Pfliigers Arch (1990) 417:120-122 Journal of Physiology 9 Springer-Verlag 1990 Short communication Cellular and segmental distribution of Ca 2 + in rat intestine -pump epitopes J. L. Borke, A. Caride, A. K. Verma, J. T. Penniston, and R. Kumar Nephrology Research, Departments of Medicine, Biochemistry and Molecular Biology, Mayo Clinic and Foundation, Rochester, MN 55905, USA Received March 14/Accepted June 27, 1990 Abstract. We used a monoclonal antibody (5F10) specific for the human erythrocyte plasma membrane Ca + +-pump to demonstrate the presence and distribution of Ca++-pump epitopes in rat intestine. In paraffin embedded tissue sections, antibody 5F10 binds to epitopes in the basolateral membranes of absorptive cells in rat duodenum and portions of jejunum but not ileum. Western blot analysis of intestinal mucosal proteins with antibody 5F10 shows binding of antibody to major bands of Mr -- 135,000 and Mr 72,000, and to lesser bands of Mr = 125,000 and Mr = 27,000. This pattern was seen in mucosal homogenates of rat duodenal and jejunal cells and to a lesser extent in ileal cells. The Mr 135,000 band corresponds to the molecular weight of Ca + +-pumps in other tissues. The other bands correspond in size to known proteolytic fragments of the Ca++-pump. Slot-blot analysis of nitrocellulose immobilized mucosal homogenates shows binding of 5F10 to be greatest in duodenum and least in ileum. Ca ++- transport studies by the everted gut sac technique show a correlation between vitamin D induction of active Ca + +-transport and the segmental distribution of Ca++-pump epitopes. Key words: Calcium pump, ATPase, calcium transport, intestine, vitamin D. INTRODUCTION Active transport of calcium in the rat intestine is a vitamin D-dependent process. While both Na +/Ca + + exchange and ATP- dependent transport processes are present in basolateral membrane preparations, only ATP-dependent Ca ++ transport is vitamin D-dependent and has the same segmental distribution as calcium transport measured in whole tissue [5]. Studies in erythrocyte membranes and sareolemma have suggested that ++ ++ Ca -Mg ATPase activity and ATP dependent plasma membrane Ca++-transport are the same [9]. Work by Moy, et al., however, has shown this not to be the ease in rat intestine [7]. In their study, the biochemical and biophysical properties of Ca ++- Mg ++ ATPase were markedly different from vesicular Ca++- uptake in intestinal basolaterai membranes and did not correlate with vitamin D-dependent active Ca++-transport [7]. Ca + +-pumps from several tissues and species have been purified (for review, see 8). Ca++-pumps in human erythrocytes, rat brain, pig smooth muscle and dog heart, for example, all have apparent molecular weights of between 130 and 150 kDa. In addition these also form Ca++-dependent, La +++ stabilized, Offprint requests to: R. Kumar, Mayo Clinic, 911 Gugenheim Build- ing, Rochester, MN 55905, USA hydroxylamine-sensitive acylptiosphate intermediates [4,9]. These enzymes also transport Ca ++ at a higher rate when bound to calmodulin [4,9]. In our study, we used monoclonal antibody 5F10 which was raised against the human erythrocyte plasma membrane Ca++-pump. This antibody cross-reacts with Ca++-pumps in a wide range of tissues and species so we were able to use it to examine the distribution of Ca++-pump epitopes in rat intestine [2]. We have found a correlation between the cellular and segmental distribution of Ca++-pump epitopes and vitamin D- dependent active Ca++-transport in rat intestine. METHODS Preparation of Western Blots Intestinal segments were obtained from anesthetized (sodium pentobarbital, 50 mg/kg i.p.) 200 g male adult Sprague- Dawley rats maintained on ad libitum water and Rodent Laboratory Chow #5001 (Purina Mills, St. Louis, MO). The intestines were perfused with saline, the mucosa was scraped, frozen in liquid nitrogen, and kept frozen at -70~ C until used. Frozen tissue fragments were placed in a homogenization solution containing 50 mmol/l Tris base pH 7.4, 5 mmol/l benzamidine, 10 mmol/1 phenylmethyl sulfonyl fluoride (PMSF), 1.5 pmol/1 pepstatin A, 2.5 pmol/1 leupeptin, and 10 pg/ml Trypsin inhibitor. Samples were homogenized on ice for 15 seconds with a Brinkman Polytron (Westbury, NY), and protein was assayed by the BCA procedure (Pierce, Rockford, IL). Protein (30 pg) from homogenates was denatured by heating to 100~ C for 5 minutes in 50 mmol/1 Tris base (pH 6.75) containing 5% B-mercaptoethanol, 2% sodium dodecylsulfate, 10% glycerol and 0.1% bromophenyl blue. Twenty pg samples of marker proteins were also treated in a similar manner. Protein samples were analyzed on 7% sodium dodecylsulfate polyacrylamide gels (SDS-PAGE). Lanes containing marker proteins were stained with 0.5% Coomassie Blue in 25% ethanol and 8% acetic acid. Other lanes were transferred to nitrocellulose using a Bio-Rad transblot apparatus (Richmond, CA) as previously described [2]. Transferred proteins on nitrocellulose were stained with mouse monoelonal antibody 5F10 by avidin-biotinmethods as previously described [3]. Briefly, endogenous peroxidase activity was inhibited by a 10-minute incubation in 3% HzO 2 plus 0.1% sodium azide. Non-specific protein binding was blocked with a 1-hour incubation in 10 mg/ml bovine serum albumin in PBS. A 1:1000 dilution of 5F10 antibody was allowed to react with the blots at room temperature for one hour. After washing in 0.5% Tween 20 (Sigma) in PBS (PBS- Tween) a second biotin conjugated antibody made in horse and directed against mouse immunoglobulins (1:500) was allowed to react for 30 minutes at room temperature with the 5F10 on the