THE ANATOMICAL RECORD 212:364-367 (1985) Functional Gap Junctions in Mouse Small Intestinal Crypts MATTHEW BJERKNES, HAZEL CHENG, AND STANLEY ERLANDSEN Department of Anatomy, University of Toronto, Toronto, Ontario, Canada M5S l A 8 (M.B., H. C.) and Department of Anatomy, University of Minnesota, Minneapolis, M N (S.E.) ABSTRACT We demonstrate intercellular transfer of Lucifer yellow and the existence of gap junctions in isolated mouse small intestinal crypts. The pattern of dye transfer approximates the normal pattern of cell proliferation and differentiation in the intestinal epithelium. These findings indicate that the cells of normal crypts form a n effective intercellular continuum. This intercellular route may function in the establishment of chemical, ionic, or electrical fields, which in turn may play a role in the control of cell proliferation, differentiation, and secretion in the crypt. It is widely held that establishment of and response to fields is responsible in large measure for many developmental phenomena (Turing, 1952; Crick, 1970; Meinhardt, 1982). In the embryo, these morphogenic fields often span many cells. As a result it is thought likely that many fields (e.g., chemical or ionic gradients) are established through a direct intercellular route of some kind. The most likely candidate for the intercellular channel is the gap junction (Wolpert, 1978; Lo, 1980; Loewenstein, 1981; Othmer, 1983; Warner et al., 1984; Guthrie, 1984). It is plausible that mechanisms similar to embryonic fields help govern cell proliferation and differentiation in the adult intestinal epithelium. Crypts do respond to injury in a manner consistent with the existence of regulatory fields, although these crypt responses have usually been ascribed largely to the effects of some form of feedback from the villus, e.g., through a chalone-like mechanism (Cairnie, 1967; Rijke et al., 1976; Wright and Al-Nafussi, 1982). To begin investigating the plausibility of field-like control mechanisms acting in the intestinal epithelium, we attempted to determine whether functional gap junctions exist in the normal adult epithelium. In this paper we report the presence of functional gap junctions in mouse intestinal crypts, as demonstrated by fluorescent dye transfer and freezefracture studies. placed in PBS, and then used immediately to study dye transfer in the villus epithelial cell population. Fluorescent Dye Transfer The ability of mouse duodenal epithelial cells to transfer the dye Lucifer yellow CH (mol. wt. 457.3, Sigma) was tested by iontophoretic microinjection (micropipettes were pulled from Omega Dot glass and filled with 4% solution of Lucifer yellow in distilled water; 2 rectangular iontophoretic current pulses of 200-msec duration, and 1- to 2- nA amplitude were applied each second; Stewart, 1978) into either villus cells in mucosal strips (20 mice were used to prepare mucosal strips, from each strip 2-3 villi were studied) or into the cells of isolated but intact crypts (usually only one or two crypts were used from any individual mouse; about 100 mice were used in this study). For any given cell, the injection period lasted about 30 sec. The injection and transfer of the dye were monitored through a Dage silicon intensified target video camera mounted on a fluorescence microscope. Freeze-Fracture Studies Five mice were used for freeze-fracture studies. The crypts were isolated directly into 2.5% glutaraldehyde, fixed for a n additional 30 min in the same fixative, and then stored in 0.1 M cacodylate buffer containing 5% sucrose. The tissue was cryoprotected, fractured, and MATERIALS AND METHODS immediately replicated without etching. The replicas Preparation of Intestinal Crypts and Villi were collected on Formvar-coated slot grids for examiMale Swiss albino mice weighing 20-25 gm were used nation. The nomenclature for identification of memthroughout the study. Intestinal crypts were isolated as brane surfaces as proposed by Branton et al. (1975) was described previously (Bjerknes and Cheng, 1981). Briefly, followed. mice were perfused through the left ventricle with 30 RESULTS mM EDTA. A segment of duodenum was removed and Dye Transfer Studies everted, and intestinal crypts were isolated by mechanWe were able to demonstrate efficient dye transfer ical vibration. The tissue was suspended in PBS and used immediately following preparation. Only viable between crypt-base columnar cells. When Lucifer yellow crypts were used for dye injection studies. In other studies, strips of mucosa, about 4-5 villi long and 1 villus Received November 16,1984; accepted March 14,1985. wide, were removed from fresh segments of duodenum, 0 1985 ALAN R. LISS, INC FUNCTIONAL GAP JUNCTIONS IN CRYPTS was injected into a cell in this region, the dye was rapidly transferred to many neighboring cells, first to those immediately adjacent and then to cells more distant (Fig. 1, a-c). At this point, the exact number of cells involved was difficult to determine because of the threedimensional nature of the tissue. The average time lapsed (ATL)between the time of injection of a given cell and the time of dye appearance in a n adjacent cell in the crypt-base columnar cells was 2.5 f 0.16 sec (X 5 SE, n = 35). Crypt-base columnar cells were not observed to transfer dye to Paneth cells and vice versa. No dye transfer was observed in most Paneth cells (n = 75), which were easily recognized from their size, shape, and granular content. However, a slow transfer of dye to a n immediately adjacent Paneth cell was observed in a few cases with a n ATL of 27.5 5 6.94 sec (n = 71, (Fig. 1,d0. (Unlike Paneth cells, mucous and enteroendocrine cells were not readily identifiable; thus we were not able to study their ability to transfer Lucifer yellow.) We observed efficient dye transfer in midcrypt epithelial cells (Fig. 1,g-i) with a n ATL of 4.8 0.17 sec (n = 28) and 9.0 0.36 sec (n = 63) in the lower and upper halves of the midcrypt, respectively. When Lucifer yellow was injected into lower midcrypt cells, the dye was rapidly transferred to many neighboring cells (Fig. 1, i). In the case of upper midcrypt cells, the dye appeared to be transferred to a smaller number of neighboring cells. Only a low degree of dye transfer was observed in crypttop epithelial cells with a n ATL of 15.2 f 1.15 sec (n = 211, (Fig. 1, j-1). Lucifer yellow transfer was observed only to immediately adjacent crypt-top cells. We were not able to demonstrate dye transfer in any region of the villus (n = 97). Freezefracture Studies In freeze-fracture studies we found clusters of intramembranous particles (IMP) resembling gap junctions in epithelial cells at all levels of the crypt. Fractures permitting identification of Paneth cell, goblet cell, or enteroendocrine cell membrane surfaces were rarely observed. Therefore the presence or absence of gap junctions on these cells could not be properly evaluated. Single gap junctions were routinely seen on the lateral surface of crypt cells and as many as three gap junctions were observed in a single cell (Fig. 1,m-0). The size of the IMP ranged from 6.8 to 10.0 nm in diameter with a mean size of 8.5 k 0.6 nm (n = 1201, and the number of IMP in a n individual gap junction varied from a few dozen to more than several hundred. In individual crypt cells, gap junctions were observed in fracture faces of the lateral surface from the level of the nucleus to just beneath the tight junction. We also found small gap junctions in villus epithelial cells. DISCUSSION Correlation of Dye Transfer With Proliferation and Differentiation We found that the pattern of dye transfer approximates the normal pattern of cell proliferation and differentiation in the intestinal epithelium. Specifically, we found that immature cells with extensive proliferative abilities, i.e., crypt-base columnar cells and lower midcrypt cells CLeblond and Cheng, 1976), demonstrated efficient dye transfer. In higher regions of the crypt, where cells begin to terminally differentiate and hence 365 cease proliferating, cells decreased in dye transfer efficiency until, in the villus (where cells are mature and do not divide), we were not able to demonstrate dye transfer. Similarly, no or very low efficiency dye transfer was observed between Paneth cells, terminally differentiated nonproliferating cells found in the crypt base (Leblond and Cheng, 1976). In summary, actively proliferating epithelial cells demonstrated a n ability to transfer Lucifer yellow. As these cells stop proliferating and terminally differentiate they gradually lose their ability to transfer the dye under our experimental conditions. Freeze-fracture studies of the intestinal epithelium have previously reported gap junctions (Staehelin, 1972; Madara et al., 1981; Curtis et al., 1984); however, the location of the junctions within the epithelium was not stated. We found gap junctions in epithelial cells a t all levels of the crypt. While we also found that villus epithelial cells possess small gap junctions, we could not demonstrate transfer of Lucifer yellow between villus epithelial cells. Consistent with the presence of gap junctions is Sheridan’s observation (personal communication) of intercellular communication as demonstrated by electrotonic coupling in villus epithelial cells in preliminary experiments. Therefore it is plausible that the gap junctions present in villus cells may be qualitatively different from those present in crypt cells, e.g., the cellto-cell channels may have been scaled down so that molecules such as Lucifer yellow can no longer pass but the channels are still permeable to the smaller ions. Such regional differences in gap junction permeability have been observed in other systems (Lo, 1980; Iwatsuki and Petersen, 1978a). Gap Junctions and Secretion Intestinal crypts are also known to be a site of cellular secretion of secretory IgA, lysozyme, and fluid (Wilson et al., 1982; Olson and Erlandsen, 1981; Welsh et al., 1982). It is plausible that intercellular communication through gap junctions has a role in regulating or possibly synchronizing crypt secretion such as may occur in pancreas (Petersen and Ueda, 19761, salivary gland (Roberts et al., 1978; Kater and Galvin, 1978), and lacrimal gland (Iwatsuki and Petersen, 1978b). In conclusion, the ability of crypt epithelial cells to transfer Lucifer yellow indicates that direct intercellular exchange of ions and small molecules may be a normal occurrence in the crypt epithelium. Our observation of gap junctions in crypt cells is consistent with the conclusion that intercellular transfer may be mediated through gap junctions. Proliferation and differentiation in the intestinal epithelium is apparently under regulation. The most common view is that the regulation results from a chalonelike feedback from the villus population (Cairnie, 1967; Rijke et al., 1976; Wright and Al-Nafussi, 1982). We would like to propose that the present findings suggest a n additional mechanism that may be important in regulation of epithelial cellular behavior (a detailed mathematical model will be presented elsewhere). The intercellular continuum formed via the gap junctions interconnecting most crypt cells could form the physical substrate within which fields (chemical, electrical, etc.) are established that in turn influence cellular behavior. Experimental perturbations that altered the size of the crypt population (e.g., irradiation with subsequent loss FUNCTIONAL GAP JUNCTIONS IN CRYPTS of many crypt cells) would alter the boundary conditions for the field, resulting in a new field and hence altered cell behavior (e.g., the pattern of cell proliferation). ACKNOWLEDGMENTS This study was supported by grants from the Medical Research Council of Canada and the Canadian Foundation for Ileitis and Colitis. LITERATURE CITED Bjerknes, M., and H. Cheng (1981) Methods for the isolation of intact epithelium from the mouse intestine. Anat. Rec., 199:565-574. Branton, D., S. Bullivant, N.B. Gilula, M.J. Karnovsky, H. Moor, K. Muhlethaler, D.H. Northcote, L. Packer, B. Satir, P. Satir, L.A. Staehlin, R.L. Steere, and R.S. Weinstein (1975) Freeze-etching nomenclature. Science, 190:54-56. Cairnie, A.B. (1967) Cell proliferation studies in the intestinal epithelium of the rat: Response to continuous irradiation. Radiat. Res., 32240-264. Crick, F. (1970) Diffusion in embryogenesis. Nature (Lond.), 225:420422. Curtis, R.L., J.S. Trier, R.A. Frizzell, N.M. Lindem, and J.L. Madara (1984) Flounder intestinal absorptive cells have abundant gap junctions and may be coupled. Am. J. Physiol., 246:C77-C83. Fig. 1. a-1. Photomicrographs showing intercellular transfer of Lucifer yellow between crypt epithelial cells. All photomicrographs are at the same magnification. a x . Dye transfer between crypt-base columnar cells: a) bright-field photomicrograph of a crypt base (intestinal crypts are roughly test-tube-shaped structures); notice the Paneth cells with secretory granules below the cell to be injected (which is indicated by the micropipette in the upper left-hand quadrant); b) fluorescent photomicrograph of the initial injection into a crypt-base columnar cell; c) fluorescent photomicrograph of the same cell at a later time showing transfer of dye to neighboring cells; notice that dye transfer to cells situated below stopped at the level of the Paneth cells. d-f. Dye transfer in Paneth cells: d) bright-field photomicrograph of crypt base; el fluorescent photomicrograph of the Paneth cells shortly after injection; fl fluorescent photomicrograph of the same cell at a later time showing transfer of dye to a n adjacent Paneth cell; a second injection into a crypt-base columnar cell can also be seen. g-i. Dye transfer in midcrypt cells: g) bright-field photomicrograph of midcrypt region; h) fluorescent photomicrograph of the injected midcrypt cell; i) fluorescent photomicrograph of the same cell at a later time showing transfer of dye to neighboring cells; two other injections in the same crypt can also be seen. j-1. Dye transfer in crypt-top cells: j) bright-field photomicrograph of crypt-top region; k) fluorescent photomicrograph of the injected crypt-top cell; 1)fluorescent photomicrograph of the same cell at a later time showing transfer of dye to neighboring cells: note that the orientation of the crypt has been shifted. m-o. Electron micrographs of freeze-fracturereplica of the P-face of mid-crypt cells (bars = 0.1 pm): m) three gap junctions on lateral surface of mid-crypt cell; n) a small eau iunction consisting of about 25 particles; 0) a large gap ~. junction consisting of several h<ndred particles. ~ 367 Guthrie, S.C. (1984) Patterns of junctional communication in the early amphibian embryo. Nature (Lond.), 311:149-151. Iwatsuki, N., and O.H. Petersen (1978a) Electrical coupling and uncoupling of exocrine acinar cells. J. Cell Biol., 79:533-545. Iwatsuki, N., and O.H. Petersen (197813) Intracellular Ca'+ injection causes membrane hyperpolarization and conductance increase in lacrimal acinar cells. Pflugers Arch., 3773185-187. Kater, S.B., and N.J. Galvin (1978) Physiological and morphological evidence for coupling in mouse salivary gland acinar cells. J. Cell Biol., 79:20-26. Leblond, C.P., and H. Cheng (1976) Identification of stem cells in the small intestine of the mouse. In: Stem Cells of Renewing Populations. A.B. Cairnie, P.K. Lala, D.G. Osmond, Eds. Academic, New York, pp. 7-31. Lo, C.W. (1980) Gap junctions and development. In: Development in Mammals, Vol. 4. M.H. Johnson, Ed. Elsevier/North-Holland,New York, pp. 39-80. Loewenstein, W.R. (1981)Junctional intercellular communication: The cell-tocell membrane channel. Physiol. Rev, 61:829-913. Madara, T.L., M.R. Neutra, and J.S. Trier (1981) Junctional complexes in fetal rat small intestine during morphogenesis. Dev. Biol., 86:170-178. Meinhardt, H. (1982) Models of Biological Pattern Formation. Academic, New York. Olson, R.E., and S.L. Erlandsen (1981) Paneth cell function: The effects of cholinergic and adrenergic drugs on lysozyme secretion. Anat. Rec., 1993186A. Othmer, H.G. (1983) A continuum model for coupled cells. J. Math. Biol.. 17:351-369. Petersen, O.H., andN. Ueda (1976) Pancreatic acinar cells: The role of calcium in stimulus-secretion coupling. J. Physiol., 254:583-606. Rijke, R.P.C., W.R. Hanson, H.M. Plaisier, and J.W. Osborne (1976) The effect of ischemic villus cell damage on crypt cell proliferation in the small intestine: Evidence for a feedback control mechanism. Gastroenterology, 71:786-792. Roberts, M.L., N. Iwatsuki, and O.H. Petersen (1978) Parotid acinar cells: Ionic dependence of acetylcholine-evoked membrane potential changes. Pflugers Arch., 376:159-167. Staehelin, L.A. (1972) Three types of gap junctions interconnecting intestinal epithelial cells visualized by freeze-etching. Roc.Natl. Acad. Sci. USA, 69:1318-1321. Stewart, W.W. (1978) Functional connections between cells as revealed by dye-coupling with a highly fluorescent naphthalimide tracer. Cell, 14:741-759. Turing, A.M. (1952)The chemical basis of morphogenesis. Phil. Trans. R. Soc. B, 237337-72. Warner, A.E., S.C. Guthrie, N.B. Gilula (1984) Antibodies to gapjunctional protein selectively disrupt junctional communication in early amphibian embryo. Nature (Lond.), 311:127-131. Welsh, M.J., P.L. Smith, M. Fromm, and R.A. Frizzell (1982) Crypts are the site of intestinal fluid and electrolyte secretion. Science, 218:1219-1221. Wolpert, L. (1978) Gap junctions: Channels for communication in development. In: Intercellular Junctions and Synapses (Receptors and Recognition, Series B, Volume 2). J. Feldman, N.B. Gilula, and J.D. Pitts, Eds. Chapman and Hall, London, pp. 83-96. Wright, N.A., and A. Al-Nafussi (1982) The kinetics of villus cell populations in the mouse small intestine. 11. Studies on erowth control after death of proliferative cells induced by cytosincarabinoside, with special reference to negative feedback mechanisms. Cell Tissue Kinet., 15:611-621.