Endoplasmic reticulum stress enhances fibroticremodeling in the lungsWilliam E. Lawsona,b,1,2, Dong-Sheng Chenga,1, Amber L. Degrysea, Harikrishna Tanjorea, Vasiliy V. Polosukhina,Xiaochuan C. Xua, Dawn C. Newcomba, Brittany R. Jonesa, Juan Roldana,c,d, Kirk B. Lanea, Edward E. Morriseye,f,g,Michael F. Beerse, Fiona E. Yullh, and Timothy S. Blackwella,b,h,i

aDepartment of Medicine, Division of Allergy, Pulmonary and Critical Care Medicine, and Departments of iCell and Developmental Biology andhCancer Biology, Vanderbilt University School of Medicine, Nashville, TN 37232; bDepartment of Veterans Affairs Medical Center, Nashville, TN 37212;cFundació Institut d’Investigació Germans Trias i Pujol and dServei de Pneumologia, Hospital Santa Caterina, Calle Dr. Castany, 17190 Salt, Spain; andeDepartment of Medicine, Division of Pulmonary and Critical Care Medicine, fDepartment of Cell and Developmental Biology, and gPenn Institute forRegenerative Medicine, University of Pennsylvania School of Medicine, Philadelphia, PA 19104

Edited by Michael J. Welsh, Howard Hughes Medical Institute and University of Iowa, Iowa City, IA, and approved May 18, 2011 (received for review May11, 2011)

Evidence of endoplasmic reticulum (ER) stress has been found inlungs of patients with familial and sporadic idiopathic pulmonaryfibrosis. We tested whether ER stress causes or exacerbates lungfibrosis by (i) conditional expression of amutant form of surfactantprotein C (L188Q SFTPC) found in familial interstitial pneumoniaand (ii) intratracheal treatment with the protein misfolding agenttunicamycin. We developed transgenic mice expressing L188QSFTPC exclusively in type II alveolar epithelium by using the Tet-On system. Expression of L188Q SFTPC induced ER stress, as deter-mined by increased expression of heavy-chain Ig binding protein(BiP) and splicing of X-box binding protein 1 (XBP1) mRNA, but nolung fibrosis was identified in the absence of a second profibroticstimulus. After intratracheal bleomycin, L188Q SFTPC-expressingmice developed exaggerated lung fibrosis and reduced static lungcompliance compared with controls. Bleomycin-treated L188QSFTPC mice also demonstrated increased apoptosis of alveolar ep-ithelial cells and greater numbers of fibroblasts in the lungs. Witha complementary model, intratracheal tunicamycin treatment fail-ed to induce lung remodeling yet resulted in augmentation ofbleomycin-induced fibrosis. These data support the concept thatER stress produces a dysfunctional epithelial cell phenotype that fa-cilitates fibrotic remodeling. ER stress pathways may serve as im-portant therapeutic targets in idiopathic pulmonary fibrosis.

S100A4 | unfolded protein response

Idiopathic pulmonary fibrosis (IPF) is the most common andsevere form of idiopathic interstitial pneumonia (IIP). IPF

is characterized by dyspnea, decreased exercise tolerance, andprogression to respiratory failure as a result of ongoing fibroticremodeling of the distal lung parenchyma (1). Although the causeof IPF remains unknown, recent cases of familial interstitialpneumonia (FIP) have begun to shed light on potential patho-genic mechanisms. FIP, which represents a small proportion ofIIP, is defined as two or more biologically related family memberswith a diagnosis of IIP (2, 3). In FIP, 85% of biopsy-proven caseshave pathology consistent with usual interstitial pneumonia, thepathological equivalent of IPF (2). In 2002, we reported a largeFIP family with a heterozygous mutation in the carboxyl-terminalregion of surfactant protein C (SFTPC) (4). This exon 5 +128T→A transversion results in substitution of glutamine for leucineat amino acid 188 (L188Q) in the carboxyl-terminal region of thepro-SFTPCprecursor protein (pro-SP-C). This region of pro-SP-Cis known as the BRICHOS domain, which is essential for proteinfolding and processing. Mutations in proteins containing BRI-CHOS domains are linked to several degenerative and proliferativediseases through mechanisms related to altered posttranslationalprotein processing (5).In cultured alveolar epithelial cells (AECs), expression of

L188Q SFTPC results in a precursor protein that cannot befolded properly in the endoplasmic reticulum (ER), leading toER stress and activation of the unfolded protein response (UPR)

(4, 6, 7). We evaluated lung tissue from individuals with FIP whocarried the L188Q mutation and found up-regulation of ERstress markers in the alveolar epithelium (6). Subsequently, westudied lung tissue from individuals with FIP and sporadic IPFwithout mutations in SFTPC and noted that ER stress markerswere also present in the alveolar epithelium in the same pattern(6), a finding confirmed by other investigators (8). Therefore, itappears that ER stress and UPR activation are common featuresof the alveolar epithelium in IPF. These findings raise a numberof important and unresolved issues, including whether ER stresscauses or exacerbates fibrosis, and, if so, how ER stress in theepithelium regulates fibrotic remodeling. To address these issues,we developed a transgenic mouse model by using the Tet-Onsystem in which mutant L188Q SFTPC can be inducibly ex-pressed in type II AECs in the adult mouse. In this model, ex-pression of L188Q SFTPC in type II AECs resulted in ER stressand UPR activation; however, no fibrosis was seen in the absenceof a second profibrotic stimulus. In contrast, enhanced bleomy-cin-induced lung fibrosis was found in mice expressing L188QSFTPC, as well as in mice treated with the ER stress-inducingagent tunicamycin, in association with increased epithelial celldeath and increased fibroblast accumulation. Our data supportthe idea that dysfunctional type II AECs facilitate lung fibrosisthrough increased susceptibility to injury, leading to excessiveand dysregulated remodeling.

ResultsExpression of Mutant L188Q SFTPC Induces ER Stress in AECs. Wegenerated an expression construct under control of the (tet-O)7promoter containing human SFTPC (six exons and introns) with anexon 5+128T→A substitution and insertion of an 11-aamyc tag inexon 2 that is expressed in the mature peptide. We also createda construct in which the murine SFTPC promoter drives the re-verse tetracycline transactivator (rtTA), mSFTPC.rtTA. Cotrans-fection of mSFTPC.rtTA and (tet-O)7–L188Q SFTPC-myc led todoxycycline (Dox)-inducible expression of mutant pro-SP-C inA549 cells (Fig. S1A). Next, we purified mSFTPC.rtTA and (tet-O)7–L188Q SFTPC-myc constructs, as well as a construct ex-pressing a tetracycline-controlled transcriptional silencer (tTS)under the murine SFTPC promoter (mSFTPC.tTS) to prevent

Author contributions: W.E.L., D.-S.C., A.L.D., K.B.L., M.F.B., F.E.Y., and T.S.B. designedresearch; W.E.L., D.-S.C., A.L.D., H.T., V.V.P., X.C.X., D.C.N., B.R.J., J.R., F.E.Y., and T.S.B.performed research; E.E.M. and M.F.B. contributed new reagents/analytic tools; W.E.L.,D.-S.C., A.L.D., H.T., V.V.P., X.C.X., D.C.N., B.R.J., J.R., K.B.L., E.E.M., M.F.B., F.E.Y., and T.S.B.analyzed data; and W.E.L., D.-S.C., and T.S.B. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.1W.E.L. and D.-S.C. contributed equally to this work.2To whom correspondence should be addressed. E-mail: william.lawson@vanderbilt.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1107559108/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1107559108 PNAS Early Edition | 1 of 6

CELL

BIOLO

GY

basal leakiness of transgene expression (9) (Fig. S1B) and per-formed simultaneous pronuclear microinjection of the three DNAconstructs into fertilized oocytes from C57BL/6J mice. This strat-egy leads to tandem integration of constructs into a single in-tegration site in themajority of cases (10).We identified 4 founders(of 52 potential founders) that possessed all three constructs (Fig.S2A). Transgenic mice developed normally, appeared healthy, andbred well, transmitting all three transgenes to their progeny. Adultmice were givenDox in drinking water, andRNAwas isolated fromwhole-lung tissue. Evaluation by real-time RT-PCR revealed thatinduction of L188Q SFTPC expression did not affect native pro-SP-C mRNA levels and was ∼10-fold less than expression of theendogenous gene in the highest expressing founder line (Fig. S2B).Furthermore, mutant L188Q SFTPC was expressed for up to 6 mowith continuous exposure to Dox (Fig. S2C). Immunohistochem-istry (IHC) for the myc-tagged transgene demonstrated that mu-tant L188Q pro-SP-C expression localized exclusively to cells incorners of alveoli, consistent with the site and appearance of type IIAECs (Fig. S2D–F). No transgene expression was identified in theabsence of Dox treatment.Formalin-fixed, paraffin-embedded lung tissue sections were

immunostained for the ER stress markers heavy-chain Ig bindingprotein (BiP) and X-box binding protein 1 (XBP1). BiP is aprotein chaperone that assists with protein folding and increasesdramatically with protein accumulation and ER stress. XBP1 isa potent transactivator of UPR gene expression, regulating a va-riety of cellular functions (11, 12). After administration of Dox for1 wk, both BiP and XBP1 were found by IHC to be up-regulatedin type II AECs in L188Q SFTPC mice (Fig. 1 A–D). In addition,BiP protein and mRNA expression was increased in whole-lungtissue samples after Dox treatment (Fig. 1 E and F). XBP1 isnormally expressed as a full-length unspliced mRNA. When pro-tein accumulation induces ER stress, unspliced XBP mRNAundergoes inositol requiring enzyme 1 (IRE1)-dependent splic-ing, yielding a spliced XBP1 isoform that permits translation ofthe biologically active XBP1 protein. The ratio of the splicedisoform to total XBP1 mRNA can be used as a marker of theIRE1-mediated ER stress response (13). Dox-treated L188QSFTPC mice had evidence of increased XBP1 splicing in lungtissue, indicative of ER stress (Fig. 1 G and H). Together, thesestudies show that transgene expression in L188Q SFTPC micecauses ER stress localized to type II AECs.After Dox treatment for 1 wk, we isolated primary type II AECs

to determine trafficking of the myc-tagged mutant pro-SP-C.In type II AECs from L188Q SFTPC mice, immunofluorescencefor the myc tag colocalized with BiP, which is expressed exclu-sively in the ER (Fig. 2 A–C). In contrast, myc expression did notcolocalize with giantin, which is found in the Golgi apparatus (Fig.2 D–F). Thus, it appears that mutant L188Q SFTPC is principallylocalized to the ER with little presence in the Golgi, suggestingthat processing of mutant pro-SP-C is impaired. Consistent withthis idea, we have been unable to detect mature myc-tagged SP-Cin supernatant of primary type II AECs or in the airway of L188QSFTPC-expressing mice. Consistent with studies using whole-lungtissue, type II AECs from mice with L188Q SFTPC expressionexhibited greater BiP protein and mRNA expression and in-creased XBP1 splicing compared with type II AECs from WTmice (Fig. 2 G–J). In the absence of Dox treatment, expression ofBiP and XBP1 splicing in type II AECs from L188Q SFTPC micewere similar to WT cells, confirming the inducibility of ER stressin our model.

Expression of Mutant L188Q SFTPC Exacerbates Lung Fibrosis. Wetreated L188Q SFTPC mice for up to 6 mo with Dox to de-termine whether induction of ER stress was sufficient to causelung remodeling; however, lungs from these mice appeared his-tologically normal despite persistent transgene expression (Fig.S3). Although L188Q SFTPC expression alone did not result inlung fibrosis, we reasoned that a second stimulus might inducegreater fibrosis in L188Q SFTPC mice as a result of vulnerabletype II AECs. Thus, we turned to intratracheal (i.t.) bleomycin,

the most commonly used model of experimental lung fibrosis.Mutant L188Q SFTPC mice and WT controls were started onDox and, 1 wk later, received low-dose i.t. bleomycin (0.04 unit);3 wk later, lung fibrosis was increased in L188Q SFTPC mice, asdetermined by evaluation of trichrome-stained sections (Fig. 3A–C). Furthermore, total lung collagen was greater in L188QSFTPC mice compared with WT mice after bleomycin treatment(Fig. 3D). In previous studies, we have used S100A4 as a fibro-blast marker in the lungs (14–17). Therefore, we evaluated lungsections by IHC for S100A4 expression and noted greaternumbers of S100A4-positive fibroblasts in L188Q SFTPC micecompared with WT controls after bleomycin treatment (Fig. S4A and B and Fig. 3E). Similarly, L188Q SFTPC mice had greaternumbers of myofibroblasts positive for α-smooth muscle actin(αSMA) in lung parenchyma after bleomycin treatment than WTcontrols did (Fig. S4 C and D and Fig. 3F).

BiP

XB

P1

WT + Dox L188Q SFTPC + Dox

A B

C D

L188

QSF

TPC

L188

QSF

TPC

L188

QSF

TPC

WT

WT

WT

mar

ker

Spliced

Unspliced

G

FE

0

0.1

0.2

0.3

0.4

0.5

WT + Dox L188Q SFTPC +Dox

BiP

mR

NA

Exp

ress

ion

(2Δ

CT )

*

0

0.1

0.2

0.3

0.4

0.5

WT + Dox L188Q SFTPC +Dox

XB

P1

Spl

icin

g R

atio

*H

WT

WT

WT

L188

QSF

TPC

L188

QSF

TPC

L188

QSF

TPC

BiP

β-actin

Fig. 1. Expression of mutant L188Q SFTPC in vivo leads to ER stress in type IIAECs. (A and B) IHC for BiP in lung sections of WT (A) and L188Q SFTPC (B)mice, both treated with Dox for 1 wk. (C and D) IHC for XBP1 in lungs fromthe same mice. (Magnification in A–D: ×400.) Arrows point to immunostain-positive cells. (E) Western blot analysis on whole-lung lysates for BiP fromWT and L188Q SFTPC mice treated with Dox for 1 wk. β-Actin is shown asloading control. (F) Results of quantitative real-time RT-PCR for expression ofBiP mRNA from lungs of mice treated with Dox for 1 wk, normalized to thehousekeeping gene RPL19. (n = 4–5 per column; *P < 0.001 between col-umns.) (G) RT-PCR gel demonstrating splice variants for XBP1 mRNA in lungsof L188Q SFTPC and WT mice treated with Dox for 1 wk. (H) XBP1 splicinganalysis by densitometry. (n = 5 per column; *P < 0.01 between columns.)Graphical data are presented as mean ± SEM.

2 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1107559108 Lawson et al.

Additionally, we wanted to determine whether lung mechanicswere different at 3 wk after bleomycin treatment in L188QSFTPC mice. We found that static lung compliance was de-creased in both WT mice and L188Q SFTPC mice treated withbleomycin compared with their respective saline-treated con-trols. In the bleomycin-treated group, static lung compliance wasfurther reduced in L188Q SFTPC mice compared with WTcontrols (Fig. 3G). No differences were noted in static lungcompliance between the two groups in the absence of bleomycin.Although our findings indicated a profibrotic effect of L188Q

SFTPC expression, we wondered whether the impact of transgeneexpression was limited by the relatively low ratio of mutant SFTPCexpression compared with the endogenous gene. Thus, we crossedL188QSFTPCmicewithmice thatuse thehumanSFTPCpromoter(hSFTPC.rtTA) (18), in hopes of increasingmutant L188Qpro-SP-C expression. Dox-inducible mRNA expression of mutant pro-SP-C did increase approximately threefold (Fig. S5A), but themice didnot develop spontaneous fibrosis, and lung fibrosis after bleomycintreatment was not different between mutant L188Q SFTPC micewith or without hSFTPC.rtTA (Fig. S5B). These studies suggestthat level of transgene expression is not a limiting factor indetermining the phenotype of L188Q SFTPC-expressing mice.

Expression of L188Q SFTPC Increases AEC Apoptosis After BleomycinTreatment. Because high levels of ER stress have been linked toincreased epithelial cell apoptosis (6, 7, 19), we wanted to de-

termine whether ER stress in L188Q SFTPC mice was associatedwith increased AEC apoptosis. In the absence of bleomycin,TUNEL+ AECs were rare in Dox-treated L188Q SFTPC miceand were present in similar numbers compared with WT litter-mates. We have previously shown that the number of TUNEL+epithelial cells peaks in the lungs at 1 wk after i.t. bleomycin in-jection (17). At this time point, lung sections from L188Q SFTPCmice showed greater numbers of TUNEL+ AECs than lungsfrom WT controls did (Fig. 4 A–C), indicative of greater AECapoptosis. This finding indicates that AECs in L188Q SFTPC

Myc BiP MergedA B C

Myc Giantin MergedD E F

HG

I

WT

WT

WT

mar

ker

L188

QSF

TPC

L188

QSF

TPC

L188

QSF

TPC

Spliced

Unspliced

J

0.00

0.05

0.10

0.15

0.20

WT + Dox L188Q SFTPC +Dox

BiP

mR

NA

Exp

ress

ion

(2Δ

CT )

*

0

0.1

0.2

0.3

WT + Dox L188Q SFTPC +Dox

XB

P1

Spl

icin

g R

atio *

L188

QSF

TPC

L188

QSF

TPC

L188

QSF

TPC

WT

WT

WT

BiP

β-actin

Fig. 2. Mutant L188Q pro-SP-C localizes to the ER in type II AECs isolatedfrom L188Q SFTPC mice after in vivo Dox treatment for 1 wk. (A–C) Dualimmunofluorescence for themyc tag (red;A) and the ERmarker BiP (green; B)demonstrated strong colocalization on merged imaging (yellow; C). Blue,DAPI. (D–F) Dual immunofluorescence for the myc tag (red; D) and the Golgimarker giantin (green; E) demonstrated minimal colocalization on mergedimaging (yellow; F). Blue, DAPI. (G) Western blot analysis for BiP from type IIAECs isolated fromWT or L188Q SFTPCmice treatedwith Dox in vivo for 1 wk.β-Actin is shown as loading control. (H) Real-time RT-PCR for expression of BiPmRNA, normalized to the housekeeping gene RPL19. (n = 3 per column; *P <0.05 between columns.) (I) RT-PCR gel demonstrating a splice variant for XBP1in type II AECs. (J) XBP1 splicing analysis by densitometry. (n = 3 per column;*P < 0.05 between columns.) Graphical data are presented as mean ± SEM.

WT + Dox + Bleo L188Q SFTPC + Dox + Bleo

A B

0

1

2

3

4

WT + Bleo WT + Dox+ Bleo

L188QSFTPC +

Bleo

L188QSFTPC +

Dox + Bleo

Fibr

osis

Sco

re

DC *

0.0

0.5

1.0

1.5

2.0

WT + Bleo WT + Dox+ Bleo

L188QSFTPC +

Bleo

L188QSFTPC +

Dox + Bleo

S10

0A4

Fibr

obla

st S

core

0.0

0.5

1.0

1.5

2.0

WT + Bleo WT + Dox+ Bleo

L188QSFTPC +

Bleo

L188QSFTPC +

Dox + Bleo

αS

MA

Fib

robl

ast S

core **E F

0

50

100

150

200

250

Saline Bleo

RLL

Col

lage

n ( μ

g)

WT + DoxL188Q SFTPC + Dox

*

0

0.01

0.02

0.03

0.04

0.05

Saline Bleo

Sta

tic lu

ng c

ompl

ianc

e(m

l/cm

H2O

)

WTL188Q SFTPC

#

#*

G

0.0

0.5

1.0

1.5

2.0

Saline Bleo

Airw

ay re

sist

ance

(cm

H2O

.s/m

l)

WTL188Q SFTPC

H

Fig. 3. Mice expressing mutant L188Q SFTPC had greater lung fibrosis afteri.t. bleomycin (Bleo). (A and B) Trichrome-stained lung sections from Dox-treatedWTmice and L188Q SFTPCmice at 3 wk after 0.04 unit of i.t. bleomycin.(Magnification: ×100.) (C) Semiquantitative fibrosis scoring of trichrome-stained lung sections fromWT and L188Q SFTPCmice at 3 wk after 0.04 unit ofi.t. bleomycin. (n = 5 per group; *P < 0.05 compared with other groups.) Resultsare representative of three separate experiments. (D) Total lung collagencontent from right lower lobe (RLL) based on hydroxyproline microplate assayat 3 wk after 0.04 unit of i.t. bleomycin or saline. (n = 4–6 per group for salineand 10 per group for bleomycin; *P < 0.05 compared with other groups.) (E)Semiquantitative scoring of S100A4+ lung fibroblasts on immunostained lungsections from WT and L188Q SFTPCmice at 3 wk after i.t. bleomycin. (n = 5 pergroup; *P < 0.05 compared with other groups.) (F) Semiquantitative scoring ofαSMA+ lung fibroblasts on immunostained lung sections from WT and L188QSFTPC mice at 3 wk after i.t. bleomycin. (n = 5 per group; *P < 0.05 comparedwith other groups.) (G) Static lung compliance in WT and L188Q SFTPC-expressing mice treated with Dox at 3 wk after i.t. bleomycin. (n = 3 per groupfor saline and 5–8 per group for bleomycin; *P < 0.05 between bleomycin-treated groups; #P < 0.001 compared with respective saline-treated group.) (H)Airway resistance in WT and L188Q SFTPC-expressing mice treated with Dox at3 wk after i.t. bleomycin. Graphical data are presented as mean ± SEM.

Lawson et al. PNAS Early Edition | 3 of 6

CELL

BIOLO

GY

mice are more susceptible to bleomycin-induced injury. Further-more, active caspase-3 expression was increased in the lungs ofL188Q SFTPC mice compared with WT controls (Fig. 4 D andE), suggesting that caspase-dependent cell-death pathways areselectively activated in the presence of ER stress in this model.ER stress has been linked to apoptosis through activation of ER-bound caspase-12 in mice (caspase-4 in humans) (20, 21). ByWestern blot analysis from whole-lung lysates, we noted thatcaspase-12 expression was increased in L188Q SFTPC mice at 1wk after bleomycin treatment compared with WT controls (Fig. 4F and G). Apoptosis mediated through expression of CCAAT/enhancer-binding protein homologous transcription factor(CHOP) has also been linked to ER stress (22), but we did notidentify increased CHOP expression in our model.Interestingly, we found that BiP mRNA expression and XBP1

splicing were increased in lungs of WT mice at 1 wk after bleo-mycin treatment (Fig. S6). In addition, BiP expression was fur-ther increased in the lungs of L188Q SFTPC mice treated withbleomycin in addition to Dox, although XBP1 splicing was un-changed. Together, these findings indicate that bleomycin itselfinduces ER stress in the lungs, potentially contributing to thepathological lung remodeling in this model.Next, we asked whether L188Q SFTPC mice had altered lung

inflammation after bleomycin. At 1 wk after bleomycin treatment,L188Q SFTPC and WT mice had similar total cells, neutrophils,macrophages, and lymphocytes in bronchoalveolar lavage (Fig.5A). To further evaluate inflammatory signaling in the lungs, wecrossed L188Q SFTPC mice to NF-κB/GFP/luciferase (NGL)reporter mice. The NGL reporter mouse expresses luciferase asa function of NF-κB activation, providing a surrogate readout ofinflammatory pathway activation (23). With Dox alone, neitherWT/NGL nor L188Q SFTPC/NGL mice had increased luciferase

expression above baseline, and values were similar between thetwo groups. After bleomycin treatment, luciferase expression in-creased similarly in both groups (Fig. 5B and Fig. S7). Further-more, whole-lung luciferase levels were similar between WT/NGLand L188Q SFTPC/NGLmice after 14 d of Dox alone (23.4 ± 22.5vs. 22.5± 11.7 relative light units per μg/μL) and after 14 d ofDox+bleomycin (285.7 ± 53.4 vs. 297.3 ± 58.7 relative light units per μg/μL). Thus, we found no evidence for differences in bleomycin-induced inflammation in L188Q SFTPC-expressing mice.

Induction of ER Stress via Tunicamycin Administration EnhancesBleomycin-Induced Lung Fibrosis. Although our data suggest thatER stress is the underlying process by which L188Q SFTPC ex-pression contributes to lung fibrosis, we sought to determinewhether ER stress induced by other means could lead to a similarphenotype. Among ER stress-inducing agents, tunicamycin, anantibiotic that induces ER stress by blocking N-linked proteinglycosylation, is themost commonly used (24).We administered i.t.tunicamycin (20 μg/mL in 100 μL of 20%DMSOdiluted in PBS) toWT (C57BL/6J background) mice. IHC on sections of lungs har-vested 2 d later showed that BiP and XBP1 were induced mostprominently in AECs after tunicamycin treatment (Fig. S8 A–D).Tunicamycin also induced BiP protein and mRNA expression andXBP1 mRNA splicing in whole-lung samples (Fig. S8 E–H). Aswith L188Q SFTPC expression, there was no evidence of lung fi-brosis or architectural change 3 wk after a single dose of tunica-mycin or after twice-weekly tunicamycin administration for 3 wk.We then tested the combination of tunicamycin followed bybleomycin treatment. WT mice received i.t. tunicamycin (or vehi-cle control) followed 48 h later by i.t. bleomycin (0.04 unit). At 2 wkafter bleomycin injection, the tunicamycin + bleomycin group hadmuch greater lung fibrosis than the vehicle + bleomycin group(Fig. 6 A–D). Thus, tunicamycin-induced ER stress resulted ina similar profibrotic phenotype as in the L188Q SFTPC model,strongly supporting the idea that ER stress facilitates lung fibrosis.

DiscussionEvaluation of IPF lung biopsies reveals evidence of ER stress inAECs lining the areas of fibrosis (6, 8), but the degree to whichER stress contributes to disease pathogenesis remains undefined.Although SFTPC mutation-associated interstitial lung disease israre in the broad scope of IPF (25), modeling such mutationsmay serve as a paradigm to better understand IPF and the role ofER stress. With these studies, we have shown that expression ofmutant L188Q SFTPC results in ER stress in type II AECsin vivo, leading to a vulnerable type II AEC population that ishighly susceptible to the effects of bleomycin. With expression ofL188Q SFTPC and resultant ER stress, AECs were more prone

A B C

*

E

WT L188Q SFTPC

Cas

pase

-12

band

den

sity

0.00.20.40.60.81.01.21.41.61.8

*F G

D

0.00.5

1.01.5

2.02.5

3.03.5

WT L188Q SFTPC

TUN

EL

Sco

re

WT L188Q SFTPC

Activ

e ca

spas

e-3/

tota

l cas

pase

-3 b

and

dens

ity

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4 *

Fig. 4. Mice expressing L188Q SPTPC had greater AEC death after i.t. bleo-mycin. (A and B) TUNEL-stained lung sections from WT (A) and L188Q SFTPC(B) mice at 1 wk after i.t. bleomycin. Arrows point to representative TUNEL+cells. (Magnification: ×600.) (C) Semiquantitative evaluation of TUNEL+ AECson lung sections fromWT and L188Q SFTPCmice at 1 wk after i.t. bleomycin. (n= 8–11 per group; *P < 0.05 between groups.) (D and E) Western blot analysis(D) and densitometry (E) for caspase-3 using whole-lung lysates from bleo-mycin-treated WT and L188Q SFTPC mice. β-Actin was used as a loading con-trol. Graphical data represent the ratio of band densities for active caspase-3and total caspase-3 (WT normalized to value of 1). (n = 5 per group; *P < 0.05compared with WT.) (F and G) Western blot analysis (F) and densitometry (G)for caspase-12 from lung lysates. Graphical data represent the band density ofcaspase-12 adjusted for β-actin (WT normalized to value of 1). (n = 5 per group;*P < 0.05 compared with WT.) Graphical data are presented as mean ± SEM.

A

0

20

40

60

80

100

Total Macrophages Lymphocytes Neutrophils

BA

L ce

ll co

unt (

x10,

000) WT+Dox

L188Q SFTPC+Dox

B

02468

101214

Baseline onDox

Dox + Bleo3 days

Dox + Bleo7 days

Pho

ton

coun

ts (x

107 )

WT NGLL188Q SFTPC NGL

Fig. 5. Mutant L188Q SFTPC expression does not enhance lung inflammationafter i.t. bleomycin (Bleo). (A) Total and differential cell counts in bron-choalveolar lavage (BAL) from WT and L188Q SFTPC mice at 1 wk after i.t.bleomycin. (n = 5–6 per group). (B) Bioluminescence imaging over the thoraxas an indicator of NF-κB activation in lungs of WT and L188Q SFTPC micecrossed with NGL mice. Baseline photon counts represent mice with Doxtreatment for 14 d. Subsequently, mice were treated with bleomycin andimaged at 3 and 7 d. Photon counting was performed after i.p. injection ofluciferin (1mg). (n= 4 per group.) Graphical data are presented asmean± SEM.

4 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1107559108 Lawson et al.

to apoptosis after bleomycin treatment, and greater numbers oflung fibroblasts were observed, findings that could explain en-hanced fibrosis in the mutant L188Q SFTPC mice. In addition,the fact that a similar phenotype was observed when tunicamycinwas administered provides a confirmatory model that ER stressin the alveolar epithelium predisposes this cell population togreater injury, driving enhanced fibrosis.Multiple cases of pediatric and adult interstitial lung di-

sease have been linked to SFTPC mutations (4, 25–29), includ-ing reports of a mutation that deletes exon 4 and its 37 aa(SFTPCΔexon4 mutation) (26) and the L188Q mutation analyzedhere (4). Both mutations have been analyzed in vitro, revealingthat expression of these mutant pro-SP-C forms results in proteinaccumulation in the ER, ER stress, UPR activation, and in-creased apoptosis through activation of ER-associated caspase-12/4 (6, 7, 19). In the first evaluation of an SFTPC mutationin vivo, Bridges et al. designed a transgenic mouse expressingmutant SFTPCΔexon4 under the human SFTPC promoter, whichwas expressed during embryogenesis and resulted in ER stressand in utero lethality (30). Although transgene expression in ourmodel was restricted to adulthood and bypassed the issue ofeffects on lung development, it is likely that the SFTPCΔexon4

mutation is more severe than the L188Q mutation becausepatients with this mutation exclusively present with interstitiallung disease in early childhood.With protein accumulation in the ER, UPR pathways are ac-

tivated in an attempt to control ER stress and protect the cell (11).The UPR has three main pathways governed by three ER trans-membrane proteins: IRE1, activating transcription factor 6(ATF6), and PKR-like ER kinase (PERK) (11). In the absence ofER stress, these sensors are maintained in their inactive statebecause of binding by BiP. With ER stress, BiP is released to serveas a folding chaperone, allowing activation of each sensor (31).IRE1 activation allows its intrinsic endonuclease activity to spliceXBP1 to its active state, leading to expression of UPR-mediated

protein degradation enzymes such as ER degradation-enhancingα-mannosidase–like protein (EDEM) and chaperone proteins(12). ATF6 activation leads to up-regulation of protein-foldingchaperone proteins such as BiP (32) and to the increased ex-pression of XBP1 (12). Activation of PERK leads to a phosphor-ylated eukaryotic initiation factor 2α (p-eIF2α)–dependent globalattenuation of protein translation and an ATF4-dependent ex-pression of redox and metabolism proteins (33). Together, thesethree UPR pathways are designed to attenuate ER stress. How-ever, with severe or prolonged ER stress, UPR mechanisms canlead to activation of apoptosis pathways (11). ER stress can triggerapoptosis by induction of CHOP/GADD153, via activation ofcaspase-12/4, and through phosphorylation of eIF2α. In ourstudies with L188Q SFTPC mice and in humans with IPF (6), wehave not identified increased p-eIF2α. In addition, we did not findup-regulation of CHOP.We did, however, note increased caspase-12 levels in the lung after bleomycin treatment in mutant L188QSFTPC mice compared with littermate controls. ER-bound cas-pase-12 (and its human homolog, casapse-4) has been linked toER stress-induced apoptosis (20, 21). Previous studies by Mulu-geta et al. detailed that mutant forms of SFTPC, including L188Q,lead to caspase-4–mediated caspase-3 activation and cell apo-ptosis in vitro (7), a finding complementary to results from ourin vivo model. AEC apoptosis has been linked to fibrosis in bothhuman IPF lung biopsy samples (34) and in animal models (35),including in a recent study by Sisson et al. in which directed ex-pression of diphtheria toxin by means of the SFTPC promoter ledto type II AEC apoptosis and lung fibrosis (36). Thus, increasedapoptosis may be one of the mechanisms contributing to exuber-ant fibrosis in L188Q SFTPC mice.In our models, expression of L188Q SFTPC and treatment with

tunicamycin induce ER stress but do not result in fibroticremodeling, suggesting that murine type II AECs are able tomanage the UPR response without obvious pathology. Rather,ER stress likely places the type II AEC population in a vulnerablestate in which a second stimulus, in this case bleomycin, exertsa prominent effect, leading to enhanced lung fibrosis. Given thefact that patients with FIP, including many with SFTPC muta-tions, frequently present with interstitial lung disease in adult-hood, one could infer that a “second hit” might be required toinduce clinical disease in at-risk individuals. In FIP (and sporadicIPF), the relevant injurious stimulus likely results from an envi-ronmental exposure, but the nature of this agent(s) is currentlyunknown. To date, only a history of cigarette smoking and thepresence of herpesvirus antigens in the lungs have been associ-ated with clinical disease in FIP. Hopefully, future studies withL188Q SFTPC-expressing mice or other genetic models treatedwith relevant environmental stimuli will help clarify the potentialsecond hits that contribute to lung fibrosis.In summary, our studies demonstrate that ER stress in AECs

predisposes the lung to greater injury and fibrosis in experi-mental models. These findings provide mechanistic informationsupporting an important role for ER stress, which is prominentin the lungs of individuals with IPF, in disease pathogenesis. ERstress and downstream UPR pathways may serve as therapeutictargets for future interventions in this devastating lung disease.

Materials and MethodsTransgenic Mice. To generate transgenic mutant L188Q SFTPC mice, threeconstructs were coinjected at the Vanderbilt Transgenic/ES Shared Resourcefacility (Fig. S1): (i) (tet-O)7–L188Q SFTPC-myc, (ii) mSFTPC.rtTA, and (iii)mSFTPC.tTS (37). Transgenic mice expressing rtTA under the human SFTPCpromoter (hSFTPC.rtTA) were obtained from Jackson Laboratories. NGLmice, which have consensus NF-κB binding sites upstream of the HSV mini-mal thymidine kinase promoter driving the GFP/luciferase construct, havebeen described previously (23). All mice were C57BL/6J background andentered experiments at 8–10 wk of age. All mouse experiments were ap-proved by the Vanderbilt Institutional Animal Care and Use Committee.

Animal Model and Drug Administration. Bleomycin (0.04 unit) (Bedford Lab-oratories) was injected i.t. in mice by intubation as previously described (17).Lungs were harvested for histology, frozen tissue, bronchoalveolar lavage,

WT + Vehicle + Bleo WT + TM + Bleo

BA

DC*

0

1

2

3

4

Vehicle + Bleo TM + Bleo

Fibr

osis

Sco

re

*

0

100

200

300

400

500

Saline Bleo

Rig

ht L

ung

Col

lage

n ( μ

g) VehicleTM

Fig. 6. Tunicamycin (TM) enhances lung fibrosis induced by bleomycin (Bleo).(A and B) Trichrome-stained lung sections at 2 wk after bleomycin from WTmice treated with a single i.t. injection of vehicle (DMSO; A) or tunicamycin (B)at 48 h before bleomycin. (Magnification: ×100.) (C) Semiquantitative fibrosisscoring of trichrome-stained lung sections from mice exposed to vehicle +bleomycin (vehicle + Bleo) compared with tunicamycin + bleomycin (TM +Bleo) at 2 wk after i.t. bleomycin. (n = 5–6 per group; *P < 0.05 betweengroups.) (D) Total collagen content of entire right lung from mice exposed tovehicle or tunicamycin followed by i.t. administration of saline or bleomycin.(n = 5 per group for saline and 7–10 per group for bleomycin; *P < 0.05compared with other groups.) Graphical data are presented as mean ± SEM.

Lawson et al. PNAS Early Edition | 5 of 6

CELL

BIOLO

GY

or cell isolation as previously described (14–17). L188Q SFTPC mice andcontrols were maintained on normal water ad libitum until transgene acti-vation was desired. Then, mice were given Dox (Sigma-Aldrich) in sterilewater (2 g/dL with additional 2% sucrose).

Tunicamycin (Sigma) was dissolved in DMSO and diluted to 20 μg/mL in20% DMSO diluted in PBS, and then 100 μL of solution was delivered to miceby i.t. intubation. A similar volume of 20% DMSO diluted in PBS was used asvehicle control.

Lung Sample Processing and Analysis. Formalin-fixed lung sections wereprepared as previously described (14–16). IHC (14, 15), immunofluorescence(16, 17), TUNEL (15, 17), and bronchoalveolar lavage with cell counts (15, 17)were performed as previously described. RNA isolation, real-time RT-PCR,and Western blot analysis were performed by using standard techniques(detailed in SI Materials and Methods).

Cell Culture/Isolation. Plasmids were transfected into A549 cells with anEffectene transfection kit (Qiagen) per the manufacturer’s instructions. Cellswere incubated with the transfection complexes under normal growthconditions for 4 h, and then 0.5–1.0 μg/mL Dox was added to the medium.After 24 h, cells were harvested for Western blot analysis. Type II AECs wereisolated as previously described (14, 16).

In Vivo Bioluminescence and Luciferase Measurements. Live bioluminescenceimaging was performed after i.p. injection of luciferin (1 mg in 100 μL of

saline), and whole-lung tissue luciferase levels were determined as pre-viously described (23).

Measurement of Airway Resistance and Compliance. For airway resistance andcompliance measurements, mice were anesthetized with i.p. pentobarbital,and tracheas were cannulated with a 20-gauge metal stub adapter. Eachmouse was placed on a small-animal ventilator, flexiVent (SCIREQ), with 150breaths per min and a tidal volume of 10 mL/kg of body weight. Airwayresistance (cm of H2O/mL per s) and static lung compliance (using a 2-s breathpause) (mL/cm of H2O) were determined with the manufacturer’s software.

Semiquantitative Scoring and Collagen Content. Scoring of lung fibrosis (15,17), fibroblasts (15), and TUNEL staining (9) and determination of collagencontent by hydroxyproline assay (38) were performed as previously described.

Statistics. Statistical analyses were performed with InStat (GraphPad Soft-ware). Differences among groups were assessed with one-way ANOVA andbetween pairs with Student’s t test. Results are presented as mean ± SEM.P values <0.05 were considered significant.

Detailed methods are available in SI Materials and Methods.

ACKNOWLEDGMENTS. We thank Linda A. Gleaves and Frank B. McMahonfor work in our histology core laboratory and the Vanderbilt Transgenic/ESShared Resource facility for assistance with generation of the L188QSFTPC mice.

1. American Thoracic Society European Respiratory Society (2000) Idiopathic pulmonaryfibrosis: Diagnosis and treatment. International consensus statement. Am J Respir CritCare Med 161:646–664.

2. Steele MP, et al. (2005) Clinical and pathologic features of familial interstitialpneumonia. Am J Respir Crit Care Med 172:1146–1152.

3. LawsonWE, Loyd JE (2006) The genetic approach in pulmonary fibrosis: Can it provideclues to this complex disease? Proc Am Thorac Soc 3:345–349.

4. Thomas AQ, et al. (2002) Heterozygosity for a surfactant protein C gene mutationassociated with usual interstitial pneumonitis and cellular nonspecific interstitialpneumonitis in one kindred. Am J Respir Crit Care Med 165:1322–1328.

5. Beers MF, Mulugeta S (2005) Surfactant protein C biosynthesis and its emerging rolein conformational lung disease. Annu Rev Physiol 67:663–696.

6. Lawson WE, et al. (2008) Endoplasmic reticulum stress in alveolar epithelial cells isprominent in IPF: Association with altered surfactant protein processing andherpesvirus infection. Am J Physiol Lung Cell Mol Physiol 294:L1119–L1126.

7. Mulugeta S, et al. (2007) Misfolded BRICHOS SP-C mutant proteins induce apoptosisvia caspase-4- and cytochrome c-related mechanisms. Am J Physiol Lung Cell MolPhysiol 293:L720–L729.

8. Korfei M, et al. (2008) Epithelial endoplasmic reticulum stress and apoptosis insporadic idiopathic pulmonary fibrosis. Am J Respir Crit Care Med 178:838–846.

9. Cheng DS, et al. (2007) Airway epithelium controls lung inflammation and injurythrough the NF-&kappaB pathway. J Immunol 178:6504–6513.

10. Lee CG, et al. (2004) Early growth response gene 1-mediated apoptosis is essential fortransforming growth factor &beta1-induced pulmonaryfibrosis. J ExpMed 200:377–389.

11. Schröder M, Kaufman RJ (2005) The mammalian unfolded protein response. AnnuRev Biochem 74:739–789.

12. Yoshida H, Matsui T, Yamamoto A, Okada T, Mori K (2001) XBP1 mRNA is induced byATF6 and spliced by IRE1 in response to ER stress to produce a highly activetranscription factor. Cell 107:881–891.

13. Shang J (2005) Quantitative measurement of events in the mammalian unfoldedprotein response. Methods 35:390–394.

14. Lawson WE, et al. (2005) Characterization of fibroblast-specific protein 1 inpulmonary fibrosis. Am J Respir Crit Care Med 171:899–907.

15. Lawson WE, et al. (2005) Increased and prolonged pulmonary fibrosis in surfactantprotein C-deficient mice following intratracheal bleomycin. Am J Pathol 167:1267–1277.

16. Tanjore H, et al. (2009) Contribution of epithelial-derived fibroblasts to bleomycin-induced lung fibrosis. Am J Respir Crit Care Med 180:657–665.

17. Degryse AL, et al. (2010) Repetitive intratracheal bleomycin models several featuresof idiopathic pulmonary fibrosis. Am J Physiol Lung Cell Mol Physiol 299:L442–L452.

18. Perl AK, Wert SE, Nagy A, Lobe CG, Whitsett JA (2002) Early restriction of peripheraland proximal cell lineages during formation of the lung. Proc Natl Acad Sci USA 99:10482–10487.

19. Mulugeta S, Nguyen V, Russo SJ, Muniswamy M, Beers MF (2005) A surfactant proteinC precursor protein BRICHOS domain mutation causes endoplasmic reticulum stress,proteasome dysfunction, and caspase 3 activation. Am J Respir Cell Mol Biol 32:521–530.

20. Hitomi J, et al. (2004) Involvement of caspase-4 in endoplasmic reticulum stress-induced apoptosis and Aβ-induced cell death. J Cell Biol 165:347–356.

21. Hitomi J, et al. (2004) Apoptosis induced by endoplasmic reticulum stress depends onactivation of caspase-3 via caspase-12. Neurosci Lett 357:127–130.

22. Ma Y, Brewer JW, Diehl JA, Hendershot LM (2002) Two distinct stress signalingpathways converge upon the CHOP promoter during the mammalian unfoldedprotein response. J Mol Biol 318:1351–1365.

23. Everhart MB, et al. (2006) Duration and intensity of NF-&kappaB activity determinethe severity of endotoxin-induced acute lung injury. J Immunol 176:4995–5005.

24. Reimertz C, Kögel D, Rami A, Chittenden T, Prehn JH (2003) Gene expression duringER stress-induced apoptosis in neurons: Induction of the BH3-only protein Bbc3/PUMAand activation of the mitochondrial apoptosis pathway. J Cell Biol 162:587–597.

25. Lawson WE, et al. (2004) Genetic mutations in surfactant protein C are a rare cause ofsporadic cases of IPF. Thorax 59:977–980.

26. Nogee LM, et al. (2001) A mutation in the surfactant protein C gene associated withfamilial interstitial lung disease. N Engl J Med 344:573–579.

27. Crossno PF, et al. (2010) Identification of early interstitial lung disease in an individualwith genetic variations in ABCA3 and SFTPC. Chest 137:969–973.

28. van Moorsel CH, et al. (2010) Surfactant protein C mutations are the basis ofa significant portion of adult familial pulmonary fibrosis in a Dutch cohort. AmJ Respir Crit Care Med 182:1419–1425.

29. Nogee LM, et al. (2002) Mutations in the surfactant protein C gene associated withinterstitial lung disease. Chest 121(3, Suppl):20S–21S.

30. Bridges JP, Wert SE, Nogee LM, Weaver TE (2003) Expression of a human surfactantprotein C mutation associated with interstitial lung disease disrupts lungdevelopment in transgenic mice. J Biol Chem 278:52739–52746.

31. Bertolotti A, Zhang Y, Hendershot LM, Harding HP, Ron D (2000) Dynamic interactionof BiP and ER stress transducers in the unfolded-protein response. Nat Cell Biol 2:326–332.

32. Wang Y, et al. (2000) Activation of ATF6 and an ATF6 DNA binding site by theendoplasmic reticulum stress response. J Biol Chem 275:27013–27020.

33. Harding HP, Zhang Y, Bertolotti A, Zeng H, Ron D (2000) Perk is essential fortranslational regulation and cell survival during the unfolded protein response. MolCell 5:897–904.

34. Uhal BD, et al. (1998) Alveolar epithelial cell death adjacent to underlyingmyofibroblasts in advanced fibrotic human lung. Am J Physiol 275:L1192–L1199.

35. Hagimoto N, Kuwano K, Nomoto Y, Kunitake R, Hara N (1997) Apoptosis andexpression of Fas/Fas ligand mRNA in bleomycin-induced pulmonary fibrosis in mice.Am J Respir Cell Mol Biol 16:91–101.

36. Sisson TH, et al. (2010) Targeted injury of type II alveolar epithelial cells inducespulmonary fibrosis. Am J Respir Crit Care Med 181:254–263.

37. Zhu Z, Ma B, Homer RJ, Zheng T, Elias JA (2001) Use of the tetracycline-controlledtranscriptional silencer (tTS) to eliminate transgene leak in inducible overexpressiontransgenic mice. J Biol Chem 276:25222–25229.

38. Brown S, Worsfold M, Sharp C (2001) Microplate assay for the measurement ofhydroxyproline in acid-hydrolyzed tissue samples. Biotechniques 30:38–40, 42.

6 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1107559108 Lawson et al.

Supporting InformationLawson et al. 10.1073/pnas.1107559108SI Materials and MethodsDevelopment of Mutant L188Q Surfactant Protein C (SFTPC) Mice.The plasmid containing mutant L188Q SFTPC was generatedas previously described (1, 2). The L188Q SFTPC construct wascloned into the EcoRV site of a modified pBluescript II SK ex-pression vector (pBSII KS/Asc). This vector contains a (tet-O)7-CMV promoter together with bovine growth hormone poly-adenylation sequences. The final plasmid, (tet-O)7–L188QSFTPC-myc–BGH.polyA, was verified by sequencing. A plasmidcontaining murine SFTPC.rtTA (mSFTPC.rtTA) was providedby E.E.M. To prevent basal leakiness, we used a third constructexpressing a tetracycline-controlled transcriptional silencer (tTS)under control of the murine SFTPC promoter (3). We purifiedthese three constructs with a GELase Agarose Gel-DigestingPreparation Kit (Epicentre) following the manufacturer’s in-structions, and these constructs were coinjected at the VanderbiltTransgenic/ES Shared Resource facility to generate transgeniclines of mice (C57BL/6 background) with integration of all threeconstructs. Genotyping of founder animals was performed bySouthern blot analysis, and later stages of genotyping were per-formed by PCR.Transgenic mice expressing the reverse tetracycline trans-

activator (rtTA) under the human SFTPC promoter (hSFTPC.rtTA) were obtained from Jackson Laboratories. NF-κB/GFP/luciferase (NGL)mice, which have consensus NF-κB binding sitesupstream of the HSVminimal thymidine kinase promoter drivingthe GFP/luciferase construct, have been described previously (4).All transgenic and WT mice for these experiments were in aC57BL/6J background and entered experiments at 8–10 wk ofage. All mouse experiments were approved by the VanderbiltInstitutional Animal Care and Use Committee.

Antibodies. The following primary antibodies were used in theseinvestigations: pro-SFTPC precursor protein (pro-SP-C) goatpolyclonal antibody (Santa Cruz Biotechnology); S100A4 rabbitpolyclonal antibody (obtained from Eric Neilson, VanderbiltUniversity); α-smooth muscle actin (αSMA) rabbit polyclonalantibody (Abcam); heavy-chain Ig binding protein (BiP) goalpolyclonal antibody (Santa Cruz Biotechnology); X-box bindingprotein 1 (XBP1) rabbit polyclonal antibody (Santa Cruz); giantinrabbit polyclonal antibody (Abcam); myc mouse monoclonal an-tibody (Invitrogen Life Technologies); myc goat polyclonal anti-body (Santa Cruz Biotechnology); active caspase-3 rabbitpolyclonal antibody (Millipore); β-actin rabbit polyclonal antibody(Sigma); caspase-12 rabbit polyclonal antibody (Cell SignalingTechnology); and CCAAT/enhancer-binding protein homologoustranscription factor (CHOP) rabbit polyclonal antibody (CellSignaling Technology).

Bleomycin Model. Bleomycin (0.04 unit; Bedford Laboratories)was injected intratracheally (i.t.) in WT and transgenic mice byusing an intubation procedure as previously described (5). Atbaseline and at designated time points after bleomycin injection,lungs were harvested for histology, frozen tissue, bronchoalveo-lar lavage, or cell isolation as previously described (5–8).

Doxycycline (Dox) Administration. All L188Q SFTPC mice andcontrols were maintained on normal water ad libitum untiltransgene activation was desired. At that time, Dox (Sigma-Aldrich) was mixed in sterile water at a concentration of 2 g/dLwith additional 2% sucrose. The bottles containing Dox were

wrapped with foil to prevent light-induced Dox degradation, andDox bottles were replaced twice per week.

Tunicamycin Administration. Tunicamycin (Sigma) was dissolved inDMSO and diluted to a concentration of 20 μg/mL in 20%DMSO diluted in PBS, and then 100 μL of solution was de-livered to experimental mice by i.t. intubation. A similar volumeof 20% DMSO diluted in PBS was used as a vehicle control.

Histology and Microscopy. Formalin-fixed, paraffin-embedded lungtissue was sectioned and stained with H&E or Masson’s tri-chrome as previously described (7, 8). Light and fluorescentmicroscopy was performed with an Olympus IX81 InvertedResearch Microscope configured with an Olympus IX2-DSUBiological Disk Scanning Unit.

Immunostaining. Immunohistochemistry (IHC) on paraffin-em-bedded lung tissue sections and immunocytochemistry on cellpreparations were performed with primary antibodies and stan-dard immunoperoxidase techniques as previously described (7, 8).Immunofluorescence staining was performed on cell preparationsby using primary antibodies followed by appropriate fluorescentsecondary antibodies (Jackson ImmunoResearch) with nuclearstaining performed with DAPI using Vectashield mounting me-dium (Vector Laboratories) as previously described (5, 6).

TUNEL Assay. TUNEL assays on lung sections were performed aspreviously described (5, 8) with a commercially available kit (InSitu Cell Death Detection Kit; Roche Molecular Biochemicals)in accordance with the manufacturer’s instructions. Counter-stains for preparations were performed with hematoxylin.

Cell Culture/Isolation. Plasmids were transfected into A549 cellswith anEffectene transfection kit (Qiagen) per themanufacturer’sinstructions. Cells were incubated with the transfection com-plexes under normal growth conditions for 4 h, and then 0.5–1.0μg/mL Dox was added to the medium for induction of transgeneexpression. After transient transfection for 24 h, cells were har-vested for determination of gene expression by Western blotanalysis. Type II alveolar epithelial cells (AECs) were isolatedfrom adult mice using techniques as previously described (6, 7).

Lung Lavage and Cell Counts. Bronchoalveolar lavage was per-formed as detailed previously (5, 8). After euthanasia, three 800-μL lavages of sterile saline were performed with a 20-gauge blunt-tipped needle inserted into the trachea. Samples were centrifugedat 400 × g for 10 min, and cells were resuspended and countedmanually under light microscopy with a hemocytometer. Ap-proximately 30,000 cells from each specimen were loaded ontoslides using a Cytospin 2 centrifuge (Shandon Southern Prod-ucts). These slide preparations were then stained with a modifiedWright stain.

RNA Isolation, Real-Time RT-PCR, and Densitometry. Total RNA fromtype IIAECs and lung tissuewas isolated by using theRNeasyMiniKit (Qiagen), according to the manufacturer’s specifications. Toremove contaminating genomic DNA, samples were incubatedwith DNase (Ambion) and then converted to cDNA using Su-perScript II reverse transcriptase (Invitrogen). For BiP evaluation,PCR amplification and quantification were performed with SYBRGreen PCR Master Mix (Ambion). Primer sequences were asfollows: BiP (forward, 5′-CCTGCGTCGGTGTGTTCAAG -3′;reverse, 5′-AAG GGT CAT TCC AAG TGC G-3′) and RPL19

Lawson et al. www.pnas.org/cgi/content/short/1107559108 1 of 6

(forward, 5′-ATG CCAACTCCCGTCAGCAG -3′; reverse, 5′-TCA TCC TTC TCA TCC AGG TCA CC-3′). PCR amplificationwas conducted on a StepOnePlus Real-Time PCR System (Ap-plied Biosystems). The relative mRNA amount in each sample wascalculated based on its threshold cycle (Ct) in comparison with theCt of the housekeeping gene RPL19. Relative mRNA expressionwas presented as 2ΔCt(housekeeping gene)−ΔCt(target gene). Evaluationof XBP1 splicing was performed by using previously describedmethods (9). Primer sequences for XBP1 were as follows: forward,5′-CTG GAA AGC AAG TGG TAG A-3′ and reverse, 5′-CTGGGT CCT TCT GGG TAG AC-3′. cDNA (2 μL) was used witha reaction volume of 50 μL. PCR products were run on a 2.5% gelwith 398-bp and 424-bp forms representing spliced and unsplicedfragments, respectively. XBP1 splicing ratio was calculated asspliced divided by total. Densitometry was performed with NIHImage J software where band densities were calculated and sub-tracted from the background.For SFTPC expression, real-time reactions were performed on

a Bio-Rad iCycler with iQ SYBR Green Supermix (Bio-Rad).Standard curves were generated by the amplification of targetsequences previously cloned into pGEM-T (Promega), in dilutionseries from 10−1 to 10−6 fmol of target sequence per well. Real-time PCR primers for mutant L188Q SFTPC were as follows:5′ (located in the inserted Myc sequence), 5′-GAG CAG AAACTC ATC TCT G-3′ and 3′ (located in the mSFTPC sequence),5′-CTG GCT TGT AGG CGA TCA GC-3′. Real-time PCRprimers for mouse native SFTPC are as follows: sense, 5′-CTCCACACC CAC CTC TAAGCT-3′ and antisense, 5′-GTA GGAGAG ACA CCT TTC CTT-3′.

Western Blot Analysis. Western blot analysis was performed oncytoplasmic extracts from whole-lung and cell preparations byusing previously described techniques (7).

In Vivo Bioluminescence and Luciferase Measurements. Live-animalbioluminescence imaging in NGL transgenic mice was performedafter i.p. injection of luciferin (1 mg per mouse in 100 μL ofisotonic saline) using an intensified CCD camera (IVIS 200;Xenogen) as previously described (4). Whole-lung tissue lucif-erase levels were determined as previously described (4).

Measurement of Airway Resistance and Compliance. For airway re-sistance and compliance measurements, mice were anesthetized

with i.p. pentobarbital, and tracheas were cannulated with a 20-gauge metal stub adapter. Each mouse was placed on a small-animal ventilator, flexiVent (SCIREQ), with 150 breaths per minand a tidal volume of 10 mL/kg of body weight. Airway resistanceand static lung compliance (using a 2-s breath pause) wereassessed with SCIREQ manufacturer-provided software, whichcalculates the resistance by dividing the change in pressure by thechange in flow (cm of H2O/mL per s) and compliance by dividingchange in volume by change in pressure (mL/cm of H2O).

Semiquantitative Scoring. Quantification of lung fibrosis on histo-logical specimens was performed by an investigator blinded to thegroup using a semiquantitative score on 10 sequential, nonover-lapping fields (magnification: ×300) as previously described (5, 8).For scoring of S100A4+ and αSMA+ lung fibroblasts, slidesimmunostained for S100A4 or αSMA were evaluated on a 0- to 4-point scale as previously described (8): 0, no positive cells; 1, few(≤3) positive cells; 2, multiple (>3) individual positive cells; 3,multiple positive cells in isolated clumps; and 4, multiple clumpsof positive cells. The mean score for the 10 sequential fieldsrepresented the score for each individual specimen. For evalua-tion of TUNEL staining, slides were evaluated on 10 sequential,nonoverlapping fields (magnification: ×600) of lung parenchymafor each specimen and were scored using a 0- to 4-point semi-quantitative scale as previously described: 0, no positive cells; 1,≤1% of cells in field positive; 2, 1–5% of cells in field positive; 3,5–10% of cells in field positive; and 4, 10–25% of cells in fieldpositive (10).

Collagen Content. Frozen lung tissue samples were hydrolyzed in6 M HCl, and hydroxyproline content was quantitated by using amicroplate assay based on Ehrlich’s reaction as previouslydescribed (11). Lung collagen content was calculated fromthese results as hydroxyproline accounts for ∼13.3% of collagenby weight.

Statistics. Statistical analyses were performed with GraphPadInStat (GraphPad Software). Differences among groups wereassessed with one-way ANOVA. Differences between pairs wereassessed with Student’s t test. Results are presented as mean ±SEM. P values <0.05 were considered significant.

1. Thomas AQ, et al. (2002) Heterozygosity for a surfactant protein C gene mutationassociated with usual interstitial pneumonitis and cellular nonspecific interstitialpneumonitis in one kindred. Am J Respir Crit Care Med 165:1322–1328.

2. Lawson WE, et al. (2008) Endoplasmic reticulum stress in alveolar epithelial cells isprominent in IPF: Association with altered surfactant protein processing andherpesvirus infection. Am J Physiol Lung Cell Mol Physiol 294:L1119–L1126.

3. Zhu Z, Ma B, Homer RJ, Zheng T, Elias JA (2001) Use of the tetracycline-controlledtranscriptional silencer (tTS) to eliminate transgene leak in inducible overexpressiontransgenic mice. J Biol Chem 276:25222–25229.

4. Everhart MB, et al. (2006) Duration and intensity of NF-&kappaB activity determinethe severity of endotoxin-induced acute lung injury. J Immunol 176:4995–5005.

5. Degryse AL, et al. (2010) Repetitive intratracheal bleomycin models several features ofidiopathic pulmonary fibrosis. Am J Physiol Lung Cell Mol Physiol 299:L442–L452.

6. Tanjore H, et al. (2009) Contribution of epithelial-derived fibroblasts to bleomycin-induced lung fibrosis. Am J Respir Crit Care Med 180:657–665.

7. Lawson WE, et al. (2005) Characterization of fibroblast-specific protein 1 inpulmonary fibrosis. Am J Respir Crit Care Med 171:899–907.

8. Lawson WE, et al. (2005) Increased and prolonged pulmonary fibrosis insurfactant protein C-deficient mice following intratracheal bleomycin. Am J Pathol 167:1267–1277.

9. Shang J (2005) Quantitative measurement of events in the mammalian unfoldedprotein response. Methods 35:390–394.

10. Cheng DS, et al. (2007) Airway epithelium controls lung inflammation and injurythrough the NF-&kappaB pathway. J Immunol 178:6504–6513.

11. Brown S, Worsfold M, Sharp C (2001) Microplate assay for the measurement ofhydroxyproline in acid-hydrolyzed tissue samples. Biotechniques 30:38–40, 42.

Lawson et al. www.pnas.org/cgi/content/short/1107559108 2 of 6

Doxycycline + - + mSFTPC.rtTA - + + tetO.L188Q SFTPC-myc - + +

myc

β-actin

A

B(tet-O)7 /CMV hL188QSFTPC-Myc BGH-poly A

0.5 kb 2.98 kb 0.3 kbAsc I Asc I

(tet-O)7/CMV-hL188QSFTPC-Myc construct

mSFTPC promoter rtTA SV40-poly A

6.5 kb 1.6 kbItoNIohX

mSFTPC-rtTA construct+

mSFTPC promoter tTS SV40 poly A

6.5 kb 1.3 kbNot IKpn I

mSFTPC-tTS construct+

Triple transgenic mutant L188Q SFTPC mice

Fig. S1. Constructs were developed to express myc-tagged mutant L188Q SFTPC in a tetracycline-dependent fashion. (A) Western blot analysis for the myc-tagged transgene in A549 cells transfected with both the mSFTPC.rtTA and tetO.L188Q SFTPC-myc constructs. β-Actin is shown as a loading control.(B) Schematic of rtTA, tTS, and L188Q myc-tagged SFTPC constructs that were designed for the purpose of generating the transgenic L188Q SFTPC mouse.

1 2 3 4 5 44 45 46 47 52 53 54PositiveControls

mSFTPC.tTS - 4.9kb

tetO.L188Q SFTPC – 1.6kb

mSFTPC.rtTA – 5.2kb

A

0.01

0.1

1

10

100

WT Line 1 Line 2 Line 45 Line 52

pro-

SP

-C m

RN

A le

vels

(fmol

per

ng

RN

A)

NativeMutant

0

1

2

3

4

5

7 days 3 months 6 months

Duration of Doxycycline Exposure

pro-

SP-C

-myc

mR

NA

leve

ls(fm

ol p

er n

g R

NA) WT+Dox

L188Q SFTPC+Dox

CB

L188Q SFTPC (-Dox)WT (+Dox)

D E F

L188Q SFTPC (+Dox)

Fig. S2. Expression of mutant pro-SP-C is detected in vivo in L188Q SFTPC mice. (A) Southern blot analysis for the three transgenic constructs demonstratingidentification of four founder lines: 1, 2, 45, and 52. (B) pro-SP-C–myc mRNA levels for native versus mutant pro-SP-C for each founder line exposed to Dox for1 wk. (n = 3 per group.) (C) pro-SP-C–myc mRNA levels in whole lung from founder line 1 mice exposed to Dox for up to 6 mo. (n = 3 per group.) Graphical dataare presented as mean ± SEM. (D–F) IHC for myc tag in WT mice with Dox (D) and L188Q SFTPC mice without Dox (E) and with Dox (F). (Magnification: ×400.)

Lawson et al. www.pnas.org/cgi/content/short/1107559108 3 of 6

WT + Dox L188Q SFTPC + Dox

A B

Fig. S3. Expression of mutant L188Q SFTPC in vivo does not lead to aberrant lung histology. H&E-stained lung tissue sections from WT (A) and L188Q SFTPC(B) mice exposed to Dox for 6 mo. (Magnification: ×200.)

BA

WT + Dox + Bleo L188Q SFTPC + Dox + Bleo

C D

S10

0A4

αS

MA

Fig. S4. Mice expressing mutant L188Q SFTPC had more lung fibroblasts after i.t. bleomycin (Bleo) than littermate controls did. (A and B) IHC for the fibroblastmarker S100A4 in lung sections from WT (A) and L188Q SFTPC (B) mice at 3 wk after bleomycin treatment. (C and D) IHC for the myofibroblast marker αSMA inlung sections from WT (C) and L188Q SFTPC (D) mice at 3 wk after bleomycin treatment. (Magnification: ×200.)

A

0.0

0.5

1.0

1.5

2.0

2.5

WT + Dox L188Q SFTPC +Dox

hSFTPC.rtTA xL188Q SFTPC +

Dox

L188

Q p

ro-S

P-C

-myc

mR

NA

(fmol

per

ng

RN

A)

**B

0.0

0.5

1.0

1.5

2.0

2.5

WT + Dox + Bleo L188Q SFTPC +Dox + Bleo

hSFTPC.rtTA xL188Q SFTPC +

Dox + Bleo

Fibr

osis

Sco

re

Fig. S5. Crossing L188Q SFTPC mice to the human SFTPC promoter-driven rtTA (hSFTPC.rtTA) did not alter lung fibrosis. (A) L188Q pro-SP-C–myc mRNA levelsin whole-lung tissue from WT, L188Q SFTPC, and hSFTPC.rtTA × L188Q SFTPC mice exposed to Dox for 1 wk. (n = 3 per group.) (B) Semiquantitative fibrosisscoring of trichrome-stained lung sections from WT, L188Q SFTPC, and hSFTPC.rtTA × L188Q SFTPC mice at 3 wk after 0.04 unit of i.t. bleomycin (Bleo). (n = 5–7per group; *P < 0.05 compared with WT.) Graphical data are presented as mean ± SEM.

Lawson et al. www.pnas.org/cgi/content/short/1107559108 4 of 6

A

WT

+Bl

eoW

T+

Bleo

WT

+Bl

eoW

T+

Bleo

WT

+Bl

eoL1

88Q

SFTP

C+

Bleo

L188

QSF

TPC

+Bl

eoL1

88Q

SFTP

C+

Bleo

L188

QSF

TPC

+Bl

eoL1

88Q

SFTP

C+

Bleo

mar

ker

Spliced

Unspliced

B

0

0.2

0.4

0.6

0.8

Dox Dox + Bleo

BiP

mR

NA

Exp

ress

ion

(2Δ

CT )

WTL188Q SFTPC

#

#*

*

C

0

0.1

0.2

0.3

0.4

0.5

0.6

Dox Dox+Bleo

XB

P1

Spl

icin

g R

atio

WTL188Q SFTPC

*

#

Fig. S6. Markers of endoplasmic reticulum (ER) stress are increased by bleomycin (Bleo) treatment in WT mice and mutant L188Q SFTPCmice. (A) Real-time RT-PCR for expression of BiP mRNA normalized to expression of the housekeeping gene RPL19 at 1 wk after bleomycin treatment. (n = 4–5 per column; *P < 0.001for WT versus L188Q SFTPC in both Dox and Dox + Bleo groups; #P < 0.05 for Dox versus Dox + Bleo for both WT and L188Q SFTPC.) (B) RT-PCR gel dem-onstrating splice variants for XBP1 mRNA in lungs of WT and L188Q SFTPC mice treated with Dox and harvested 1 wk after bleomycin injection. (C) XBP1splicing analysis by densitometry. (n = 5 per column; *P < 0.01 for WT versus L188Q SFTPC in Dox group; #P < 0.001 for Dox versus Dox + Bleo for WT.) Graphicaldata are presented as mean ± SEM.

L188Q SFTPC NGLWT NGL

A B

C D

Bas

elin

e on

Dox

Dox

+ B

leo

3 da

ys

Fig. S7. (A and B) Representative in vivo photon-capture images from WT/NGL (A) and L188Q SFTPC/NGL (B) mice after 14 d of Dox treatment. (C and D)Representative in vivo photon-capture images from WT/NGL (C) and L188Q SFTPC/NGL (D) mice at 3 d after i.t. bleomycin (Bleo) with continued Dox.

Lawson et al. www.pnas.org/cgi/content/short/1107559108 5 of 6

A B

BiP

XB

P1

WT + Vehicle WT + TM

C D

F

Vehi

cleVe

hicle

Vehi

cleTM TM TM

Spliced

Unspliced

mar

kerG

0

0.2

0.4

0.6

0.8

Vehicle TM

BiP

mR

NA

Exp

ress

ion

(2Δ

CT )

*

0

0.1

0.2

0.3

0.4

Vehicle TM

XBP

1 S

plic

ing

Rat

io

*H

E

Vehi

cle

Vehi

cle

TM TM

BiP

β-actin

Fig. S8. Tunicamycin (TM) leads to ER stress in the lungs. (A and B) IHC for the ER stress marker BiP in lung sections from WT mice at 48 h after i.t. ad-ministration of vehicle (DMSO; A) and tunicamycin (B). (C and D) IHC for the ER stress marker XBP1. (Magnification: ×600.) (E) Western blot analysis for BiPusing whole-lung lysates from WT mice at 48 h after i.t. administration of vehicle or tunicamycin. β-Actin is shown as loading control. (F) Real-time RT-PCR forexpression of BiP mRNA normalized to expression of the housekeeping gene RPL19. (n = 3 per column; *P < 0.05 between columns.) (G) RT-PCR gel dem-onstrating splice variants for XBP1 mRNA in lungs of WT mice treated with either vehicle or tunicamycin. (H) XBP1 splicing analysis by densitometry. (n = 3 percolumn; *P < 0.0001 between columns.) Graphical data are presented as mean ± SEM.

Lawson et al. www.pnas.org/cgi/content/short/1107559108 6 of 6