Craig Henke, Peter Bitterman,. Urvashi Roongta, David Ingbar, and ...... Hu T, Klier G, Cheresh DA: Integrin aV13 antagonists promote tumor regression by ...
American Journal of Pathology, Vol. 149, No. 5, November 1996 Copynight © American Society for Investigative Pathology
Induction of Fibroblast Apoptosis by Anti-CD44 Antibody Implications for the Treatment of Fibroproliferative Lung Disease
Craig Henke, Peter Bitterman, Urvashi Roongta, David Ingbar, and Vitaly Polunovsky From the Department of Medicine, University of Minnesota, Minneapolis, Minnesota
Fibroblast migration and proliferation within the alveolar waUl and airspace after lung injury can lead to the development of fibrosis. Fibroblast cell surface CD44 is an adhesion receptor for provisional matrix proteins and mediates fibroblast invasion into fibrin matrices. Here we show that incubation of culturedftbroblasts with an anti-CD44 monoclonal antibody induces fibroblast detachment from the substratum and morphological changes compatible with apoptosis. In addition, we show that anti-CD44 monoclonal antibody rapidly induces fibroblast apoptosis within fibrin matrices. The effect of anti-CD44 antibody on induction offibroblast apoptosis occurred within 8 hours and was dose dependent. Anti-CD44 antibody also inducedfibroblast apoptosis in suspension. Furthermore, fibroblasts plated on anti-CD44-antibody-coated surfaces initially attached and spread on the antibody; however, after an 8-hour incubation time, many of the ceUs developed characteristic morphological features of apoptosis. Collectively, these data indicate that apoptosis did not result solely due to detachment from the substratum. Our results identify a new function for fibroblast ceU surface CD44 related to the control of ceU viability. We suggest this function may be important in fibroblast population control and could potentially be exploited in designing anti-fibrotic therapies. (AmJ Pathol 1996, 149:1639-1650)
The fibroproliferative lung diseases often lead to progressive alveolar fibrosis culminating in respiratory failure and death. In the early stages of fibroproliferation, activated myofibroblasts move from the lung interstitium across the disrupted epithelial barrier into a provisional matrix at the air-lung interface. This provisional matrix consists of polymerized fibrin and entrapped plasma proteins.1-3 Clinical and experimental findings indicate that such lesions have the potential to regress, with restoration of alveolar structure and gas exchange.4 - However, in patients with progressive fibroproliferative lung disease, permanent physiological derangements result from the deposition of excessive collagenous matrix by fibroblasts within the alveolar wall and airspace. Current therapeutic modalities for the treatment of pulmonary fibrosis that target lung inflammation have marginal benefit in terms of halting progressive fibrosis.e Fibroblast cell surface CD44 functions as an adhesion receptor for provisional matrix proteins and mediates fibroblast migration and invasion into a fibrin provisional matrix.9 In this report we show that disruption of CD44 receptor function with anti-CD44 monoclonal antibody induces apoptosis of fibroblasts cultured on tissue culture plastic. We have also examined the effect of anti-CD44 monoclonal antibody on fibroblast viability in an in vitro context that more closely simulates early fibroproliferative lesions by utilizing a fibrin gel model of fibrosis. Here we show that anti-CD44 monoclonal antibody rapidly Supported by grant HL 50152-01 (Specialized Center of Research in Acute Lung Injury from the National Institutes of Health and Minnesota Medical Foundation Special Equipment grant SMF 952-94. Accepted for publication June 17, 1996. Address reprint requests to Dr. Craig A. Henke, Pulmonary and Critical Care Division, Department of Medicine, UMHC Box 276, University of Minnesota, 420 Delaware Street SE, Minneapolis, MN 55455.
1639
1640
Henke et al
AJP November 1996, Vol. 149, No. 5
induced fibroblast apoptosis in a dose-dependent manner within the fibrin gel. Anti-CD44 antibody induced fibroblast apoptosis in suspension culture and when cells were cultured on anti-CD44-anti-
body-coated surfaces. This indicated that fibroblast apoptosis did not solely result from disruption of fibroblast adhesion to the substratum. We suggest that induction of fibroblast apoptosis by modulating CD44 function may be a reasonable strategy in the treatment of fibrotic lung disorders.
Materials and Methods Cells Primary cultures of lung mesenchymal cells were developed from the lungs of patients with severe acute lung injury (adult respiratory distress syndrome) and were characterized as myofibroblasts as previously described.9,10 The myofibroblasts were maintained in culture (370C, 10% C02) in Dulbecco's modified Eagle's medium (DMEM) containing 10% heat-inactivated fetal calf serum, subcultivated weekly, and used before the eighth subcultivation. Two diploid fibroblast strains, normal human lung fibroblasts (ATCC 210) and human fetal skin fibroblasts (ATCC 1475) were also utilized. Cell cultures were stored frozen until ready for use. Cells were used within 12 subcultivations.
Antibodies The following antibodies were used: mouse anti-human CD44 antibody BU52 (The Binding Site, San Diego, CA; Ancell, Bayport, MN); rat anti-mouse CD44 antibody IM7 (PharMingen, San Diego, CA); mouse anti-human CD44 antibody A3D8 (Sigma Chemical Co., St. Louis, MO); mouse anti-rabbit CD44 antibody W4/86 (Serotec, Oxford, UK); mouse anti-rabbit CD44 antibody MW5-B (Spring Valley Labs, Woodbine, MD); isotype-matched control antibody (mouse IgG1lK; Sigma).
Anti-CD44 Antibody Treatment of Fibroblasts Cultured on Tissue Culture Plastic Fibroblasts were plated on tissue culture plastic dishes coated with fibronectin (10 ,ug/ml) and incubated with DMEM plus 2.5% calf serum (24 hours at 370C). The medium was then removed from the cells and replaced with DMEM plus 2.5% calf serum containing varying
concentrations of anti-CD44 monoclonal antibody and incubated overnight (16 hours at 370C).
Fibrin Gels Gels composed of fibrin were prepared under sterile conditions as described by Dvorak.11 Cells were harvested with trypsin-EDTA, washed, and resuspended in a HEPES-saline buffer (0.13 mol/L NaCI, 0.025 mol/L N-2-hydroxyethylpiperazine-N'-2 ethanesulfonic acid, 0.005 mol/L CaCI2, pH 7.4) containing human fibrinogen (3 mg/ml; Sigma) at a concentration of 1 x 105 cells/ml. The cell suspension was placed in Falcon six-well dishes (1.5 ml/well; Becton-Dickson, Lincoln Park, NJ), clotted with human thrombin (0.2 U/ml; Sigma), allowed to solidify (1 hour at 370C), and overlaid with 1 ml of DMEM containing 2.5% heat-inactivated calf serum. The gels were incubated (370C,10% C02 for 48 hours) in a humidified incubator. Medium was removed, and fibrin gels were incubated (24 hours at 370C) with DMEM + 2.5% heat-inactivated calf serum containing 20 ,ug/ml mouse anti-human CD44 monoclonal antibody or isotype-matched control monoclonal antibody.
Morphological and Histochemical Evaluation of Fibroblasts Cultured on Tissue Culture Plastic and in Fibrin Gels To identify morphological changes induced by antiCD44 antibody, cells cultured on tissue culture plastic or within the fibrin gels were examined using phase contrast microscopic techniques employing an Olympus IMT-2 inverted microscope. Cells possessing morphological changes consistent with apoptosis were identified using previously defined criteria including cytoplasmic condensation, chromatin compaction, nuclear fragmentation, and blebbing of the plasma membrane.12-14 To confirm these changes, gels were fixed with 70% ethanol (30 minutes at 22°C) and stained with acridine orange as previously described.12 To estimate the percentage of cells within fibrin gels that had undergone apoptosis, nuclei containing degraded DNA were identified by in situ labeling using the terminal transferase nick end translation assay (TUNEL) as described by Gavrieli.15 Before TUNEL, the fibrin gels were fixed in 10% buffered formalin, embedded in paraffin, adhered to slides, deparaffinized (30 minutes at 600C), and rehydrated using, sequentially, xylene, 96% ethanol, 90% ethanol, 80% ethanol, and double-distilled water. Slides were incu-
Anti-CD44 Antibody Induces Fibroblast Apoptosis
1641
AJP November 1996, Vol. 149, No. 5
bated (15 minutes at 370C) with 20 jug/ml proteinase K (Sigma), washed four times with double-distilled water, treated with 2% H202 (10 minutes 370C), and rinsed with double-distilled water. Slides were immersed in TdT buffer (30 mmol/L Tris base, pH 7.2, 140 mmol/L sodium cacodylate, 1 mmol/L cobalt chloride). The tissue sections were covered with TdT buffer containing 12 ,umol/L biotinylated dUTP plus 0.2 U/Iul TdT, and the reaction was allowed to proceed (60 minutes at 370C). The reaction was terminated by incubation (15 minutes at 370C) in buffer (300 mmol/L sodium chloride, 30 mmol/L sodium citrate). Slides were rinsed in double-distilled water, blocked (10 minutes at 220C) with 2% bovine serum albumin, rinsed in double-distilled water, and immersed in phosphate-buffered saline (PBS) for 5 minutes at 220C. Incorporated nucleotide was detected using strepavidin-horseradish peroxidase (Gibco BRL, Gaithersburg, MD) diluted 1:10 in water in a humidified chamber. Slides were washed with double-distilled water, immersed with PBS (5 minutes at 22°C), and color developed with aminoethylcarbazole (Sigma).
Agarose Gel Electrophoresis of DNA To identify DNA fragments after exposure to antiCD44 antibody, lung fibroblasts (ATCC 210, 2 x 106 cells) adhered to tissue culture flasks were incubated (24 hours at 370C) in DMEM plus 2.5% calf serum containing 20 ,ug/ml anti-CD44 antibody. Nuclear DNA degradation was detected by agarose gel electrophoresis. Fibroblasts were released with trypsin-EDTA, washed with ice-cold PBS, and incubated (18 hours at 500C) in lysis buffer (100 mmol/L NaCI, 10 mmol/L Tris/HCI, pH 8, 25 mmol/L EDTA, 0.5% sodium dodecyl sulfate, 0.1 mg/ml proteinase K). Samples were extracted twice with phenol/chloroform (1:1), treated with 20 p,g/ml RNAse A, and precipitated with 7.5 mol/L CH3COONH4 and 2 volumes of 100% ethanol. DNA was recovered by centrifugation and pellets were washed with 70% ethanol, air dried, and resuspended in TE buffer (10 mmol/L Tris/HCI, pH 8.0, 1 mmol/L EDTA). DNA samples were separated on 1.2% agarose gels with 0.25 jAg/ml ethidium bromide, visualized by UV fluorescence, and photographed as previously described. 12
Determination of Kinetics and Dose Dependency of Fibroblast Apoptosis Lung fibroblasts (ATCC 210) were seeded into sixwell dishes (1 x 105 cells/well) and incubated in
DMEM containing 2.5% calf serum for 24 hours. For kinetic studies, medium was removed and fibroblasts incubated for the specified time intervals with DMEM plus 2.5% calf serum containing 10 ,ug/ml anti-CD44 monoclonal antibody. To examine dose dependency, fibroblasts were incubated (12 hours at 370C) with the indicated concentration of antiCD44 antibody in DMEM plus 2.5% calf serum. Control assays were performed using untreated cells and cells treated with isotype-matched control monoclonal antibody. After antibody treatment, the percentage of apoptotic cells was determined by quantification of DNA fragmentation by analysis of relative DNA content using flow cytometry after propidium iodide staining as previously described.12
Suspension Cultures To distinguish the ability of anti-CD44 antibody to directly influence cell viability from its effects on cell attachment, cells in suspension were evaluated. Fibroblasts in late log-phase culture were rinsed three times with PBS, incubated in PBS (15 minutes at 370C), and removed from the dish with a rubber policeman. Detached cells were washed three additional times with PBS and pelleted after each wash by centrifugation (600 x g for 10 minutes). The resultant pellet of washed cells was resuspended in DMEM containing 0.1 mg/ml bovine serum albumin, 3 x 10-8 mol/L selenium and 10 ,gg/ml transferrin, and filtered through a 100-p.m nylon mesh before use. Cells were added to Falcon polypropylene tubes (10 ml; Becton Dickinson) containing the indicated concentration of anti-CD44 antibody. Isotypematched control antibody served as a negative control. At the indicated times, cells were washed twice with PBS (40C), filtered through nylon mesh, and fixed in 70% ethanol (added dropwise at 40C) for propidium iodide staining and flow cytometric analysis of cellular DNA distribution.
Anti-CD44 Antibody Coating of Tissue Culture Chamber Slides To further examine whether anti-CD44 antibody induction of fibroblast apoptosis was dependent upon detachment of adhesion-dependent fibroblasts from the substratum, cells were plated on tissue culture chamber slides (Nunc, Naperville, IL) coated with anti-CD44 antibody. Anti-CD44 antibody (50 ,ug/ml) in PBS was added to eight-well tissue culture chamber slides and incubated (12 hours at 40C). Wells were washed three times with PBS, and fibroblasts
1642
Henke et al
AJP November 1996, Vol. 149, No. 5
Figure 1. Anti-CD44 antibody induces morphological changes consistent with apoptosis. Acute lung injury fibroblasts were allowed to attach to fibronectin-coated dishes (24 hours at 370C. They wvere then incubated overnight ( 16 hours at 370() with anti-CD44 antibody (20 gg/ml). To identify morphological chanlges consistentt with apoptosis, the cells uerefixed (10 minutes at 2200) with 700o ethbaniol, stained (3 houirs at 220() u'ith 6 mglml acrldine orange and examined underphase contrast and green fluorescence usinzg an) inverted mnicroscope. A: Characteristic morphological chaniges of apoptosis were present in fibroblasts treated u'ith anti-CD44 anttibody including cytoplasmic condenisation anid plasmna membrane blebbing. B: Additionally, nuiclei of dead cells displayed chromatin condensation andfragmentation typicalforprogrammed cell death. C and D: Shown also are the phase contrast and fluorr,sc-cent appearances of cells treated uith i.sotype-matched control antibody displaving normal inorpho-
logicalftatures. Magniification,
X 200.
(25 x 103 cells/300 ,ul) in DMEM plus 0.1% calf serum were added to the wells and incubated (370C, 10% C02) in a humidified incubator. At the indicated times, morphological analysis of cells was performed using phase contrast microscopy with Hoffman modulation optics and fluorescence microscopy after acridine orange staining. Coating of chamber slides with anti-f31 integrin subunit antibody (50 ,ug/ml;
Chemicon, Temecula, CA), poly-D-lysine (50 ,ug/ml; Sigma), and PBS served as controls.
Data Analysis All data are expressed as mean ± SEM. Experiments were performed in duplicate and performed at least three times unless otherwise indicated. For assessment of the percentage of apoptotic cells within fibrin gels using the TUNEL assay, microscopic analysis of at least 100 cells per gel was performed.
Results Anti-CD44 Antibody Induces Fibroblast Apoptosis During studies examining the role of CD44 in fibroblast attachment to the underlying substratum we observed that acute lung injury fibroblasts, which had attached and spread on the surface of tissue culture dishes as well as fibronectin-coated surfaces, became round in appearance when incubated in the presence of anti-CD44 antibody. Over 24 hours, many of the cells detached from the surface of the dishes. As blockage of integrin function with anti-integrin antibodies causes endothelial cells to detach from the underlying matrix and undergo apoptosis, we sought to determine whether antiCD44 antibodies might function in a similar manner. Lung injury fibroblasts incubated (16 hours at 370C) with BU52 anti-CD44 antibody (20 gg/ml) underwent
Anti-CD44 Antibody Induces Fibroblast Apoptosis 1643 AJP November 1996, Vol. 149, No. 5
V.
*1
is
.t
I
Figure 2. Myofibroblasts within fibrin gels morphologicallv resemble gratntulation tissue and undergo apoptosis when treated with anti-CD44 antibody. Myofibroblasts obtaitnedfrom lung injury patients u'ere incorporated into fibrin gels. A: Shown is the histological appearance of the fibrin gel after 1 week in culture. The anatomic patteni closely resembles loose graniulation tissue. Hematoxylin stain; magnification, X 100. B: Phase contrast microscopic appearance showing myofibroblasts exteniditng long processes within thefibrinz matrix. Magnificationz, X 40. C: Phase contrast microscopic appearance of mnyofibroblasts wcithin a fibrin gel incubated in the presence of aniti-CD44 monoclonal antibody (24 hours; 20 pug/ml). Magnification, X 100. Note plasma menmbranie blebbing an-id cytoplasmic condensation. D: After acridine orange staining, pronounced chromatin fragmentation anid condensation u'ithin nluclei qf myofibroblasts treated with anti-CD44 antibody was apparent. Magnification, X 100. E and F: Normal norphological appearanice of 7nyojibroblasts treated uwith isotype-matcbed control antibody by phase contrast microscopy and acridine oranige staining, respectively. MagfiJicationi, X 100.
typical apoptotic morphological changes including cytoplasmic condensation and plasma membrane blebbing (Figure 1A). Analysis of fibroblast morphology after staining with acridine orange revealed pronounced chromatin condensation and fragmentation characteristic of programmed cell death (Figure 1 B). In addition, we tested the effect of four additional anti-CD44 monoclonal antibodies (IM7, A3D8, W4/
86, and MW5-B) on fibroblast viability. All four additional anti-CD44 antibodies tested induced fibroblast apoptosis but required longer incubation times (48 to 72 hours) for apoptosis to occur (data not shown). All subsequent apoptosis assays were performed using the BU52 antibody. IgG1 isotype-matched control antibody at a maximal concentration of 20 gg/ml did not induce fibroblast apoptosis (Figure 1, C and D).
1644
Henke et al
A/P November 1996, Vol. 149, No. 5
Figure 3. TUNEL of myofibroblasts undergoing apoptosis within fibrin gels in response to anti-CD44 antibody. The nuclei of myofibroblasts with DNA degradation stained bright red (A) and were easily distinguishedfrom normalfibroblasts (B). Shown are apoptotic cells in fibrin gels treated u.ith anti-CD44 antibody (A) and isotype-matched control antibody (B).
Anti-CD44 Antibody Induces Fibroblast Apoptosis in a Fibrin Gel Fibrin gels containing myofibroblasts obtained from patients with acute lung injury morphologically re-
1
2
4
3
Figure 4. Agarose gel electrophoresis qf DNA. DNA
was
isolated anid
subjected to gel electrophoresis oni a 1.2% agarose gel. Lane 1, 123-bp ladder; lane 2, DNA isolated from large v'e.ssel enldothelial cells treated u'ith cyclobheximide and TNF-a (positive coot rol); lane 3, DNA isolated from fibroblasts trceated with isotWpe conitrol anitibodv (24 hours; 20 ug/ml); lane 4, DNA isolatce fromn fibroblasts treated with anti-CD44 antibody (24 hours; 20 p.g/mln) demonstrating the typical internucleasomal cleavage pattern. The gels contained 0.25 ,ug/mnl ethidium brimide antd were photographed using ItV fluorescence.
semble loose granulation tissue (Figure 2A). The cells extend long thin cytoplasmic processes and are interspersed throughout the fibrin matrix (Figure 2B). To evaluate whether CD44 might be an important regulator of cell viability, the effect of anti-CD44 antibody was examined using this culture system. Myofibroblasts incorporated into fibrin gels and cultured in the presence of anti-CD44 antibody underwent morphological changes typical of apoptosis. By phase contrast microscopy, plasma membrane blebbing and cytoplasmic condensation were readily apparent (Figure 2C). Analysis of fibroblast nuclear morphology after acridine orange staining revealed chromatin condensation and fragmentation (Figure 2D). In contrast, myofibroblasts incorporated into fibrin gels and treated with isotype control antibody exhibited normal morphological features (Figure 2, E and F). To determine whether the ability of anti-CD44 antibody to induce apoptosis was restricted to myofibroblasts obtained from patients with lung injury or whether other fibroblast strains were similarly susceptible, two additional strains of fibroblasts (ATCC 210 and ATCC 1475) were used in this culture system. Anti-CD44 antibody induced apoptosis in both strains of fibroblasts indicating that the ability of antiCD44 antibody to induce fibroblast apoptosis was not restricted to lung injury myofibroblasts (data not shown). The percentage of apoptotic lung injury myofibroblasts within fibrin gels was analyzed microscopically after in situ labeling of DNA fragmentation by TUNEL. Positive nuclei stained bright red (example shown in Figure 3A) and could be easily distinguished from negative nuclei (example shown in
Anti-CD44 Antibody Induces Fibroblast Apoptosis 1645 A/P November 1996, Vol. 149, IVo. 5
L-
o
50
n
E
z 0
50
240
720
240
720
240
720
20
10
0
Fluorescence Intensity
Time (hours)
100
50
80
40
I--
-o
o
1=
60
U
a)
30
a)
0 0
CL
C.
40
0
Q
20 _-
0. CL Q
20
110
0
0
.01
.1
1
10
100
Antibody Concentration (g/rml) Figure 5. Kinietics and dose dependency of fibroblast apoptosis in response to anti-CD44 antibocdy. A: Adherent long fibroblasts (ATCC cultured in tbe presence of anti-CD44 antibody ( 10 tg/nil) 210) jobr various tinie periods (0, 2, 4, 8, 16, an1d 24 hours). Shown are the per-cenitaiges of apoptotic cells as determined bv quantification of DNA firagmentation by analy,sis of relative DNA conitent icsing flon', cytomery after propiditni iodidec staininlg. Similar results were obtainedc sing skin fibrohlasts (ATCC 1475; data niot shouln). B: Adherenlt fibroblasts were cultured ( 12 boors) in ivariotis conicetntrationis of anti-CD44 antibody (0, 0.25, 2. 5, 10, anid 20 gig/mb. The percentage apoptotic cells was determined by flou' cytlometry after propidium iodidle stcininig. were
Figure 3B). After anti-CD44 antibody treatment, 66 + 6% of myofibroblast nuclei were positive in contrast to 6 + 2% in cultures treated with isotype control antibody. In fibrin gels treated with anti-CD44 antibody, some cells ( 12%) contained no identifiable nucleus and only the cytoplasm remained. Such cells were not included as positive, and therefore TUNEL may have underestimated the actual percentage of cells undergoing apoptosis in response to anti-CD44 antibody. Fibroblast nuclear DNA degradation induced by anti-CD44 antibody was analyzed by agarose gel
I 11 1111111
O
0
.01
.1
I
I
1111111
1
I
I
1l 1111|tl
10
100
Antibody Concentration (,ug/ml) Figure 6. Anti-CD44 antibody induces apoptosis offibroblasts cultured in suspension. A: Detachedfibroblasts were resuspended in sen- ii-fiee niediumi and placed in pol propylen1e tubes and cultuirced./r 'arioois tinie periods (0, 2, 4, 6, 8, 16, and 24 hours) in suspension in the prensece of/anti-GD44 antibody (20 jig/ml) or isotype conitrol an-tibodly (20 ig/ml). B: Detached fibroblasts iere cultnred (24 hours) in sitspenision uith various concentrations o?f anti-CD44 antibody, (0, 0.25, 2.5, 10, and 20 jig/ml). Shoun are the percentages ofapoptotic cells as determined byfloun cytometr afterpropidium iodide staining. ii
electrophoresis. DNA isolated from lung injury fibroblasts cultured on tissue culture plastic demonstrated the typical internucleosomal cleavage pattern in anti-CD44-antibody-treated cells but not in cells treated with isotype-matched control antibody (Figure 4). Large-vessel endothelial cells treated with cycloheximide and tumor necrosis factor (TNF)-a undergo apoptosis and also demonstrate the typical internucleosomal cleavage pattern and were used as a positive control.16 These results provide further support of the ability of anti-CD44 antibody to induce fibroblast apoptosis as demonstrated morphologically in the fibrin gels.
1646 Henke et al AJP November 1996, Vol. 149, No. 5
FE$ N% v*'>t< X
lPr... ! 1W.s t--
I..
:; It. '4.
I
S.
..
,;'?b
..
I ".¢
X.
....
.4S,;',
4x
-
U
IM
.
_*
-N
|
N
-Iw-.;.___
,.4 '
I
Anti-CD44 Antibody Induces Fibroblast Apoptosis 1647 AJP November 1996, Vol. 149, No. 5
Kinetics and Dose Dependency of Apoptosis in Response to Anti-CD44 Monoclonal Antibody
dent
The kinetics of anti-CD44 antibody induction of fibroblast apoptosis was analyzed using 210 and 1475 fibroblasts seeded onto tissue culture plastic (Figure 5A). Apoptosis was quantified by flow cytometry after propidium iodide staining. The percentage of apoptotic cells in untreated cultures was 2 + 1%. After a 2-hour incubation with 10 ,ug/ml anti-CD44 antibody, the percentage of apoptotic cells was 6 + 3%. This value increased to 23 + 5% by 8 hours, and exceeded 50% by 16 hours. At 24 hours, approximately 60% of cells were apoptotic. Fibroblast apoptosis induced by anti-CD44 antibody occurred in a dose-dependent manner (Figure 5B). An effect was first reproducibly detected using anti-CD44 antibody at a concentration of 2.5 jug/ml. The proportion of apoptotic cells progressively increased until approximately 70% of the cells underwent apoptosis at 20 ,ug/ml anti-CD44 antibody. The level of fibroblast apoptosis in cultures treated with identical concentrations of isotype-matched control antibody was 4 + 2%.
Fibroblasts Cultured on Anti-CD44Antibody-Coated Surfaces Undergo Apoptosis
Anti-CD44 Antibody Induces Fibroblast Apoptosis in Suspension Cultures Our results raised the possibility that the induction of apoptosis by anti-CD44 antibody was tightly connected to its ability to detach adhesion-dependent fibroblasts from the substratum. To examine this issue, acute lung injury fibroblasts in suspension culture were exposed to anti-CD44 antibody (20 ,tg/ml; 24 hours). This resulted in an approximately twofold increase in the percentage of apoptotic cells compared with cells treated with control antibody (80% in anti-CD44-antibody-treated cultures versus 42% in control cultures). Similar to cells plated on tissue culture plastic, the effect was both time (Figure 6A) and dose (Figure 6B) dependent. Similar results were obtained using normal human lung fibroblasts and human fetal skin fibroblasts (data not shown). These results indicated that anti-CD44 antibody induction of fibroblast apoptosis was not solely depen-
on
disruption of fibroblast adhesion to the sub-
stratum.
To further examine the issue of whether anti-CD44 antibody induction of fibroblast apoptosis occurs independent of its effect on cellular adhesion, lung injury fibroblasts were cultured on surfaces on which anti-CD44 antibody served as the substratum. After a 2-hour incubation period, most cells plated on anti-CD44-antibody-coated surfaces had attached and spread (Figure 7A). Similarly, the majority of cells cultured on anti-,81 integrin subunit antibodycoated surfaces for 2 hours also attached and visually displayed a greater degree of spreading as compared with cells plated on the anti-CD44-antibody-coated surface (Figure 7B). In contrast, during the same interval, most fibroblasts plated on poly-Dlysine-coated surfaces had not yet attached. Moreover, the degree of cell spreading of the cells that had attached was less than that seen with either anti-CD44 antibody or anti-,1 integrin antibodycoated surfaces (Figure 7C). Very few cells plated on surfaces exposed only to PBS attached to the surface and most cells were clumped together (Figure 7D). After an additional 6-hour incubation period, many fibroblasts that had initially attached and spread on anti-CD44-antibody-coated surfaces began to display characteristic morphological features of apoptosis. Analysis of fibroblast morphology by Hoffman phase contrast microscopy revealed many cells with condensed cytoplasm and pronounced plasma membrane blebbing characteristic of cells undergoing apoptosis (Figure 7, E and H). Additionally, fluorescence microscopy after acridine orange staining revealed cells with marked chromatin condensation and fragmentation (Figure 7H). In contrast, cells incubated on anti-f3l integrin subunit antibody-coated surfaces for the same interval remained attached, spread, and viable as assessed by Hoffman phase contrast and fluorescence microscopy (Figure 7, F and J). After 8 hours of incu-
Figure 7. Fibroblasts cultured on anti-CD44-antibody-coated surfaces undergo apoptosis. Fibroblasts were plated on glass surfaces coated with ,ug/ml), poly-D-lysine (50 llg/mnl), or PBS and cultured (3 7C; 10% GO2). The anti-,3P integrin subunit antibody (50 cells were analyzed by phase contrast microscopy (Hoffman modulation optics) andfluorescence microscopy after acridine orange staining. A to D: Phase contrast microscopy offibroblasts cultured for 2 hours on anti-CD44 antibody, anti-A,B integrin antibody, poly-D-lysine, and PBS-coated surfaces, respectively. E to G: Phase contrast microscopy offibroblasts cultured for 8 hours on anti-CD44 antibody (arrowheads denote cells with prominent plasma membrane blebbing), anti-A,B integnin antibody, and poly-D-lysine-coated surfaces, respectively. H, J, and K: Fluorescence microscopy after acridine orange staining offibroblasts culturedfor 8 hours on anti-CD44 antibody, anti-f, integrin antibody, andpoly-D-lysinecoated surfaces, respectively.
anti-CD44 antibody (50 pg/ml),
1648 Henke et al AJP November 1996, Vol. 149, No. 5
bation, cells cultured on surfaces coated with polyD-lysine had attached and spread. Visually, the degree of cell spreading on poly-D-lysine was less than that of cells plated on anti-,1l integrin antibodycoated surfaces. However, cells plated on poly-Dlysine did not display apoptotic features (Figure 7, G and K). Similar results were obtained using normal human lung fibroblasts (data not shown). Together with the results of the suspension cultures, these data not only indicate that induction of fibroblast apoptosis by anti-CD44 antibody is not solely dependent on disruption of adhesion but also suggest a direct agonist function of the antibody in initiating programmed cell death.
Discussion The fibroproliferative lung diseases are characterized by the accumulation of fibroblasts and the deposition of a collagenous matrix within the alveolar wall and airspace that may lead to permanent physiological dysfunction and, if progressive, respiratory failure and death. Intra-alveolar fibrosis results from the migration and invasion of myofibroblasts into the fibrin provisional matrix that covers the injured alveolar epithelial surface.1-39 Our previous studies have focused on identification of fibroblast cell surface matrix receptors that mediate fibroblast migration and invasion. We have identified one such receptor, CD44, as important in mediating myofibroblast migration and invasion into a fibrin matrix.9 During these studies, we noted that antiCD44 antibody inhibited fibroblast attachment to the substratum. In this study, we report that anti-CD44 antibody induces fibroblast detachment from the substratum and morphological changes consistent with apoptosis. We also report that anti-CD44 antibody is capable of inducing fibroblast apoptosis in fibrin gels, an in vitro model that is designed to simulate early fibroproliferative lesions. Anti-CD44 antibody also induced apoptosis of fibroblasts in suspension cultures. Moreover, fibroblasts cultured on anti-CD44-antibodycoated surfaces initially adhered and spread before undergoing apoptosis, indicating that its effect on induction of apoptosis was not solely dependent on its ability to disrupt fibroblast adhesion to the substratum. Data derived from patients who recover from severe lung injury and from experimental animal studies indicate that regression of intra-alveolar fibrosis can occur with restoration of normal alveolar structure and function.4-7 Apoptotic cells are present within airspace granulation tissue indicating that apoptosis is one biological process involved in airspace remodeling after lung injury.10 Mesenchymal
cells constituting newly formed fibrotic tissue after lung injury express CD44.9 These data, combined with the results from the current study showing that modulation of CD44 function with a monoclonal antibody induces fibroblast apoptosis, suggest that fibroblast cell surface CD44 and its interaction with the extracellular matrix has the potential to play an important role in fibroblast population control. Fibroblast cell surface CD44 functions as an adhesion receptor for provisional matrix proteins including fibrin, fibronectin, and hyaluronic acid.9 The 85-kd isoform of CD44 is present in lung myofibroblasts and mediates their invasion into fibrin matrices.9 Recent data indicate that CD44 also mediates cellular functions other than cell adhesion. For example, ligation of monocyte CD44 with anti-CD44 antibodies triggers TNF-a and interleukin-1lB release.17 Macrophage CD44 activation by hyaluronate also induces cytokine expression.18 In addition, antibodies to CD44 trigger a variety of effector functions in T cells including proliferation and cytolytic activity.19`21 Our current study indicates that ligation of CD44 by antibody activates a cell death program resulting in fibroblast apoptosis. In contrast, recently it has been shown that anti-CD44 antibody protects T lymphocytes from apoptosis.22 The reason for this apparent dichotomy is not known, but importantly, this study supports our finding that CD44 may influence cell survival. Although the mechanism by which ligation of CD44 with monoclonal antibody triggers the activation of programmed cell death in fibroblasts remains to be elucidated, additional investigation in this area may shed light on this dichotomy. The concept of cell surface receptor-mediated control of apoptosis is firmly established. The TNF receptor superfamily includes Fas/TNF receptors that, when ligated by Fas ligand or TNF, respectively, rapidly induce cell death in certain cell types.2324 Furthermore, blocking epithelial and endothelial cell integrin function by RGD synthetic peptides or monoclonal antibodies induces apoptosis of these cell types.25 27 Therefore, it is not without precedent that a cell surface receptor such as CD44 may regulate cell death. Extracellular matrix provides survival signals for both epithelial and endothelial cells.2526 Disruption of endothelial or epithelial cell adhesion to the underlying substratum by synthetic RGD peptides rapidly induces apoptosis. As fibroblast CD44 functions as an adhesion receptor for extracellular proteins, and treatment of fibroblasts with anti-CD44 antibody inhibits cell adhesion, interaction of CD44 with the underlying substratum likely regulates, at least in part, fibroblast survival. However, the results of our
Anti-CD44 Antibody Induces Fibroblast Apoptosis
1649
A/P November 1996, Vol. 149, No. 5
suspension culture experiments, which show that anti-CD44 antibody induces an approximately twofold increase in the level of apoptosis compared with control cells, indicate that induction of fibroblast apoptosis does not solely result from disruption of fibroblast adhesion to the substratum. To further dissociate the apoptotic effect of the antibody from its adhesion-disrupting effect, we cultured fibroblasts on surfaces coated with anti-CD44 antibody. Interestingly, the cells initially attached and spread on the antibody; however, over an 8-hour period, many of the cells that had attached to the antibody began to undergo apoptosis. This effect on fibroblasts appears to be specific for CD44 and not other matrix receptors such as integrins, as fibroblasts plated on anti-f1 integrin antibody attach, spread, and remain viable. Collectively, these data indicate that induction of fibroblast apoptosis by anti-CD44 antibody is not solely a function of detachment from the substratum, but rather, modulation of CD44 function per se appears to influence the decision to survive or die. It has been previously reported that epithelial cells and endothelial cells in suspension culture undergo apoptosis; however, in both of these reports, the fibroblasts studied appeared to be resistant to apoptosis when detached from the substratum.25'26 The explanation for this discrepancy between these two previous reports and our results is not apparent. One possible difference may lie in how the assays were performed. In both of the previous studies, the fibroblasts were plated on either dishes coated with agarose gel/bovine serum albumin or polyHEMA plates. Although the fibroblasts cannot attach to these substances, it is possible that contact with a surface may in some way (possibly by serving as a platform for endogenous matrix protein deposition) alter their susceptibility to detachment-induced apoptosis. In contrast, in our assays, the fibroblasts were kept in suspension in polypropylene tubes. A second possible explanation may lie in differences between fibroblast phenotypes. Much of our work was performed using primary fibroblast cultures. However, this does not appear to be the sole answer for this discrepancy because two additional fibroblast cell types (ATCC 210 and 1475) were tested in our assays and they also underwent apoptosis in suspension culture. Additional mechanistic studies examining the molecular basis for induction of fibroblast apoptosis by the anti-CD44 antibody may help to shed light on the susceptibility of particular cells to the antibody and to culture in suspension. Our results identify a new function for fibroblast cell surface CD44 in the regulation of cell viability. Identification of the signaling pathways activated by
CD44 could prove important in understanding basic aspects of fibroblast population control. The ability to induce fibroblast death in vitro suggests that modulation of CD44 activity using molecules such as antiCD44 antibody may provide a means of arresting the fibroproliferative response after injury in which excessive fibrosis results in organ dysfunction. In vivo studies in preclinical model systems will be needed to determine whether modulation of CD44 function will attenuate the fibrotic response.
References 1. Fukuda Y, Ishizake M, Masuda Y, Kimura G, Kawanami 0, Masugi Y: The role of intraalveolar fibrosis in the process of pulmonary structural remodeling in patients with diffuse alveolar damage. Am J Pathol 1987,126:171-182 2. Katzenstein AA, Myers JL, Mazur MU: Acute interstitial pneumonia: a clinicopathologic, ultrastructural and cell kinetic study. Am J Surg Pathol 1986, 10:256-267 3. Henke C, Marinelli W, Jessurun J, Fox J, Harms D, Peterson M, Chiang L, Doran P: Macrophage production of basic fibroblast growth factor in the fibroproliferative disorder of alveolar fibrosis after lung injury. Am J Pathol 1993, 143:1189-1199 4. Suchyta MR, Eliott CG, Colby T, Rasmussen BY, Morris AH, Jensen RL: Open lung biopsy does not correlate with pulmonary function after the adult respiratory distress syndrome. Chest 1991, 99:1232-1237 5. Meduri GU, Belenchia JM, Estes RJ, Wunderink RG, El Torky M, Leeper KV: Fibroproliferative phase of ARDS: clinical findings and effects of corticosteroids. Chest 1991, 100:943-952 6. Meduri GU, Chinn AJ, Leeper KV, Wunderink RG, Tolley E, Winer-Muran HT, Khare V, El Torky M: Corticosteroid rescue treatment of progressive fibroproliferation in late ARDS: patterns of response and predictors of outcome. Chest 1994, 105:1516-1527 7. Kapanci Y, Weibel ER, Kaplan HP, Robinson FR: Pathogenesis and reversibility of the pulmonary lesions of oxygen toxicity in monkeys. Lab Invest 1969, 20:101-118 8. Hunninghake GW, Kalica AR: Approaches to the treatment of pulmonary fibrosis. Am J Respir Crit Care Med 1995, 151:915-918 9. Svee K, Roongta U, Henke C: Cell surface CD44 related chondroitin sulfate proteoglycan mediates acute lung injury fibroblast migration and invasion into a fibrin matrix. Am J Respir Crit Care Med 1995, 151:A592 10. Chen B, Polunovsky V, White J, Blazar B, Nakhleh R, Jessurun J, Peterson M, Bitterman P: Mesenchymal cells isolated after acute lung injury manifest an enhanced proliferative phenotype. J Clin Invest 1992, 90:1778-1785 11. Brown LF, Lamir N, McDonagh J, Tognozzi K, Dvorak AM, Dvorak HF: Fibroblast migration in fibrin gel matrices. Am J Pathol 1993, 142:273-283 12. Polunovsky VA, Chen B, Henke C, Snover D, Wendt C,
1650
Henke et al
AJP November 1996, Vol. 149, No. 5
13. 14.
15.
16.
17.
18.
19.
Ingbar DH, Bitterman PB: Role of mesenchymal cell death in lung remodelling after injury. J Clin Invest 1993, 92:388-397 Wyllie AH, Kerr JFR, Curie AR: Cell death: the significance of apoptosis. Int Rev Cytol 1980, 68:251-305 Wyllie AH: Cell death. Int Rev Cytol 1987, 17(Suppl): 755-785 Gavrieli Y, Sherman Y, Ben-Sasson SA: Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation. J Cell Biol 1992, 119: 493-501 Polunovsky VA, Wendt CH, lngbar DH, Peterson MS, Bitterman PB: Induction of endothelial cell apoptosis by TNFa: modulation by inhibitors of protein synthesis. Exp Cell Res 1994, 214:584-594 Webb DSA, Shimizu Y, Van Seventer GA, Shaw S, Gerrard TI: LFA-3, CD44, CD45: physiologic triggers of human monocyte TNF and IL-1 release. Science 1990, 249:1295-1297 Noble PW, Lake FR, Henson PM, Riches DWH: Hyaluronate activation of CD44 induces insulin-like growth factor-1 expression by a tumor necrosis factor-a-dependent mechanism in murine macrophages. J Clin Invest 1993, 91:2368-2377 Denning SM, Le PT, Singer KH, Haynes BF: Antibodies against the CD44 p80, lymphocyte homing receptor
20.
21.
22.
23.
24. 25.
26.
27.
molecule augment human peripheral blood T cell activation. J Immunol 1990, 144:7-15 Galandrini R, Albi N, Tripodi G, Zarcone D, Terenzi A, Moretta A, Grossi CE, Velardi A: Antibodies to CD44 trigger effector functions of human T cell clones. J Immunol 1993, 150:4225-4235 Sconocchia G, Titus JA, Segal DM: CD44 is a cytotoxic triggering molecule in human peripheral blood NK cells. J Immunol 1994, 153:5473-5481 Ayroldi E, Cannarile L, Migliorati G, Bartoli A, Nicoletti I, Riccardi C: CD44 (Pgy-1) inhibits CD3 and dexamethasone-induced apoptosis. Blood 1995, 86:2672-2678 Smith CA, Farrah T, Goodwin RG: The TNF receptor superfamily of cellular and viral proteins: activation, costimulation, and death. Cell 1994, 76:959-962 Nagata S, Goldstein P: The Fas death factor. Science 1995, 267:1449-1456 Meredith JE, Fazeli B, Schwartz MA: The extracellular matrix as a survival factor. Mol Biol Cell 1993, 4:953-961 Frisch SM, Francis H: Disruption of epithelial cell-matrix interactions induces apoptosis. J Cell Biol 1994, 124: 619-626 Brooks PC, Montgomery AMP, Rosenfeld M, Reisfeld RA, Hu T, Klier G, Cheresh DA: Integrin aV13 antagonists promote tumor regression by inducing apoptosis of angiogenic blood vessels. Cell 1994, 79:1157-1164
