Renal tubular transport and catabolism of proteins and peptides

Transcription

1 Kidney International, Vol. 16 (1979), pp Renal tubular transport and catabolism of proteins and peptides FRANK A. CARONE, DARRYL R. PETERSON, SUZANNE OPARIL, and THEODORE N. PULLMAN Departments of Pathology, Physiology and Medicine, Northwestern University Medical School and Veterans Administration Lakeside Hospital, Chicago, Illinois, and Department of Medicine, University of Alabama, Birmingham, Alabama It is established that the kidney plays an important role in the metabolism of a number of protein, polypeptide, and small peptide molecules, including plasma proteins, growth hormone, L-chains of immunoglobulins, /32-microglobulin, lysozyme, insulin, proinsulin, parathyroid hormone, glucagon, and small vasoactive peptides. Absorption, transport, and/or degradation of proteins or peptides are functions of the proximal tubule; there is little evidence that other segments of the nephron have the mechanisms for uptake or transport of these substances. Indirect and direct studies indicate that a variety of proteins and polypeptides filtered at the glomerulus are absorbed by the proximal tubule by luminal endocytosis and hydrolyzed by lysosomal enzymes. Our recent studies suggest that small linear peptides, consisting of eight to ten amino acids, are handled by the proximal tubule by a different mechanism. We have demonstrated that small linear peptides microinfused into proximal tubules are hydrolyzed at the luminal surface of the brush border, which is rich in a variety of enzymes, by the process of membrane or contact digestion with reabsorption of most of the breakdown products. Proximal tubular handling of proteins and large peptides Qualitative morphologic and indirect functional studies have demonstrated that proximal tubular cells absorb protein from the luminal fluid and have suggested a pathway by which protein undergoes intracellular digestion. These processes have been quantified directly. Although some findings suggest that certain proteins are transported intact across proximal epithelial cells, other investigations do not support this conclusion. Direct studies on isolated tubular segments indicate that albumin and insulin are not transported into tubular cells across the peritubular cell membrane, presumably due to the absence of an endocytic mechanism on this side of the tubular epithelium. Luminal uptake. Many morphologic, micropuncture, microinfusion, and microperfusion studies have demonstrated that a wide variety of proteins are absorbed across the luminal aspect of proximal tubular cells [1, 2]. We quantified the uptake of '251-labeled rabbit serum albumin microperfused into isolated segments of the rabbit nephron [3]. Proximal convoluted and proximal straight segments accumulated 1251-albumin nearly linearly as a function of time (Fig. 1), whereas cortical collecting segments did not accumulate measurable amounts of protein. The rate of accumulation of albumin in the proximal convoluted tubule was 3.2 x 1-2 ng/mm/min, which was 2.6 times as great as that in the proximal straight tubule. Assuming an albumin concentration of.3 mg/dl in normal gbmerular filtrate [4], we can calculate that the reabsorptive capacity of the entire proximal tubule for albumin exceeds the amount of albumin filtered by the normal glomerulus. The ultrastructural basis for reabsorption of labeled albumin by proximal tubules was investigated autoradiographically by Maunsbach with in vivo microinfusion methods in the rat kidney [5] and in our laboratory in isolated rabbit tubules microperfused with '251-albumin [3]. Sequential studies revealed that silver grains were initially located Received for publication March 13, / $ by the International Society of Nephrology 271

2 272 Carone et al PCT I E C, ( C E -C 1. PST Duration of perfusion, mlvi CCL 2 Fig. 1. Accumulation of iodinated albumin as a function of time in proximal convoluted (PCT) proximal straight (PST) and cortical collecting (CCT) segments of the rabbit nephron. All tubules were perfused with a 21-mg/dl solution of iodinated rabbit albumin at 18 to 2 ni/mm at 37 C. (Reprinted with permission of J Cell Biol [3]) S - - 'ç A Fig. 3. Electron microscope radioautograph with radioactive label in dense bodies (arrows) in cells of the proximal tubule after 85 mm's perfusion with "51-albumin. (X 15,2) (Reprinted with permission off Cell Biol [3]) '4- Fig. 2. Electron microscopic radioautograph of a proximal tubule perfused 1 mm with "51-albumin. Grains are located at the base of the brush border (small arrow), in small apical vesicles (medium arrow), and in apical vacuoles (large arrow). (X26,51) (Reprinted with permission of f Cell Biol [3]) over tubular invaginations at the base of the brush border and over small apical vesicles and later in larger membrane-bound apical vacuoles (Fig. 2). Finally, grains becoming concentrated in cytoplasmic dense bodies (Fig. 3) were associated with acid phosphatase positive bodies, indicative of lysosomes. Similar ultrastructural findings were observed with labeled insulin in our laboratory [6] and with a number of other proteins in several experimental animals [1, 2]. There is evidence that the first step in endocytosis involves binding of protein to the luminal plasma membrane [2]. There is wide variation in the affinity of different proteins for the plasma membrane, which may be related to the number and chemical structure of membrane binding sites and to the net charge on the protein molecules. Small amounts of certain proteins bind largely to the plasma membrane, whereas large amounts in tubular fluid appear mainly in endocytic vesicles unbound to the plasma membrane [2]. Thus, endocytosis may be largely specific when small quantities of protein are reabsorbed due to membrane

3 binding and less specific when large amounts of protein are reabsorbed. Several studies suggest that the membrane of apical invaginations and vesicles in the endocytic process is replaced by a de novo synthesis of plasma membrane [7, 8] and not by downward flow of brush border membrane [9]. Most evidence favors the conclusion that protein is transferred to lysosomes by fusion of endocytic vacuoles and preexisting lysosomes. Lysosomes contain many hydrolytic enzymes which have been shown to digest a wide variety of proteins. Lysosomal extracts isolated from renal cortical homogenates have been used to quantify hydrolysis of albumin and other proteins. Hydrolysis of '251-albumin is maximal at a low ph, and the major labeled product of digestion is monoiodotryosine [1]. Other in vitro studies have demonstrated hydrolysis of absorbed protein within isolated intact lysosomes [11] or within lysosomes of intact cells in kidney slice preparations [12]. Small-molecularweight metabolites of proteins diffuse out of lysosomes into the cell cytoplasm and interstitial fluid. There is no evidence for nonendocytic reabsorption of protein on the luminal side of proximal tubular cells or for release of intact reabsorbed protein from lysosomes on the contraluminal side of the cells. Contraluminal uptake of protein. Although the uptake of proteins from the luminal aspect of the proximal tubules is well established, uptake from the contraluminal aspect is uncertain. Studies in intact kidneys, however, provide evidence for contraluminal uptake of certain proteins such as f32-microglobulin [13] and insulin [14]. Because the basement membranes of isolated perfused tubules are moderately permeable to albumin, as demonstrated by Welling and Grantham [15], it is possible that interstitial protein is in contact with the basilar membranes of tubular cells. We studied the contraluminal uptake of albumin directly in isolated perfused segments of proximal thick ascending limb and cortical collecting tubules of the rabbit incubated in '251-albumin [16]. After 2 to 9 mi contraluminal uptake of albumin was negligible in all segments compared to luminal uptake in the proximal tubules (Fig. 4). In a companion study [6], we also found that contraluminal uptake of '251-insulin was negligible compared to luminal uptake (Fig. 5). It is known that albumin and insulin accumulate within proximal tubular cells from the luminal side where endocytosis is prominent, but not in cells of collecting tubules, where endocytosis is not pronounced. Absence of significant contraluminal uptake of albumin, or insulin, or both is presumably Tubular handling of proteins and pept ides Cox c::i contraiumnai PCT PST TAL CCT Fig. 4. Comparison of luminal and contraluminal accumulation rates of albumin in isolated renal tubules. Luminal uptake data is taken from Fig. I above. (Reprinted with permission of Am J Physiol [16]) Ca a C 2 1 Luminal Contraluminal Fig. 5. Luminal and contraluminal 1251-insulin accumulation by isolated perfused proximal convoluted tubules. (Reprinted with permission of Am J Physiol [6]) due to the absence of a prominant endocytic mechanism or the lack of specific receptor sites on the basilar side of tubular cells. Intercellular and transcellular transport of intact proteins. In normal mature kidneys, intercellular transport of proteins has not been demonstrated. In the immature kidney [17] and in certain abnormal states, such as elevated tubular hydrostatic pressure [18], proteins may pass between cells through cellular junctional complexes. On the other hand, transcellular transport of intact protein by renal tubules remains a controversial topic. A number of studies provide evidence both for and against a mechanism for the transcellular transport of intact proteins [2]. This problem may be resolved by direct studies that quantify and characterize protein

4 274 Carone et a! and protein catabolites absorbed and released by tubular cells. Renal handling of small peptides The kidney has been shown to degrade circulating angiotensin II (All) rapidly. Although 4 to 7% of 14C-AII infused into the renal artery is extracted by the dog kidney, little labeled material appears in the urine. The high extraction ratio suggests that renal handling involves more than glomerular filtration alone [19]. Similar studies have shown that 75% of infused 14C-AII is metabolized in a single passage through the kidney, and 98.7% of injected radiolabeled material is recovered in renal venous blood [2]. These findings indicate that extensive renal hydrolysis occurs, and that tissue sequestration of '4C- All or its metabolites is not prolonged. Thus, renal hydrolysis of All allows for the rapid return of cleavage-products to the general circulation. Recently, we have assessed the role of individual nephrons or isolated nephron segments in the transport and hydrolysis of radiolabeled angiotensin I (Al), All, bradykinin (BKN), and oxytocin (OT). The techniques for in vivo microinfusion of surface tubules in rats [21], and in vitro microperfusion of isolated rabbit nephron segments [22] were used. Reabsorption of radiolabeled material was measured, and the intact peptide or its metabolites were identified and quantified in urine, or bathing medium and collection fluid. In addition, peptides were incubated in the presence of isolated membrane preparations to localize a probable cellular site of hydrolysis. Characterization of labeled material was accomplished by two-dimensional peptide mapping involving high voltage paper electrophoresis in combination with decending paper chromatography [23] or high-voltage paper electrophoresis alone. Standard peptide fragments were generated by digesting labeled peptide in the presence of several purified enzymes. '4C-AII labeled in the fifth position on isoleucine (Fig. 6) was microinfused in vivo into individual surface nephrons of the rat kidney [23]. Following proximal infusion, 11% of '4C was recovered in the urine, and most (95%) of this was present as metabolites, including the carboxyl terminal tetrapeptide as the principal hydrolytic end-product. In contrast, recovery of 14C was 95% when distal tubules were infused, and virtually all was intact octapeptide. The data suggest that the peptide is rapidly hydrolyzed and reabsorbed in the proximal tubule but not in the distal tubule. Further in vivo studies in the rat suggested that constituent free amino acids influence proximal tubular handling of 14C-AIl [24]. When the labeled peptide was microinfused into proximal nephrons with excess unlabeled L-lle, urinary recovery of 14C greatly exceeded that seen with 14C-AII alone and increased directly with distance of the infusion site from the glomerulus (Fig. 7). Because '4C-Ile appeared as the predominant labeled material in urine and only 5% of the labeled material excreted was in the form of intact All, these results suggest that excess unlabeled Ile interfered with the reabsorption of labeled Ile derived from All. When it is considered, however, that the total recovery of radiolabeled material was much greater when isoleucine was infused with 14C-AII than when 14C-AII was infused alone, the 5% of radioactivity as unaltered All assumes greater prominence. Thus, suppression of the reabsorption of labeled Ile by excess unlabeled Ile was the predominant effect and overshadowed the inhibition of hydrolysis of All by excess unlabeled Ile. From similar experiments and analogous reasoning, we concluded that excess aspartic acid also affected both hydrolysis and reabsorption of All but ANGIOTENSIN H Asp - Arg Vat - Tyr lieu His - Pro - Phe - OH "14C I Tyrpsin Chymotrypsir, Carboxypepticlase Fig. 6. Angiotensin II labeled at fifth position with '4C. Digestion with trypsin, chyrnotrypsin, and carboxypeptidase A (arrows) yielded standard breakdown products C Afl-isoleucine 1 8 -o > C C) S....., c%ocp6o Proximal length, % Fig. 7. Percent recovery of 14C-AII as a function of tubular length. Closed circles refer to infusion of '4C-AII plus unlabeled isoleucine in molar ratio of 1:2. Open circles refer to infusion of 14C-AII alone. (Reprinted with permission of Am J Physiol [24])

5 Tubular handling of proteins and peptides 275 I Time, rn/n Fig. 8. Reabsorption of tritium from proximal straight tubular segments microperfused with tritia ted Al as a function of time. Reabsorption is expressed as picograms tritiated Al per millimeter of tubular length. (Reprinted with permission of Am J Physiol [27]) to a different degree than isoleucine. These data are consistent with the observation that aspartic acid is a much stronger aminopeptidase inhibitor than is isoleucine [25] and suggest that the effect of these constituent amino acids on hydrolysis of All may be accomplished by amino-peptidase suppression. In vitro microperfusion of rabbit proximal straight nephron segments with 14C-AII provided direct evidence for proximal hydrolysis of the peptide, accompanied by rapid and extensive reabsorption of 14C-labeled material across the tubular epithelium [26]. Approximately 3% of perfused '4C was reabsorbed per millimeter of tubular length over a broad range of delivery rates. Most of the labeled material was rapidly transported across the tubular epithelium into the bathing medium, and less than 1% of perfused label remained sequestered by the tubular cells following the 35-mm perfusion period. Electrophoresis of collected perfusate demonstrated that '4C-AII was hydrolyzed to 14C-Ile. The foregoing data indicate that All is quickly degraded upon passage through the renal proximal tubule and suggest that cleavage occurs at the level of the tubular luminal membrane. This hypothesis was confirmed by incubations of 14C-AII directly in the presence of isolated membranes from rabbit renal brush border [27]. Hydrolysis of the peptide yielded '4C-Ile as the only labeled end-product appearing in the incubation medium. In vitro studies with tritiated Al labeled in the 1th position on leucine have provided further evidence that angiotensin is degraded at the luminal membrane of the proximal tubule, followed by reabsorption of hydrolytic products [27]. Upon microperfusion of isolated rabbit nephron segments for 3 mi tritium was rapidly reabsorbed (Fig. 8). Electrophoretic analysis of the collection fluid and bathing medium revealed that tritiated Leu appeared as the predominant labeled material. Incubation of tntiated AT directly in the presence of isolated membranes from renal brush border yielded tritiated Leu as the major labeled hydrolytic end-product. Thus, hydrolysis at the brush border was directly related to reabsorption of a hydrolytic end-product. There is considerable evidence that renal handling of bradykinin (BKN) is similar to that of Al and All. Surface nephrons in the rat kidney were microinfused in vivo with tritiated BKN labeled in the second position on proline [28]. After proximal infusion, urinary recovery of tritium was 24%, 85% of which was in the form of metabolites (81% Pro2 and 4% Arg1-Phe5). Distal infusion resulted in recovery of 98% of tritium, all of which appeared in the urine as intact tritiated BKN. In addition, tntiated BKN and 14C-inulin were simultaneously infused into proximal and distal surface nephrons, and recovery of the respective labels was plotted as a function of time at 3-second intervals following microinfusion (Fig. 9). For both proximal and distal tubules, the urinary concentration-time curves for tritium derived from tritiated BKN did not differ appreciably from those for '4C-inulin (Fig. 9). Because tritium represents both intact tritiated BKN and hydrolytic products in varying proportions depending upon the site of microinfusion, tubular transit time Proximal Distal 3...c3 H i 234 2:34 Time, rn/n Time, rn/n Fig. 9. Appearance of tritium and 14C in urine plotted against time. Tritium and '4C are quantified as percents of total excretions of each isotope, respectively. Isotopes were simultaneously microinfused into rat tubules as tritiated bradykinin and 14Cinulin. (Reprinted with permission of Am J Physiol [28]) 2 1

6 276 Carone et al for the metabolites of tritiated BKN is essentially as rapid as that of 14Cinulin or intact tritiated BKN. These data suggest that cleavage of tritiated BKN in the proximal tubule is a rapid process and favor the interpretation that tritiated BKN, like tritiated AT or 14C-AII, is hydrolyzed at the luminal brush border membrane. Indeed, Ward et al [29] have demonstrated kininase activity in rat brush border membranes. The concept of membrane or contact digestion has been developed to explain the hydrolysis and absorption of proteins, peptides, and other substrates by cell membranes, particularly the mucosa of the small intestine [3]. The mechanism for proximal reabsorption of BKN and metabolites appears to be of high capacity but not high specificity, and the processes of hydrolysis and reabsorption may be characterized by different capacities and specificities [31]. Upon simultaneous microinfus ion of tritiated BKN with excess unlabeled BKN or AT, urinary recovery of tritium was increased to the same extent. Only unlabeled BKN, however, and not unlabeled Al, effectively inhibited the hydrolysis of tritiated BKN, as determined by identification and quantification of products appearing in the urine. A nonlinear molecular configuration may restrict the hydrolysis and uptake of small peptide hormones at the luminal membrane of the proximal tubule. Oxytocin (OT) (Fig. 1) and vasopressin both contain a disulfide bridge. Although the kidney has been implicated in the hydrolysis of both peptides [32, 33], in vitro microperfusion of tritiated OT (labeled in second position on tryosine) (Fig. 1) through rabbit proximal straight nephron segments resulted in no detectable hydrolysis, and the reabsorption rate of labeled material was low compared to corresponding measurements for 54C-AII under similar conditions [26]. In vivo microinfusion of 525J arginine vasopressin in rat nephrons yielded similar results [34]. Collectively, these studies demonstrate that small linear peptides are cleaved in the proximal tubule of the rat and rabbit kidney when presented to the lu- OXYTOCI N Cys- Tyr- Tie -Gin- Asn - Cys Pro Leu Gly NH2 Fig. 1. Molecular structure qf oxytocin labeled with tritium in the second position (asterisk) demonstrating the disulfide bridge between cysteines in the one and six positions. minal aspect of the tubular cells. The mechanism for tubular handling appears to involve enzymatic hydrolysis at the brush border, followed by rapid reabsorption of metabolites (Fig. 11). It is reasonable to believe that amino acids released by hydrolysis at the luminal membrane are actively reabsorbed by amino acid pumps known to exist there. The distal tubule appears to lack this property. Furthermore, the mechanism of this process appears to differ from that for the proximal reabsorption of proteins and large peptides, which involves endocytosis and lysosomal digestion (Fig. 11). It remains to be determined whether these larger molecules are partially degraded at the luminal membrane of the proximal tubule prior to or during the endocytic process. Our studies indicate that the proximal tubules of the mammalian kidney possess a high capacity mechanism for the rapid hydrolysis of small linear peptides. This mechanism may be important biologically to (1) conserve amino acids, (2) inactivate toxic peptides, and (3) help regulate the circulating levels of small peptide hormones because there is evidence that serum levels of these hormones are determined more by rates of degradation than they are by synthesis. Lumen Residual digestive body Endocytic vesicle Phagosome Lysosome Phagolysosome Amino acids Amino acids FIg. 11. Schemata comparing the cellular mechanisms of the proximal tubule for reabsorption and catabolism of protein or large polypeptide molecules to that described in our laboratory for small, linear peptides. The left figure illustrates that protein is taken up by endocytic vesicles, which fuse to form phagosomes into which primary lysosomes empty their hydrolytic enzymes. Enzymatic cleavage of proteins occurs in the phagolysosomes; liberated amino acids diffuse into the interstitium and are returned to the renal circulation. The right figure depicts tubular handling of small, linear peptides. Hydrolysis occurs at the site of enzymes associated with the brushborder of the proximal tubule. Liberated amino acids are rapidly transported across the epithelium, probably involving active amino acid pumps located at the apical cell membrane. Partially hydrolyzed peptide fragments may be reabsorbed intact or undergo further intracellular cleavage prior to reabsorption.

7 Tubular handling of proteins and peptides 277 Summary The kidney plays an important role in the metabolism of proteins and peptides. Current evidence indicates that only the proximal tubule possesses the mechanism for degradation or transport of these substances and reabsorption of metabolic products. Proteins and large polypeptides filtered at the gbmerulus are absorbed from proximal tubular fluid by luminal endocytosis into apical vacuoles. These fuse with primary lysosomes, where hydrolysis occurs followed by diffusion of metabolites out of the cells and into the blood. Recent evidence indicates that small linear peptides are handled by a different mechanism. It is likely that small peptides are degraded at the luminal surface of the brush border of proximal tubules, which contains many hydrolytic enzymes, by the process of membrane or contact digestion with reabsorption of the breakdown products. The probable biological significance of proximal tubular mechanisms for handling of proteins and peptides are conservation of amino acids, inactivation of toxic substances, and participation in the regulation of the circulating level of protein and peptide hormones. References 1. BOURDEAU JE, CARONE FA: Protein handling by the renal tubule. Nephron 13:22 34, MAUNSBACH AB: Cellular mechanisms of tubular protein transport, in Kidney and Urinary Tract Physiology II, edited by THURAU K, Baltimore, University Park Press, 1976, Vol. 11, pp , 3. BOURDEAU JE, CARONE FA, GANOTE CE: Serum albumin up take in isolated perfused renal tubules. J Cell Biol 54: , OAKEN DE, COTES SC, MENDE CW: Micropuncture study of tubular transport of albumin in rats with aminonucleoside nephrosis. Kidney mt 1:3 11, MAUNSBACH AB: Absorption of I 125-labeled homologous albumin by rat kidney proximal tubule cells: A study of microperfused single proximal tubules by electron microscopic autoradiography and histochemistry. J Ultrastruct Res 15: , BOURDEAU JE, CHEN ERY, CARONE FA: Insulin uptake in the renal proximal tubule. Am J Physiol 225: , BODE F, POCKRANDT-HEMSTEDT H, BAUMANN K, KINNE R: Analysis of the pinocytic process in rat kidney: I. Isolation of pinocytic vesicles from rat kidney cortex. J Cell Biol 63:998 18, BODE F, BAUMANN K, KINNE R: Analysis of the pinocytic process in rat kidney: II. Biochemical composition of pinocytic vesicles compared to brush border microvilli, lysosomes and basolateral plasma membranes. Biochim Biophys Acta 433: , BENNETT HS: The concepts of membrane flow and membrane vesiculation as mechanism for active transport and ion pumping. J Biophys Biochem Cytol 2:99-13, MAUNSBACH AB: Ultrastructure and digestive activity of lysosomes from proximal tubular cells. Proc 4th Cong Nephrol, Stockholm, Basel, Karger, 197, Vol. 1, p DAVIDSON SJ: Protein absorption by renal cells: II. Very rapid lysosomal digestion of exogenous ribonuclease in vitro. J Cell Biol 59: , CHRISTENSEN El, MAUNSBACH AB: Intralysosomal digestion of lysozyme in renal proximal tubule cells. Kidney mt 6:396 47, RAVNSKOV U, JOHANSSON BG, GoTHLIN J: Renal extraction of $2-microglobulin. Scand J Clin Lab Invest 3:71-75, KATZ AT, RUBENSTEIN AH: Metabolism of proinsulin, insulin and C-peptide in the rat. J Clin Invest 52: , WELLING LW, GRANTHAM JJ: Physical properties of isolated perfused renal tubules and tubular basement membranes. J Clin Invest 51: , BOURDEAU JE, CARONE FA: Contraluminal serum albumin uptake in isolated perfused renal tubules. Am J Physiol 224:399-44, LAR55N L: Ultrasti-ucture and permeability of intercellular contacts of developing proximal tubule in the rat kidney. J Ultrastruct Res 52:1 113, OTTOSEN PD: Effect of intratubular pressure on the ultrastructure and protein transport in the proximal tubule. Kidney mt 9: , BAILIE MD, RECTOR FC JR, SELDIN DW: Angiotensin II in arterial and renal venous plasma and renal lymph in the dog. J Clin Invest 5: , OPARIL 5, BAILIE MD: Mechanism of renal handling of angiotensin II in the dog. Circ Res 33:5 57, GOTTSCHALK CW, MOREL F, MYLLE M: Tracer microinjection studies of renal tubular permeability. Am J Physiol 29: , BURG J, GRANTHAM J, ABRAMOW M, ORLOFF J: Preparation and study of fragments of single nephrons. Am J Physiol 21: , PULLMAN TN, OPARIL 5, CARONE FA: Fate of labeled angiotensin II microinfused into individual nephrons in the rat. Am J Physiol 229: , PULLMAN TN, CARONE FA, OPARIL S, NAKAMURA S: Effects of constituent amino acids on tubular handling of microinfused angiotensin II. Am J Physiol 234:F325 F33l, PFLEIDERER G: Particle-bound aminopeptidase from pig kidney, in Methods in Ezymology: Proteolytic Enzymes, edited by PERLMANN GF, LORAND L, New York, Academic Press, 197, Vol. 19, pp PETERSON DR, OPARIL 5, FLOURET G, CARONE FA: Handling of angiotensin II and oxytocin by renal tubular segments perfused in vitro. Am J Physiol 232:F319-F324, PETERSON DR, CHRABASZCZ G, PETERSON WR, OPARIL S: Mechanism of renal tubular handling of angiotensin. Am J Physiol 236:F365 F372, CARONE FA, PULLMAN TN, OPARIL 5, NAKAMURA 5: Micropuncture evidence of rapid hydrolysis of bradykinin by rat proximal tubule. Am J Physiol 23: , WARD PE, ERDOS EG, GEDNEY CD, DOWDEN RM, REY- NOLDS RC: Isolation of membrane-bound enzymes that metabolize kinins and angiotensins. Biochem J 157:643-65, UGOLEV AM, DELAEY P: Membrane digestion: A concept

8 278 Carone et a! of enzymic hydrolysis on cell membranes. Biochim Biophys Acta 3:15 128, OPARIL S, CARONE FA, PULLMAN TN, NAKAMURA S: Inhibition of proximal tubular hydrolysis and reabsorption of bradykinin by peptides. Am J Physiol 23 1: , WALTER R, BOWMAN RH: Mechanism of inactivation of vasopressin and oxytocin by the isolated perfused rat kidney. Endocrinology 92: , WALTER R, SHANK H: In vivo inactivation of oxytocin. Endocrinology 89:99-995, LINDHEIMER MD, RE!NHARZ A, GRANDCHAMP A, VALLOT- TON MB: Fate of arginine vasopressin (AVP) perfused into nephron of Wistar (W) and Brattleboro (DI) rats. Clin Res 25:595A, 1977