Expression and Compartmentalization of Integral Plasma Membrane Proteins by Hepatocytes and Their Progenitors in the Rat Pancreas
Total Page：16
File Type：pdf, Size：1020Kb
Expression and compartmentalization of integral plasma membrane proteins by hepatocytes and their progenitors in the rat pancreas JAMES R. BARTLES1'*, M. SAMBASIVA RAO2, LIQIN ZHANG1, BARBARA E. FAYOS1, CHERYL L. NEHME1 and JANARDAN K. REDDY2 Departments oflCell, Molecular and Structural Biology and 2Pathology, Northwestern University Medical School, 303 East Chicago Avenue, Chicago, Illinois 60611, USA * Author for correspondence Summary A combination of Western blotting, Northern blotting analogy to their respective localizations on hepato- and immunofluorescence was used to examine the cytes in liver, rat hepatic lectin-1 was concentrated expression and compartmentalization of plasma on those surfaces exposed to the pancreatic matrix at membrane proteins by those hepatocyte-like cells the periphery of the hepatocyte clusters (the basal that arise in the pancreases of rats subjected to surface equivalent), whereas HA 321 was concen- sequential dietary copper depletion and repletion. trated on those surfaces exposed to adjacent hepato- The pancreatic hepatocytes were found to: (1) ex- cytes within the clusters. The hepatocyte plasma press several integral membrane proteins known to membrane proteins were found to be expressed in be concentrated within the apical, lateral or basolat- the pancreas at different times during the copper eral domains of the plasma membranes of hepato- depletion/repletion protocol: for example, rat hep- cytes in liver; and (2) compartmentalize the mem- atic lectin-1 and the bulk of the HA 4 were expressed brane proteins to equivalent plasma membrane relatively late in the protocol, only after large domains, despite the organization of these cells into numbers of pancreatic hepatocytes had appeared; clusters instead of highly vascularized plates. The whereas dipeptidylpeptidase IV was induced > 10- apical plasma membrane proteins dipeptidylpepti- fold early in the protocol and proved to be an apical- dase IV and HA 4 were found to line bile canaliculus- specific marker for those ductular epithelial cells like openings between adjacent pancreatic hepato- that are believed to be the progenitors of the cytes; these openings were shown to be continuous pancreatic hepatocytes. with the pancreatic exocrine duct by India ink infusion. In contrast, the basolateral plasma mem- brane protein rat hepatic lee tin-1 and lateral plasma membrane protein HA 321 were detected elsewhere Key words: epithelial cells, cell surface polarity, plasma membrane domains, plasma membrane proteins, hepatocytes, about the surfaces of the pancreatic hepatocytes: by cell differentiation, pancreatic stem cells. Introduction detected changes in their expression and localization during liver regeneration (Bartles and Hubbard, 1986) One hallmark of cellular differentiation is the expression and upon exposure to a group of hepatocarcinogens known of the distinct complement of integral plasma membrane as the peroxisome proliferators (Bartles et al. 1990). (PM) proteins - receptors, enzymes, transporters, adhesion In the present study, we have used our antibodies to molecules, etc. - necessary to mediate specific interactions these PM proteins to examine the characteristics of those with a particular extracellular environment. This aspect hepatocyte-like cells that appear in abundance in the of cellular phenotype is highly developed in the case of pancreases of rats that have been subjected to sequential polarized cells, such as epithelial cells, whose PMs are dietary copper depletion and repletion (Eao et al. 1986, compartmentalized into domains that themselves contain 1988, 1989). These pancreatic hepatocytes are believed to distinct subsets of these functionally important integral arise through the proliferation and differentiation of membrane proteins (reviewed by Simons and Fuller, 1985; pluripotent pancreatic stem cells, most likely of ductular Rodriguez-Boulan and Nelson, 1989). or periductular origin (Rao et al. 1989). In many ways, this We have been using a predominantly imrnunological process is thought to be analogous to the differentiation of approach to study the PM of the hepatocyte, the major biliary epithelial cells and oval cells into hepatocytes, epithelial cell of the liver. We have identified a number of which can be observed in the liver under a variety of domain-specific integral proteins of the rat hepatocyte PM experimental-pathological conditions (reviewed by Sell, (Hubbard et al. 1985; Bartles et al. 1985a,6), compared the 1990; and Sirica et al. 1990). The pancreatic hepatocytes pathways along which they are sorted en route to the PM not only resemble liver hepatocytes by light and electron (Bartles et al. 1987; Bartles and Hubbard, 1988) and have microscopy (Scarpelli and Rao, 1981; Rao et al. 1982, 1983, Journal of Cell Science 98, 46-54 (1991) Printed in Great Britain © The Company of Biologists Limited 1991 45 1989), but they have been shown to exhibit a number of the Sections were stained with uranyl acetate and lead citrate and functional attributes of hepatocytes in liver. For example, examined on a JEOL-100 electron microscope. they: (1) proliferate following partial hepatectomy (Rao et For Western blotting, frozen tissue specimens were homogen- al. 1983); (2) display an increase in the number of ized in 4.5 volumes (ml g~' of moist tissue) of 0.25 M sucrose, 3 mM imidazole-HCl, pH7.4, containing protease inhibitors (lira peroxisomes upon dietary exposure to the peroxisome phenylmethylsulfonyl fluoride, l/zgrnl"1 each of antipain and proliferators methyl clofenopate and ciprofibrate (Rao et leupeptin and 25^gml~1 of aprotinin). SDS-polyacrylamide gel al. 1982; Reddy et al. 1984); and (3) express albumin and samples were prepared from the homogenates under reducing or certain other hepatocyte-specific proteins (Rao et al. 1983, nonreducing (for HA 321 and the epidermal growth factor 1989; Usuda et al. 1988; Dwivedi et al. 1990; Yeldandi et al. receptor, EGF-R) conditions and electrophoresed in 7.5% polyac- 1990). rylamide-SDS gels. The proteins were transferred to nitrocellu- lose and labeled sequentially with affinity-purified rabbit We have been intrigued by the possibility of using the 12s appearance of hepatocytes within the pancreas as a model polyclonal anti-PM protein antibodies and I-labelled protein A. system for examining the expression and compartmental- Since the specificities of these antibodies have been demonstrated in other articles (Bartles et al. 1985a,6, 1987; Bartles and ization of PM proteins during epithelial cell differen- Hubbard, 1986; Bartles et al. 1990), only the relevant segments of tiation and histogenesis. Here we have investigated the Western blot autoradiograms are shown. Estimates of relative whether the pancreatic hepatocytes express the proper protein concentration were made following scanning laser complement of integral PM proteins and whether these densitometric analysis (Ultroscan XL, Pharmacia-LKB Biotech- proteins are organized into compositionally distinct PM nology, Inc., Piscataway, NJ) of autoradiograms obtained at domains. reduced exposure relative to those shown; for each of the PM proteins examined, the peak area was found to vary with protein concentration in a roughly linear fashion over a wide range of protein input (data not shown). Materials and methods In preparation for Northern blotting, frozen tissue specimens were homogenized in guanidinium thiocyanate and the total RNA A more detailed account of the methods employed may be was isolated as outlined by Chirgwin et al. (1979). Total RNA was electrophoresed in 1 % agarose gels, transferred to nylon obtained by consulting Rao et al. (1989), Bartles et al. (1990) and 32 references cited therein. Male F344 rats (80-90 g; Charles River membranes and hybridized with nick-translated P-labeled Breeding Laboratories, Wilmington, MA) were housed as indi- cDNA for rat hepatic lectin 1 (Leung et al. 1986). viduals in hanging cages with wire-mesh bottoms in a tempera- ture- and humidity-controlled room and maintained on a 12-h light/12-h dark cycle. The experimental rats were given a copper- deficient diet (Copper Test Diet Catalogue no. 13835; United Results States Biochemical Corp., Cleveland, OH) to which the copper chelator triethylene-tetramine tetrahydrochloride (Aldrich Eight to nine weeks on the copper-deficient diet was Chemical Co., Milwaukee, WI) had been added to a final sufficient to cause a >90 % loss of pancreatic acinar cells concentration of 0.6 % (w/w) and deionized water ad Libitum for (cf. Fig. 1A and B). In place of the densly packed acini, we up to 8-9 weeks before being returned to normal (i.e. copper- containing) rat chow (Ralston Purina Co., St Louis, MO) for the observed collections of relatively small, nondescript cells designated period of time. Rats maintained on the copper- interspersed among adipocytes (Fig. IB). Examples of deficient diet gained weight at a slower rate than did control rats pancreatic ducts and islets of Langerhans could still be maintained on a diet of normal rat chow; compared to age- identified after 9 weeks on the copper-deficient diet matched controls, the rats maintained on the copper-deficient diet (Fig. IB). Within the first several days of starting copper for 8-9 weeks typically showed a 30 % lower body weight, a 75 % repletion, we began to detect small numbers of pancreatic reduction in pancreatic mass, a >90% loss in pancreatic acinar hepatocytes; it was common to observe just one or a few cells and<10% mortality. At the designated intervals, rats