Expression of transforming growth factors alpha and beta in colonic mucosa in inflammatory bowel disease

GASTROENTEROLOGY 1996;110:975–984

ALIMENTARY TRACT Expression of Transforming Growth Factors a and b in Colonic Mucosa in Inflammatory Bowel Disease MARK W. BABYATSKY,*,‡ GUILLERMO ROSSITER,* and DANIEL K. PODOLSKY* *Gastrointestinal Unit and Center for the Study of Inflammatory Bowel Disease, Massachusetts General Hospital, and Harvard Medical School, Boston, Massachusetts; and ‡Gastrointestinal Division, Mount Sinai Medical Center, New York, New York

Background & Aims: Transforming growth factors (TGFs) a and b are key regulatory peptides that modulate mucosal cell populations critical to inflammatory bowel disease. The aim of this study was to assess TGF-a and TGF-b expression in human colonic mucosa. Methods: TGF-a and TGF-b expression was assessed in colonic mucosa from patients with ulcerative colitis, patients with Crohn’s disease, and controls by Northern blot analysis, in situ hybridization, and bioassay. Results: TGF-a messenger RNA expression localized to the villous tips of the small intestine and the surface epithelium of the colon. TGF-a expression was enhanced 2.3-fold in inactive ulcerative colitis mucosa relative to active ulcerative colitis, Crohn’s disease, or normal controls. Enhanced expression correlated with duration of disease. TGF-b expression was increased in affected mucosa from both patients with ulcerative colitis and Crohn’s disease with active disease. TGF-b1 messenger RNA expression in ulcerative colitis and Crohn’s disease localized mostly to cells of the lamina propria with the highest concentration in inflammatory cells closest to the luminal surface. Conclusions: TGF-a may contribute to epithelial hyperproliferation and the increased risk of malignancy in long-standing ulcerative colitis. TGF-b may be a key cytokine during periods of active inflammation, modulating epithelial cell restitution and functional features of cells within the lamina propria.

D

uring the past several years, a number of proteins have been recognized for the potentially important roles they may play in modulating growth and functional differentiation of diverse cell populations within gastrointestinal tract mucosa. Some have been recognized initially as products of cellular constituents of the immune systems and/or for their ability to modulate the growth and functional status of these populations and therefore have been designated cytokines or more specifically monokines, interleukins, and lymphokines.1,2 Other growth-modulating proteins have been identified initially by their direct effects on proliferative activity of / 5e0b$$0013

03-11-96 17:04:05

gasa

some cell populations and thus were designated peptide growth factors.3 – 13 Among this network of regulatory peptides, the transforming growth factors (TGFs) a and b may be especially important in the economy of the colonic and intestinal mucosa. TGF-a, a member of the epidermal growth factor (EGF) family, may be the most significant physiological ligand of the EGF in intestinal mucosa.5,10,14 Thus, it has been possible to show significant expression of this peptide in the mucosa, although EGF itself could not be found. Direct evidence has shown that TGF-a can stimulate proliferation of rat intestinal and human colonic epithelial derived cell lines. It is possible that the trophic effects of EGF on intestinal and colonic mucosa found in many experimental models reflect the physiological role of TGF-a in contributing to growth and proliferation.9,11 In addition, studies on enterocytes have localized EGF receptors to the basolateral cell surfaces.15 Other findings indicate that TGF-a may have functional effects on intestinal epithelial populations beyond those on proliferative activity, including stimulation of electrolyte and nutrient uptake and processing of brush border enzymes within the small intestine.10,11 Although initially identified through a bioassay similar to the one that led to the recognition of TGF-a, TGF-b is structurally unrelated.16 TGF-b1 and other closely related forms of TGF-b are 25-kilodalton proteins composed of identical or nearly identical subunits.17 – 20 During the past several years, TGF-b has been found to have a wide variety of effects on both proliferation and functional features of many cells present in the intestinal mucosa. Many of these effects may reflect the paracrine action of this peptide, which is produced by several cell populations present in the mucosa; others may be truly Abbreviations used in this paper: EGF, epidermal growth factor; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; SDS, sodium dodecyl sulfate; SSC, sodium chloride sodium citrate buffer; TGF, transforming growth factor. 䉷 1996 by the American Gastroenterological Association 0016-5085/96/$3.00

WBS-Gastro

976 BABYATSKY ET AL.

GASTROENTEROLOGY Vol. 110, No. 4

autocrine with the observed production of this protein by cell populations that also respond to the factor. Indeed, we have shown interrelated autocrine regulation and response to both TGF-a and TGF-b in an intestinal epithelial cell line IEC-6.21 Importantly, TGF-b has been found to inhibit proliferation of this intestinal epithelial cell line. Despite this property, paradoxically TGF-b seems to play a central role in promoting rapid re-epithelialization after mucosal wounding in wounds in vitro models.22 Within lamina propria cell populations, TGFb has been found to have potent effects on isotype switching among B lymphocytes, activation and pattern of cytokine expression by T lymphocytes and macrophages, as well as effects on other cells.13,23 – 25 Interestingly, TGFb has also been noted to have profound chemotactic activity for neutrophils, using pathways that supersede those of other potent chemotactic factors.26 It is apparent that TGF-a and TGF-b affect several cells and cellular processes that may be important in the pathogenesis of inflammatory bowel disease (IBD). These include epithelial response to injury, activation and modulation of a variety of lamina propria cell populations, and recruitment of cells to the site of disease activity. Graham et al. showed that TGF-b promotes collagen deposition by isolated smooth muscle cells, suggesting that this factor may play a significant role in stricture formation in Crohn’s disease (CD).27 As described in a preliminary report from this laboratory28 and by McCabe et al., increased content of TGF-b messenger RNA has been suggested to be present in the mucosa of patients with IBD.29 However, comprehensive assessment of the expression of these factors in mucosa in the context of IBD has not been undertaken. In this report, we note that enhanced expression of TGF-a is present in the mucosa of patients with inactive ulcerative colitis (UC) and is localized to the superficial epithelial cells, whereas TGF-b expression is increased in the lamina propria clos-

Figure 2. Relative expression of TGF-a in colonic mucosa. Relative abundance of TGF-a in samples of colonic mucosa was quantitated by normalization of hybridization relative to constitutive transcript (either actin or GAPDH) using scanning densitometry derived from Northern blot analysis as shown in Figure 1. *P õ 0.001.

est to the luminal surface during active UC and CD, suggesting distinct and disparate roles for these factors in the pathogenesis of IBD.

Materials and Methods Tissue Samples Colonic mucosal biopsy samples were obtained using standard pinch biopsies in parallel with samples obtained for diagnostic histological evaluation from patients undergoing colonoscopy at the Medical Endoscopy Unit of Massachusetts General Hospital. Samples were obtained from a total of 90 sites in 64 different patients, including 12 patients with active pan-UC, 10 patients with left-sided active UC, 12 patients with inactive UC, 14 patients with active or inactive segmental CD of the colon, and 16 normal controls. Classification was confirmed by review of official pathology diagnostic reports after completion of the study. Six biopsy samples were obtained from each sampling site and snap frozen on dry ice. Later samples were placed into 4 mL guanidine isothiocyanate buffer, homogenized in a polytron, and stored at 070⬚. These studies were approved by the Subcommittee on Human Studies of Massachusetts General Hospital.

Northern Blot Analysis

Figure 1. TGF-a expression in colonic mucosa. Representative Northern blot analysis of TGF-a in colonic mucosa; messenger RNA was prepared from mucosal biopsy samples (n Å 6 per lane; approximately 1.5 mg) electrophoresed with probe for TGF-a and the constitutively expressed transcript GAPDH as detailed in Materials and Methods. A, active disease; IA, inactive disease.

/ 5e0b$$0013

03-11-96 17:04:05

gasa

Samples were centrifuged in a CsCl gradient and poly(A)/ RNA obtained by oligodeoxythymidylic acid chromatography after phenol/chloroform extraction.30 The messenger RNA (approx 1.5 mg/lane) was electrophoresed in 1% formaldehydeagarose and transferred to nylon membrane using standard techniques.31 Northern blots were serially hybridized for TGF-a and TGF-b as well as for g actin or glyceraldehyde-3-phosphate dehydrogenase (GAPDH) using techniques essentially as previously described.5,20,32 Residual radiolabel probe was removed by boiling blots (1–2 minutes) before rehybridization. Briefly, TGF-a was evaluated using either a riboprobe prepared from

WBS-Gastro

April 1996

TGF-a AND TGF-b IN IBD 977

Table 1. TGF-a and TGF-b Bioactivity in Colonic Mucosa Bioactivity equivalence Colonic mucosa

TGF-a (ng/mg protein)

TGF-b (ng/mg protein)

2.4 { 0.2

0.3 { 0.2

1.8 { 0.6 5.5 { 0.6 2.1 { 0.5

1.7 { 0.3 0.5 { 0.3 0.4 { 0.2

1.9 { 0.5 2.1 { 0.4

1.9 { 0.5 0.4 { 0.2

Normal UC Active Inactive Uninvolved CD Involved Uninvolved

In Situ Hybridization of TGF-a and TGF-b1 Messenger RNA Figure 3. TGF-a expression in UC: effect of activity and location. Northern blot analysis of TGF-a RNA in mucosa from left colon (LC) and right colon (RC) with active or inactive colitis of patients with UC confined to the left colon.

an insert obtained by EcoRI digestion of the plasmid TGF-C1 subcloned into the SP65 plasmid and linearized with HindIII or by the insert labeled by nick translation. Hybridization was performed in 50% formamide, 51 sodium chloride sodium citrate buffer (SSC), 50 mmol/L P/Na (pH 7), 1 mmol/L ethylenediaminetetraacetic acid, 2.51 Denhardt’s solution, 220 mg/ mL salmon sperm DNA, and 0.1% sodium dodecyl sulfate (SDS) for 20 hours at 55⬚C, and blots were then washed 41 in 0.11 SSC and 0.1% SDS at 65⬚C before autoradiography. TGF-b was assessed using a probe prepared by nick translation of a 1.08-kilobase insert of mouse TGF-b1 as described previously.5 Hybridization was performed at 42⬚C in 40% formamide, 51 SSC, 51 Denhardt’s solution, 10% dextran sulfate, and 20 mmol/L P/Na at pH 7, and blots were then washed in 21 SSC and 0.1% SDS followed by 11 SSC / 0.1% SDS, 0.11 SSC / 0.1% SDS, and 0.11 SSC at 65⬚C before autoradiography. g Actin and GAPDH were evaluated precisely as described in earlier studies.5,21 Hybridization was quantitated by scanning densitometry, normalizing expression of TGF-a and TGF-b to the constitutive probes. No disparity was observed when normalized to either of the two constitutive transcripts. Results for each of the TGFs was evaluated using a one-way analysis of variance.

The tissue localization for the messenger RNA’s encoding TGF-a and TGF-b1 were analyzed using in situ hybridization. Fresh surgical specimens from patients with IBD were fixed in neutral-buffered formaldehyde, cut in 4-mm sections, and floated onto aminopropylethoxysilane (2%, vol/vol)treated slides. The sections were subsequently dewaxed, rehydrated with phosphate-buffered saline, treated with proteinase K (20 mg/mL), postfixed in 4% paraformaldehyde, acetylated with 0.1 mol/L triethanolamine with acetic anhydride (0.25% vol/vol), and dehydrated before hybridization. Hybridization was performed overnight at 55⬚C in a 25-mL solution containing 0.02% Denhardt’s solution, 10% dextran sulfate, 50% formamide, 0.3% mg/mL bovine ribosomal RNA, 10 mmol/ L Na2HPO4 , 10 mmol/L Tris-HCl, 5 mmol/L ethylenediaminetetraacetic acid, and 1 1 106 cpm of [35S]uridine triphosphate–labeled TGF-a and TGF-b1 complementary RNA probes (antisense or sense). Complementary RNA probes were prepared by ligating a 0.93-kilobase EcoR1 insert of a human TGF-a complementary DNA clone designated PHTGF1-10925 and of a 2.14-kilobase EcoR1 insert of a human TGFb1 complementary DNA clone obtained from American Type Culture Collection (Rockville, MD) into the pBluescript KS// vector (Stratagene, La Jolla, CA). [35S]uridine triphos-

Table 2. Expression of TGF-a in UC: Effect of Disease Activity and Duration TGF-a/actin

Bioassays for TGFs

Duration of disease

Mucosal biopsy samples were placed in 20 vol of phosphate-buffered saline, pH 7.4, containing 1 mmol/L phenylmethylsulfonyl fluoride and 0.1% NaN3 , and were sonicated. Solubilized protein present in the supernatant after centrifugation (105,000g for 60 minutes) was dialyzed against water using Spectraphor III dialysis tubing (Spectrum Medical, Los Angeles, CA) and was lyophilized. The lyophilized sample was then assessed for TGF-a– and TGF-b–related stimulation of anchorage independent growth using NRK or BHK nontransformed fibroblast cell lines as previously reported.33

/ 5e0b$$0013

03-11-96 17:04:05

gasa

Disease activity

õ5 yr

Active Inactive

0.9 { 0.2 1.4 { 0.3

5–10 yr

10–15 yr

1.1 { 0.3 1.1 { 0.2 1.7 { 0.4 2.4 { 0.6 P õ 0.05 P õ 0.01

ú15 yr 1.3 { 0.3 2.6 { 0.5

NOTE. Relative abundance of TGF-a assessed by scanning densitometry of Northern blots and normalization to the constitutively expressed transcript actin as described in Materials and Methods and the legend to Figures 1 and 2. Four to eight samples were evaluated for each condition.

WBS-Gastro

978 BABYATSKY ET AL.

GASTROENTEROLOGY Vol. 110, No. 4

Figure 4. TGF-b expression in colonic mucosa. Representative Northern blot of messenger RNA (approximately 1.5 mg/lane) from colonic mucosal biopsy samples of patients with UC or CD or normal controls hybridized sequentially with probes for TGF-b1 and GAPDH as described in the legend to Figure 1 and Materials and Methods. A, active disease; IA, inactive disease.

phate–labeled strand-specific complementary RNA probes were then generated after linearization with an appropriate restriction enzyme and incubation with T3 or T7 RNA polymerase. After hybridization, the tissues were treated with ribonuclease A to remove excess probe, washed with increasing stringency to 0.51 SSC at 65⬚C, and dehydrated in alcohols containing 0.3 mol/L ammonium acetate. The sections were then coated with autoradiographic emulsion (NTB-2; Eastman Kodak Co., Rochester, NY) and exposed for 6–8 weeks at 4⬚C. The sections were then developed and counterstained with H&E. Tissue blocks were prepared from at least four different surgical specimens per group, and patterns of hybridization were uniformly consistent within different groups.

Results In the past several years, characterization of the biological activities of TGF-a and TGF-b has shown that these proteins possess many activities that could play a role in the major forms of IBD. As an initial approach to assessing the role these factors may play in these disorders, we have evaluated the expression of TGFa and TGF-b1, the most abundant member of the TGFb family, in colonic mucosa of patients with UC or CD as well as controls. As shown in Figure 1, evaluation of expression of TGFa messenger RNA in mucosal biopsy samples showed enhanced expression in the mucosa of patients with UC. However, enhanced expression of this transcript was not significant in mucosal biopsy samples obtained from sites of active disease but was only notable in biopsy samples from sites of inactive or uninvolved mucosa (Figure 2). The failure to observe increased expression of the TGFa transcript in the active mucosa could reflect, in part, the epithelial destruction present in this disorder. In contrast to members of the TGF-b family (see below), TGF-a has been exclusively or predominantly found as an epithelial cell product. The observed consistent increase in TGF-a expression in inactive mucosa suggests that this association is not directly linked to the acute / 5e0b$$0013

03-11-96 17:04:05

gasa

inflammatory response. Although the role of a cytokine or other factor resulting from the persistent mild chronic inflammatory and immune cells typically present in the lamina propria of patients with inactive UC cannot be entirely excluded, it is notable that a lesser enhancement of TGF-a was also present in the right colon of patients with disease activity confined to the left colon, suggesting that the constitutive increase in TGF-a is not directly related to UC inflammatory activity (Figure 3). This finding did not reflect any inherent difference in the level of expression of TGF-a between the right and left colon. Essentially equivalent amounts of TGF-a messenger RNA were present in mucosa from the right and left colon from normal individuals in whom samples were taken from both sides of the large intestine. Interestingly, alterations in the relative expression of TGF-a were not observed in mucosa obtained from patients with CD of the colon. As indicated in Figures 1 and 2, relative expression of TGF-a in mucosa obtained from active, inactive, or uninvolved segments of colon of patients with CD, as well as a limited number of patients with other inflammatory disorders, were essentially the same as that observed in normal mucosa. It should be noted that the relative amount of TGF-a among the different classes of samples examined was the same when either of two constitutive markers, actin and GAPDH, were used to normalize expression. Relative expression among patient groups at the level of messenger RNA was generally mirrored in bioassays for the TGF-a protein itself (Table 1). Thus, extracts of mucosal biopsy samples from patients with inactive UC showed an ability to stimulate anchorage independent growth of the indicator fibroblast line BHK in soft agar equivalent on mean to 5.5 ng of TGF-a or EGF protein standard. In contrast, normal mucosa and active UC as

Figure 5. Relative expression of TGF-b in colonic mucosa quantitated by normalization of hybridization relative to constitutive transcript using scanning densitometry derived from Northern blot analysis as shown in Figure 4. *P õ 0.005.

WBS-Gastro

April 1996

TGF-a AND TGF-b IN IBD 979

Figure 6. Localization of TGF-a messenger RNA in small intestine and colon by in situ hybridization. (A ) Representative section of noninflamed colon hybridized with TGF-a antisense complementary RNA probe (bright field). (B ) Dark-field illumination of identical section shows specific hybridization on the surface epithelium. (C ) Colonic mucosa from patient with active UC; antisense complementary RNA probe. Increased expression of TGF-a enables detection on bright field. (D ) Dark-field illumination of identical section. (E ) Normal colonic mucosa; TGF-a sense cRNA probe (bright field). (F ) Identical section in dark field after 8 weeks of exposure (original magnification 601).

well as CD mucosa contained on mean 1.8–2.4 ng TGFa equivalents. It should be noted that we were unable to detect any EGF messenger RNA in mucosa from any patients in Northern blot analysis, despite use of as much as 40 mg poly(A)/ RNA from pooled colonic mucosa. / 5e0b$$0013

03-11-96 17:04:05

gasa

Although it is apparent that mucosa from patients with inactive UC consistently show increased levels of TGF-a expression, this relationship is accentuated by increasing duration of disease. As shown in Table 2, the highest levels of TGF-a in mucosa from patients with WBS-Gastro

980 BABYATSKY ET AL.

GASTROENTEROLOGY Vol. 110, No. 4

inactive UC were found in patients with a history of 20 years or more of known disease. However, significant increases in TGF-a compared with normal or diseasecontrol mucosal samples were present in inactive mucosa of patients with UC within 5 years of diagnosis. The relationship between relative expression of TGFb1 and IBD contrasted with that of TGF-a. As shown in Figure 4, enhanced levels of expression of TGF-b1 were present in actively inflamed mucosa of both patients with UC and CD when compared with normal controls. The relative increase in TGF-b1 expression in actively involved mucosa seemed to be the same in tissue from patients with the two disorders. Although TGF-b1 messenger RNA was easily detectable in IBD-involved mucosa, TGF-b1 messenger RNA was present in low to nearly undetectable levels in normal mucosa when assessed by Northern blot analysis, despite using substantial amounts of messenger RNA. The same low level of TGF-b messenger RNA was found in mucosa from both the right and left colon (not shown). Although TGF-b1 messenger RNA expression was enhanced in association with active inflammation, levels of TGF-b1 transcript in uninvolved or inactive mucosa from patients with IBD were the same as that observed in normal mucosa (Figure 5). The level of TGF-b1 expression was also enhanced in the small number of mucosal samples obtained from patients with non-IBD inflammatory states. The levels of bioactive TGF-b1 in colonic mucosal biopsy samples closely paralleled the relative levels of messenger RNA (Table 1). Thus, colonic mucosa from patients with active UC and CD contained the equivalent of 1.7 and 1.9 ng/mg mucosal protein of TGF-b standard. In contrast, normal mucosa and inactive mucosa from patients with IBD contained õ0.5 ng/mg mucosal protein TGF-b equivalent. It should be noted that these bioactivity determinations encompass the aggregate TGF-b protein present in the mucosa. It seems that the association of enhanced TGF-b with active IBD determined through Northern blot analysis for the dominant species TGF-b1 is representative of the overall content of TGF-b species in mucosa. In situ hybridization was used to determine the cellular distribution of TGF-a and TGF-b1 expression. In both inflamed and normal human tissues, the low abun-

dance of TGF-a messenger RNA required exposure of the tissue sections to the 35S-labeled complementary RNA probe for a minimum of 6 weeks to detect specific expression. Even at 6–8 weeks, dark-field photography was necessary to highlight the spatial pattern of TGF-a expression. As shown in Figure 6A and B, TGF-a messenger RNA localized to the epithelial cells of the villus in normal human small intestine. In normal colon, TGFa messenger RNA localized to the superficial epithelial cell layer. In IBD, all specimens examined showed the same pattern of superficial epithelial expression as in normal tissues. Thus, as shown in Figure 6C and D, TGF-a was expressed in a rim of superficial epithelium in a patient with active UC. The low abundance of TGF-b1 messenger RNA also necessitated a long exposure time (6 weeks) for in situ hybridization, but in contrast to TGF-a, bright-field examination localized TGF-b1 gene expression. The pattern of TGF-b1 RNA localization contrasted with that of TGF-a. In normal mucosa, predominant TGF-b1 messenger RNA expression was present in the lamina propria and not in the mucosal epithelium. As shown in Figure 7A and B, TGF-b1 was expressed predominantly in the lamina propria and in an underlying lymphoid aggregate. At high power, TGF-b1 messenger RNA expression was observed in multiple inflammatory cells surrounding an isolated crypt and in a subset of epithelial cells at the crypt base (Figure 7C). In mild UC, TGF-b1 messenger RNA was also most evident in inflammatory cells near the lumen (Figure 7D). TGF-b1 was also expressed primarily in the lamina propria in severe CD tissue (Figure 7E–G), including a gradient of expression with the highest expression closest to the luminal surface.

Discussion The TGF-a and TGF-b have been recognized for a number of functional activities that could play a role in IBD. It is possible that the growth-promoting effects of TGF-a for epithelial populations could contribute to epithelial regeneration after mucosal injury in IBD. It has been well shown that human colonic epithelium and presumably small intestinal epithelium express receptors for TGF-a paralleling observations in rats and other experimental animals. Although both EGF and TGF-a

䉴 Figure 7. Localization of TGF-b1 messenger RNA in colonic mucosa by in situ hybridization. (A ) Representative section of noninflamed colonic mucosa with TGF-b1 antisense complementary RNA probe (bright field). (B ) Identical section in dark field. Note the expression of TGF-b1 in cells of the lamina propria and underlying lymphoid aggregate. (C ) High-power magnification showing expression predominantly in inflammatory cells surrounding a single crypt; some expression may also be present in the epithelium (bright field). (D ) Mildly active UC with a specific focus on silver grains. Note the gradient of expression toward the luminal surface at the top left of the figure (bright field). (E ) Severe CD; lumen toward top of figure (dark field). (F ) Higher power magnification of active CD (bright field). (G ) Identical section in dark field. (H ) Normal ileum with TGF-b1 sense complementary RNA probe (bright field) after 6 weeks of exposure (original magnification: A, B, and E, 401; C, 4001; D and H, 651; and F and G, 1601).

/ 5e0b$$0013

03-11-96 17:04:05

gasa

WBS-Gastro

April 1996

TGF-a AND TGF-b IN IBD 981

/ 5e0b$$0013

03-11-96 17:04:05

gasa

WBS-Gastro

982 BABYATSKY ET AL.

GASTROENTEROLOGY Vol. 110, No. 4

seem to bind to the same receptor, in the rat it has been suggested that TGF-a may be the more important physiological ligand in intestinal and colonic mucosa. Although significant levels of TGF-a were present in intestinal epithelial cells, EGF could not be shown.5 The findings shown in Figures 1 and 2 and in Table 1 suggest that TGF-a may also be the physiological ligand of the so-called EGF receptor in human colonic mucosa. Although relative expression of the two ligands were not dissociated at the protein level in this study, Northern blot analysis showed substantial content of TGF-a messenger RNA, whereas EGF could not be detected. This is consistent with studies of human colonic epithelial cell lines that showed that TGF-a in conjunction with amphiregulin, another related peptide, may be the most abundant members of the EGF peptide family in colonic mucosa.34 As shown in Figures 1 – 3, the levels of messenger RNA encoding the proliferation-promoting factor TGF-a were significantly enhanced in UC mucosa during disease quiescence. It should be noted that actual TGF-a bioactivity closely paralleled messenger RNA expression. These observations suggest that TGF-a may contribute to the well-documented increase in epithelial proliferation in UC mucosa. 35 – 37 Although the relationship of the persistent increase in epithelial turnover in UC to the recognized increased incidence of colon cancer in this setting is unproven, it is notable that the overall enhanced expression in TGF-a increases progressively with increasing disease duration. This relationship suggests that a dynamic feature of the disease process promotes TGFa expression. Although TGF-a expression is found in the absence of significant active inflammation, it is possible that a factor such as a cytokine, produced by the residual cellular infiltrate present during disease quiescence, contributes to stimulation of TGFa expression. It should be emphasized that TGF-a expression and responsiveness in mucosa is confined to the epithelial cell compartment in contrast to TGF-b. The observed increase in TGF-a in quiescent mucosa and essentially normal content in actively inflamed mucosa suggest that the factor may not be essential to the reconstitution of epithelial cells after injury in this disease. However, the estimation of the ligand content, particularly if normalized to cell number, may be artifactually low in the context of active disease in which epithelial cells are destroyed. Furthermore, the constitutive level of expression may be sufficient for adequate stimulation of epithelial proliferation. It is notable that the enhanced expression of TGF-a is selectively associated with UC and / 5e0b$$0013

03-11-96 17:04:05

gasa

not found in patients with CD. These observations imply that the enhanced expression is not simply a concomitant of inflammation but a distinctive feature of UC. It is noteworthy that Sottili et al. have recently observed up-regulation of TGF-a binding sites in experimental rabbit colitis.38 These studies showed increased TGF-a binding sites in mucosa, perhaps enabling TGFa and related peptides to facilitate mucosal protection and repair. Increased binding was also present in deeper muscle. The present studies used biopsy samples and reflect expression in human mucosa. The potential roles of TGF-b in IBD are more diverse. This protean factor is known to modulate functional characteristics of the entire spectrum of cells present in the inflammatory infiltrate in CD, including B and T lymphocytes and macrophages. TGF-b has also been found to be a potent chemotactic factor for neutrophils. Furthermore, Graham et al. have shown that TGF-b enhances collagen deposition by smooth muscle cells isolated from CD tissue, a process that may contribute to fibrosis and stricture formation in CD. 27 In addition, TGF-b is known to inhibit intestinal epithelial cell proliferation, a process that may inhibit reepithelialization after surface injury.5,6,21 However, paradoxically, TGF-b may nonetheless promote healing through stimulation of migration of epithelial cells into a wounded surface. Recent studies using rat intestinal epithelial IEC-6 cell line have shown TGF-b promotion of healing of wounded monolayers, despite inhibition of cell proliferation.22,39,40 As shown in Figures 4 and 5, TGF-b expression, minimal in uninflamed mucosa, is expressed in significant amounts in actively involved mucosa of patients with IBD. TGF-b expression is associated with inflammation, irrespective of the nature of the underlying disease process, suggesting it plays a role in processes present if not identical in both UC and CD. Determination of the actual role of TGF-b in these disorders is more problematic than that for TGF-a. Although Northern blot analysis focused on TGF-b1, presumably other forms of TGF-b such as TGF-b2 and TGF-b3 found in rodent intestine are also present in colonic mucosa.41 However, the functional distinctions and their importance between these TGF-b species is still unclear, and the bioassay shows that aggregate content of the collective species of TGF-b present is increased in association with inflammatory activity in IBD mucosa. The presence of the highest levels of TGF-b messenger RNA in the lamina propia cells closest to the surface epithelium suggests that TGF-b may indeed play a role in promoting healing of the overlying epithelium by enhancing the process of restitution, which serves to reestablish surface continuity after mucosal injury. WBS-Gastro

April 1996

TGF-a AND TGF-b IN IBD 983

References 1. Shirota K, LeDuy L, Yuan SY, Jothy S. Interleukin-6 and its receptor expression in human intestinal epithelial cells. Virchows Arch B Cell Pathol Incl Mol Pathol 1990;58:303–308. 2. Madara JL, Stafford J. Interferon-gamma directly affects barrier function of cultured intestinal epithelial monolayers. J Clin Invest 1989;83:724–727. 3. Kurokawa M, Lynch K, Podolsky DK. Effects of growth factors on an intestinal epithelial cell line: transforming growth factor b inhibits proliferation and stimulates differentiation. Biochem Biophys Res Commun 1987;142:775–782. 4. Menard D, Arsenalt P, Pothier P. Biologic effects of epidermal growth factor in human fetal jejunum. Gastroenterology 1988;94: 656–663. 5. Koyama S, Podolsky DK. Differential expression of transforming growth factors a and b in rat intestinal epithelial cells. J Clin Invest 1989;83:1768–1773. 6. Barnard JA, Beauchamp RD, Coffey RJ, Moses HL. Regulation of intestinal epithelial cell growth by transforming growth factor type b. Proc Natl Acad Sci USA 1989;86:1578–1582. 7. Coffey RJ Jr, Shipley GD, Moses HL. Production of transforming growth factors by human colon cancer cells. Cancer Res 1986;46:1164–1169. 8. Coffey RJ Jr, Goustin AJ, Soderquist AM, Shipley GD, Wolfshohl J, Carpenter G, Moses HL. Transforming growth factors a and b expression in human colon cancer lines. Cancer Res 1987;47:4590–4594. 9. Nathan C, Sporn M. Cytokines in context. J Cell Biol 1991;113:981–986. 10. Goodlad RA, Raja KB, Peters TJ, Wright NA. Effects of urogastrone-epidermal growth factor on intestinal brush border enzymes and mitotic activity. Gut 1991;32:994–998. 11. Opleta-Madsen K, Hardin J, Gall DG. Epidermal growth factor upregulates intestinal electrolyte and nutrient transport. Am J Physiol 1991;260:G807–G814. 12. Wahl SM, Wong H, McCartney-Francis N. Role of growth factors in inflammation and repair. J Cell Biochem 1989;40:193–199. 13. Wahl SM, McCartney-Francis N, Mergenhagen SE. Inflammatory and immunomodulatory roles of TGF-beta. Immunol Today 1989;10:258–261. 14. Winkler ME, O’Connor L, Winget M, Fendly B. Epidermal growth factor and transforming growth factor a bind differently to the epidermal growth factor receptor. Biochemistry 1989;28:6373– 6378. 15. Scheving LA, Shiurba RA, Nguyen TD, Gray GM. Epidermal growth factor receptor of the intestinal enterocyte. Localization to laterobasal but not brush border membrane. J Biol Chem 1989; 264:1735–1741. 16. Derynck R, Jarrett JP, Chen EY, Eaton DH, Bell JR, Assoian RK, Roberts AB, Sporn MB, Goeddel DV. Human transforming growth factor b complementary DNA sequence of expression in normal and transformed cells. Nature 1985;316:701–705. 17. Massague J, Cheifetz S, Ignotz RA, Boyd FT. Multiple type b transforming growth factors and their receptors. J Cell Physiol 1987;5:43–47. 18. Lyons RM, Gentry LE, Purchio AF, Moses HL. Mechanism of activation of latent recombinant transforming growth factor b1 by plasmin. J Cell Biol 1990;110:1361–1367. 19. McCaffrey TA, Domenick JF, Brayton CF, Agarwal LA, Welt FGP, Weksler BB. Transforming growth factor–b activity is potentiated by heparin via dissociation of the transforming growth factor–b/ a2 -macroglobulin inactive complex. J Cell Biol 1989;109:441– 448. 20. Yamaguchi Y, Mann DM, Ruoslahti E. Negative regulation of transforming growth factor– b by the proteoglycan decorin. Nature 1990;346:282–284.

/ 5e0b$$0013

03-11-96 17:04:05

gasa

21. Suemori S, Podolsky DK. Regulation of transforming growth factor expression in rat intestinal epithelial cell lines. J Clin Invest 1991;87:2216–2221. 22. Ciacci C, Lind SE, Podolsky DK. Transforming growth factor b regulation of repair in intestinal epithelial monolayers. Gastroenterology 1993;105:527–532. 23. Wahl SM, Hunt DA, Wakefield LM, McCartney-Francis N, Wahl LM, Roberts AB, Sporn MB. Transforming growth factor type beta induces monocyte chemotaxis and growth factor production. Proc Natl Acad Sci USA 1987;84:5788–5792. 24. Wahl SM, McCartney-Francis N, Allen JB, Dougherty EB, Dougherty SF. Macrophage production of TGF-beta and regulation by TGF-beta. Ann NY Acad Sci 1990;593:188–196. 25. Grotendorst GR, Smale G, Pencev D. Production of transforming growth factor beta by human blood monocytes and peripheral neutrophils. J Cell Physiol 1989;140:396–402. 26. Reibman J, Meixler S, Lee TC, Gold LI, Cronstein BN, Haines KA, Kolasinski SL, Weissmann G. Transforming growth factor b1, a potent chemoattractant for human neutrophils, bypasses classic signal-transduction pathways. Proc Natl Acad Sci USA 1991; 88:6805–6809. 27. Graham MF, Bryson GR, Diegelmann RF. Transforming growth factor beta 1 selectively augments collagen synthesis by human intestinal smooth muscle cells. Gastroenterology 1990;99:447– 453. 28. Rossiter G, Podolsky DK. Expression of transforming growth factor alpha and beta in colonic mucosa in ulcerative colitis (abstr). Gastroenterology 1990;98:A421. 29. McCabe RP, Secrist H, Botney M, Egan M, Peters MG. Cytokine mRNA expression in intestine from normal and inflammatory bowel disease patients. Clin Immunol Immunopathol 1993; 66:52–58. 30. Chirgwin JM, Przybyla AF, Rutters JW. Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 1979;18:5295–5299. 31. Sambrook J, Fritsch EF, Maniatis T. Molecular cloning: a laboratory manual. 2nd ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory, 1989. 32. Melton DA, Krieg DA, Rebagliati MR, Maniatis T, Zinn H, Green MR. Efficient in vitro synthesis of biologically active RNA and RNA hybridization probes from plasmids containing a bacteriophage sp6 promoter. Nucleic Acid Res 1984;12:7055–7056. 33. Kurokawa M, Lynch K, Podolsky DK. Multiplicity of transforming growth factors in human malignant effusions: presence of ‘‘new’’ TGF. Cancer Res 1988;48:1791–1797. 34. Barnard JA, Beauchamp RD, Russell WG, Dubois RN, Coffey RJ. Epidermal growth factor: related peptides and their relevance to gastrointestinal pathophysiology. Gastroenterology 1995;108: 564–580. 35. Eastwood GL, Trier JS. Epithelial cell renewal in cultured rectal biopsies in ulcerative colitis. Gastroenterology 1973;64:383– 390. 36. Allan A, Bristol JB, Williamson RCN. Crypt cell production rate in ulcerative proctocolitis: differential increments in remission and relapse. Gut 1985;26:999–1003. 37. Serafini EP, Kirk AP, Chambers TJ. Rate and pattern of epithelial cell proliferation in ulcerative colitis. Gut 1981;22:618–652. 38. Sottili M, Sternine C, Reinshager M, Brecka NL, Nast CC, Walsh JH, Eysselein VI. Up-regulation of transforming growth factor a binding sites in experimental rabbit colitis. Gastroenterology 1995;109:24–31. 39. Dignass AU, Tsunekawa S, Podolsky DK. Fibroblast growth factors modulate intestinal epithelial cell growth and migration. Gastroenterology 1994;106:1254–1262. 40. Dignass A, Podolsky DK. Cytokine modulation of intestinal epithelial cell restitution: central role of transforming growth factor b. Gastroenterology 1993;105:1323–1332.

WBS-Gastro

984 BABYATSKY ET AL.

GASTROENTEROLOGY Vol. 110, No. 4

41. Barnard JA, Coffey RJ, Pelton RW, Moses HL. Distribution of transforming growth factor b (TGFb) mRNA and protein in the small intestine (abstr). Gastroenterology 1990;98:A483.

Received July 27, 1995. Accepted December 8, 1995. Address correspondence to: Daniel K. Podolsky, M.D., Gastrointes-

/ 5e0b$$0013

03-11-96 17:04:05

gasa

tinal Unit, Massachusetts General Hospital, Fruit Street, Boston, Massachusetts 02114. Fax: (617) 726-3673. Supported by grants DK41557 and DK43351 from the National Institutes of Health, the Fundacion Gran Mariscal de Ayacucho, Latin American Scholarship Program of American Universities, Inc. (to G.R.), and a Lucille Markey research scholarship (to M.W.B.). The authors thank Michelle deBeaumont for technical assistance in performing bioassays.

WBS-Gastro