Transforming Growth Factor & Selectively Augments Collagen Synthesis by Human Intestinal Smooth Muscle Cells MARTIN F. GRAHAM, GENE R. BRYSON, and ROBERT F. DIEGELMANN Divisions of Pediatric Gastroenterology [Children’s Medical Center) and Plastic Surgery (MCV Wound Healing Center), Medical College of Virginia, Richmond, Virginia
Intestinal smooth muscle cells play a major role in the stricture formation that complicates chronic intestinal inflammation, by proliferating and producing collagen. Transforming growth factor & has been identified as an important inflammatory mediator in the fibrotic response of human tissue to inflammation. To determine whether this mediator might be involved in intestinal fibrosis, the effect of transforming growth factor & on collagen production and proliferation by human intestinal smooth muscle cells was studied in vitro. Cells in the second passage were grown to subconfluence in medium containing 10% Nu-Serum (Collaborative Research Inc., Bedford, MA), after which the concentration of NuSerum was decreased. Forty-eight hours later, transforming growth factor & was added to the culture medium to achieve concentrations of 1-500 pmol/L. After 24 hours exposure to the transforming growth factor &, cellular collagen synthesis was determined by the uptake of [SH]proline into collagenase-sensitive protein. Transforming growth factor & caused a 100% increase in collagen production and a 40% increase in noncollagen protein production per cell, reflecting an increase in relative collagen synthesis of 58%. This effect was maximal at a concentration of 10 pmol/L. Epidermal growth factor, by comparison, had no significant effect on relative collagen synthesis. Transforming growth factor & caused a signiflcant increase in the uptake of methylaminoisobutyric acid, a nonmetabolized amino acid analog, into the cells at IO pmol/L. However, this effect was small (20% increase] compared with the effect on the uptake of proline into collagen (loo?70 increase] at this concentration. When cell proliferation was examined by the uptake of [Wlthymidine, transforming growth factor fll had no effect, whereas epidermal growth factor (1000 pmol/L) caused a 94% increase. Transforming growth factor & selectively augments
collagen production by human intestinal smooth muscle cells in vitro. This effect is potent and is not related to effects on either cell proliferation or amino acid uptake. These data suggest that transforming growth factor & has an important role as an inflammatory mediator in the pathogenesis of intestinal fibrosis.
S
tricture formation, a common complication of chronic intestinal inflammation, leads to significant morbidity. Our recent studies have shown that intestinal smooth muscle cells play a major role in the pathogenesis of strictures (1). The evidence suggests that the lesions result from proliferation of and collagen production by the smooth muscle cells, a response on the part of these local mesenchymal cells that is common to many tissues in the body. To define further the pathogenesis of intestinal fibrosis, we have developed and characterized an in vitro model of human intestinal smooth muscle (HISM) cells. When grown in culture, HISM cells produce large amounts of collagen compared with other human mesenchymal cells such as fibroblasts (2). Growth in culture is inhibited by heparin (3), and collagen production is selectively downregulated by cyclic adenosine monophosphate (4) and by protamine sulfate (5). Having characterized the behavior of intestinal smooth muscle cells in vitro, we are now using this model to characterize the effects of defined inflammatory mediators on the cells, and hence determine which of these mediators are involved in the fibrotic
Abbreviations used in this paper: DMEM, Dulhecco’s mod&d Eagle medium; EGF, epidermal growth factor: HISM, human intestinal smooth muscle; TGF-&, transform@ growth factor 8,. 0 IBSOby the American Gastroenterologkal Aseodation 0018-5085/80/$3.00
448
GRAHAM ET AL.
GASTROENTEROLOGY
response of smooth muscle cells to inflammation in vivo. The mediator(s) in question would be either mitogenic or fibrogenic or both, and would be produced by one or more of the inflammatory elements that characterize chronic inflammation including platelets, lymphocytes, macrophages, and mast cells. Mesenchymal cell mitogens have been well defined and HISM cells proliferate in vitro in the presence of fetal bovine serum (3, the major mitogen in serum being platelet-derived growth factor (6). As far as fibrogenic factors are concerned, nothing is known at present about the effect of defined inflammatory mediators on the capacity of intestinal mesenchymal cells to produce collagen. Transforming growth factor & (TGF-P,) has been identified as one of the few mediators that specifically induces a fibrotic response in tissue (7). We were therefore intrigued to determine whether TGF-P, regulated collagen synthesis by HISM cells, and whether this mediator might be involved in the pathogenesis of intestinal strictures. Materials and Methods Cell Isolation and Culture Human intestinal smooth muscle cells were isolated from the muscularis propria of normal human jejunum using crude bacterial collagenase digestion as previously described (8). Cells were maintained in primary culture for 3 weeks in Dulbecco’s modified Eagle medium (DMEM) containing 10% Nu-Serum (Collaborative Research Inc., Bedford, MA: DMEM-lo), after which they were passaged by trypsinization for use in the experiments described below.
Determination of Nu-Serum
of the Optimal
Concentration
in Culture Medium for Mediator
Experiments Because fetal bovine serum contains inflammatory mediators, this growth supplement is unsuitable as an additive to culture medium in experiments designed to test the effect of inflammatory mediators on the cells. These problems can be overcome by employing a more defined cell growth supplement, and by decreasing the concentration of the supplement in the culture medium to the extent that the cells are quiescent yet metabolically stable and responsive. In these experiments, therefore, fetal bovine serum was replaced by a commercially produced, partially defined culture medium supplement-Nu-Serum. Initial experiments demonstrated that Nu-Serum was a satisfactory replacement for fetal bovine serum; HISM cells proliferated and produced collagen in a similar fashion when cultured with Nu-Serum as compared to fetal bovine serum [data not shown]. An experiment was then performed to determine the optimal minimal concentration of Nu-Serum for the TGF-& experiments. The HISM cells were grown to confluence in loo-mm culture dishes in DMEM-10. The concentra-
Vol. 99, No. 2
tion of Nu-Serum in the culture medium was then decreased in tenfold dilutions from 9.1% to 0.0001% and to 0% (time 0). The culture medium was changed daily with fresh ascorbate added (0.1 mmol/L). Each plate received 5 mL of pulsing medium containing L-[5-3H]proline (4 mCi/mL, 22 Ci/ mmol/L: Amersham Corp., Arlington Heights, IL) for 5 hours at time 0, and 48 hours and 96 hours later for the determination of collagen synthesis at those times. Collagen synthesis was assayed as previously described (21.
Effect of Transforming Growth Factor ,& on Collagen Synthesis by Human Intestinal Smooth Muscle Cells This experiment was performed three times, twice using cells cultured in loo-mm culture plates and a standard assay (9) and once using cells in microwell culture plates and a newly developed modification of the standard assay (10). The experimental protocols were identical in all three experiments. The reason for miniaturizing the standard assay was to facilitate an increase in the number of replicates for each data point (n = 6), an increase in the number of data points and a decrease in the cost of cells, mediators, isotopes, and media. For the standard macroassay, readers should refer to previous papers (2). The microassay was performed as follows: HISM cells were plated at a density of 3 x lo4 per well in 24-well culture plates (Costar, Cambridge, MA). After growth to subconfluence in DMEM-10 (3 days), the concentration of Nu-Serum in the medium was decreased to 0.0001% (DMEM-10e4) for a further 48 hours. Human platelelet-derived TGF-& [a kind gift from Drs. Michael Sporn and Anita Roberts, NIH, Bethesda, MD) was then added to the cultures to achieve final concentrations of l-500 pmol/L. Epidermal growth factor (EGF, 10-1000 pmol/L, Collaborative Research Inc., Bedford, MA) was added to an additional set of wells. After another 24 hours, all the cultures were incubated with L-[5-3H]proline (40 mCi/mL, 22 Ci/mmol/L) in the presence of the above agonists for 5 hours. The culture plates were then heated to 90°C to stop isotope incorporation and then frozen. After thawing, the cells were sonicated and 200 mL of the contents of each well were removed for determination of DNA content (11). A 1 mL solution of chick embryo carrier protein (2 mg/mL in 10 mmol/L proline) was then added to each well. Radioactive protein was separated from unincorporated L-[5-3H]proline by repeated precipitation with 5% trichloracetic acid at 4°C and the remaining trichloracetic acid was removed by ethanol/ether (3:1). The dried residue in each well was incubated for 90 minutes with 0.5 mL of buffer containing HEPES (60 mmol/ pH 7.21, CaCl, 10.25 mmol], and N-ethylmaleimide (1.25 mmol). Protein was reprecipitated with trichloracetic acid and the radioactivity in the solute was measured as an incubation blank. Thrichloracetic acid was removed, and the dried residue was suspended in the incubation buffer with purified bacterial collagenase (9). After a go-minute incubation, both the soluble [3H]collagen peptides and the [3H]noncollagen protein were counted. Collagen synthesis was expressed both on a per-cell basis (per nanogram DNA) and relative to total protein synthesis (relative collagen synthesis] according to a
TGF-8, INCREASES
August 1990
standard formula that accounts for the enriched content of collagen (12).
amino acid
COLLAGEN
SYNTHESIS
449
IN HISM CELLS
1100
900
Effect of Transforming Growth Factor & on Amino Acid Uptake by Human Intestinal Smooth Muscle Cells The protocol for this experiment was identical to that described for the collagen synthesis experiment above. The assay for amino acid uptake was performed as published (13). After addition of mediators to the culture medium for 24 hours, cells were washed twice with Hank’s balanced salt solution (pH 7.5) containing 25 mmol/L HEPES and 0.1% bovine serum albumin. The cells were then incubated for 25 minutes in the above salt solution containing both [14C]methylaminoisobutyric acid (0.5 mCi/mL, 48 mCi/mmol/ L, NEN Products, Boston, MA), cold (20 mmol/L) methylaminoisobutyric acid [Sigma, St. Louis, MO), and mediators. After washing, the cells were solubilized in 0.2 N sodium hydroxide, neutralized with an equal volume of 0.2 N hydrochloric acid, and the radioactivity of the contents of each well was determined by liquid scintillation counting.
Effect of Transforming Growth Factor & on /%/Thymidine Uptake by Human Intestinal Smooth Muscle Cells
Human intestinal smooth muscle cells were plated at a density of 3 x lo4 per well in 24-well culture plates and cultured in DMEM-10. After 2 days, DMEM-10 was replaced by DMEM-10m4. Forty-eight hours later, mediators were added to the culture medium and after 24 hours in the presence of the mediators, cells were incubated for 4 hours with culture medium containing [3H]thymidine (1 mCi/mL, mCi/mmol/L, NEN Products, Boston, MA], and mediators. The cells were then precipitated and washed with 5% trichloracetic acid at 4°C and solubilized in 0.2 N sodium hydroxide. The contents of each well were then counted for radioactivity. Statistical
Analysis
Experimental values were compared to control values by Student’s t test and were considered significantly different at the P < 0.05 level.
Results Optimal Concentration
of Nu-Serum
Analysis of collagen production by HISM cells in culture medium supplemented with progressive log dilutions of Nu-Serum demonstrated that 0.0001% was the optimal concentration of Nu-Serum for the TGF-& experiments that followed. At that concentration of Nu-Serum, collagen synthesis decreased significantly for the first 48 hours but then remained unchanged for the subsequent 48 hours (Figure 1). During
700
500
300
** I””
6
43
9’8
HOURS Figure 1. The effect of low concentrations of Nu-Serum on collagen synthesis by HISM cells. HISM cells were grown to confluence in culture medium supplemented with 10% NuSerum. At time 0,the concentration of NuSerum in the culture medium was decreased to O.OOl%,O.OOOl%,or 0%. Cells were pulsed with b[5-sII]proline for 5 hours at time 0,at 46 hours, and at 96 hours for the determination of collagen synthesis at those times. The values are the mean + SEM, n = 4. ?? Signf!kantly greater than, **signiEcantly less than the previous value at 46 hours (P < 0.05)
the second 48-hour period, TGF-/3, would be added to the cultures and collagen synthesis would be assayed. Stable baseline conditions would be essential during that time. In medium supplemented with 0.001% Nu-Serum, collagen production decreased similarly for the first 24 hours, but then increased significantly during the second 48-hour period (Figure 1). In medium without Nu-Serum (O?%),collagen production continued to decrease during the second period (Figure 1). Therefore, both 0.001% Nu-Serum and 0% Nu-Serum were unsuitable whereas 0.0001% was optimal. Similar findings were noted for noncollagen protein synthesis (data not shown]. Collagen Synthesis Exposure of HISM cells to TGF-& resulted in a 100% increase in absolute collagen synthesis per cell [Figure 2A). This effect was seen at 10 pmol/L concentration and did not increase at higher concentrations of TGF-/3, (100, 500 pmol/L). Epidermal growth factor had a similar effect on absolute collagen synthesis, causing a 60% increase. However, the effect of EGF was seen at a log higher concentration (100 pmol/L). Exposure of HISM cells to both TFG-& (100 pmol/L) and EGF (1000 pmol/L) had no additive effect on collagen synthesis (T + E, Figure 2A). Transforming growth factor-p, and EGF had similar effects on the production of noncollagen protein.
450 GRAHAM
GASTROENTEROLOGY
ET AL.
A
* *3
* 4
*
&
4
I
$ z s
1500
zi 9 i g
1000
10
1
100
500 1000 T+E
B
*
*
*
??
H
??
5oc
and EGF (100 pmol/L and 1000 pmol/L, respectively) had no additive effect (T+E, Figure 2B). When collagen synthesis was calculated relative to total protein synthesis, TGF-& was shown to cause a 58% increase. This effect was seen at 10 pmol/L [Figure 2CJ and did not increase at higher concentrations of TGF-& (100, 500 pmol/L). This increase in relative collagen synthesis resulted from a specific increase in collagen production rather than from an inhibition of noncollagen production. In contrast, EGF had no effect on relative collagen synthesis (Figure 2C). This was because EGF caused parallel increases in both collagen and noncollagen synthesis. These data represent the results of one experiment performed with a newly developed microassay for collagen synthesis using 24-well culture plates. Similar results were obtained in two other experiments performed with the standard assay using 100 mm culture dishes and HISM cells isolated from different donors. In those experiments, relative collagen synthesis was maximally increased 100% and 87% by TGF-
** +=d@ 0
z 2000 zi
Vol. 99, No. 2
I 0
I 1
I 10
I 100
I
TGF EGF
I I I 500 1000 T+E
??
EGF
31 0
1
10
100
500 1000 T+E
CONCENTRATION (PM)
Figure 2. The effect of TGF-/I, on (A) absolute collagen synthesis, (B] absolute noncollagen protein synthesis, and (c) relative collagen synthesis by HISM cells. Cells were grown to subconfluence, placed in a quiescent medium for 48 hours, and exposed to the various mediators for 24 hours, after which the assay for collagen synthesis was performed. T+E, TGF-8, (100 pmol/L) + EGF (1000 pmol/L); values are mean * SEM, n = 6 (*signBlcantly greater than control, P < 0.05). The Bgures represent data from one experiment performed in &well culture plates using a newly developed modification of the standard assay. A similar response was seen in two other experiments using lOO-mm culture plates and the standard assay.
There was a small (40%) but statistically significant (P < 0.05) increase in this parameter that was maximal at 100 pmol/L that did not increase further at 500 pmol/L (TGF-8,) or 1000 pmol/L (EGF). Again, TGF-
61.
Amino Acid Uptake Transforming growth factor & has been reported to increase the cellular uptake of amino acids (13). The assay used in the current studies depends, in part, on the uptake of a radiolabeled amino acid. Although the computation of relative collagen synthesis utilized above obviates the potential variable that might result from fluxes in amino acid uptake induced by the agonist under test (121, it was intriguing to determine whether the effects of TGF-P, observed in these experiments were associated with or independent of changes in the rate of amino acid uptake. Transforming growth factor & caused a significant (P < 0.05) increase in amino acid uptake at 10 pmol/L and 100 pmol/L [Table 1). However, the increase at 10 pmol/L was small (20%) compared to the large increase (100%) seen in the incorporation of isotope into collagen at that concentration. Epidermal growth factor had no effect on amino acid uptake (Table 1). These data demonstrate that TGF-6, does increase the uptake of amino acids into HISM cells. However, the data also demonstrate that the effect of TGF-0, on collagen synthesis by HISM cells is not, to any significant degree, caused by the effect of this mediator on amino acid uptake.
rH/Thymidine
Uptake
Transforming growth factor & had no effect on the uptake of [3H]thymidine by HISM cells (Figure 3).
TGF-6, INCREASES COLLAGEN SYNTHESIS IN HISM CELLS
August 1990
Growth Factor &, and Epidermal Growth Factor on the Uptake of “/C]Methylaminoisobutyric Acid lnto Human Intestinal Smooth Muscle Cells
Table 1. Effect of Transforming
Mediator None 10% Nu-Serum TGF-j3, 1 pmol/L 10 pmol/L 100pmol/L EGF 10pmol/L 100pmol/L 1000pmol/L
mAIB uptake (dpm/weJI]
% of control
3100* 110 5010+ 280b
100 161b
3350f 190 3720f 180" 4580+ 230"
108 120b 148b
3180+ 220 3120k 070 3250 + 140
103 101 105
NOTE. After growth to confluence in medium containing 10% Nu-Serum, cells were incubated in medium containing 0.0001% Nu-Serum for 48 hours (control conditions). The cells were then exposed to the mediators for 24 hours, after which isotope was added for the determination of AIB uptake. “Mean + SEM. bSignificantlygreater than control; P < 0.05.
This contrasted with an approximately 95% increase in this parameter produced by either 1000 pmol/L EGF (Figure 3) or 10% Nu-Serum. Moreover, TGF-& did not augment or inhibit the EGF-induced increase in [3H]thymidine uptake (T+ E, Figure 3).
inhibitors (7). An additional, indirect effect of TGF-PI on tissue repair is thought to be the recruitment to the area of inflammation, of macrophages (24) and then the stimulation of those inflammatory cells to produce interleukin 1 (24) and TGF-& (25). Tissue repair in the intestine is a process that has not been investigated systematically in the past. Our recent work has demonstrated that intestinal smooth muscle cells are involved in this repair process in that they, rather than any other mesenchymal cell type, proliferate and produce collagen in response to chronic inflammation in the intestinal wall (1). In addition, our in vitro studies have demonstrated that these cells have the capacity to produce large amounts of collagen (2). An important question that remains is the identity of the inflammatory mediators that are responsible for inducing the healing response on the part of the smooth muscle cells. Degranulating platelets in thrombosed submucosal vessels and activated macrophages are a major feature of chronic inflammatory infiltrates in the intestine (26). It is therefore reasonable to postulate that TGF-/3,, a major product of platelet degranulation (14) and of macrophages (251, and now
12000 rig TGF-b + EGF
Discussion Transforming growth factor & is a homodimeric peptide that belongs to a large family of genes responsible for the growth and differentiation of tissue. In addition to its role in development, TGF-& appears to be particularly important in tissue repair (7, for review). The highest concentration of TGF-& in the body is in platelets (14), attesting to its predominant role in wound healing. The peptide is secreted in a latent form thought to be activated locally by the clotting process (15). Systemic effects are thought to be prevented by binding in plasma to cu,-macroglobulin (16). With regard to tissue repair, the mode of action of TGF-& is to increase the production of extracellular matrix by mesenchymal cells, and to decrease the turnover of the matrix. The peptide induces fibrosis when injected subcutaneously in vivo, selectively increases collagen production in chick, rat, and human fibroblasts (17,18), and increases cellular levels of collagen mRNA (19-22). A nuclear factor 1 binding site has been identified as the promoter that mediates the transcriptional activation of the collagen gene (mouse, a$]) by TGF-0, (23). Decreased matrix turnover is accomplished by decreasing the production of proteases and increasing the production of protease
451
*
8000
I
6000
4000
2000
0
I I
I
I
I
0
1
IO
100
I
I
500
1000
CONCENTRATION
(PM)
I
T+E
Figure 2. The effect of TGF-~9,on the uptake of [QJthymidine by HISM cells. Cells were grown to suboonfluenco, placed in a quiescent medium for 48 hours, and eqnmed to the various conditions for 24 hours after which the assay for thymidine uptake was performed. T+E, TGF-fl, (lOa pmol/L) + EGF (1000pmol/L); values are mean + SEM, n = 6 (*signikantly greater than control, P < 0.05).
452
GRAHAM ET AL.
recognized as an important factor in the healing of injured tissue, is one of those mediators in the intestine. The data presented in this paper support that hypothesis. Transforming growth factor & augmented collagen production by HISM cells to a significant degree and in a selective fashion. The effect was potent in that its threshold was at concentrations of TGF-P, between 1 and 10 pmol/L. The selectivity for collagen production was demonstrated by the 100% increase produced in this parameter vs. a 40% increase in noncollagen protein production. This resulted in a 58% increase in relative collagen synthesis. By comparison, EGF produced a 50% increase in both collagen and noncollagen protein synthesis with no net effect on relative collagen synthesis, In addition, the marked effect of TGF-8, on matrix synthesis was exclusive of any effect on DNA synthesis. The lack of stimulation of HISM cell proliferation by TGF-& is consistent with previous reports studying bovine aortic smooth muscle cells and normal rat kidney fibroblasts (27). In those studies, in addition to the lack of mitogenic effect, TGF-& antagonized the mitogenic effect of EGF. In the current study, no such inhibitory effect was seen. Therefore, TGF-/3, appears to have an important regulatory role in collagen production by HISM cells, but no role in regulating the proliferation of these cells. It is conceivable, therefore, that the mechanism whereby TGF-/3, promotes healing in the intestine is twofold: a stimulation of matrix production by local smooth muscle cells; and a recruitment of inflammatory cells such as macrophages into the area for the production of smooth muscle-cell mitogens. In addition, TGF-/3, may decrease the turnover of the intestinal extracellular matrix, but this question is yet to be examined. Therefore, on the basis of the data from these in vitro studies, we would postulate that TGF-& is an important inflammatory mediator in the pathogenesis of intestinal fibrosis. Further studies of pathological specimens to localize this mediator in situ will be necessary to confirm this hypothesis. References 1. Graham MF. Dienelmann RF, Elson CO, Lindblad WI, Gotschalk N, Gay S, Gay R Collagen content and types in the intestinal strictures of Crohn’s disease. Gastroenterology 1988;94:257-285. Graham MF, Drucker DEM, Diegelmann RF, Elson CO. Collagen synthesis by human intestinal smooth muscle cells in culture. Gastroenterology 1987;92:400-405. Graham MF, Drucker DEM. Perr HA, Ebrlich HP, Diegelmann RF. Heparin modulates human intestinal smooth muscle cell proliferation, protein synthesis and lattice contraction. Gastroenterology 1987;93:801-809. Perr HA, Graham MF, Diegelmann RF, Downs RW. Cyclic nucleotides regulate collagen production by human intestinal smooth muscle cells. Gastroenterology 1989;96:1521-1528. Perr HA, Drucker DEM, Co&ran DL. Diegelmann RF, Lind-
GASTROENTEROLOGY
Vol. 99, No. 2
blad WJ, Graham MF. Protamine selectively inhibits collagen synthesis by human intestinal smooth muscle cells and other mesenchymal cells. J Cell Physiol1989;140:483-470. 6. Ross R, Raines EW, Bowen-Pope DF. The biology of plateletderived growth factor. Cell 1986;46:155-159. 7. Sporn MB, Roberts AB, Wakefield LM, de Crombugghe B. Some recent advances in the chemistry and biology of transforming growth factor-& J Cell Biol1987:105:1039-1045. 8. Graham MF, Diegelmann RF, Elson CO, Bitar KN, Ehrlich HP. Isolation and culture of human intestinal smooth muscle cells, Proc Sot Exp Biol Med 1984;176:503-507. 9. Peterkofsky B, Diegelmann RF. The use of a mixture of proteinase-free collagenase for the specific assay of radioactive collagen in the presence of other proteins. Biochemistry 1971;10:988994. 10. Diegelmann RF, Bryson GR, Flood LC, Graham MF. Microassay to quantitate collagen synthesis by cells in culture. Anal Biochem 1990;186:296-300. 11. Labarca C, Paigen K. A simple, rapid and sensitive DNA assay procedure. Anal Biochem 1980;102:344-352. 12. Diegelmann RF, Peterkofsky B, Collagen biosynthesis during connective tissue development in chick embryo. Dev Biol 1972;28:443-453. 13. Boerner P, Resnick RJ, Racker E. Stimulation of glycosis and amino acid uptake in NRK-49F cells by transforming growth factor-@ and epidermal growth factor. Proc Nat1 Acad Sci USA 1985;82:1350-1353. 14. Assoian RK, Komoriya A, Meyers CA, Miller DM, Sporn MB. Transforming growth factor-8 in human platelets: identification of a major storage site, purification and characterisation. J Biol Chem 1983;258:7155-7160. 15. Wakefield LM, Smith DM. Flanders KC, Sporn MB. Latent transforming growth factor-8 from human platelets. J Biol Chem 1988;263:7646-7654. 16. O’Connor-McCourt MD, Wakefield LM. Latent transforming growth facto@ in serum. J Biol Chem 1987;262:14090-14099. 17. Roberts AB, Sporn MB, Assoian RK, Smith JM, Roth NS, Wakefield LM, Heine UI, Liotta LA, Falanga V, Kehrl JH, Fauci AS. Transforming growth factor type-p: rapid induction of fibrosis and angiogenesis in vivo and stimulation of collagen synthesis in vitro. Proc Nat1 Acad Sci USA 1986;83:4167-4171. 18. Ignotz R, Massague J. Transforming growth factor+3 stimulates the expression of fibronectin and collagen and their incorporation into the extracellular matrix. J Bioi Chem 1986;261:43374345. 19. Raghow R. Postlethwaite AE, Keski-Oja J, Moses HL, Kang AH. Transforming growth factor-8 increases steady state levels of type I procollagen and fibronectin messenger RNAs posttranscriptionally in cultured human dermal fibroblasts. J Clin Invest 1987;79:1285-1288. 20. Ignotz RA, Endo T, Massague J. Regulation of fibronectin and type I collagen mRNA levels by transforming growth factor-@. J Biol Chem 1987;262:6443-6446. 21. Varga J, Rosenbloom J, Jimenez SA. Transforming growth factor+ (TGF 8) causes a persistent increase in steady state amounts of type I and type III collagen and fibronectin mRNAs in normal human dermal fibroblasts. Biochem J 1987;247:597604. 22. Pentinnen RP, Kobayashi S, Bornstein P. Transforming growth factor-8 increases mRNA for matrix proteins both in the presence and in the absence of changes in mRNA stability. Proc Nat1 Acad Sci USA 1988:85:1105-1108. 23. Rossi P, Karsenty G, Roberts AB, Roth NS, Sporn MB, deCrombugghe B. A nuclear factor 1 binding site mediates the transcriptional activation of a type I collagen promoter by transforming growth factor-& Ceil 1988;52:405-414.
August 1990
TGF-j3, INCREASES COLLAGEN SYNTHESIS IN HISM CELLS
N, 24. Wahl SM, Hunt DA, Wakefield LM, McCartney-Francis Wahl LM, Roberts AB, Sporn MB. Transforming growth factor @ (TGF-0) induces monocyte chemotaxis and growth factor production. Proc Nat1 Acad Sci USA 1987;84:5788-5792. 25. Assoian RK. Fleurdelys BE, Stevenson HC, Miller PJ, Madtes DK, Raines EW. Ross R. Sporn MB. Expression and secretion of type /3 transforming growth factor by activated human macrophages. Proc Nat1 Acad Sci USA 1987;84:6020-6024. 26. Dvorak AM, Osage JE, Monahan RA, Dickersin DR. Crohn’s disease: transmission electron microscopic studies. Hum Path01 1980;11:806-634. 27. Assoian RK, Sporn MB. Type fl transforming factor in human platelets: release during platelet degranulation and action on vascular smooth muscle cells. J Cell Biol1986;102:1217-1223.
453
Received June 18.1989. Accepted February 18,199O. Address requests for reprints to: Martin F. Graham, M.D., Division of Pediatric Gastroenterology, Children’s Medical Center, Box 529 MCV Station, Richmond, Virginia 23298-0529. Supported by NIH Grants DK 34151 and GM 20298. The authors would like to thank Drs. Michael Sporn and Anita Roberts for their generous gift of human platelet-derived transforminggrowth factor p, and Drs. Harvey Sugerman and John Kellum for their invaluable help in procuring surgical specimens. Part of this work was presented in preliminary form at the annual meeting of the American Gastroenterological Association, May 1988, New Orleans, Louisiana; and at the New York Academy of Sciences conference on the Structure, Molecular Biology, and Pathology of Collagen, April 1989, Bethesda, Maryland.
Intestinal smooth muscle cells play a major role in the stricture formation that complicates chronic intestinal inflammation, by proliferating and producing collagen. Transforming growth factor & has been identified as an important inflammatory mediator in the fibrotic response of human tissue to inflammation. To determine whether this mediator might be involved in intestinal fibrosis, the effect of transforming growth factor & on collagen production and proliferation by human intestinal smooth muscle cells was studied in vitro. Cells in the second passage were grown to subconfluence in medium containing 10% Nu-Serum (Collaborative Research Inc., Bedford, MA), after which the concentration of NuSerum was decreased. Forty-eight hours later, transforming growth factor & was added to the culture medium to achieve concentrations of 1-500 pmol/L. After 24 hours exposure to the transforming growth factor &, cellular collagen synthesis was determined by the uptake of [SH]proline into collagenase-sensitive protein. Transforming growth factor & caused a 100% increase in collagen production and a 40% increase in noncollagen protein production per cell, reflecting an increase in relative collagen synthesis of 58%. This effect was maximal at a concentration of 10 pmol/L. Epidermal growth factor, by comparison, had no significant effect on relative collagen synthesis. Transforming growth factor & caused a signiflcant increase in the uptake of methylaminoisobutyric acid, a nonmetabolized amino acid analog, into the cells at IO pmol/L. However, this effect was small (20% increase] compared with the effect on the uptake of proline into collagen (loo?70 increase] at this concentration. When cell proliferation was examined by the uptake of [Wlthymidine, transforming growth factor fll had no effect, whereas epidermal growth factor (1000 pmol/L) caused a 94% increase. Transforming growth factor & selectively augments
collagen production by human intestinal smooth muscle cells in vitro. This effect is potent and is not related to effects on either cell proliferation or amino acid uptake. These data suggest that transforming growth factor & has an important role as an inflammatory mediator in the pathogenesis of intestinal fibrosis.
S
tricture formation, a common complication of chronic intestinal inflammation, leads to significant morbidity. Our recent studies have shown that intestinal smooth muscle cells play a major role in the pathogenesis of strictures (1). The evidence suggests that the lesions result from proliferation of and collagen production by the smooth muscle cells, a response on the part of these local mesenchymal cells that is common to many tissues in the body. To define further the pathogenesis of intestinal fibrosis, we have developed and characterized an in vitro model of human intestinal smooth muscle (HISM) cells. When grown in culture, HISM cells produce large amounts of collagen compared with other human mesenchymal cells such as fibroblasts (2). Growth in culture is inhibited by heparin (3), and collagen production is selectively downregulated by cyclic adenosine monophosphate (4) and by protamine sulfate (5). Having characterized the behavior of intestinal smooth muscle cells in vitro, we are now using this model to characterize the effects of defined inflammatory mediators on the cells, and hence determine which of these mediators are involved in the fibrotic
Abbreviations used in this paper: DMEM, Dulhecco’s mod&d Eagle medium; EGF, epidermal growth factor: HISM, human intestinal smooth muscle; TGF-&, transform@ growth factor 8,. 0 IBSOby the American Gastroenterologkal Aseodation 0018-5085/80/$3.00
448
GRAHAM ET AL.
GASTROENTEROLOGY
response of smooth muscle cells to inflammation in vivo. The mediator(s) in question would be either mitogenic or fibrogenic or both, and would be produced by one or more of the inflammatory elements that characterize chronic inflammation including platelets, lymphocytes, macrophages, and mast cells. Mesenchymal cell mitogens have been well defined and HISM cells proliferate in vitro in the presence of fetal bovine serum (3, the major mitogen in serum being platelet-derived growth factor (6). As far as fibrogenic factors are concerned, nothing is known at present about the effect of defined inflammatory mediators on the capacity of intestinal mesenchymal cells to produce collagen. Transforming growth factor & (TGF-P,) has been identified as one of the few mediators that specifically induces a fibrotic response in tissue (7). We were therefore intrigued to determine whether TGF-P, regulated collagen synthesis by HISM cells, and whether this mediator might be involved in the pathogenesis of intestinal strictures. Materials and Methods Cell Isolation and Culture Human intestinal smooth muscle cells were isolated from the muscularis propria of normal human jejunum using crude bacterial collagenase digestion as previously described (8). Cells were maintained in primary culture for 3 weeks in Dulbecco’s modified Eagle medium (DMEM) containing 10% Nu-Serum (Collaborative Research Inc., Bedford, MA: DMEM-lo), after which they were passaged by trypsinization for use in the experiments described below.
Determination of Nu-Serum
of the Optimal
Concentration
in Culture Medium for Mediator
Experiments Because fetal bovine serum contains inflammatory mediators, this growth supplement is unsuitable as an additive to culture medium in experiments designed to test the effect of inflammatory mediators on the cells. These problems can be overcome by employing a more defined cell growth supplement, and by decreasing the concentration of the supplement in the culture medium to the extent that the cells are quiescent yet metabolically stable and responsive. In these experiments, therefore, fetal bovine serum was replaced by a commercially produced, partially defined culture medium supplement-Nu-Serum. Initial experiments demonstrated that Nu-Serum was a satisfactory replacement for fetal bovine serum; HISM cells proliferated and produced collagen in a similar fashion when cultured with Nu-Serum as compared to fetal bovine serum [data not shown]. An experiment was then performed to determine the optimal minimal concentration of Nu-Serum for the TGF-& experiments. The HISM cells were grown to confluence in loo-mm culture dishes in DMEM-10. The concentra-
Vol. 99, No. 2
tion of Nu-Serum in the culture medium was then decreased in tenfold dilutions from 9.1% to 0.0001% and to 0% (time 0). The culture medium was changed daily with fresh ascorbate added (0.1 mmol/L). Each plate received 5 mL of pulsing medium containing L-[5-3H]proline (4 mCi/mL, 22 Ci/ mmol/L: Amersham Corp., Arlington Heights, IL) for 5 hours at time 0, and 48 hours and 96 hours later for the determination of collagen synthesis at those times. Collagen synthesis was assayed as previously described (21.
Effect of Transforming Growth Factor ,& on Collagen Synthesis by Human Intestinal Smooth Muscle Cells This experiment was performed three times, twice using cells cultured in loo-mm culture plates and a standard assay (9) and once using cells in microwell culture plates and a newly developed modification of the standard assay (10). The experimental protocols were identical in all three experiments. The reason for miniaturizing the standard assay was to facilitate an increase in the number of replicates for each data point (n = 6), an increase in the number of data points and a decrease in the cost of cells, mediators, isotopes, and media. For the standard macroassay, readers should refer to previous papers (2). The microassay was performed as follows: HISM cells were plated at a density of 3 x lo4 per well in 24-well culture plates (Costar, Cambridge, MA). After growth to subconfluence in DMEM-10 (3 days), the concentration of Nu-Serum in the medium was decreased to 0.0001% (DMEM-10e4) for a further 48 hours. Human platelelet-derived TGF-& [a kind gift from Drs. Michael Sporn and Anita Roberts, NIH, Bethesda, MD) was then added to the cultures to achieve final concentrations of l-500 pmol/L. Epidermal growth factor (EGF, 10-1000 pmol/L, Collaborative Research Inc., Bedford, MA) was added to an additional set of wells. After another 24 hours, all the cultures were incubated with L-[5-3H]proline (40 mCi/mL, 22 Ci/mmol/L) in the presence of the above agonists for 5 hours. The culture plates were then heated to 90°C to stop isotope incorporation and then frozen. After thawing, the cells were sonicated and 200 mL of the contents of each well were removed for determination of DNA content (11). A 1 mL solution of chick embryo carrier protein (2 mg/mL in 10 mmol/L proline) was then added to each well. Radioactive protein was separated from unincorporated L-[5-3H]proline by repeated precipitation with 5% trichloracetic acid at 4°C and the remaining trichloracetic acid was removed by ethanol/ether (3:1). The dried residue in each well was incubated for 90 minutes with 0.5 mL of buffer containing HEPES (60 mmol/ pH 7.21, CaCl, 10.25 mmol], and N-ethylmaleimide (1.25 mmol). Protein was reprecipitated with trichloracetic acid and the radioactivity in the solute was measured as an incubation blank. Thrichloracetic acid was removed, and the dried residue was suspended in the incubation buffer with purified bacterial collagenase (9). After a go-minute incubation, both the soluble [3H]collagen peptides and the [3H]noncollagen protein were counted. Collagen synthesis was expressed both on a per-cell basis (per nanogram DNA) and relative to total protein synthesis (relative collagen synthesis] according to a
TGF-8, INCREASES
August 1990
standard formula that accounts for the enriched content of collagen (12).
amino acid
COLLAGEN
SYNTHESIS
449
IN HISM CELLS
1100
900
Effect of Transforming Growth Factor & on Amino Acid Uptake by Human Intestinal Smooth Muscle Cells The protocol for this experiment was identical to that described for the collagen synthesis experiment above. The assay for amino acid uptake was performed as published (13). After addition of mediators to the culture medium for 24 hours, cells were washed twice with Hank’s balanced salt solution (pH 7.5) containing 25 mmol/L HEPES and 0.1% bovine serum albumin. The cells were then incubated for 25 minutes in the above salt solution containing both [14C]methylaminoisobutyric acid (0.5 mCi/mL, 48 mCi/mmol/ L, NEN Products, Boston, MA), cold (20 mmol/L) methylaminoisobutyric acid [Sigma, St. Louis, MO), and mediators. After washing, the cells were solubilized in 0.2 N sodium hydroxide, neutralized with an equal volume of 0.2 N hydrochloric acid, and the radioactivity of the contents of each well was determined by liquid scintillation counting.
Effect of Transforming Growth Factor & on /%/Thymidine Uptake by Human Intestinal Smooth Muscle Cells
Human intestinal smooth muscle cells were plated at a density of 3 x lo4 per well in 24-well culture plates and cultured in DMEM-10. After 2 days, DMEM-10 was replaced by DMEM-10m4. Forty-eight hours later, mediators were added to the culture medium and after 24 hours in the presence of the mediators, cells were incubated for 4 hours with culture medium containing [3H]thymidine (1 mCi/mL, mCi/mmol/L, NEN Products, Boston, MA], and mediators. The cells were then precipitated and washed with 5% trichloracetic acid at 4°C and solubilized in 0.2 N sodium hydroxide. The contents of each well were then counted for radioactivity. Statistical
Analysis
Experimental values were compared to control values by Student’s t test and were considered significantly different at the P < 0.05 level.
Results Optimal Concentration
of Nu-Serum
Analysis of collagen production by HISM cells in culture medium supplemented with progressive log dilutions of Nu-Serum demonstrated that 0.0001% was the optimal concentration of Nu-Serum for the TGF-& experiments that followed. At that concentration of Nu-Serum, collagen synthesis decreased significantly for the first 48 hours but then remained unchanged for the subsequent 48 hours (Figure 1). During
700
500
300
** I””
6
43
9’8
HOURS Figure 1. The effect of low concentrations of Nu-Serum on collagen synthesis by HISM cells. HISM cells were grown to confluence in culture medium supplemented with 10% NuSerum. At time 0,the concentration of NuSerum in the culture medium was decreased to O.OOl%,O.OOOl%,or 0%. Cells were pulsed with b[5-sII]proline for 5 hours at time 0,at 46 hours, and at 96 hours for the determination of collagen synthesis at those times. The values are the mean + SEM, n = 4. ?? Signf!kantly greater than, **signiEcantly less than the previous value at 46 hours (P < 0.05)
the second 48-hour period, TGF-/3, would be added to the cultures and collagen synthesis would be assayed. Stable baseline conditions would be essential during that time. In medium supplemented with 0.001% Nu-Serum, collagen production decreased similarly for the first 24 hours, but then increased significantly during the second 48-hour period (Figure 1). In medium without Nu-Serum (O?%),collagen production continued to decrease during the second period (Figure 1). Therefore, both 0.001% Nu-Serum and 0% Nu-Serum were unsuitable whereas 0.0001% was optimal. Similar findings were noted for noncollagen protein synthesis (data not shown]. Collagen Synthesis Exposure of HISM cells to TGF-& resulted in a 100% increase in absolute collagen synthesis per cell [Figure 2A). This effect was seen at 10 pmol/L concentration and did not increase at higher concentrations of TGF-/3, (100, 500 pmol/L). Epidermal growth factor had a similar effect on absolute collagen synthesis, causing a 60% increase. However, the effect of EGF was seen at a log higher concentration (100 pmol/L). Exposure of HISM cells to both TFG-& (100 pmol/L) and EGF (1000 pmol/L) had no additive effect on collagen synthesis (T + E, Figure 2A). Transforming growth factor-p, and EGF had similar effects on the production of noncollagen protein.
450 GRAHAM
GASTROENTEROLOGY
ET AL.
A
* *3
* 4
*
&
4
I
$ z s
1500
zi 9 i g
1000
10
1
100
500 1000 T+E
B
*
*
*
??
H
??
5oc
and EGF (100 pmol/L and 1000 pmol/L, respectively) had no additive effect (T+E, Figure 2B). When collagen synthesis was calculated relative to total protein synthesis, TGF-& was shown to cause a 58% increase. This effect was seen at 10 pmol/L [Figure 2CJ and did not increase at higher concentrations of TGF-& (100, 500 pmol/L). This increase in relative collagen synthesis resulted from a specific increase in collagen production rather than from an inhibition of noncollagen production. In contrast, EGF had no effect on relative collagen synthesis (Figure 2C). This was because EGF caused parallel increases in both collagen and noncollagen synthesis. These data represent the results of one experiment performed with a newly developed microassay for collagen synthesis using 24-well culture plates. Similar results were obtained in two other experiments performed with the standard assay using 100 mm culture dishes and HISM cells isolated from different donors. In those experiments, relative collagen synthesis was maximally increased 100% and 87% by TGF-
** +=d@ 0
z 2000 zi
Vol. 99, No. 2
I 0
I 1
I 10
I 100
I
TGF EGF
I I I 500 1000 T+E
??
EGF
31 0
1
10
100
500 1000 T+E
CONCENTRATION (PM)
Figure 2. The effect of TGF-/I, on (A) absolute collagen synthesis, (B] absolute noncollagen protein synthesis, and (c) relative collagen synthesis by HISM cells. Cells were grown to subconfluence, placed in a quiescent medium for 48 hours, and exposed to the various mediators for 24 hours, after which the assay for collagen synthesis was performed. T+E, TGF-8, (100 pmol/L) + EGF (1000 pmol/L); values are mean * SEM, n = 6 (*signBlcantly greater than control, P < 0.05). The Bgures represent data from one experiment performed in &well culture plates using a newly developed modification of the standard assay. A similar response was seen in two other experiments using lOO-mm culture plates and the standard assay.
There was a small (40%) but statistically significant (P < 0.05) increase in this parameter that was maximal at 100 pmol/L that did not increase further at 500 pmol/L (TGF-8,) or 1000 pmol/L (EGF). Again, TGF-
61.
Amino Acid Uptake Transforming growth factor & has been reported to increase the cellular uptake of amino acids (13). The assay used in the current studies depends, in part, on the uptake of a radiolabeled amino acid. Although the computation of relative collagen synthesis utilized above obviates the potential variable that might result from fluxes in amino acid uptake induced by the agonist under test (121, it was intriguing to determine whether the effects of TGF-P, observed in these experiments were associated with or independent of changes in the rate of amino acid uptake. Transforming growth factor & caused a significant (P < 0.05) increase in amino acid uptake at 10 pmol/L and 100 pmol/L [Table 1). However, the increase at 10 pmol/L was small (20%) compared to the large increase (100%) seen in the incorporation of isotope into collagen at that concentration. Epidermal growth factor had no effect on amino acid uptake (Table 1). These data demonstrate that TGF-6, does increase the uptake of amino acids into HISM cells. However, the data also demonstrate that the effect of TGF-0, on collagen synthesis by HISM cells is not, to any significant degree, caused by the effect of this mediator on amino acid uptake.
rH/Thymidine
Uptake
Transforming growth factor & had no effect on the uptake of [3H]thymidine by HISM cells (Figure 3).
TGF-6, INCREASES COLLAGEN SYNTHESIS IN HISM CELLS
August 1990
Growth Factor &, and Epidermal Growth Factor on the Uptake of “/C]Methylaminoisobutyric Acid lnto Human Intestinal Smooth Muscle Cells
Table 1. Effect of Transforming
Mediator None 10% Nu-Serum TGF-j3, 1 pmol/L 10 pmol/L 100pmol/L EGF 10pmol/L 100pmol/L 1000pmol/L
mAIB uptake (dpm/weJI]
% of control
3100* 110 5010+ 280b
100 161b
3350f 190 3720f 180" 4580+ 230"
108 120b 148b
3180+ 220 3120k 070 3250 + 140
103 101 105
NOTE. After growth to confluence in medium containing 10% Nu-Serum, cells were incubated in medium containing 0.0001% Nu-Serum for 48 hours (control conditions). The cells were then exposed to the mediators for 24 hours, after which isotope was added for the determination of AIB uptake. “Mean + SEM. bSignificantlygreater than control; P < 0.05.
This contrasted with an approximately 95% increase in this parameter produced by either 1000 pmol/L EGF (Figure 3) or 10% Nu-Serum. Moreover, TGF-& did not augment or inhibit the EGF-induced increase in [3H]thymidine uptake (T+ E, Figure 3).
inhibitors (7). An additional, indirect effect of TGF-PI on tissue repair is thought to be the recruitment to the area of inflammation, of macrophages (24) and then the stimulation of those inflammatory cells to produce interleukin 1 (24) and TGF-& (25). Tissue repair in the intestine is a process that has not been investigated systematically in the past. Our recent work has demonstrated that intestinal smooth muscle cells are involved in this repair process in that they, rather than any other mesenchymal cell type, proliferate and produce collagen in response to chronic inflammation in the intestinal wall (1). In addition, our in vitro studies have demonstrated that these cells have the capacity to produce large amounts of collagen (2). An important question that remains is the identity of the inflammatory mediators that are responsible for inducing the healing response on the part of the smooth muscle cells. Degranulating platelets in thrombosed submucosal vessels and activated macrophages are a major feature of chronic inflammatory infiltrates in the intestine (26). It is therefore reasonable to postulate that TGF-/3,, a major product of platelet degranulation (14) and of macrophages (251, and now
12000 rig TGF-b + EGF
Discussion Transforming growth factor & is a homodimeric peptide that belongs to a large family of genes responsible for the growth and differentiation of tissue. In addition to its role in development, TGF-& appears to be particularly important in tissue repair (7, for review). The highest concentration of TGF-& in the body is in platelets (14), attesting to its predominant role in wound healing. The peptide is secreted in a latent form thought to be activated locally by the clotting process (15). Systemic effects are thought to be prevented by binding in plasma to cu,-macroglobulin (16). With regard to tissue repair, the mode of action of TGF-& is to increase the production of extracellular matrix by mesenchymal cells, and to decrease the turnover of the matrix. The peptide induces fibrosis when injected subcutaneously in vivo, selectively increases collagen production in chick, rat, and human fibroblasts (17,18), and increases cellular levels of collagen mRNA (19-22). A nuclear factor 1 binding site has been identified as the promoter that mediates the transcriptional activation of the collagen gene (mouse, a$]) by TGF-0, (23). Decreased matrix turnover is accomplished by decreasing the production of proteases and increasing the production of protease
451
*
8000
I
6000
4000
2000
0
I I
I
I
I
0
1
IO
100
I
I
500
1000
CONCENTRATION
(PM)
I
T+E
Figure 2. The effect of TGF-~9,on the uptake of [QJthymidine by HISM cells. Cells were grown to suboonfluenco, placed in a quiescent medium for 48 hours, and eqnmed to the various conditions for 24 hours after which the assay for thymidine uptake was performed. T+E, TGF-fl, (lOa pmol/L) + EGF (1000pmol/L); values are mean + SEM, n = 6 (*signikantly greater than control, P < 0.05).
452
GRAHAM ET AL.
recognized as an important factor in the healing of injured tissue, is one of those mediators in the intestine. The data presented in this paper support that hypothesis. Transforming growth factor & augmented collagen production by HISM cells to a significant degree and in a selective fashion. The effect was potent in that its threshold was at concentrations of TGF-P, between 1 and 10 pmol/L. The selectivity for collagen production was demonstrated by the 100% increase produced in this parameter vs. a 40% increase in noncollagen protein production. This resulted in a 58% increase in relative collagen synthesis. By comparison, EGF produced a 50% increase in both collagen and noncollagen protein synthesis with no net effect on relative collagen synthesis, In addition, the marked effect of TGF-8, on matrix synthesis was exclusive of any effect on DNA synthesis. The lack of stimulation of HISM cell proliferation by TGF-& is consistent with previous reports studying bovine aortic smooth muscle cells and normal rat kidney fibroblasts (27). In those studies, in addition to the lack of mitogenic effect, TGF-& antagonized the mitogenic effect of EGF. In the current study, no such inhibitory effect was seen. Therefore, TGF-/3, appears to have an important regulatory role in collagen production by HISM cells, but no role in regulating the proliferation of these cells. It is conceivable, therefore, that the mechanism whereby TGF-/3, promotes healing in the intestine is twofold: a stimulation of matrix production by local smooth muscle cells; and a recruitment of inflammatory cells such as macrophages into the area for the production of smooth muscle-cell mitogens. In addition, TGF-/3, may decrease the turnover of the intestinal extracellular matrix, but this question is yet to be examined. Therefore, on the basis of the data from these in vitro studies, we would postulate that TGF-& is an important inflammatory mediator in the pathogenesis of intestinal fibrosis. Further studies of pathological specimens to localize this mediator in situ will be necessary to confirm this hypothesis. References 1. Graham MF. Dienelmann RF, Elson CO, Lindblad WI, Gotschalk N, Gay S, Gay R Collagen content and types in the intestinal strictures of Crohn’s disease. Gastroenterology 1988;94:257-285. Graham MF, Drucker DEM, Diegelmann RF, Elson CO. Collagen synthesis by human intestinal smooth muscle cells in culture. Gastroenterology 1987;92:400-405. Graham MF, Drucker DEM. Perr HA, Ebrlich HP, Diegelmann RF. Heparin modulates human intestinal smooth muscle cell proliferation, protein synthesis and lattice contraction. Gastroenterology 1987;93:801-809. Perr HA, Graham MF, Diegelmann RF, Downs RW. Cyclic nucleotides regulate collagen production by human intestinal smooth muscle cells. Gastroenterology 1989;96:1521-1528. Perr HA, Drucker DEM, Co&ran DL. Diegelmann RF, Lind-
GASTROENTEROLOGY
Vol. 99, No. 2
blad WJ, Graham MF. Protamine selectively inhibits collagen synthesis by human intestinal smooth muscle cells and other mesenchymal cells. J Cell Physiol1989;140:483-470. 6. Ross R, Raines EW, Bowen-Pope DF. The biology of plateletderived growth factor. Cell 1986;46:155-159. 7. Sporn MB, Roberts AB, Wakefield LM, de Crombugghe B. Some recent advances in the chemistry and biology of transforming growth factor-& J Cell Biol1987:105:1039-1045. 8. Graham MF, Diegelmann RF, Elson CO, Bitar KN, Ehrlich HP. Isolation and culture of human intestinal smooth muscle cells, Proc Sot Exp Biol Med 1984;176:503-507. 9. Peterkofsky B, Diegelmann RF. The use of a mixture of proteinase-free collagenase for the specific assay of radioactive collagen in the presence of other proteins. Biochemistry 1971;10:988994. 10. Diegelmann RF, Bryson GR, Flood LC, Graham MF. Microassay to quantitate collagen synthesis by cells in culture. Anal Biochem 1990;186:296-300. 11. Labarca C, Paigen K. A simple, rapid and sensitive DNA assay procedure. Anal Biochem 1980;102:344-352. 12. Diegelmann RF, Peterkofsky B, Collagen biosynthesis during connective tissue development in chick embryo. Dev Biol 1972;28:443-453. 13. Boerner P, Resnick RJ, Racker E. Stimulation of glycosis and amino acid uptake in NRK-49F cells by transforming growth factor-@ and epidermal growth factor. Proc Nat1 Acad Sci USA 1985;82:1350-1353. 14. Assoian RK, Komoriya A, Meyers CA, Miller DM, Sporn MB. Transforming growth factor-8 in human platelets: identification of a major storage site, purification and characterisation. J Biol Chem 1983;258:7155-7160. 15. Wakefield LM, Smith DM. Flanders KC, Sporn MB. Latent transforming growth factor-8 from human platelets. J Biol Chem 1988;263:7646-7654. 16. O’Connor-McCourt MD, Wakefield LM. Latent transforming growth facto@ in serum. J Biol Chem 1987;262:14090-14099. 17. Roberts AB, Sporn MB, Assoian RK, Smith JM, Roth NS, Wakefield LM, Heine UI, Liotta LA, Falanga V, Kehrl JH, Fauci AS. Transforming growth factor type-p: rapid induction of fibrosis and angiogenesis in vivo and stimulation of collagen synthesis in vitro. Proc Nat1 Acad Sci USA 1986;83:4167-4171. 18. Ignotz R, Massague J. Transforming growth factor+3 stimulates the expression of fibronectin and collagen and their incorporation into the extracellular matrix. J Bioi Chem 1986;261:43374345. 19. Raghow R. Postlethwaite AE, Keski-Oja J, Moses HL, Kang AH. Transforming growth factor-8 increases steady state levels of type I procollagen and fibronectin messenger RNAs posttranscriptionally in cultured human dermal fibroblasts. J Clin Invest 1987;79:1285-1288. 20. Ignotz RA, Endo T, Massague J. Regulation of fibronectin and type I collagen mRNA levels by transforming growth factor-@. J Biol Chem 1987;262:6443-6446. 21. Varga J, Rosenbloom J, Jimenez SA. Transforming growth factor+ (TGF 8) causes a persistent increase in steady state amounts of type I and type III collagen and fibronectin mRNAs in normal human dermal fibroblasts. Biochem J 1987;247:597604. 22. Pentinnen RP, Kobayashi S, Bornstein P. Transforming growth factor-8 increases mRNA for matrix proteins both in the presence and in the absence of changes in mRNA stability. Proc Nat1 Acad Sci USA 1988:85:1105-1108. 23. Rossi P, Karsenty G, Roberts AB, Roth NS, Sporn MB, deCrombugghe B. A nuclear factor 1 binding site mediates the transcriptional activation of a type I collagen promoter by transforming growth factor-& Ceil 1988;52:405-414.
August 1990
TGF-j3, INCREASES COLLAGEN SYNTHESIS IN HISM CELLS
N, 24. Wahl SM, Hunt DA, Wakefield LM, McCartney-Francis Wahl LM, Roberts AB, Sporn MB. Transforming growth factor @ (TGF-0) induces monocyte chemotaxis and growth factor production. Proc Nat1 Acad Sci USA 1987;84:5788-5792. 25. Assoian RK. Fleurdelys BE, Stevenson HC, Miller PJ, Madtes DK, Raines EW. Ross R. Sporn MB. Expression and secretion of type /3 transforming growth factor by activated human macrophages. Proc Nat1 Acad Sci USA 1987;84:6020-6024. 26. Dvorak AM, Osage JE, Monahan RA, Dickersin DR. Crohn’s disease: transmission electron microscopic studies. Hum Path01 1980;11:806-634. 27. Assoian RK, Sporn MB. Type fl transforming factor in human platelets: release during platelet degranulation and action on vascular smooth muscle cells. J Cell Biol1986;102:1217-1223.
453
Received June 18.1989. Accepted February 18,199O. Address requests for reprints to: Martin F. Graham, M.D., Division of Pediatric Gastroenterology, Children’s Medical Center, Box 529 MCV Station, Richmond, Virginia 23298-0529. Supported by NIH Grants DK 34151 and GM 20298. The authors would like to thank Drs. Michael Sporn and Anita Roberts for their generous gift of human platelet-derived transforminggrowth factor p, and Drs. Harvey Sugerman and John Kellum for their invaluable help in procuring surgical specimens. Part of this work was presented in preliminary form at the annual meeting of the American Gastroenterological Association, May 1988, New Orleans, Louisiana; and at the New York Academy of Sciences conference on the Structure, Molecular Biology, and Pathology of Collagen, April 1989, Bethesda, Maryland.
