immunolocalization of a new intestinal antiproliferative factor in human intestinal epithelial cells

Immunolocalization of a New IntestinalAntiproliferative Factor in Human Intestinal

Epithelial CellsCHRISTIAN LAVAGNA, PhD,* CARINE STRUP, PhD,† AGNES RAMPAL, MD,‡PAUL HOFMAN, MD,‡ SYLVIE BARDON, PhD,§ PATRICK RAMPAL, MD,* and

JEAN-CLAUDE POIREE, MD*

A new intestinal antiproliferative factor (IAF) with an approximate molecular weight of 120kDa has been purified from the human small intestine. This factor blocks the progression ofhuman colon adenocarcinoma cells HT-29 from the G1 to the S phase. IAF, specific of thelower part of the digestive tract, was detected rather late in mouse embryonic development.For determination of the specific intestinal cell producing IAF, long-term differentiatedmucus-secreting HT-29 Cl 16E and enterocytic HT-29 Cl 19A cell lines were used. IAF issynthesized exclusively in the intestinal goblet cells; it is processed in the RER and Golgicomplex before being excreted in secretory vesicles independently of mucin secretion. IAFcan be considered a growth inhibitor of intestinal proliferation for the same reason as TGF-�.However, two features differentiate it from TGF-�: (1) the intestinal cell type synthesizing it,and (2) the delay in its expression in embryonic development. Particular interest was paid toIAF expression in pathological conditions using human colon biopsies. IAF was consistentlyrecovered in biopsies from patients with inflammatory bowel diseases and benign tumors, butit was never detected in malignant tumors. IAF could represent a marker of colon cancerowing to its absence from malignant tumors.

KEY WORDS: proliferation; growth factor; immunology; goblet cell; intestine; human.

Cell growth is regulated by positive and negativecontrols. Besides tumor suppressor genes that displayan antiproliferative activity; (1) other genes code forextracellular growth-inhibitory proteins, such astransforming growth factor-� (TGF-�) (2) and leuke-mia inhibiting factor (LIF) (3). Among tumor-

suppressor genes, those that most frequently undergomutations in colorectal cancer are APC (adenoma-tous polyposis coli); (4) DCC (deleted in colorectalcancer); (5) and p53 (6). In colorectal carcinogenesis,mutation of the APC gene is thought to mark one ofthe earliest events, whereas mutation of p53 appearsto be a late phenomenon (7). The decrease in DCCexpression may have an important role in the progres-sion of colorectal cancer (8). Of the extracellulargrowth inhibitory proteins, TGF-� is the only growthfactor known to inhibit proliferation in the lowerdigestive tract (9). In a previous work, we demon-strated that a cytosolic fraction obtained from humansmall intestinal mucosa decreased labeled thymidineincorporation in the undifferentiated human adeno-

Manuscript received August 30, 2001; accepted January 28, 2002.From the *Laboratory of Gastroenterology and Nutrition, Fac-

ulty of Medicine, Nice, France; †IPSN, B.P.6, Fontenay aux RosesCedex, France; ‡Laboratory of Anatomic Pathology, CHU, Nice,France; and §Laboratory of Nutrition, INRA, Jouy en Josas,France.

This work was supported by grants from the Hospital ClinicalResearch Program (PHRC 97) and the Cancer Research Associa-tion (ARC 99).

Address for reprint requests: Dr Jean-Claude Poiree, Laboratoryof Gastroenterology and Nutrition, Faculty of Medicine, 28 Ave. deValombrose, 06107 Nice Cedex 2, France.

Digestive Diseases and Sciences, Vol. 47, No. 11 (November 2002), pp. 2446–2453 (© 2002)

2446 Digestive Diseases and Sciences, Vol. 47, No. 11 (November 2002)0163-2116/02/1100-2446/0 © 2002 Plenum Publishing Corporation

carcinoma HT-29 cell line. Using nondenaturing sep-aration techniques, we purified a protein migratingwith an apparent molecular weight of 120-kDa. Mo-lecular evidence has been obtained showing that thisfactor is different from TGF-� and has no inhibitoryeffect on nonintestinal epithelial cell lines (10). Be-cause our preliminary comparative studies on IAFexpression in malignant and benign colon tumorsrevealed a correlation with the loss of IAF phenotypeand colon cancer malignancy, intestinal cell localiza-tion of IAF was investigated with regard to its pro-duction in normal and pathological conditions. Con-sidering its organ specificity and loss of expression incolon cancer, IAF could represent a new marker of aparticular interest for colorectal cancer, which is aprominent public health problem.

MATERIALS AND METHODS

Materials. Culture materials including medium, antibiot-ics, and serum were purchased from Intermed (France).Preimmune and immune sera against intestinal antiprolif-erative factor (IAF) were prepared in our laboratory. En-hanced chemiluminescence (ECL) reagent was obtainedfrom Amersham (France). All other chemicals were ob-tained from commercial sources and were of the highestpurity available.

Cell Culture. Differentiated colon cancer cell lines HT-29Cl 16E and HT-29 Cl 19A were kindly supplied by Dr.Laboisse (11). The HT-29 D4 cell line was obtained in ourlaboratory in an undifferentiated state in the presence ofglucose-containing medium, according to Fantini et al (12).Cells were grown at 37°C in a 95% air–5% CO2 humidifiedatmosphere in a standard medium consisted of H21/F12medium supplemented with 10% (v/v) fetal bovine serum,penicillin (50 units/ml) and streptomycin (50 mg/ml). Themedium was changed three times a week.

Western Blot Analysis. Protein samples solubilized in 4%SDS (v/v) and boiled for 3 min were separated on a 7.5%polyacrylamide gel (w/v) using the Mini-Protean II Cell(Biorad). Western blots were obtained with the antiserumraised in rabbit against the purified IAF fraction using aMini Trans Blot apparatus (Biorad). Electrophoresis andwestern blot buffers and procedures have been describedpreviously (10).

Immunohistochemistry. Mice were killed by cervical dis-location, then immediately frozen in liquid nitrogen. Frozensections (5 �m) were cut on a cryostat and rinsed with PBS.Human colon tissue samples were fixed and embedded inparaffin; 4-�m sections were prepared in the AnatomicPathology Laboratory of the Archet Hospital (Nice,France). Immunohistochemical staining for IAF was per-formed with a rabbit polyclonal antibody against IAF (ob-tained in the laboratory). The 4-�m-thick sections of fixed,paraffin-embedded blocks were mounted on polylysine-coated slides. Following deparaffinization with xylene andethanol, and rehydration through decreasing concentra-tions of ethanol, endogenous peroxidase activity was elim-inated with 3% hydrogen peroxide in PBS. Nonspecific sites

were saturated using 2% bovine serum albumin in PBS for2 hr at room temperature. Sections were incubated for 1 hrat room temperature with the primary antibody at a dilutionof 1:200. Secondary incubation was performed with goatanti-rabbit IgG antibody conjugated to peroxidase at adilution of 1:500 for 45 min at room temperature. Thesections were stained at room temperature for 10 min withdiaminobenzidine� (DAB�, Dako, France). The tissueswere counterstained with differentiated Mayer’s hemalumsolution (Merck, Germany). After the slides had been coun-terstained, they were dehydrated through graded alcoholand xylene and mounted on cover-slipping film (Tissue-TekSCA, Bayer, France). Negative controls were obtained us-ing DAB� alone or the goat anti-rabbit IgG coupled toperoxidase followed by DAB� staining.

Immunoelectron Microscopy. For immunoelectron mi-croscopy, HT-29 Cl 16E mucus-secreting cells were fixed in3.7% paraformaldehyde and embedded at low temperatureinto LR White resin (Hard LR White, London, UK). Ul-trathin sections were put on 300-mesh nickel grids, washedwith PBS buffer (pH 7.4), then incubated for 1 hr at roomtemperature with anti-IAF antibodies (diluted 1:100). Afterwashing with PBS, the grids were incubated for 1 hr with 10nM colloidal gold-conjugated rabbit anti-mouse secondaryantibody (Tebu, Paris, France). Negative controls wereperformed using 10 nM colloidal gold-conjugated rabbitanti-mouse secondary antibody alone. The grids werewashed with PBS, then with distilled water, and stained withuranyl acetate. Sections were examined with a Jeol 1200EXII electron microscope.

Distribution of IAF and Mucin 2 Using Immunochemis-try and Confocal Microscopy. HT-29 Cl 16E cell monolay-ers were washed extensively with PBS and fixed with meth-anol for 10 min. Polyclonal rabbit anti-IAF antibody(dilution 1:200) and mouse monoclonal IgG antibodyagainst human mucin 2 (dilution 1:100) (Tebu, Paris,France) were incubated with the cells for 45 min at 37°C.The monolayers were washed and then treated with tetra-methylrhodamine isothiocyanate-conjugated anti-rabbitIgG (Sigma) and fluorescein isothiocyanate-conjugated an-ti-mouse IgG (Dakopatts F205) for 45 min at 37°C. Afterwashes with PBS, observation was carried out with a Zeissconfocal laser scanning microscope.

RESULTS

Expression of IAF in Mouse Embryo. Intestinalantiproliferative factor has been purified from humanintestinal mucosa, allowing preparation of specificpolyclonal antibodies. The immune serum cross-reacted with both rat and mouse proteins migratingwith the same molecular weight. To study the antigenexpression, immunohistochemistry experiments wereperformed on mouse embryos, focusing on the diges-tive tract. Positive staining appeared in the 16-day-oldfetus (Figure 1).

Specific Localization of IAF Among Cells of Intes-tinal Mucosa. Four main differentiated cell types arepresent along the crypt–villus axis of the intestine:

NEW INTESTINAL ANTIPROLIFERATIVE FACTOR

2447Digestive Diseases and Sciences, Vol. 47, No. 11 (November 2002)

enterocytes (80%), goblet cells (15%), endocrine, andPaneth cells. All of these cells derive from the sameundifferentiated stem cell present in the crypts (13).Our aim was to determine which of these differenti-ated cell types synthesizes the IAF protein.

Immunohistochemistry performed on human smallintestinal tissue sections revealed positive staining forIAF from the crypts to the top of the villi with zonesof intense staining along the villi (Figure 2A). Thislocalization seems to be similar to the distribution ofgoblet cells on the intestinal villi. In addition, thestaining for IAF coincided with mucin staining byAlcian blue; IAF was also found in the intestinallumen with the bulk of mucins (Figure 2B,C).

To obtain direct evidence on the IAF productivecell type, two stable differentiated populations of en-terocytic and mucus-secreting phenotypes were se-lected. By exposing the human HT-29 adenocarci-noma cell line to sodium butyrate, Augeron andLaboisse (11) isolated a stable differentiated popula-tion named HT-29 Cl 16E with the characteristics ofmucus-secreting cells. Immunostaining of these cellswith IAF antibodies revealed an obvious positive in-

tracellular labeling 6 days after seeding (Figure 3). Toconfirm that the staining was due to the presence ofIAF, western blot analysis using IAF antibodies wascarried out on the bulk of proteins isolated from apostnuclear fraction prepared by sonication of HT-29Cl 16E grown to confluence (day 6 after seeding). Asshown in Figure 4, a protein band of 120 kDa, theapproximate molecular weight of the purified IAF,was detected. The use of a preimmune serum con-firms the nonspecific background of the other bandsseen on the gel. In contrast, another population ofbutyrate selected cells with the enterocytic pheno-type, named HT-29 Cl 19A, displayed weak stainingby immunocytochemistry (Figure 3) and no immuno-labeling of the IAF protein by western blotting (Fig-ure 4). This result agrees with our previous observa-tion that Caco2 cells, which are known to exhibittransport properties of the enterocytic phenotype(14), are negative when immunostained with IAFantibodies (10). The 120-kDa IAF protein was alsoabsent in a postnuclear fraction from undifferentiatedHT-29 cells (not shown). Moreover IAF expressionbegan earlier than that of mucins as observed by

Fig 1. Expression of IAF in mouse embryo. Sections from 13 day-old to 16-day-old fetus were obtained from Novagen (MadisonWisconsin, USA). Immunohistochemical staining was performed as described in Materials and Methods using anti-IAF antibodies.Positive staining (arrows) was not detected before the day 16. Magnification was �40 in all slides, focused on the intestine.

LAVAGNA ET AL

2448 Digestive Diseases and Sciences, Vol. 47, No. 11 (November 2002)

immunolabeling of IAF and mucin-2 on mucus-secreting HT-29 Cl 16E at different times after seed-ing. The staining of IAF became visible at day 4 whilemucin detection was only obvious at day 8 (notshown).

Subcellular Localization of IAF. As IAF appar-ently belongs to the family of extracellular growthinhibitory proteins, it might be secreted by gobletcells before binding to its target cell. If so, IAFwould have to be synthesized on rough endoplasmicreticulum and processed in the Golgi complex be-fore exocytosis. When rough endoplasmic reticu-lum and Golgi fractions were isolated from humansmall intestinal mucosa extract by fractionation ondiscontinuous sucrose gradient according to Mok-tari et al, (15), IAF was detected in both fractionsby western blotting (data not shown). Protein gly-cosylation is one of the major activities of thelumen of endoplasmic reticulum and Golgi com-plexes. IAF would thus be glycosylated during pro-tein processing. IAF was purified by affinity chro-matography on an immunoadsorbent gelconstituted with IgG purified from the immuneantiserum against IAF coupled to Sepharose (10).The retained fraction was eluted and concentrated,then incubated with N-glycosidase F for 24 hr at37°C; N-glycosidase F (EC 3.5.1.52) is an endogly-cosidase commonly used to deglycosylate N-linkedoligosaccharides from N-glycoproteins. After glyco-

sidase digestion, the molecular weight of IAF waslower than that of IAF incubated without the en-zyme under the same conditions (Figure 5).

Immunoelectron microscopy was carried out onHT-29 Cl 16E mucus-secreting cells to determine theintracellular localization of IAF. Analysis of anti-IAFantibody-labeled cells revealed numerous gold beads,predominantly inside the cytoplasm, and in particularinside the secretory vesicles (Figure 6). No gold beadswere seen in the cytosol, the nucleus, or the mem-branes. No staining was noted after use of the sec-ondary antibody alone. Confocal microscopy on thesame cells at day 10 also revealed that IAF andmucin-2 were not localized in the same compartmentof the cell (Figure 7).

Fate of IAF in Bowel Diseases. About 30 tissuesamples representative of various colon diseaseswere also tested for the presence of IAF. Tissuesections were prepared as described in Materialsand Methods, using rabbit anti-IAF serum as theprimary antibody and anti-rabbit IgG conjugated toperoxidase as the secondary antibody. As shown inFigure 8 on three representative samples, IAF im-munostaining was positive in inflammatory diseasessuch as scar regenerating areas of Crohn’s disease(A), in celiac disease (not shown), and mild dyspla-sia (B), whereas it was always negative in adeno-carcinoma (C).

Fig 2. Immunolocalization of IAF in normal human small intestinal mucosa. Tissue sections from human small intestinal mucosawere probed with the antiserum against IAF. Sections were stained with diaminobenzidine (A,B) and counterstained withhemalum solution (A). Mucins were visualized with Alcian blue dye (C). Magnification was �10 in A and �40 in slides B andC.



DISCUSSION

In a previous study, we described the purification of anew growth inhibitory factor from human small intes-tine that we named intestinal antiproliferative factor(IAF). IAF has an approximate molecular weight of 120kDa and blocks the proliferation of the undifferentiatedintestinal cell line HT-29 (10). When IAF immunostain-ing was performed on intestinal tissue sections obtained

at different days during the development of mouse em-bryos, IAF appeared in the late stage of the embryonicdevelopment, when the small intestine shows numerousrelatively thick villi and the large intestinal crypts areforming. This observation reveals that the antigen isassociated with well-differentiated structures. Histo-chemical localization studies have shown that TGF-�,the only growth factor known to inhibit proliferation in

Fig 3. Expression of IAF in differentiated cells of enterocytic and mucus-secretingphenotypes. HT-29 Cl 16E mucus-secreting cells (A,B) and HT-29 Cl 19A enterocyticcells (C,D) were seeded at day 0 on slides in the standard medium. At day 3 (A,C) and6 (B,D) the cells were stained as indicated in Materials and Methods using anti-IAFantibodies. An increase in the positive staining was observed between days 3 and 6 onlyin mucus-secreting cells. Magnification was �10.

LAVAGNA ET AL


the lower digestive tract (9), is expressed throughout itsembryonic development (16).

Preferential immunostaining of IAF was obvious onthe villi of human intestinal tissue sections. This rulesout IAF production by Paneth cells, since these differ-entiated cells remain in the crypts. The main cell typespresent in the intestinal villi are enterocytes and gobletcells. Analysis of IAF immunostaining from the proxi-mal to the distal lower digestive tract revealed thatstaining increased from the duodenum to the colon(data not shown). The fact that goblet cells have asimilar cell density gradient (17) is in favor of IAFproduction by these cells. To demonstrate the role ofmucus-secreting cells in IAF synthesis, we used perma-nently differentiated clones from the human coloniccancer cell line HT-29 (11). It has been well establishedthat human colon cancer cells as such HT-29 exhibitdifferentiation characteristics very similar to those asso-ciated with the terminal differentiation of normal en-terocytes or goblet cells (14). In our experiments, onlythe mucus-secreting HT-29 Cl 16E cells displayed obvi-ous labeling with anti-IAF antibodies, whereas the en-terocytic HT-29 CL 19A cells were not labeled. Thisfinding further highlights the difference between IAFand TGF-�; the latter is expressed at a higher level inenterocytes than in goblet cells (18).

However, in addition to their inhibitory effect oncell proliferation in the G1 phase of the cell cycle, IAFand TGF-� share the characteristics of secretorypolypeptides. We, in fact, demonstrated that IAF isprocessed in the RER and Golgi apparatus beforebeing excreted in secretory vesicles. In this field it canbe noted that IAF and mucins are probably segre-gated in different secretory vesicles. Moreover, inlight of the hydrolysis of the glycan moiety by N-glycosidase F, it is obvious that IAF is a N-glycoprotein, unlike the mucins, which are O-glycoproteins synthesized by the same cells.

Certain inflammatory bowel diseases such as Crohn’sdisease result in abrasion of the intestinal mucosa, withshrinkage of the epithelial cells; IAF is always expressedin such cases, although at a lower level related to thereduction in goblet cells. Similarly, IAF production isnot affected by benign tumor growth, as observed intubular polyadenoma. This is not the case for neoplasticprocesses. The absence of IAF in tissue sections wasobserved in malignant adenocarcinoma. The IAF genemight be partially or completely deleted by tumorigen-esis, as observed for the anti-oncogenes APC (4), DCC(5), and MCC (19). The absence of the 120-kDa proteinin western blot analysis of cytosolic extracts from biop-sies of malignant tumors (10) seems to confirm thishypothesis. Our studies have potential implications forthe diagnosis of colorectal cancer. Until now a mutation

Fig 4. Western blot of IAF in differentiated cells. HT-29 Cl 16Emucus-secreting cells (lanes 1 and 3) and HT-29 Cl 19A enterocyticcells (lane 2) were grown in standard medium for 6 days. Afterthree freeze–thaw cycles, the cells were sonicated three times for 30sec at the maximal intensity in PBS (pH 7.4), then the homogenatecentrifuged for 30 min at 10,000 g in a Microfuge (Eppendorf).Twenty micrograms of each supernatant was loaded on a 7.5%SDS-polyacrylamide slab gel under reducing conditions. Immuno-blotting of IAF was performed with anti-IAF immune (lanes 1 and2) or preimmune (lane 3) sera. Using an ECL kit (Amersham) IAF(approximate mw of 120 kDa) was only detected on mucus-secreting cell extract. The protein markers used were myosin (200kDa), �-galactosidase (116 kDa), phosphorylase b (97 kDa), andovotransferrin (76–78 kDa).

Fig 5. Deglycosylation of IAF by N-glycosidase F. IAF was purifiedfrom human small intestinal mucosa by affinity chromatography aspreviously described (10). Samples of 2 �l of IAF at 1 mg/ml werediluted in 0.1 ml final volume of 0.1 M sodium phosphate buffercontaining 25 mM EDTA adjusted at pH 7.4; after boiling for 3 minN-glycosidase F (20 units/ml) was added to sample 2. Controlwithout glycosidase was prepared in the same conditions (sample1). Both samples were stirred for 24 hr at 37°C, then solubilized in4% SDS. IAF was detected by western blotting using anti-IAFantibodies.



Fig 6. Intracellular localization of IAF by immunoelectron microscopy. HT-29 Cl 16E mucus-secreting cells weretreated as described in Materials and Methods for observation by electron microscope. Bars: 1 �m. Anti-IAFantibodies: beads were only observed inside vesicles. Original magnification was �30,000.

Fig 7. Representative confocal laser scanning micrographs of HT-29 Cl 16E cell monolayers immunostained for IAF and mucin 2 withspecific antibodies against these proteins followed by secondary antibodies conjugated to fluorescein isothiocyanate (mucin) ortetramethylrhodamine isothiocyanate (IAF). Original magnification �2400.

LAVAGNA ET AL


in the APC gene was the only potential marker suscep-tible to provide information in the early stages of intes-tinal tumorigenesis, even though translation of biologi-cal observations into clinically useful markers takes along time (20). IAF could be a good parameter forevaluation of colorectal cancer: it was absent in all of thehuman colon cancers we studied, and it is easy to detect.An exhaustive study on additional biopsy material andsurgical specimens is still needed, however, before IAFcan be recommended as a marker for diagnosis ofcolorectal cancer.

ACKNOWLEDGMENTS

We are grateful to Mrs. V. Foussard and S. Fournel fortechnical assistance.

REFERENCES

1. Weinberg RA: Tumor suppressor genes. Science 254:1138–1146, 1991

2. Massague J: The transforming growth factor � family. AnnuRev Cell Biol 6:597–641, 1990

3. Stahl J, Gearing DP, Willson TA, Brown MA, King JA, GoughNM: Structural organization of the genes for murine and humanleukemia inhibitory factor. Evolutionary conservation of codingand non-coding regions. J Biol Chem 265:8833–8841, 1990

4. Groden J, Thliveris A, Samowitz W, Carlson M, Gelbert L,Albertsen H, Joslyn G, Stevens J, Spirio L, Robertson M:Identification and characterization of the familial adenoma-tous coli gene. Cell 66:589–600, 1991

5. Fearon ER, Cho KR, Nigro JM, Kern SE, Simons JW, RuppertJM, Hamilton SR, Preisinger AC, Thomas G, Kinzler KW,Vogelstein B: Identification of a chromosome 18q gene that isaltered in colorectal cancer. Science 247:49–56, 1990

6. Baker SJ, Markowitz S, Fearon ER, Willson JKV, VogelsteinB: Suppression of human colorectal carcinoma cell growth bywild-type p53. Science 249:912–915, 1990

7. Kinzler KW, Vogelstein B: Lessons from hereditary colorectalcancer. Cell 87:159–170, 1996

8. Saito M, Yamaguchi A, Goi T, Tsuchiyama T, Nakagawara G,Urano T, Shiku H: Expression of DCC protein in colorectal

tumors and its relationship to tumor prometastasis. Oncology56:134–141, 1999

9. Barnard JA, Beauchamp RD, Coffey RJ, Moses HL: Regula-tion of intestinal cell growth by TGF �. Proc Natl Acad SciUSA 86:1578–1582, 1989

10. Lavagna C, Poiree JC, Fournel S, Rampal P: Purification of anew intestinal antiproliferative factor from normal humansmall intestine. Eur J Biochem 259:821–828, 1999

11. Augeron C, Laboisse CL: Emergence of permanently differen-tiated cell clones in a human colonic cancer cell line in cultureafter treatment with sodium butyrate. Cancer Res 44:3961–3969, 1984

12. Fantini J, Abadie B, Tirard A, Remy L, Ripert JP, El Battari A,Marvaldi J: Spontaneous and induced dome formation by twoclonal cell populations derived from a human adenocarcinomacell line HT29. J Cell Sci 83:235–249, 1986

13. Cheng H, Leblond P: Origin, differentiation and renewal of thefour main epithelial cell types in the mouse small intestine.Am J Anat 141:461–562, 1974

14. Zweibaum A, Laburthe M, Grasset E, Louvard D: The gastro-intestinal system. Intestinal absorption and secretion. In Hand-book of Physiology, Vol IV. M Field, RA Frizzell (eds). Be-thesda, American Physiological Society, 1991 pp 223–255

15. Moktari S, Feracci H, Gorvel JP, Mishal Z, Rigal A, Maroux S:Subcellular fractionation and subcellular localization of ami-nopeptidase N in the rabbit enterocytes. J Membr Biol 89:53–63, 1986

16. Thompson NL, Flanders KC, Smith JM, Ellingsworth LR,Roberts AB, Sporn MI: Expression of TGF-�1 in specificcells and tissues of neonatal mice. J Cell Biol 108:661– 669,1989

17. Neutra MR: The gastrointestinal tract. In Histology: Cell andTissue Biology. L Weiss (ed). New York, Elsevier Biomedical,1983, pp 642–683

18. Hafez MM, Infante D, Winawer S, Friedman E: TGF-�1 actsas an autocrine-negative growth regulator in colon enterocyticdifferentiation but not in goblet cell maturation. Cell GrowthDiffer 1:617–626, 1990

19. Kinzler KW, Nilbert MC, Su LK, Vogelstein B, Bryan MT,Levy DB, Smith KJ, Preisinger AC, Hedge P, McKechnie D:Identification of FAP locus genes from chromosome 5q21.Science 253:661–665, 1991

20. Banerjee AK: DCC expression and prognosis in colorectalcancer. The Lancet 349:968, 1997

Fig 8. Immunohistochemical staining for IAF of fixed and paraffin-embedded human colon tissues. Preparation of tissue sections andimmunohistochemical staining are described in Materials and Methods. (A) Colonic mucosa from Crohn’s disease showing glands ofLieberkuhn positively stained on the left and pyloric metaplastic glands unstained on the right. (B) Tubular polyadenoma showing milddysplasia stained (arrows). (C) Differentiated tubular colon adenocarcinoma unstained compared with normal colon mucosa positivelystained at the right bottom of the slide. Magnification was �100 in all cases.



immunolocalization of a new intestinal antiproliferative factor in human intestinal epithelial cells

Documents