Extracellular matrix organization modulates fibroblast growth and growth factor responsiveness

Experimental Cell Research 182 (1989) 512-582

Extracellular Matrix Organization Modulates Fibroblast Growth and Growth Factor Responsiveness SHIGENORI Department

NAKAGAWA, PAMELA PAWELEK, and FREDERICK GRINNELL’

of Cell Biology

and Anatomy, UT Southwestern Dallas, Texas 75235

Medical

Center,

To learn more about the relationship between extracellular matrix organization, cell shape, and cell growth control, we studied DNA synthesis by fibroblasts in collagen gels that were either attached to culture dishes or floating in culture medium during gel contraction. After 4 days of contraction, the collagen density (initially 1.5 mg/ml) reached 22 ms/ml in attached gels and 55 mg/ml in floating gels. After contraction, attached collagen gels were well organized; collagen fibrils were aligned in the plane of cell spreading; and fibroblasts had an elongated, bipolar morphology. Floating collagen gels, however, were unorganized; collagen fibrils were arranged randomly; and tibroblasts had a stellate morphology. DNA synthesis by tibroblasts in contracted collagen gels was suppressed if the gels were floating in medium but not if the gels were attached, and inhibition was independent of the extent of gel contraction. Therefore, growth of fibroblasts in contracted collagen gels could be regulated by differences in extracellular matrix organization and cell shape independently of extracellular matrix density. We also compared the responses of fibroblasts in contracted collagen gels and monolayer culture to peptide growth factors including fibroblast growth factor, platelet-derived growth factor, transforming growth factor-/I, and interleukin 1. Cells in floating collagen gels were generally unresponsive to any of the growth factors. Cells in attached collagen gels and monolayer culture were affected similarly by fibroblast growth factor but not by the others. Our results indicate that extracellular matrix organization influenced not only cell growth, but also fibroblast responsiveness to peptide growth factors. 0 1989 Academic Press, hc.

Most research on fibroblast growth and function has utilized monolayer cell cultures. Recently, however, we and others have become interested in fibroblast activities as they are expressed by cells cultured in hydrated collagen gels. One reason for this interest is that fibroblasts in collagen gels, unlike fibroblasts in monolayer culture, have morphological features similar to those of fibroblasts in uiuo [l-3], which suggests that the collagen gel environment mimics the in uiuo situation more closely than monolayer culture. Another important feature of collagen gels is that they can be physically reorganized by fibroblasts. This extracellular matrix remodeling process, which was not appreciated from studies of cells in monolayer culture, results in gel contraction [4-6], and has been likened to the in uiuo process of wound contraction [7, 81. The force of contraction is developed through cell motility [9] and depends on the actin cytoskeleton, since contraction can be inhibited by cytochalasins [4, 10, 111. ’ To whom all reprint requests should be addressed. Copy&t @ 1989 by Academic Press, Inc. AII rights of reproductmn in any form reserved 0014-4827/89$03.00

512

Matrix organization and cell growth control

573

Previous studies showed that DNA synthesis and collagen synthesis by tibroblasts decreased after collagen gel contraction [12, 131,and that the extent of this decrease could be correlated with the degree of gel contraction [14, 151.Lower synthetic activity was suggestedto occur as a result of the changed cell shape that accompanied gel contraction [16, 171, an idea previously proposed to explain growth inhibition of lens epithelial cells cultured on contracted collagen gels [ 18, 191.In part, decreased DNA synthesis also might have been a direct consequence of collagen tibril binding to the cells [ 1, 201. In the studies cited above, there were simultaneous changes in cell shape and the extent of gel contraction. Moreover, most investigators analyzed only contraction of floating collagen gels because they measured gel diameter and not gel volume. Attached collagen gels contract without changing gel diameter 16, 1II. To clarify and extend the previous studies, it was necessary to compare DNA and collagen synthesis by fibroblasts in contracted collagen gels under conditions permitting the extent of gel contraction and cell shape to be varied independently. We have performed such experiments and also have tested the responsivenessof libroblasts in contracted collagen gels to peptide growth factors. The results of our studies are reported herein. MATERIALS AND METHODS Cell culture in hydrated collagen gels. Human skin fibroblast monolayer cultures were established from foreskins obtained at circumcisions and maintained and harvested as before [ 111.Preparation of hydrated collagen gels prepared from Vitrogen “100” collagen (Collagen Corp., Palo Alto, CA) has been described previously [ll]. Fibroblasts were added to neutralized collagen solutions (1.5 mg/ml) at a concentration of ld cells/O.2 ml. Aliquots (0.2 ml) of the cell/collagen mixtures warmed to room temperature were placed in Costar 24-well culture plates. Each aliquot occupied an area outlined by a 12 mm diameter circular score within the well. Gels were polymerized by raising the temperature to 37°C and incubating the samples for 60 min, after which 1.0 ml of culture medium (Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum (FBS) and 50 &ml ascorbic acid) was added to each well. After polymerization, the tibroblasts were dispersed throughout the gels. To obtain cultures of floating collagen gels, the attached gels were gently lifted off the bottom of the wells with a spatula. In experiments to test the effects of peptide growth factors, monolayer cell cultures were prepared in Costar 24-well culture plates with 0.6x lo6 libroblasts/well and 1.Oml of culture medium as above. Cells in monolayer culture, attached collagen gels, and floating collagen gels were cultured for 2 days in 10% serum-containing medium after which fresh medium containing 2% FBS (instead of 10% FBS) and growth factors as indicated were added. It was not possible to perform the experiments in the absence of serum or with low serum in the medium throughout the culture period because serum is required for gel contraction to proceed normally [ll, 2 I]. The growth factors tested were transforming growth factor-p (TGF-B) [22], platelet-derived growth factor (PDGF) [23], tibroblast growth factor (FGF) [24], and interleukin 1 (IL-l) [25], which all have been proposed to be important in wound repair and shown to modulate growth of tibroblasts in monolayer culture. Highly purified growth factors were purchased as follows: TGF-p (human platelet) from Calbiochem; FGF (bovine pituitary) and PDGF (human, 420,000 half-maximal U/mg) from Collaborative Research; and IL-1 (human, 8x lo6 U/ug) from Genzyme. The growth factors were used in the incubations at the concentrations recommended by the suppliers. Measurement ofgel contraction. Gel contraction was determined by measuring gel volume [14]. At the beginning of the incubations, we added 1 uCi of ‘HZ0 (1 mCi/g, New England Nuclear) to the culture medium. At the times indicated, the medium was removed, and the gels were rinsed quickly and dissolved in 0.5 ml of 1 M NaOH. The samples were neutralized with HCl, mixed with 10 ml Budget Solve (RPI Corp.), and counted in a Beckman scintillation spectrometer. Initial gel volume was determined in gels without fibroblasts, and equilibration of radioactivity in these gels required 30 min.

514 Nakagawa, Pawelek, and Grinnell Cell number and DNA synthesis. DNA synthesis was measured by thymidine incorporation. Cultures were radiolabeled metabolically for 3 h with 2 l&i/ml [‘Hlthymidine (20 Wmmol, New England Nuclear) added to the culture medium. At the end of the incubations, the medium was discarded. The gels were solubilized by treatment for 2 h at 37’C with 3 mg/ml collagenase (Type I, Sigma Chemical Co.) in 130 mM NaCl, 10 mM Ca acetate, 20 mM Hepes, pH 7.2. Single cell preparations were obtained by incubating the samples for an additional 20 min at 37°C with 0.05% trypsin (GIBCO) and 10 mM EDTA. Aliquots of the samples were mixed with trypan blue, and cell number was measured with a hemocytometer. The remaining portions of the samples were subjected to three precipitations with 10% TCA at 4°C. The final precipitates were solubiliied with Protosol tissue solubilizer (New England Nuclear), and radoactivity was determined as above. Each experiment was performed in triplicate, and the results were normalized by cell number. The experiments described in this paper were performed over a period of several months with fibroblasts ranging from 3rd to 10th passage, so there was some variability from experiment to experiment in the rates of DNA synthesis. Collagen synthesis. Collagen synthesis was measured by the protease-free collagenase method [26]. Cultures were radiolabeled metabolically for 24 h with 2 uCi/ml L-[3H]proline (126.9 Ci/mmol, New England Nuclear) added in fresh culture medium supplemented with SO&ml /I-aminopropionitrile and 50 &ml ascorbic acid. At the end of the incubations, the medium was removed. The gel (or cell matrix) samples were dissolved by incubating them for 90 min at 37°C with 0.5 ml of 10 mM HCl, and then the solubiliied samples were neutralized with NaOH. One half of each medium and solubilized gel (matrix) sample was brought to 1 ml containing 50 mM Tris, 10 mM Ca acetate, pH 7.2, and treated with collagenase (5 BTC U/ml, Form III, Advanced Biofactures Corp.) for 3 h at 37°C. Bovine serum albumin (0.5 mg/ml) was added to collagenase-treated and untreated samples to act as carrier protein, after which all of the samples were subjected to three precipitations with 10% TCA at 4°C. The precipitates were dissolved with Protosol, and radioactivity was measured as above. Each experiment was performed in triplicate, and the results were normalized by cell number. Collagen synthesis was calculated as total less collagenase-insensitive TCA-precipitable radioactivity. Collagen degradation. Radiolabeled collagen was prepared by acetylation with [‘HIacetic anhydride (2.5 mCi, 50 mCi/mmol, New England Nuclear Corp.) as has been described [ll, 271. Gels containing radiolabeled collagen and fibroblasts were polymerized as described above for nonradiolabeled collagen. At the times when collagen degradation was measured, the culture medium was removed from the incubations. The gels were rinsed briefly and dissolved with 1 M NaOH and neutralized with 1 M HCl, and radioactivity was counted with a scintillation spectrophotometer as above. Microscopy. Samples for microscopic analysis were fmed for 2 h at 4°C with 2% glutaraldehyde, 1% paraformaldehyde, 1% tannic acid, in 0.1 M Na cacodylate (pH 7.4), postfixed for 30 min at 22°C with 1% aqueous OsO,, and stained en bloc with 1% aqueous uranyl acetate for 30 min at 4°C. After dehydration, specimens were embedded in Epon 812. Thick sections were stained with 1% toluidine blue and examined and photographed using a Zeiss Photomicroscope III. Thin sections were observed and photographed with a Philips 300 electron microscope.

RESULTS Contraction of Floating and Attached Collagen Gels We compared the rates of fibroblast contraction of attached and floating gels by measuring gel volume. In either case, the decrease in gel volume showed a biphasic pattern in which there was an initial period of rapid contraction (2-4 h) followed by a subsequent period of slower contraction (4 h-4 days) (Fig. 1). Contraction of attached gels occurred more slowly than contraction of floating gels (e.g., -50% compared to -75 % after 4 h). The time course of gel contraction was influenced by distribution of fibroblast in the gels. In the experiments described in this report, the fibroblasts were distributed uniformly throughout the gels. If the collagen/fibroblast mixtures were at 4°C when polymerization was initiated, then most cells settled to the bottom of the gels during polymerization,

Matrix organization and cell growth control

5

2 a ca

75

o---o

Attached

0-e

Floating

575

gel gel

50

1 0 s R

25

0

4

6

12

16

20

24

46

96

HOUR3

Fig. 1. Contraction of attachedand floating collagengels. Collagengels containing fibroblasts were attachedto culture dishesor floating in culture medium. At the times indicated, the gel volume was measured.Other details are describedunder Materials and Methods.

and the rate of gel contraction was markedly lower than with uniformly distributed cells (data not shown). Experiments with radiolabeled collagen showed that during the first 24 h of contraction, about 12 and 13% of the starting collagen was lost from the attached and released gels, respectively. After 4 days 30% of the starting collagen was lost from the gels regardless of whether they were attached or floating. From these data and the volume measurements, it could be calculated that the collagen concentration in attached gels increased from about 1.5 to about 7 mg/ml after 1 day of contraction and to about 22 mg/ml after 4 days of contraction. In floating collagen gels the density was about 28 mg/ml after 1 day and 55 mg/ml after 4 days. Morphology of Cells in Contracted Collagen Gels Morphological features of fibroblasts in contracted collagen gels were examined by light and electron microscopy. After 24 h of culture, cells in attached gels were elongated parallel to the surface of the underlying culture dishes (Figs. 2A and 2 B). Collagen fibrils were aligned in the plane of cell spreading (Fig. 2A), and were bound individually and in small clusters all over the cell surfaces (Fig. 2B). On the other hand, cells in floating gels had a stellate morphology (Figs. 2 C and 20). Collagen fibrils were randomly organized in the floating gels (Fig. 2 C) but bound all over the cell surfaces (Fig. 20). The morphological features of cells in attached gels was independent of cell position in the collagen gels. The morphological features of cells in floating gels was similar for cells surrounded by collagen, but cells that were situated on the outer edges of the gels often spread in a bipolar morphology. Also, the morphologies of cells in attached and floating collagen gels were essentially the same after 4 days of culture as after 24 h (data not shown). DNA and Collagen Synthesis by Fibroblasts in Contracted Collagen Gels DNA synthesis by fibroblasts in attached collagen gels occurred at similar rates regardless of whether the gels were contracted for 1 or 4 days (Fig. 3). Also, DNA

516 Nakagawa, Pawelek, and Grinnell

Fig. 2. Morphological appearance of cells and collagen in contracted collagen gels. Collagen gels containing fibroblasts were attached to culture dishes (A and B) or floating in medium (C and 0) for 24 h. (A and B) After contraction, cells in attached gels were elongated parallel to the surface of the culture dishes and collagen tibrils were aligned similarly. The cell nuclei were fusiform and collagen fibrils were bound individually and in small clusters all over the cell surface. (C and D) After 24 h of contraction in floating gels, the cells had a stellate morphology and collagen tibrils were randomly organized. Collagen fibrils were bound individually and in small clusters all over the cell surface. Other details are described under Materials and Methods. (A and C) Bar= 10 pm; (S and 0) bar= 1 pm.

synthesis by fibroblasts in floating collagen gels occurred at similar rates after 1 or 4 days of contraction (Fig. 3). On the other and, if attached and floating gels were compared to each other, then DNA synthesis was found to be much lower in fibroblasts in floating collagen gels compared to fibroblasts in attached gels. For instance, after 1 day, the level of DNA synthesis in fibroblasts in floating gels was only 15% of that in attached gels. After 4 days, the level of DNA synthesis by cells in floating gels was 21% of that in attached gels. Moreover, DNA synthesis was much higher in attached collagen gels that were contracted 4 days than in floating collagen gels that were contracted for 1 day, even though the extent of gel contraction was similar in these samples. Based on these observations, it could be concluded that suppression of DNA synthesis was dependent not on the extent of gel contraction, but rather on whether the contracted collagen gels were floating or attached.

Matrix organization and cell growth control 5 8

577

3 DNA synthem

Collagen

synthesis ---

D

1 Day

4 Days Culture

1 Day

1

16

2 L)

4 Days

Period

Fig. 3. DNA and collagen synthesis by cells in contracted collagen gels. Attached or floating collagengels were contracted by fibroblasts for the time periods shown. Cultures were radiolabeled with f’H]thymidine to determineDNA synthesisor with [3H]prolineto determinecollagensynthesis. Other details are described under Materials and Methods. 0, Attached gel; n , floating gel; 8, attached3 days+ floating day.

Collagen synthesis also was lower in floating gels than in attached gels, but the difference was not as great as observed for DNA synthesis (Fig. 3). The amount of collagen synthesis by cells in floating gels relative to attached gels was 55% after 1 day and 41% after 4 days. Collagen synthesis decreased as the extent of gel contraction increased, regardless of whether the gels were attached or floating. It should be noted that cells recovered at the end of the incubations from either floating or attached gels were greater than 90% viable based on trypan blue exclusion. As expected from the low rates of DNA synthesis, there was no increase in cell number in floating gels, but cells in attached gels underwent about two doublings during 4 days of culture (data not shown). After contraction, attached collagen gels could be converted rapidly to floating gels by using a spatula to release the attached gels from their underlying substratum. After contracted, attached gels were released from the substratum, cell rounding occurred within 60 min. By 24 h after release, fibroblasts developed a stellate morphology, and the collagen fibrils became randomly arranged and more densely packed (data not shown). In short there was a rapid transition from attached to floating gel morphology. Concomitant with these morphological changes,DNA synthesis was suppressed, e.g., when attached gels were contracted 3 days and then released and allowed to float in medium for one additional day (Fig. 3). Effect of Growth Factors on DNA Synthesis by Fibroblasts in Collagen Gels

The above results indicated that DNA and collagen synthesis by fibroblasts was modulated in response to a change from bipolar to stellate cell shape that accompanied extracellular matrix reorganization. Various growth factors also have been shown to modulate DNA and collagen synthesis by tibroblasts in monolayer culture, but the effects of these factors on cells in collagen gels have not be reported. Therefore, we selected severalof the factors and measured their

518 Nakagawa, Pawelek, and Grinnell

Att-G

Fl-G Culture

Mono

Condltrons

Fig. 4. Effect of FGF on DNA synthesis by cells in monolayer culture and contracted collagen gels. Fibroblasts were in monolayer culture (mono), floating collagen gels (fl-g), or attached collagen gels (att-g) for 4 days. During the last 2 days, the medium contained 2% FBS and FGF at the concentrations indicated (0, none; LTI, 0.06 I1M; gS, 0.3 nM; W, 1.5 n&f). Cultures were radioIabeled with [rH]thymidine to determine DNA. Other details are described under Materials and Methods.

effects on DNA and collagen synthesis by cells in attached gels and floating gels. The growth factors, all highly purified, were obtained commercially and used at the concentrations recommended by the suppliers. For the purposes of comparison, parallel experiments were performed to determine the-effects of the-growth factors on human fibroblasts in monolayer culture. None of the growth factors had a noticeable effect on cell morphology nor restored the rate of DNA synthesis by cells in contracted, floating collagen gels to the levels observed in contracted, attached collagen gels (Figs. 4-7). Therefore, even though the the cells in floating collagen gels were attached and partially spread (i.e., in a stellate shape), they were unresponsive to the factors. Also, the overall responses of cells in monolayer culture and attached collagen gels to the growth factor were much different. Addition of FGF promoted DNA synthesis by fibroblasts in monolayer culture and by cells in attached collagen gels (Fig. 4).

Fl-C

Att-C Culture

Mono

Condltmns

Fig. 5. Effect of PDGF on DNA synthesis by cells in monolayer culture and contracted collagen gels. Fibroblasts were in monolayer culture (mono), floating collagen gels @‘l-g),or attached collagen gels (at&g), for 4 days. During the last 2 days, the medium contained 2 % FBS and PDGF at the concentrations indicated (Cl, none; LH,0.008 nM; @I,0.04 nM, n , 0.2 r&f). Cultures were radiolabeled with [3H]thymidine to determine DNA. Other details are described under Materials and Methods.

Matrix organization and cell growth control

579

-2 e4 do z 2 22 zm f PO Att-G

Fl-G Culture

Mono

Condbms

Fig. 6. Effect of IL-l on DNA synthesis by cells in monolayer culture and contracted collagen gels. Fibroblasts were in monolayer culture (mono), floating collagen gels (fl-g), or attached collagen gels W-g) for 4 days. During the last 2 days, the medium contained 2% FBS and IL-1 at the concentrations indicated (Cl, none; IXI, 0.1 U/ml; q l, 0.5 U/ml; n , 1 U/ml). Cultures were radiolabeled with [‘Hlthymidine to determine DNA. Other details are described under Materials and Methods.

Addition of PDGF promoted DNA synthesis by fibroblasts in monolayer culture but had relatively little effect on fibroblasts in attached collagen gels (Fig. 5). Finally, addition of IL-l resulted in a modest increase in DNA synthesis by cells in monolayer culture, but a marked suppression of DNA synthesis by cells in contracted collagen gels (Fig. 6). In other experiments, we found that these growth factors had little effect on collagen synthesis by cells in monolayer culture or in collagen gels under the conditions tested (data not shown). With added TGF-/?, DNA synthesis by cells in monolayer culture was inhibited slightly whereas DNA synthesis by cells in attached collagen gels was increased (Fig. 7). Similar concentrations of TGF-B also promoted collagen synthesis by tibroblasts in monolayer culture and attached collagen gels, but not by fibroblasts in floating collagen gels (Fig. 8, top). In the absence of TGF-/? about 50% of the collagen synthesized in attached collagen gels and 75% of the collagen synthe-

Att-G

FI-C Culture

Mono

Condhms

Fig. 7. Effect of TGF-b on DNA synthesis by cells in monolayer culture and contracted collagen gels. Fibroblasts were in monolayer culture (mono), floating collagen gels (fl-g), or attached collagen gels (att-g) for 4 days. During the last 2 days, the medium contained’2% FBS and TGF;B at the concentrations indicated (Cl, none; q , 0.016 nM, lB, 0.08 nM; W, 0.4 nit4). Cultures were radiolabeled with [‘H]thymidine to determine DNA. Other details are described under Materials and Methods.

580 Nakagawa, Pawelek, and Grinnell

Att-G Culture

Fl-G Condltmns

Mono

Fig. 8. Effect of TGF-/I on collagen synthesis by cells in monolayer culture and contracted collagen gels. Fibroblasts were in monolayer culture (mono), floating collagen gels (fl-g), or attached collagen gels (att-g) for 4 days. During the last 2 days, the medium contained 2% FBS and TGF-/J at the concentrations indicated (0, none; 83, 0.16 nM; q l, 0.08 IN; W, 0.4 t&f). Cultures were radiolabeled with [)H]proline to determine collagen synthesis. Other details are described under Materials and Methods.

sized in floating collagen gels were incorporated into the gels, whereas less than 25% of the collagen synthesized in monolayer culture was incorporated into the extracellular matrix unless TGF-fi was added (Fig. 8, bottom). DISCUSSION To learn more about the relationship between extracellular matrix organization, cell shape, and oell growth control, we studied DNA synthesis by fibroblasts in collagen gels that were either attached to culture dishes or floating in culture medium during gel contraction. Fibroblasts contracted collagen gels under both conditions although contraction of floating gels occurred at a slightly faster rate. After 4 days of contraction, the collagen density (initially 1.5 mg/ml) reached 22 mg/ml in attached gels and 55 mg/ml in floating gels. DNA synthesis by tibroblasts in contracted collagen gels was suppressed if the gels were floating in medium but not if the gels were attached to a surface, and inhibition was independent of the extent of gel contraction. Therefore, growth of tibroblasts in contracted collagen gels can be regulated by extracellular matrix organization and cell shape independently of extracellular matrix density. Previous investigators showed that DNA synthesis by fibroblasts in collagen gels decreased after gel contraction [12, 141, but they did not distinguish between the extent of gel contraction on the one hand, and extracellular matrix organization and cell shape on the other. After contraction, attached collagen gels were well organized; collagen fibrils were aligned in the plane of cell spreading; and fibroblasts were elongated, bipolar cells. In marked contrast, floating collagen gels were unorganized; colla-

Matrix organization and cell growth control

581

gen fibrils were arranged randomly; and fibroblasts were stellate cells. It has been shown previously that cell shape plays a role in growth control of monolayer cells P81, and DNA synthesis by nontransformed cells was found to be tightly coupled to cell spreading [291. In those experiments, however, the cells were attached on a planar substratum, and most of the cell surface was unattached. Moreover, anchorage-dependentcells that are unable to grow in agar or similar polymers are completely unattached. For some cells, however, growth stimulation depends on cell attachment independently of cell spreading [30]. The situation in collagen gels is much different because the cells have collagen tibrils bound all over their surfaces regardless of cell shape, bipolar or stellate. The precise mechanism by which cell shape influences DNA synthesis is still a matter of speculation. The collagen gel contraction model is uniquely suited to study this problem because changes in cell shape can be induced without enzymatic treatment simply by releasing attached gels from the underlying substratum [cf. 181. Collagen synthesis by fibroblasts in contracted collagen gels was suppressed not only in floating gels compared to attached gels, but also in gels contracted 4 days compared to gels contracted 1 day. Therefore, unlike DNA synthesis, the rate of collagen synthesis appered to correlate with the extent of gel contraction, so there may be a direct feedback inhibition mechanism. The difference between collagen synthesis measured in attached and floating gels also might be a consequence of collagenaseactivity since synthesis of procollagenase was increased by tibroblasts in contracted, floating collagen gels [ 161. In addition to studying the effect of collagen gel contraction on DNA synthesis by fibroblasts in the gels, we also determined the responsivenessof these cells to severalpeptide growth factors including fibroblast growth factor, platelet-derived growth factor, transforming growth factor-p, and interleukin 1. Cells in floating collagen gels were generally unresponsive to any of the growth factors.. Cells in attached collagen gels and monolayer culture responded similarly to fibroblast growth factor but not to the other growth factors. The basis for the differences is currently unknown. One possibility is that extracellular matrix organization and cell shape influence expression of growth factor receptors. Recently, for instance it was shown that fibroblasts express receptors for B type PDGF in vitro but not in uiuo [31]. Another possibility is that serum mitogens bind to collagen gels and influence interactions with purified growth factors. For instance, some studies have shown that the extracellular matrix can serve as a reservoir for growth factors such as PDGF and FGF [32-341. We hope to clarify these points in the future. Regarding the inhibition of DNA synthesis by fibroblasts in collagen gels in response to IL-l, it is noteworthy that IL-1 had an inhibitory effect on the formation of granulation tissue in uiuo when IL-l was injected into implanted sponges1351,and this effect would not have been predicted considering that IL-l stimulates fibroblasts in monolayer culture [25]. Also, interesting was the ability of TGF$ to promote DNA and collagen synthesis by fibroblasts in attached collagen gels, which supports the idea that TGF-/I directly stimulatesgranulation tissue formation during wound repair [22]. Moreover, collagen that was newly

582 Nakagawa, Pawelek, and Grinnell

synthesizedby fibroblasts in collagen gels was incorporated into the extracellular matrix to a greater extent than collagen synthesized by cells in monolayer culture [cf. 131.These fmdings illustrate the potential usefulnessof attached collagen gels contracted by fibroblasts as a model system for studying features of extracellular matrix organization and wound repair in vitro. We are grateful to Drs. William Snell and George Bloom for their advice and suggestions. This research was supported by grants from the Kendall Health Care Products Co. and the NIH (DM31321).

REFERENCES 1. Elsdale, T., and Bard, J. (1972) J. Cell Biol. 84, 626-637. 2. Chinnell, F., and Bennett, M. H. (1981) J. Cell Sci. 48, 19-34. 3. Tomasek, J. J., Hay, E. D., and Fujiwara, K. (1982) Deu. Biol. 92, 107-122. 4. Bell, E., Ivarsson, B., and Merrill, C. (1979) Proc. N&l. Acad. Sci. USA 76, 12741278. 5. Bellows, C. G., Melcher, A. H., and Aubin, J. E. (1981) J. Cell Sci. 58, 299-314. 6. GrinneIl, F., and Lamke, F. R. (1984) .I. Cell Sci. 66, 51-63. 7. Bellows, C. G., Melcher, A. H., Bhargava, U., and Aubin, J. E. (1982) J. Ultrastruct. Res. 78, 178-192. 8. Ehrlich, H. P., and Wyler, D. J. (1983) J. Cell. Physiol. 116, 345-351. 9. Harris, A. K., Stopak, D., and Wild, P. (1981) Nature (London) 298, 249-251. 10. Bellows, C. G., Melcher, A. H., and Aubin, J. E. (1982) J. CeN Sci. 58, 125-138. 11. Guidry, C., and Grinnell, F. (1985) J. Cell Sci. 79, 67-81. 12. Sarber, R., Hull, B., Merrill, C., Soranno, T., and Bell, E. (1981)Me&. Aging Deu. 17,107-l 17. 13. Nusgens, B., Merrill, C., Lapiere, C., and Bell, E. (1984) Coll. Rel. Res. 4, 351-363. 14. Van Bockxmeer, F. M., Martin, C. E., and Constable, I. J. (1984) Exp. Cell Res. 155, 413-421. 15. Paye, M., Nusgens, B. V., and Lapiere, C. M. (1987) Eur. J. Cell Biol. 45, 44-50. 16. Unemori, E. N., and Werb, Z. (1986) J. Cell Biol. 183, 1021-1031. 17. Mauch, C., Hatamochi, A., Scharffetter, K., and Krieg, T. (1988) Exp. Cell Res. 178, 493-503. 18. Iwig, M., Glaesser, D., and Bethge, M. (1981) Exp. Cell Res. 131, 47-5’5. 19. Iwig, M., and Glaesser, D. (1985) Cell Tissue Kinet. 18, 169-182. 20. Yoshizato, K., Taira, T., and Yamamoto, N. (1985) Biomed. Res. 6, 61-71. 21. Steinberg, B. hf., Smith, K., Colozzo, M., and Pollack, R. (1980) J. Cell Biol. 87, 306308. 22. Spom, M. B., Roberts, A. B., Wakefield, L. M., and De Crombrugghe, B. (1987) J. Cell Biol. 105, 1039-1045. 23. Ross, R., Raines, E. W., and Bowen-Pope, D. F. (1986) Cell 46, 155-169. 24. Gospodarowicz, D., Neufeld, G., and Schweigerer, L. (1986) Mol. Cell. Endocrinol. 46, W-204. 25. Dim&lo, C. A. (1988) Fed. Proc. Fed. Amer. Sot. Exp. Biol. 2, 108-115. 26. Peterkofsky, B. (1982) in Methods in Enzymology (Cunningham, L. W., and Frederiksen, D. W., Eds.), Vol. 82, pp. 453471, Academic Press, New York. 27. Schor, S. L., Allen, T. D., and Harrison, C. J. (1980) J. Cell Sci. 46, 171-186. 28. Maroudas, N. G. (1973) Exp. Cell Res. 81, 104-110. 29. Folkman, J., and Moscona, A. (1978) Nature (London) 273, 345-349. 30. O’Neill, C., Jordan, P., and Ireland, G. (1986) Cell 44, 489-496. 31. Terracio, L., Ronnstrand, L., Tingstrom, A., Rubin, K., Claesson-Welch, L., Funa, K., and Heldin, C.-H. (1988) J. Cell Biol. 107, 1947-1957. 32. Smith, J. C., Singh, J. P., Lillquist, J. S., Goon, D. 8, and Stiles, C. D. (1982) Nature (London) 296, 154-156. 33. Roberts, R., Gallagher, J., Allen, T. D., Bloomfleld, F., and Dexter, T. M. (1988) Nature

(London) 332, 376378. 34. Folkman, J., Klagsbrun, M., Sasse, J., Wadziiski, M., Ingber, D., Vlodavsky, I. (1988)Amer. J. Pathol. 130, 393-400. 35. Laato, M., and Heino, J. (1988) Experientiu 44, 32-34. Received December 20, 1988 Revised version received February 8, 1989 Rinted

in Sweden