Introduction

Top Abstract Introduction Methods Results Discussion References

The potential of small peptides as therapeutic agents is limited by a number of factors. For example, even when the problem of proteolytic degradation by peptidases is overcome, difficulties remain in obtaining and maintaining an appropriate target organ level within the body, especially after oral administration. One reason for this is the rapid hepatic clearance of oligopeptides from the splanchnic circulation, which, in many instances, can be attributed to biliary excretion of the intact drug (Ruwart, 1995). Such elimination indicates vectorial translocation of the oligopeptide across the sinusoidal and canalicular membrane surfaces of the hepatocyte. Previous rat liver perfusion studies using ditekiren [U-71038; Boc-Pro-Phe-N-MeHis-Leu[CHOHCH₂]Val-Ile-(aminomethylpyridine)] as a model compound indicated a potential role for carrier-mediated transport at both of these membranes (Adedoyin et al., 1993). In particular, extensive biliary excretion occurred against a concentration gradient and was markedly dose-dependent.

A number of export transporters located at the canalicular membrane have been shown to be involved in the biliary secretion of both endogenous and exogenous compounds and have been primarily classified according to their biochemical characteristics, such as substrate specificity and the involved driving force(s) (Meier, 1995). In many cases, the involved proteins appear to be members of the ABC superfamily of transporters because several of these processes exhibit ATP dependency (Greenberger and Ishikawa, 1994). For example, bile acids such as taurocholate are translocated across the canalicular membrane by a specific cBST, which is a saturable, vanadate-sensitive and unidirectional process (Arias et al., 1993; Gatmaitan and Arias, 1995; Vore, 1993) that is distinct from a similar but electrogenically driven system (Kast et al., 1994). An ATP-dependent nonbile acid organic transporter, or cMOAT, is also present at the canaliculus and appears to be responsible for the transport of a large number of compounds, including cysteinyl leukotrienes (leukotriene C₄), glutathione disulfide and glutathione S-conjugates and glucuronides of xenobiotics (Arias et al., 1993; Gatmaitan and Arias, 1995; Vore, 1993). A saturable but electropotential-dependent process has also been demonstrated for the transport of such compounds (Adachi et al., 1991; Fernández-Checa et al., 1992). Finally, organic cations are considered to be secreted into bile via an ATP-dependent P-glycoprotein-type transporter present at the canalicular membrane (Arias et al., 1993; Gatmaitan and Arias, 1995; Vore, 1993).

The mechanism or mechanisms by which oligopeptides are translocated across the canalicular membrane and secreted into bile against a concentration gradient are not known. It could involve one or more of the established transporters, or a distinctly different system or systems may be responsible. Regardless, further understanding of this potentially rate-limiting step in the rapid removal process would be useful in the design of therapeutic peptides. Accordingly, we undertook characterization of the transport of a model pseudohexapeptide, ditekiren, by rat hepatic canalicular membrane vesicles.

	Methods

Top Abstract Introduction Methods Results Discussion References

Chemicals. Unlabeled ditekiren, U-71013 [Boc-Pro-Phe-N-MeHis-Leu[CHOHCH₂]Ile-(aminomethylpyridine)], U-77436 [Tham-Pro-Phe-N-MeHis-Leu[CHOHCH₂]Val-Ile-(aminomethylpyridine-N-oxide)] and radiolabeled ([³H]prolyl) ditekiren were obtained from The Upjohn Co. (Kalamazoo, MI). Radiopurity of the labeled peptide was >98% by high performance liquid chromatography (specific activity, 32.7 Ci/mmol). Angiopeptin

Preparation and characterization of canalicular membrane vesicles. Livers were obtained from male Wistar rats (200-250 g; Harlan Sprague-Dawley, Indianapolis, IN), and canalicular plasma membrane vesicles were prepared according to a two-step method described by Kobayashi et al. (1990). This consisted of an initial Percoll gradient procedure to obtain a mixed hepatic plasma membrane fraction from which a canalicular-enriched preparation was obtained using sucrose-density gradient centrifugation. The vesicles were suspended in buffer A (0.25 M sucrose, 10 mM HEPES·Tris and 0.2 mM CaCl₂, pH 7.4) at a protein concentration of 2 to 3 mg/ml and stored at 70°C for 14 days before use in transport studies.

Transport studies. Membrane vesicle transport of [³H]ditekiren was determined by a rapid filtration technique. Frozen vesicles were quick-thawed by immersion in a 37°C water bath, revesiculated by being passed through a 25-gauge hypodermic needle and then kept on ice until use. A 10-µl aliquot of the suspended vesicles (20-30 µg of protein) that had been preincubated for 5 min at 37°C was added to 50 µl of standard medium containing [³H]ditekiren (0.22 µCi; 0.011 µM), 1 to 25 µM unlabeled ditekiren (depending on the particular study), 5 mM ATP and an ATP-regenerating system (3 mM creatine phosphate and 100 µg/ml creatine phosphokinase) in buffer B (0.25 M sucrose, 10 mM HEPES·Tris, 10 mM MgCl₂ and 0.2 mM CaCl₂, pH 7.4) that had been preincubated for 10 min at 37°C. Radiolabeled ditekiren uptake was terminated after the desired time period by the addition of 1 ml ice-cold buffer C (0.25 M sucrose, 10 mM HEPES·Tris, 10 mM MgCl₂, 0.2 mM CaCl₂, 100 mM NaCl, and 50 µM ditekiren, pH 7.4). The suspension was immediately filtered through a 1.2-µm GF/C glass fiber filter (Whatman International Ltd., Maidstone, UK) that had been previously presoaked with buffer C. The filter was then washed three times with 9 ml of ice-cold buffer C and subsequently dissolved in 5 ml of BCS scintillation fluid (Amersham Corp., Arlington Heights, IL). The vesicle-associated radioactivity was determined in a 1219-Rackbeta liquid scintillation counter (Pharmacia-LKB Nuclear, Gaithersburg, MD) using an automatic quench correction procedure. All incubations were performed in triplicate, and the reported results (mean ± S.E.M.) represent the mean observations from at least three different vesicle preparations.

(1)

(2)

(3)

(4)

	Results

Top Abstract Introduction Methods Results Discussion References

The vesicle preparation was markedly enriched (45-163-fold) relative to the four marker enzymes associated with the hepatic canalicular membrane. In contrast, Na⁺,K⁺-ATPase, which is exclusively localized in the basolateral membrane, was absent, and there was minimal contamination by enzymes indicative of various intracellular organelles. Approximately one-third (32.3 ± 3.9%, n = 3) of the canalicular vesicles had an inside-out orientation, thus allowing transporters that normally function in an export fashion to translocate substrate from the incubation medium into the vesicle. Such functional potential was confirmed by measuring the ATP-dependent transport of taurocholate, when a classic "overshoot" phenomenon was observed in the time course of uptake for this bile acid (data not shown).

Initial studies showed that the time profile of [³H]ditekiren uptake also indicated an ATP-dependent component that was initially rapid, reached a maximum at ~1 to 2 min, and then declined (fig. 1). This profile was similar to the loss of ATP in the incubation medium (data not shown). Furthermore, the uptake of the oligopeptide was found to be dependent on the ATP concentration in the incubation medium, and this could be described by a Michaelis-Menten relationship with a K_m value of 129 µM and with maximal stimulation being attained at an ATP concentration of ~1 mM. Additional confirmation for the critical role of ATP was also obtained by the addition (5 mM) of ATP, ADP, AMP, GTP and a nonhydrolyzable ATP analog, 5-adenylylimido-diphosphate, to the incubation medium in the absence of an ATP-regenerating system. Only ATP significantly enhanced, by ~2-fold, the initial uptake rate of [³H]ditekiren. In addition, vanadate (100 µM), which inhibits P-type ATPase activity, reduced the ATP-dependent uptake of [³H]ditekiren by more than one third (P < .05). Accordingly, all subsequent studies were undertaken at an initial medium ATP concentration of 5 mM along with an ATP-regenerating system. Under these conditions, [³H]ditekiren uptake was linear over 20 sec, and therefore subsequent initial rate of uptake studies routinely used a 15-sec incubation period.

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Time course of the initial [3H]ditekiren (9 µM) uptake rate by canalicular membrane vesicles at 37°C. ATP-dependent uptake () represents the different between transport in the presence () and absence () of ATP. Data represents the mean ± S.E.M. values for three different vesicle preparations.

When the osmolarity of the intravesicular volume was altered by preincubation with different concentrations of raffinose, [³H]ditekiren uptake was reduced with increasing osmolarity in the presence of ATP; in contrast, no such effect was observed when ATP was omitted from the incubation medium (fig. 2). Extrapolation of both sets of data to infinitely high osmolarity indicated that ~50% of the initial uptake of [³H]ditekiren reflected ATP-dependent transport into an osmotically sensitive intravesicular space and binding to membrane components contributed to a similar extent.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Effect of intravesicular osmolarity on the initial [3H]ditekiren (7.5 µM) uptake rate by canalicular membrane vesicles at 37°C in the presence () and absence () of ATP. Data represent the mean ± S.E.M. values for three different vesicle preparations.

The total uptake of [³H]ditekiren at 37°C in the presence of ATP and an ATP-regenerating system was nonlinear with respect to the ditekiren concentration in the incubation medium (fig. 3A). This was in contrast to the linear relationship observed when the incubation medium was maintained at 4°C. The latter uptake was considered to reflect passive diffusion of ditekiren into the vesicle and nonspecific binding. Accordingly, the difference between the initial uptake rates at the two temperatures was considered to indicate carrier-mediated uptake. The ATP-dependent component of such transport was estimated by also measuring uptake in the absence of ATP and its regenerating system and subtracting this from the carrier-mediated process (fig. 3B). Both the ATP-dependent (K_m = 19.1 ± 4.26 µM; mean ± S.E.M., V_max = 140 ± 29.4 pmol/mg of protein/15 sec) and ATP-independent (K_m = 17.2 ± 9.58 µM, V_max = 62.9 ± 24.5 pmol/mg of protein/15 sec) processes were concentration-dependent and exhibited saturable-type kinetics. Based on the V_max/K_m ratios at low oligopeptide concentration, the ATP-dependent process (7.33 ± 1.45 µl/mg of protein/15 sec) was about twice as effective at transporting ditekiren as that of the ATP-independent transporter (3.66 ± 5.06 µl/mg of protein/15 sec) and comparable in magnitude to the nonsaturable component (7.51 ± 2.05 µl/mg of protein/15 sec).

View larger version (10K): [in this window] [in a new window]   Fig. 3.   Effect of ditekiren concentration on the initial uptake rate of [3H]ditekiren by canalicular membrane vesicles. A, Total uptake of [3H]ditekiren in the presence of ATP at 37°C () and 4°C (). B, Carrier-mediated transport () obtained from the difference in uptake at 4°C and 37°C; ATP-dependent transport () obtained from the difference in uptake at 37°C in the presence and absence of ATP; ATP-independent transport () obtained from the difference between active and ATP-dependent transport at 37°C. Data represent mean ± S.E.M. values for five separate vesicle preparations.

Specificity of the ATP-dependent transport at a ditekiren concentration of 1 µM was investigated by determining the ability of other oligopeptides to inhibit the process when present in 100-fold excess. Related compounds as well as other renin inhibitor peptides and angiopeptin reduced transport to varying extents (table 1). EMD-55068 was the most effective inhibitor, inhibiting [³H]ditekiren transport by ~80%. A related compound, EMD-51921, and angiopeptin were also inhibitory but to a lesser extent. In contrast, other similar peptides that had been synthesized by The Upjohn Co., such as U-77436 and U-71013, did not statistically affect [³H]ditekiren transport, although the data suggested a trend with regard to the latter compound. No inhibition was observed with the dipeptide Ala-Asp (data not shown). Additional studies with EMD-51921 indicated that the inhibition of [³H]ditekiren was competitive with a K_i value of 46 µM (fig. 4A).

View larger version (12K): [in this window] [in a new window]   Fig. 4.   Dixon plots demonstrating the competitive inhibition of the initial, ATP-dependent [3H]ditekiren uptake rate by EMD-51921 (A) and daunomycin (B). Data represent mean ± S.E.M. values for three to six separate vesicle preparations.

To determine whether ditekiren transport involved one of the known canalicular membrane transport systems, the effect of prototypic substrates (100 µM) of these processes on ATP-dependent [³H]ditekiren (1 µM) transport was investigated. Glutathione disulfide, which is transported by cMOAT, had no effect on the initial uptake rate of ditekiren. Taurocholate, however, caused marked inhibition of transport (6.33 ± 0.98 vs. 2.29 ± 1.11 pmol/mg of protein/15 sec, P < .01), but this was found to be noncompetitive in nature. Similar kinetics were also obtained when the ability of ditekiren to reduce taurocholate transport was studied (data not shown). In contrast, daunomycin reduced the initial transport rate of ditekiren by approximately two thirds (6.96 ± 1.02 vs. 2.30 ± 1.23 pmol/mg of protein/15 sec); furthermore, such inhibition was competitive (fig. 4B), with a K_i value of ~16 µM. Because these findings suggested the involvement of a P-glycoprotein-type transporter, the effects of typical substrates of this system on [³H]ditekiren transport were subsequently studied. Both nicardipine (100 µM) and cyclosporin A (10 µM) markedly reduced the ATP-dependent transport of ditekiren (1 µM) (table 2). More modest (40-60%) inhibition was noted with vinblastine, verapamil, rhodamine-123 and reserpine (all at a concentration of 100 µM).

	Discussion

Top Abstract Introduction Methods Results Discussion References

Biliary excretion is a major pathway for the elimination of many oligopeptides, including those with renin-inhibiting activity such as ditekiren (Ruwart, 1995). In the case of this pseudohexapeptide, >90% of an administered dose in the rat is rapidly excreted intact into the bile in vivo (Greenfield et al., 1989); similar results were obtained in an isolated perfused liver preparation (Adedoyin et al., 1993). The transport of a number of oligopeptides from the blood into hepatocytes (i.e., across the sinusoidal membrane) has recently been investigated. In many instances, a carrier-mediated uptake process is involved, the nature of which appears to depend on the structure and lipophilicity of the compound; however, the roles of recently identified specific transporter proteins such as Ntcp and Oatp1 (Meier, 1995) have yet to be defined (Ziegler et al., 1996). Considerably less information exists concerning the translocation of oligopeptides across the canalicular membrane into bile, despite the likelihood of this being the rate-limiting step in the excretion process.

The experimental findings using ditekiren as a model oligopeptide indicate that several processes contribute to its initial uptake into inside-out vesicles prepared from rat hepatic canalicular membrane and, presumably, its biliary excretion. First, a concentration-independent process was present that was identified through the study of uptake at 4°C and is consistent with passive diffusion into the canalicular vesicles and/or nonspecific binding. This is not unexpected given the high lipophilicity of ditekiren (e.g., the logarithm of its partition coefficient between octanol and pH 7.4 buffer is >4) (Burton et al., 1991). In addition, temperature- and concentration-dependent transport into an osmotically sensitive, intravesicular space occurred. This appeared to involve two separate systems that could be differentiated according to their ATP-dependency using several different experimental approaches. On the basis of the kinetics of the initial uptake process (i.e., V_max/K_m), the transport efficiency of the ATP-independent carrier-mediated system was approximately half that of the ATP-dependent process. Studies to further characterize the non-ATP-dependent transport were not undertaken other than to note that it was not influenced by the external buffer Na⁺ concentration or pH over the range of 6.6 to 7.9. This probably precludes a potential-dependent mechanism for such transport because ditekiren is a bivalent cation with pK_a values of 3.95 and 6.3.²

Several ATP-dependent transporters are now known to be present in the canalicular membrane (Arias et al., 1993; Gatmaitan and Arias, 1995; Vore, 1993). One of these, cBST, is involved in the secretion of bile acids such as taurocholate (Meier, 1995). Although a high taurocholate concentration (100 µM) markedly inhibited ditekiren vesicular uptake and the reverse phenomenon also occurred (i.e., the oligopeptide reduced taurocholate transport), in neither case was the inhibition competitive in nature. Because the presence of bile acids is known to alter the function of other ATP-dependent transporters (e.g., P-glycoprotein) without themselves being substrates (Mazzanti et al., 1994), it was considered unlikely that cBST was involved in ATP-dependent transport of ditekiren. A similar conclusion was made with respect to a second major canalicular transporter that has relatively broad specificity for organic anions other than bile acids (i.e., cMOAT) (Arias et al., 1993; Gatmaitan and Arias, 1995; Vore, 1993). This was based on the observation that glutathione disulfide, a prototypic substrate for this carrier protein, did not inhibit ditekiren transport, even at high concentrations. However, competitive inhibition was noted when the canalicular vesicles were incubated in the presence of daunomycin with a K_i value (16 µM) similar to the K_m value for ATP-dependent transport of this drug by canalicular membrane vesicles (Bohme et al., 1993). In addition, other established substrates/inhibitors of P-glycoprotein, such as cyclosporin A, nicardipine, reserpine, rhodamine-123 and verapamil, also reduced ditekiren uptake, suggesting that this transporter was importantly involved in the uptake of the oligopeptide.

P-glycoprotein localized in the liver is primarily the product of the mdr1a gene and functions as an efflux pump at the canalicular membrane for a broad range of hydrophobic, cationic substrates with molecular weights of 400 to 1200 and containing at least two planar rings (Arias et al., 1993; Gatmaitan and Arias, 1995; Vore, 1993); the physicochemical properties of ditekiren fulfill these criteria. The importance of P-glycoprotein-mediated transport was first recognized with regard to the phenomenon of multidrug resistance in tumor cells (Gottesman and Pastan, 1993); however, the presence of the transporter in various normal tissues (Gottesman and Pastan, 1993) has led to speculation concerning its possible physiological substrate(s). Several members of the ABC superfamily in both prokaryotic and eukaryotic cells appear to be importantly involved in the transport of oligopeptides (e.g., the gene products of the Opp operon that translocate oligopeptides in bacteria; that of STE6 in yeast, which secretes the dodecapeptide -factor pheromone; and the translocation of foreign antigens into the endoplasmic reticulum for interaction with the major histocompatibility complex) (Kuchler and Thorner, 1990). Furthermore, a linear hydrophobic tripeptide, N-acetyl-Leu-Leu-Norleu, was shown to be transported by P-glycoprotein (Sharma et al., 1992), and the undecapeptide cyclosporin A is also a well-established substrate (Saeki et al., 1993). It has also been recently demonstrated that various biologically active, hydrophobic peptides stimulate ATPase activity associated with P-glycoprotein transport (Sarkadi et al., 1994). Accordingly, the finding with ditekiren, which is the first to directly determine the role of the transporter in translocating a linear oligopeptide across the hepatic canalicular membrane, is consistent with the possibility that one of the physiological functions of P-glycoprotein may be the cellular secretion of hydrophobic oligopeptides.

The range and diversity of substrates transported by P-glycoprotein are pronounced, and insights into the structure-function relationship are limited. Nevertheless, it is clear that some selectivity must be present to account for the fact that not all oligopeptides are transported (Arias et al., 1993) and that closely related analogs such as U-71013 and U-77436 are not as effective inhibitors of ditekiren transport as are other renin inhibitors (EMD-51921 and EMD-55068) and angiopeptin. However, the determinants of this selectivity are presently unknown but could involve lipophilicity, molecular size and charge and structure (Ruwart, 1995).

Finally, it is possible that modulation of the P-glycoprotein-mediated canalicular transport of oligopeptides, in a fashion analogous to reversing multidrug resistance in tumor cells (Gottesman and Pastan, 1993), might be a possible approach to reducing their rapid biliary excretion and thus enhancing their therapeutic potential.