Intracellular variation of rat intestinal mucin granules localized by monoclonal antibodies.код для вставкиСкачать
THE ANATOMICAL RECORD 230:513-518 (1991) lntracellular Variation of Rat Intestinal Mucin Granules Localized by Monoclonal Antibodies MARY G. OLIVER AND ROBERT D. SPECIAN Department of Cellular Biology and Anatomy, Louisiana State University Medical Center, Shreveport, Louisiana 71130 ABSTRACT Monoclonal antibodies produced against rat small intestinal mucins were utilized to study variability of stored mucin granules within rat ileal goblet cells. Eleven antibody-secreting hybridoma cultures were produced; six of these uniformly labeled stored mucin granules in virtually all goblet cells, suggesting that some antigenic features are common to all granules. The other five stained goblet cells in the rat small intestinal epithelium nonuniformly. R803, R805, and R807 localized within almost all goblet cells but revealed differential labeling of centrally and peripherally located mucin granules. R804 uniformly labeled the mucin granules of most villous goblet cells; some of the crypt goblet cells were uniformly labeled, but the majority were only partially labeled, resulting in a mottled staining pattern. R808 stained only a small portion of crypt goblet cells; there is, however, a n increase in both number of goblet cells labeled and in uniformity of staining of the stored granule mass from the base of the crypt to the surface, resulting in uniform labeling of virtually all goblet cells at the villus tip. This study demonstrates for the first time that rat small intestinal mucin granules are immunologically heterogeneous and nonuniformly distributed within the epithelium. Additionally, staining patterns within the stored granule mass suggest that structurally distinct subpopulations of mucin granules may exist within a single goblet cell. Goblet cells are highly polarized exocrine cells distributed along the length of the intestinal tract. They synthesize mucins, high molecular weight (-2 x lo6 daltons) glycoproteins that are packaged into mucin granules that are contained in a tightly packed apical granule mass. Preliminary evidence suggests that a mannose-containing glycoprotein tightly associated with mucins is also co-packaged into mucin granules (Sibley and Specian, 1990). It has been suggested (Florey, 1962)that mucins have three functions in the gastrointestinal tract: 1)to protect the mucosa by forming a physical barrier to chemical and physical irritants; 2) to lubricate luminal contents and facilitate their movement; and 3) to entrap and agglutinate parasites in the gut for elimination. Histochemical studies in normal human colon (Filipe, 1969)and in normal rat colon (Thomopoulos et al., 1983) have shown that more than one distinct mucin type can be distinguished based upon carbohydrate composition. These studies have also shown that the population of goblet cells is heterogeneous with respect to the type of stored mucin and that there are qualitative and quantitative differences in stored mucins between different regions of the gastrointestinal tract. Biochemical analyses of rat colonic (Lamont and Ventola, 1980), human colonic (Podolsky and Isselbacher, 19831, and human small intestinal mucins (Wesley e t al., 1985) have demonstrated that more than one chemically distinct species of mucin glycoprotein exists and that these mucin species vary in both carbohydrate and amino acid composition. UnfortuQ 1991 WILEY-LISS, INC nately, the differences between biochemically distinct mucin species are not apparently dramatic enough to account for differential staining properties, and no direct correlation has so far been made between histochemical and biochemical analyses. Biochemical analyses of mucins in ulcerative colitis (Podolsky and Isselbacher, 1984) and in cystic fibrosis (Wesley et al., 1983) have demonstrated that in these disease states there are qualitative and quantitative alterations in the mucin glycoproteins produced. These studies suggest that the individual mucin species may serve specific physiological functions in the intestinal tract. Biochemical analysis, however, leaves unclear whether these glycoproteins are normally produced by all goblet cells, or whether the production of individual mucin species is relegated to distinct subpopulations of goblet cells. To attempt to localize distinct mucin glycoproteins within the mucosa, current studies have now focused upon detecting these mucin species by immunocytochemical methods. Podolsky and co-workers (1986), using specific monoclonal antibodies to human colonic mucins, have demonstrated distinct subpopulations of goblet cells, containing specific species or combination of species of mucin glycoproteins. The goals of the present study are to demonstrate, by using monoclonal Received June 14, 1990; accepted December 17, 1990 514 M.G. OLIVER AND R.D. SPECIAN antibodies against rat small intestinal glycoproteins, the existence of immunologically distinct subpopulations of goblet cells in rat small bowel. Furthermore, mucin granules within a n individual goblet cell are immunologically heterogeneous, and there are unique storage patterns of structurally distinct mucin granules. The purpose of this report is to present the preliminary results of this study. METHODS Preparation of Rat Intestinal Mucin All reagents are from Sigma Chemical Corporation, St. Louis, MO, unless otherwise specified. Male Wistar rats (150-300 g) were fasted overnight and anesthetized by a n intraperitoneal injection of sodium pentobarbital. The intestinal epithelium was separated from the underlying tissue by vascular perfusion with 20 mM EDTA (ethylenediaminetetraacetic acid) in 0.9% sodium chloride in phosphate buffer (PBS) a t pH 7.3 for 5 minutes (modified after Bjerknes and Cheng, 1981).The small intestine from the ligament of Trietz to the cecum was removed, bisected, and then slit open. The mucosal epithelium was gently scraped from the basement membrane with a microscope slide. This results in large quantities of intestinal epithelium with virtually no contamination from underlying tissues. Enrichment of rat intestinal mucin glycoproteins was performed by modification of the method of Lamont and Ventola (1980). Isolated epithelium was homogenized in 0.1 M dibasic potassium phosphate buffer with 5 mM EDTA at pH 7.3 for 6 minutes at 4°C. The homogenate was centrifuged 2 x 60 minutes a t lOO,OOOg, dialyzed against distilled water, resuspended in 0.1 M phosphate buffer at a concentration of 10 mgl ml, and centrifuged for one hour and 100,OOOg. The high molecular weight glycoproteins were then enriched by gel filtration through a Sepharose 4B (Pharmacia Fine Chemicals, Piscataway, NJ) column (2.4 x 40 cm). The void volume was determined independently with blue dextran. Void volume fractions were pooled, dialyzed against distilled water, and lyophilized. Antibodies prepared against glycoproteins purified by this method localize exclusively within mucin granules in goblet cells (Specian and Oliver, 1990). Preparation of Antiserum Against Rat Small Intestinal Mucins Balblc mice were immunized with 200 pg of purified mucins suspended in 0.2 ml of PBS emulsified with Freund's complete adjuvant via three subcutaneous injections over 6 to 12 weeks. Mice were then bled by cutting the distal region of the tail vein and collecting blood in a microcentrifuge tube. Blood was allowed to clot at room temperature for 30 min and was centrifuged in a clinical centrifuge for 10 min. Serum was carefully decanted into another microcentrifuge tube and centrifuged again for 10 min. Indirect immunofluorescence of the antiserum demonstrated localization of the antiserum to the apical granule mass of intestinal goblet cells, results similar to antiserum produced in rabbit against this antigen (Specian and 01iver, 1990). No brush border or membrane localization was evident. Production of Hybridomas Balblc mice were immunized with 200 yg of purified mucins suspended in 0.2 ml of PBS emulsified with Freund's complete adjuvant via three subcutaneous injections over 6 to 12 weeks and then were intravenously boosted with 200 pg of purified mucins suspended in 0.9% sodium chloride 4 days before fusion. Four days after boosting, mice were sacrificed by cervical dislocation. The spleens were removed under sterile conditions and the cells were dispersed by passage through a wire-mesh screen. Immunized mouse spleen cells (4.5 x lo7) were fused with NS-1 myeloma cells ( 3 x 1 0 7 in RPMI 1640 (GIBCO, Grand Island, NY) containing 40% polyethylene glycol by the method of Gefter et al. (1977). Fusion products were dispensed into 96-well culture plates a t a density of 7.5 x lo5 cells per well. Microcultures were then maintained in RPMI medium supplemented with 10% fetal calf serum and containing 1x lo5 M hypoxanthine, 4 x lo7 M aminopterin, and 1.6 x lo5 M thymidine. Cultures producing antibody were cloned by serial dilutions and expanded. Screening for Antibody Production At 14 days, microcultures were assayed for antibody production by a n ELISA (modified from Pool et al., 1983). Supernatants from each microculture were incubated overnight a t 4°C with a polystyrene ball coated with antigen and blocked with bovine serum albumin (BSA) and ovalbumin. A positive control (mouse antiserum) and a negative control (2% BSA in PBS) were run with each set of supernatants. The balls were rinsed with PBS/0.2% BSA/0.05% Tween and incubated with peroxidase conjugated to goat antimouse IgG for 2 hours at 24°C. Balls were rinsed and incubated with substrate (0.04% o-phenylenediaminel 0.001% H202in citratelphosphate buffer, pH 5.0) for 20 min. The reaction was stopped by addition of 2.4 M H2S04.Absorbance was read on a Hitachi spectrophotometer a t 492 nm. Supernatants with absorbance readings at 60% of the positive control or higher were judged to be producing antibody. lmmunocytochemistry Segments of ileum from male Wistar rats were immersion fixed with 2% paraformaldehyde in 0.1 M potassium phosphate buffer, rinsed, dehydrated to 95% ethanol, and embedded in glycol methacrylate (JB-4, Polysciences, Warrington, PA). This preparation results in preservation of intact mucin granules in both villous and crypt goblet cells with little loss in antigenicity (Oliver et al., 1990). Immunocytochemical localizations were performed on 1.5 pm sections pretreated with 0.3% H202in methanol to block endogenous peroxidase (Streefkrek, 1972) and ovalbumin and BSA to block nonspecific staining. Sections were incubated overnight a t 4°C with supernatant; a positive control (mouse antiserum) and a negative control (2% BSA in PBS) were run with each set of supernatants. Sections were rinsed with PBSIBSAlTween and incubated with peroxidase conjugated to goat antimouse IgG for 2 hours a t 24°C. Sections were rinsed and incubated with substrate (0.05% diaminobenzidine with 0.001% H202 in 0.1 M Tris-HC1 buffer) for 20 min (Nakane and Pierce, 1966). Sections were washed with distilled wa- MONOCLONAL ANTIBODIES AGAINST RAT MUCINS Figs. 1-8. All figures are immunoperoxidase localizations of either mouse antiserum to rat small intestinal glycoproteins or a monoclonal antibody to rat small intestinal glycoproteins. All slides have been counterstained with Mayer’s hematoxylin. 515 Fig. 1. Localization of mouse antiserum results in uniform staining of the stored granule mass in all goblet cells in the crypts and on the villi. x 240. Fig. 2. Localization of R801 is indistinguishable from antiserum. All goblet cells have their stored granule mass completely labeled. x 240. ter and counterstained with Mayer’s hematoxylin. Slides were coverslipped and photographed on a n Olympus Vanox microscope. RESULTS Immunocytochemistry of antisera from mice immunized for spleen-harvest demonstrated uniform localization in the mucin granule mass of goblet cell throughout the epithelium, including those deep within the crypt (Fig. 1).The two fusions performed with immunized mouse spleens and mouse myeloma cells produced 11 antibody-producing hybridoma cell lines. Of these 11, i t was found that six of these cultures produced antibody that uniformly labeled secretory granules in all villous goblet cells and virtually all crypt goblet cells except those near the crypt base (Fig. 21, a staining pattern indistinguishable from that of mouse antisera (Fig. 1).The other five produced antibodies against epitopes that demonstrated a large degree of heterogeneity between surface and crypt goblet cells, between adjacent crypts, and between secretory granules in a single goblet cell. None of the antibodies localized within the brush border. R803 localizes within the apical granule mass of all surface goblet cells and most crypt cells except those found deep within the crypt (not illustrated). The majority of goblet cells had patches of stain dispersed throughout the granule mass. Additionally, there were goblet cells that contained labeled mucin granules within only the center of the granule mass. Immunocytochemical localization of R805 reveals that the majority of goblet cells contain mucin granules that are immunoreactive with this antibody. Within the crypts and on the villi the majority of the goblet cells have their apical granule mass labeled uniformly. Interspersed among the uniformly labeled cells exists a subpopulation of goblet cells that contain few or no immunoreactive granules (Figs. 3, 4).Interestingly, i t is only the centrally stored granules in those partially stained cells that react to R805, suggesting that there is spatial segregation of glycoproteins that contain this antigenic determinant. R807 labels the majority of goblet cells both on the villi and within the crypts; however, the granules do not label uniformly. In the goblet cells, R807 localizes sparsely but uniformly within the granule mass of crypt goblet cells (Fig. 5). but in some villous cells i t localizes more heavily within the peripherally located granules (Fig. 6). This staining pattern suggests that there are organizational distinctions between the immature crypt goblet cells and the more mature surface goblet cells. In a smaller fashion R804 stains both villous and crypt goblet cells irregularly (not illustrated). In the Figs. 3-8. MONOCLONAL ANTIBODIES AGAINST RAT MUCINS 517 crypts some of the goblet cells are stained uniformly, crypts, and sulphated mucins in colonic crypts. Biobut many of them have only a portion of their granules chemical analyses of human colonic (Podolsky and Isstained. This results in the partially stained cells hav- selbacher, 1983) and small intestinal mucins (Wesley ing a “mottled” granule mass. On the villi, the majority et al., 1985) a s well as r a t colonic mucins (Lamont and of the goblet cells are uniformly stained; the minority Ventola, 1980) have demonstrated multiple mucin speof villous goblet cells contain both labeled and unla- cies. The differences that defined a fraction as a mucin species, however, were more numerous and more subtle beled granules. All crypts contain a subpopulation of goblet cells that than was suggested by histochemical studies. Previous are labeled by R808, but rarely are all cells stained. studies in human colon have demonstrated, by using Cells a t the base of the crypt are either unstained or monoclonal antibodies, that distinct subpopulations of contain very little immunoreactive material. Goblet goblet cells exist (Podolsky et al., 1986). Results in this cells in the middle to upper crypt vary greatly in the study strongly suggest that, in rat ileum, goblet -cells amount of labeling within the granule mass. Uni- exist as a heterogeneous population. This variability, formly labeled goblet cells are found adjacent to those however, is not as dramatic a s suggested by histochemthat are either partially unstained or unstained (Fig. ical studies. Although a large degree of antigenic 7). At the villous base, cells that are only partially similarity of mucin granules stored in the mucosa is stained have the labeled granules located either within evident, specific differential staining patterns are dethe central region or a t the apex of the granule mass monstrable, not only within the epithelium as de(Fig. 8). On the upper region of the villus all cells are scribed by Podolsky and co-workers (Podolsky et al., uniformly labeled. Thus, it appears that the reactivity 1986), but also within distinct subpopulations of granto R808 increases from the base of the crypt to the tip ules within a single cell. Goblet cells arise by mitosis from multipotential of the villus. stem cells a t the base of the crypt and migrate to the DISCUSSION villus tip, where they die and are sloughed into the Histochemical studies on intestinal goblet cells have lumen (Chang and Leblond, 1971). During the migrasuggested that goblet cell mucins in both human colon tion the cells mature and undergo profound morpho(Filipe, 1969) and rat colon (Thomopoulos et al., 1983) logic changes that ultimately produce the characterisare heterogeneous due to variable carbohydrate con- tic, highly polarized shape (Radwan et al., 1990). Due tent. These studies demonstrated that not only are mu- to the common ancestry, i t would be reasonable to ascins present a s acidic, neutral, and sulphated classes, sume that goblet cells within a crypt would synthesize but also t h a t they are distributed unequally along the the same glycoprotein(s); however, there are marked length of the intestinal tract. Acidic mucins predomi- immunologic differences a s recognized by monoclonal nate on the upper villus and colonic surfaces, neutral antibody between adjacent cells. Four of the 11 antimucins on lower villus surfaces and in intestinal bodies produced did not stain stored mucins within all goblet cells of a n individual crypt; i t thus appears then that although goblet cells within a crypt may have originated with the ability to produce the same products, the synthetic mechanism within a n individual Fig. 3. R805 localization in crypt goblet cells. The majority of goblet cell can vary which chemical structures it will incorcells in the crypt label uniformly. A small subpopulation of goblet cells exist within the epithelium that have only the centrally stored porate into its apical granule mass. Not only do there appear to be different subpopulagranules labeled (arrows). x 500. tions of goblet cells in the mucosa, there also appear to Fig. 4. R805 localization in villous goblet cells. On the villus, the be immunologically distinct populations of mucin granmajority of goblet cells contain uniformly stained mucin granules. ules within a single goblet cell. Under baseline condiInterspersed among these are goblet cells that only contain labeled granules within the center ofthe apical granule mass (arrows). X 500. tions goblet cells preferentially secrete peripherally located granules; centrally located granules appear to Fig. 5. R807 localization in crypt goblet cells. Virtually all goblet not turn over, and remain a static population. In concells within the crypt are labeled sparsely, with labeled glycoproteins trast, during a n accelerated secretory event, such a s threaded through the granule mass. An occasional unlabeled goblet stimulation by cholinergic agents or luminal irritants, cell (asterisk) is found in the lower to middle crypt. X 500. mucus release is accomplished by compound exocytosis Fig. 6. R807 localization in villous goblet cells. All goblet cells on of predominantly centrally stored granules (Specian the villus are found to contain immunoreactive glycoproteins. Only a and Neutra, 1980,1982). In the present study the proportion of the total stored glycoproteins are labeled and are intermixed with the unlabeled glycoproteins. While labeled granules are duction of a n antibody that preferentially localizes found throughout the granule mass, the antibody localizes more within the central granule mass or along the periphery heavily within the peripheral granules. X 500. suggests that mucin granules secreted under baseline conditions to maintain the mucus layer are antigeniFig. 7. R808 localization in crypt goblet cells. Within the crypt, cally different than mucin granules that are secreted in there exists a distinct subpopulation of goblet cells that are entirely nonreactive to R808 (asterisks). This subset of goblet cells coexists response to a stimulus. Previous studies on human cowith one that is uniformly labeled with the antibody. In addition, a n lonic mucins have demonstrated that, under baseline intermediate group containing only a portion of their stored mucin conditions, mucosal explants maintained in organ culgranules labeled can be found. X 500. ture preferentially secrete certain mucin species into Fig. 8. R808 localization in villous goblet cells. At the base of the the media while retaining other species in the epithevillus the population of goblet cells that are intermediate between lium (Smith and Podolsky, 1987). These data suggest those labeled and those unlabeled. In these two goblet cells only the centrally and apically located granules are immunoreactive to R808. that intestinal goblet cells, under defined physiological conditions, secrete mucins with a specific structure and x 500. 518 M.G. OLIVER AND R.D. SPECIAN chemistry t h a t will be functionally advantageous to the organism. In conclusion, we have demonstrated that rat goblet cells, as has been previously documented for human goblet cells, constitute a heterogeneous population based on the products that they synthesize. Additionally, antigenically distinct granules can be found stored within a single goblet cell, with the granules segregated into structurally, and perhaps, functionally distinct subpopulations. ACKNOWLEDGMENTS The authors thank Ms. Kimberly Robinson for her technical assistance and Dr. Jerry W. Pickering for his invaluable guidance in establishing a hybridoma facility. This study was supported by NIH grant #DK 33720 and a grant from the Cystic Fibrosis Foundation. LITERATURE CITED Bjerknes, M., and H. Cheng 1981 Methods for isolation of intact epithelium from the mouse intestine. Anat. Rec., 199565-574. Chang, W.W.L., and C.P. Leblond 1971 Renewal of the epithelium in t h e descending colon of the mouse. Am. J . Anat., 131t73-100. Filipe, M.I. 1969 Value of histochemical reactions for mucosubstances in the diagnosis of certain pathological conditions of the colon and rectum. Gut, 10577. Florey, H.W. 1962 Secretion and functional of intestinal mucus. Gastroenterology, 43r326-329. Gefter, M.L., D.H. Margulies, and M.D. Scharff 1977 A simple method for polyethylene glycol-promoted hybridization of mouse myeloma cells. Somatic Cell Genet., 3t231-236. Lamont, J.T., and A S . Ventola 1980 Purification and composition of colonic epithelial mucin. Biochim. Biophys. Acta, 626t234-343. Nakane, P.K., and G.B. Pierce 1966 Enzyme-labeled antibodies: preparation and application for the localization of antigens. J . Histochem. Cytochem., 14t929-93 1. Oliver, M.G., S.S. Wiggins, and R.D. Specian 1990 Tissue preparation for immunocytochemical localization of goblet cell mucin. Trans. Am. Microsc. SOC.,IO9r205-212. Podolsky, D.K., D.A. Fournier, and K.E. Lynch 1986 Human colonic goblet cells: Demonstration of distinct subpopulations defined by mucin-specific monoclonal antibodies. J. Clin. Invest., 77t12631271. Podolsky, D.K., and K.J. Isselbacher 1983 Composition of human colonic mucin. Selective alteration in inflammatory bowel disease. J . Clin. Invest., 72t142-153. Podolsky, D.K., and K.J. Isselbacher 1984 Glycoprotein composition of colonic mucosa. Specific alterations in ulcerative colitis. Gastroenterology, 87t991-998. Pool, Chr.W., R.M. Buijs, D.F. Swaab, G.J. Boer, and F.W. Van Leeuwen 1983 On the way to a specific immunocytochemical localization. In: Immunohistochemistry. A.C. Cuello, ed. John Wiley and Sons, New York, pp. 1-46. Radwan, K.A., M.G. Oliver, and R.D. Specian 1990 Cytoarchitectural reorganization of rabbit colonic goblet cells during baseline secretion. Am. J. Anat., I89:in press. Sibley, D.A., and R.D. Specian 1990 Intestinal peroxidase, a secretory product of intestinal goblet cells. J. Cell Biol., lIlt441A. Smith, A.C., and D.K. Podolsky 1987 Biosynthesis and secretion of human colonic mucin glycoproteins. J. Clin. Invest., 80t300-307. Specian, R.D., and M.R. Neutra 1980 Mechanism of rapid mucus secretion in goblet cells stimulated with acetylcholine. J. Cell Biol., 85t626-640. Specian, R.D., and M.R. Neutra 1982 Regulation of goblet cell secretion. I. Role of parasympathetic stimulation. Am. J . Physiol., 242(Gastrointest. Liver Physiol. 5itG370-379. Specian, R.D., and M.G. Oliver 1990 Rat intestinal mucins: a n immunocytochemical study. Trans. Am. Microsc. SOC.,I09t213-222. Streefkerk, J.G. 1972 Inhibition of erythrocyte pseudoperoxidase activity by pretreatment with hydrogen peroxidase following methanol. J. Histochem. Cytochem., 202329-831. Thomopoulos, G.N., B.A. Schulte, and S.S. Spicer 1983 Light and electron microscopic cytochemistry of glycoconjugates in t h e rectosigmoid colonic epithelium of the mouse and rat. Am. J . Anat., 168t239-256. Wesley, A,, J. Forstner, R. Qureshi, M. Mantle, and G. Forstner 1983 Human intestinal mucin in cystic fibrosis. Pediatr. Res., 17:6569. Wesley, A,, M. Mantle, D. Man, R. Qureshi, G. Forstner, and J. Forstner 1985 Neutral and acidic species of human intestinal mucin: evidence for different core peptides. J. Biol. Chem., 260t79557959.