o and9.Z ± 0.8 nmoles O^/io* cells/min, respectively) andhigher than that of Group A. Similar behavior wasobserved when superoxide generation was stimulatedwith I mg/mL OZ or lo"'' M FMLP, as shown in Table i.

Since superoxide generation was reduced only inneutrophils of drug-induced LV patients (amoxycillin

Tabie 1 Neutrophii superoxide generation in patients withIeukocytociastic vasculitis

Stimulant Superoxide generation(nmoles 02/10° neutrophils/min)

Group A Controls Group B Controls

PMA (50 ng/rtiL) 6.2 ± 0.5 9.25 ± 0.8SignificanceOZ (1 mg/mL)SignificanceFMLP (10-^ M)Significance

P < 0.0015.8 ± 0.45 8.2 ± 0.3

P < 0.0012.2 ± 0.4 3.7 + 0.5

P < 0.01 ^^

8.4 ±1 .0 9.35 + 1.2P = n.s. (0.4)

8.1 ± 0.7 8.32 ± 0.4P = n.s. (0.4)

3.4 ± 0.6 3.2 ± 0.7P = n.s. (0.4) 3

The values are the mean ± SE of the maximal rates ofsuperoxide generation in neutrophils of patients andmatched controls. f

trihydrate, cloxacillin sodium or paracetamol), we exam-ined the effect of these drugs on superoxide generationin neutrophils of healthy volunteers. As shown in Fig. 4,these drugs, added in vitro, caused a dose-dependentinhibition of neutrophii superoxide generation stimulatedby PMA. Similar results were obtained by stimulationwith OZ or FMLP (Table 2).

In order to evaluate the effect of the patients' sera onneutrophii functions, neutrophils from healthy donorswere incubated with sera from either patients with LVor from healthy controls for 3omin, washed, and thensuperoxide generation was measured. As shown inTable 2, preincubation of the neutrophils with sera fromthe two groups of patients caused an increase insuperoxide production, but the effect of the sera fromGroup B was significantly higher.

IL-8 levels were determined in the sera of the twogroups of patients with LV. The levels of IL-8 in thesera of both groups of patients is significantly higher

© 1997 Blackwell Science Ltd International Journal of Dermatology 1997, 36, 509-513

512 Report Impaired neutrophii functions in patients with Ieukocytociastic vasouiitis Grunwaid el ai.

100 -

80 -

60 ..

40 -

20 .

0

Oo

co

1oa>'x2I

100 -

80 -

60 -

40 -

20 -

0

100 -

80 -

60 -

40 -

20 -

0

0.25 2.5 10

Paracetamol

100 250

0.5 1 10 20 50

Cioxaciiiin (pg/mL)

100

0 0.5 1 10 20 50 100

Amoxyciilin trihydrate (|jg/mL)

Figure 4 The dose-dependent effect of paracetamol (A),cloxacillin (B) or amoxyciilin trihydrate (C) on neutrophilsuperoxide generation stimulated by 50 ng/mL PMA orI mg/mL OZ

(P < o.ooi) than that of the matched controls: 12.8 ± z.3pg/mL in Group A compared with i.z ± 0.5 pg/mL inthe controls, and 14.8 ± 4.1 pg/mL in Group B comparedwith i.o ±0 .1 pg/mL in the controls.

Tabie 2 The effect of preincubation of normal neutrophilswith sera from patients with Ieukocytociastic vasculitis onneutrophil superoxide generation s

Stimulant

Serum

Superoxide generation(nmoles 02/10° neutrophlls/nnln)

Group A Group B Control

PMA (50 ng/mL)Significance0 2 (1 mg/mL)SignificanceFMLP (10-^ M)Significance

10.2 ±0.5 13.25 ±0 .8 8.7 ±1 .2P < 0.05 P < 0.001

10.8 ± 0.45 12.9 ± 0.3 8.3 ± 0.5P < 0.05 P < 0.005

4.2 ± 0.4 5.7 ± 0.5 3.4 ± 0.5P < 0.05 P < 0.01

The values are the mean ± SE of the maximal rates ofsuperoxide generation in neutrophils incubated with sera offive patients from each group of five controls.Three different experiments were performed.The significance was determined between the patients and |the cont ro l s . •-•j<ff'y . ;;;- ;^-' ' ' i |

Discussion

LV is a disorder which causes segmental inflammationof hlood vessels resulting in ischemic damage to thesurrounding tissue. It is thought to be caused by animmune-complex mechanism, type II • hypersensitivityreaction. The elements necessary for the expression ofthis process include soluble immune complexes, increasedvascular permeability accompanied by passive adhesionof complexes in the vessel wall, activation of thecomplement system (mainly C5a) with subsequent attrac-tion of neutrophils to the site of immune complexdeposition, and release of inflammatory mediators. Allthese effects result in the destruction of vascular integrity.*Other potential mechanisms have been related to thepathogenesis of vasculitis, including aberrant regulationof T and B cells, monocyte macrophage and endothelialcell functions, interleukin-8 (lL-8), IFN-y and IL-6vasodilator substances, prostacyclin (PGI^), endothelialleukocyte adhesion molecule-i (ELAM-i), and vascularcell adhesion molecule (VCAM-i).5

The results of this study demonstrate that neutrophilsfrom patients with LV show decreased chemotaxis andphagocytosis. The nature of the defect in neutrophil func-tion in LV patients is not clear. We showed that incubationof neutrophils from healthy donors with serum of patientswith LV increased superoxide production (Table z), sug-gesting that the patients' serum induces priming of neutro-phils. Other studies have also shown increased hydroxylradical generation by normal neutrophils incubated in serafrom patients with LV.''' Circulating immune complexes inthose sera were implicated in the enhanced production of

International Journal of Dermatology 1997, 36, 509-513 © 1997 Blach;weli Science Ltd

t'-

Grunwaid et al. Impaired neutrophil functions in patients with leukooytoclastic vasoulitis Report 513

hydroxyl radicals. Oxidants produced by primed neutro-phils may cause auto-oxidation of the cells themselves,which results in the inhibition of phagocytosis, chemotaxisand respiratory burst. 5̂ Thus, it is possible that the primedneutrophils of patients with LV produce high levels ofoxidants which caused auto-oxidation, resulting in a tem-porary defect in chemotaxis and phagocytosis. In supportof our hypothesis, alteration in neutrophil functions in theacute phase of Sweet's Syndrome has been demonstrated.'^Preincubation of neutrophils with patients' sera caused anincrease in superoxide generation (Table z), indicating thatthe sera may prime the neutrophils. It is possible that, inthe early stages of the disease, patients may show enhancedneutrophil activity. The reduced activity shown in our studyreflects a late stage of the disease as a result of their auto.-oxidation.

The generation of superoxide was low only in neutrophilsof patients with drug-induced vasculitis. Thus, stimulatedsuperoxide generation may serve as a marker for patientswith drug-induced vasculitis. A possible explanation forthe reduced superoxide production in the neutrophils ofthese patients may be the effect of the drug treatment sincethese drugs caused a dose-response inhibition of superoxidegeneration in vitro by neutrophils from healthy donors(Fig. 4). Preincubation of neutrophils from healthy controlswith serum of patients with-drug induced LV (Group A),was less effective than with serum of patients from GroupB for increasing superoxide production, although the levelsof IL-8 were similar in the sera of the two groups of patients.

In conclusion, our study shows impairment in neutrophilfunction in patients with leukocytoclastic vasculitis. It islikely that factors present in LV plasma may chronicallyactivate neutrophils, which become refractory to furtherstimulation. Additional studies may elucidate the connec-tion between this phenomenon and the different etiologiesof vasculitis.

References1 Klebanoff SJ. Oxygen intermediates and microbicidal

events. In Van Furth R, ed., Mononuctear Phagocytes:Functional Aspects. Part 11. Boston: Martinus NijhoftPublishers, 1980: 1105-112.1.

2 Babior BM. The respiratory burst of phagocytes. / ClinInvest 1984; 73: 599-605.

3 Rossi SJ: The O^-forming NADPH oxidase of the

phagocytes nature: mechanisms and functions. BiocbimBiophys Acta 1986; 853: 65-89.

4 McCord JM, Fridovich I. The biology and pathology ofoxygen radicals. Ann Intern Med 1978; 89: 122-IZ7.

5 Gravaclese EM. In Fitzpatrick TB, Eisen AZ, Wolff K,eds, Dermatology in General Medicine, 4th edn. NewYork: McGraw-Hill, 1994: 2173-Z177.

6 Tosca N, Stratigos JD. Possible pathogenic mechanismsin allergic cutaneous vasculitis. Int J Dermatol 1988; 27:291-296.

7 O'Duffy JD, Lehner T, Barnes CG. Summary of theThird International Conference on Behcet's Disease,Tokyo, Japan, October 13-24, 1^81. J Rheumatol 1983;10: 154-158.

8 Ono S, Onimaru T, Kawakami K, et al. Impairedgranulocyte chemotaxis and increase circulatingimmunocomplexes in Kawasaki disease. / Pediatr 1985;106: 567-570.

9 Casellas AM, Prat A, Llera A, et al. Increased superoxideproduction by polymorphonuclear leukocytes in systemiclupus erythematosus. Clin Exp Rheumatol 1991; 9:511-514-

10 Marhoffer W, Stein M, Federlin K. Impaired function ofpolymorphonuclear neutrophilic granulocytes inrheumatoid arthritis. Infect Immun 1993; z i : 15-16.

11 Levy R, Rotrosen D, Nagauker O et al. Induction ofrespiratory burst in HL-60 cells: correlation of activityand protein expression. / Immunol 1990; 145:Z595-2601.

IZ Levy R, Malech HL, Rotrosen D. Production of myeloidcell cytosols functionally and immunochemicallydeficient in the 47 KDa or 67 KDa NADPH oxidasecytosolic factors. Biocbem Biophys Res Commun 1990;170: H14-1120.

13 Levy R, Schlaeffer F, Keynan A, etal. Increasedneutrophil function induced by bile duct ligation in a ratmodel. Hepatology 1993; 17: 908-914.

14 Miyachi Y, Yanase K, Imamura S, Niwa Y. Increasedhydroxylradical generation by normalpolymorphonuclear leukocytes incubated in sera frompatients with leukocytoclastic vasculitis. Arcb Dermatol1982; 274: 65-71.

15 Baehner RL, Boxer LA, Allen JM, DavisJ. Autooxidation as a basis for altered function bypolymorphonuclear leukocytes. Blood 1977; 50:

32.7-335-16 Driesh P, Simon M, Gomez RS, Horenstein OP.

Impairment of some granulocyte functions in Sweet'ssyndrome. Acta Derm Venereol (Stockbolm) 199Z; 7Z:109-111.

SPIRIT OF LAVENDER.H. J. MANFULL,

IpiBficushig luib ^nmili; CJjtmist,8, AlUvWRIGHT STREET, NOTTINGHAM.

From the collection of Lawrence CharlesParish, MD, Philadelphia, Pennsylvania.

© t997 Blackwell Science Ltd intemationai Journai of Dermatoiogy 1997, 36, 509-513