Introduction

Abstract Introduction Results Discussion References

Bestatin (Ubenimex), a dipeptide analog, is a potent aminopeptidase inhibitor, first isolated from a culture filtrate of Streptomyces olivoreticuli (1). It has immunoenhancing properties with low toxicity in humans and experimental animals (see refs. 2-4). We became interested in this antibiotic because of our finding that, in mammalian cells, bestatin permits degradation of cellular proteins to di- and tripeptides, but inhibits the further degradation of these peptides to amino acids (5). This effect results in the cytosolic accumulation of intermediates in the breakdown of proteins by either the cytosolic or autophagic pathways, during normal protein turnover, degradation of abnormal proteins, or stimulated degradation of long-lived cellular components under catabolic conditions (6-8). More recently, we proposed that, after intravenous injection of bestatin, the rapid accumulation of peptide intermediates in the liver (9) and skeletal muscle (10) of live mice can be used to estimate instant rates of cellular protein breakdown in these tissues.

When assayed with purified cytosolic exopeptidases, bestatin is effective at nanomolar concentrations (11). The accumulation of peptide intermediates in mammalian cells, however, requires much higher extracellular concentrations. For instance, in reticulocyte lysates, bestatin resulted in accumulation of peptides during degradation of abnormal globin, with concentrations as low as 0.5 µg/ml. By contrast, in intact reticulocytes, it required concentrations in the incubation medium three orders of magnitude higher (5, 6). We assumed that the reticulocyte plasma membrane has a limited permeability to bestatin (6), but until now we had not measured it directly.

Information on cellular uptake of bestatin is limited. Because stimulation of lymphocyte proliferation is thought to result from the interaction of bestatin with enzymes exposed to the surface of cells (12, 13), the primary interest has been to reach effective concentrations of the drug in the extracellular fluid (3). After its intravenous injection into mice, bestatin disappears from the circulation rapidly (3). When given orally, bestatin is absorbed effectively in the intestine, via the H⁺/dipeptide transport system of the brush border (14, 15). After oral administration to humans, bestatin is largely recovered in the urine after 1 or 2 days (16, 17), most of it unmodified, except for a minor fraction (<5%), which is recovered as the active para-hydroxy-derivative. We only know of one autoradiographic study, one or more hours after the intravenous injection of ³H-bestatin to mice, which reports higher concentrations of silver grains in macrophages, in hepatocytes, in bile duct epithelium, and in the brush border of proximal straight renal tubules (18).

To measure the effects of bestatin in cellular protein degradation in the liver (9) and skeletal muscle (10) of mice, we collect the tissues 10-15 min after the intravenous injection of the drug. It is during this short interval that peptide intermediates accumulate linearly and permit an estimate of the rates of cellular protein breakdown. Maximal accumulation is observed in liver after injection of 1 mg of bestatin/mouse. For muscle, doses 5 times larger are necessary. In the present study, we add to these injections ³H-bestatin, and ¹⁴C-sucrose. We use ¹⁴C-sucrose as a nonpermeant marker of the extracellular space accessible to ³H-bestatin, and calculate by difference the cellular uptake of bestatin in the liver, kidney, intestine, red cells, and skeletal muscle.

Materials and Methods

Synthetic bestatin (Ubenimex), N-[(2S,3R)-3-amino-2-hydroxy-4-phenylbutanoyl)-L-leucine, was a generous gift of Nippon Kayaku Co. (Tokyo, Japan). Tritiated bestatin, labeled on a portion of the 4-(2-³H) phenyl moiety (343 µCi/µmol), was purchased from the same company. [¹⁴C(U)]-Sucrose (4.95 µCi/µmol) was obtained from Dupont-NEN (Boston, MA). Dulbecco's phosphate-buffered saline, in powder form, was purchased from Gibco Laboratories (Grand Island, NY).

Mice. Male CD-1 mice (36-38 g) were purchased from Charles River Laboratories (Wilmington, MA). They were kept in a room illuminated from 6 a.m. to 6 p.m., with water and food ad libitum, and fasted for 6 hr (see fig. 1 and table 1) or overnight (see table 2) before the experiment.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Disappearance of 3H-bestatin and 14C-sucrose from the plasma after their intravenous injections. A mouse (37 g) received the intravenous injection of 1 mg of bestatin in 0.3 ml of phosphate-buffered saline, which also contained 3H-bestatin (4.8 µCi) and 14C-sucrose (1.5 µCi). At the indicated times, 30-40 µl of blood was collected in heparinized microhematocrit capillary tubes, and the cells were separated by centrifugation. Plasma 3H and 14C radioactivities were determined in the same 5-µl aliquot. For comparison, 3H and 14C radioactivities are each normalized for the respective radioactivities of the injected bestatin and sucrose, and expressed as cpm/ml per injected cpm (see Materials and Methods).

Injections. Bestatin was dissolved for injection in phosphate-buffered saline at a concentration of 5 mg/ml. For intravenous bolus injection, each mouse was placed in a 50-ml conical plastic tube with the end cut out, just enough to permit protrusion of the tail. The tail was immersed in water at 45°C for 1 min or more, until the lateral veins were clearly visible. The solution was injected into one of these veins at a rate of 100 µl/sec.

Disappearance of ³H-Bestatin and ¹⁴C-Sucrose from the Plasma. Before the experiment shown in fig. 1, a 2-mm incision was made near the tip of the tail, along one of the lateral veins. Bleeding was stopped by pressing the wound gently, and resumed for collection by rubbing it open and squeezing the vein from the base of the tail toward the tip. The intravenous injection was done in the other tail vein, nearer to the base. At the indicated times, 30-40 µl of blood were collected in heparinized microhematocrit capillary tubes (Fisher Scientific, Pittsburgh, PA). These were sealed with plastic putty (Critoseal, Oxford Labware, St. Louis, MO), and the cells were immediately separated by centrifugation in a microhematocrit rotor (International-IEC #927). The capillary was cut at the interphase between plasma and red cells. Plasma ³H and ¹⁴C radioactivities were determined in the same 5-µl aliquot.

Collection of Urine. After intravenous injection, each mouse was placed in a square stainless-steel metabolic cage (13 cm each side). To ensure an abundant urinary flow, the mice received 1 ml of water intragastrically, through a #18 gavage needle (Pepper and Sons, Inc., New Hide Park, NY) 1 and 2 hr afterward. At the end of a 3-hr collection, the bladder was pressed gently to stimulate urination, the belly and genital regions were washed with distilled water, and the whole surface of the cage was rinsed by spraying it with a water mist, which was allowed to drain with occasional tapping. The urine and rinses were collected into a vial and made up to 20 ml. To ascertain the effectiveness of the collection, the cage was rinsed again into a second vial. This second rinse was found to contain <2% of the ³H and ¹⁴C radioactivities found in the first collection. The mice were killed immediately afterward, and their bladders were found to be empty.

Collection of Tissues. In the experiments of table 1, two mice were killed 10 min after injection, by decapitation under ether anesthesia. The resulting blood (~1 ml) was collected with 10 µl of sodium heparin (1,000 USP units/ml; LiphoMed, Inc., Rosemont, IL) into a 1.5 ml Eppendorf tube, and the red blood cells and plasma were separated by centrifugation. The first 8 cm of the small intestine were cut and its lumen washed clean by forcing water through its length with a syringe and gavage needle. The mucus was then expressed gently with a blunt instrument. Fifty microliters of plasma, the intestines (0.11-0.12 g), the red cell pellets (0.46-0.48 g), the right abdominal wall muscles (0.60-0.64 g), the liver (1.50-1.53 g), and both kidneys (0.52-0.58 g) were extracted overnight with 2 volumes of 5% trichloroacetic acid. The acid-soluble extract was cleared by centrifugation and 0.1 ml of the plasma and intestinal extracts, and 0.5 ml of the other extracts were used for scintillation counting.

Perfusion of Muscle. In the experiments of table 2, we sought to minimize muscle extracellular radioactivity by perfusion of the hindquarters through the abdominal aorta (just below the renal arteries), with a solution containing phosphate buffered-saline, 11 mM glucose, 100 mM sucrose, and 10 µM sodium nitroprusside (as a vasodilator agent). The perfusion started 7-8 min after death and continued for 7 min at a rate of 7 ml/min. The muscle groups indicated in the table were dissected separately and acid-extracted as described.

Total and Diffusible ³H-Bestatin in Plasma. The trichloroacetic acid-soluble extract contained all of the plasma ³H and ¹⁴C radioactivities. Preliminary experiments, however, suggested that not all the ³H-bestatin in plasma is free to diffuse, possibly because of reversible binding of a fraction of the drug to plasma macromolecules. The diffusible fraction, relative to sucrose, was determined by differential dialysis as follows. We used an ad hoc microcell assembly consisting of two pieces of Plexiglas separated by a dialysis membrane (Spectra/por molecular weight cutout of 12,000-14,000; Spectrum Medical Industries, Inc., Los Angeles, CA). Carved into each half were matching circular cells, 2.3 mm deep and 6.1 mm in diameter, that were accessible from the top through a small channel. Using a microsyringe, 60 µl of plasma was injected into one cell and the same volume of phosphate-buffered saline into the opposite cell. Diffusion through the dialysis membrane was allowed to proceed at room temperature for 1 hr. At this time, the fluids were withdrawn, mixed with 2 volumes of 5% trichloroacetic acid (which precipitated the proteins in the plasma), and centrifuged; 100 µl of the acid-soluble extract was used for scintillation counting. The diffusible fraction was calculated as the ³H/¹⁴C ratio in the saline side divided by the ³H/¹⁴C ratio in the plasma side. It was determined in duplicates, which agreed within 1.5% of their mean. We then calculated the plasma concentration of diffusible bestatin as the product of this diffusible fraction and the total bestatin content in plasma. In seven different plasma samples, the diffusible fraction ranged from 60 to 74% of the total bestatin. This range is consistent with the results of bestatin uptake by erythrocytes in vitro; uptake of bestatin from plasma was 61-73% of the uptake from a saline solution (see Results).

Entry of ³H-Bestatin into Red Blood Cells in vitro. In the experiments shown in fig. 2 and table 3, 100 µl of whole blood, or a 1:1 suspension of red blood cells in buffer (phosphate-buffered saline, 10 mM glucose, and 50 mM sucrose), was incubated at 37°C with ³H-bestatin, ¹⁴C-sucrose, and nonradioactive bestatin as described in the figure legends. At the end of incubation, the extracellular fluid was diluted 100-fold with buffer, and 1 ml of the diluted cell suspension was layered in a 1.5 ml Eppendorf tube, over a cushion of 0.5 ml of 0.5 M sucrose, and centrifuged for 30 sec in a high-speed microcentrifuge (model 3200; Brinkman Instruments, Westbury, NY). The walls were carefully wiped with a cotton-tipped swab, and the red cell pellet lysed with 100 µl of water and transferred to another tube with 0.4 ml of 5% trichloroacetic acid. The acid-soluble extract was used for scintillation counting. Recovery of nonpermeant ¹⁴C-sucrose in the pellet (0.14-0.20% of that in the original extracellular fluid) was used to estimate the residual extracellular ³H-bestatin; this was subtracted from the total ³H-bestatin in the pellet to calculate the cellular uptake of the drug.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Time course of entry of 3H-bestatin into red blood cells. Two separate experiments are shown. In the first (circles), 100 µl of whole blood was incubated at 37°C with 3H-bestatin (0.24 µCi), 14C-sucrose (0.1 µCi), and nonradioactive bestatin (0.1 mg/ml of plasma). At the indicated times, the extracellular fluid was diluted 100-fold by pipetting 20 µl of the incubated suspension in 1 ml of buffer (phosphate-buffered saline, 10 mM glucose, and 50 mM sucrose), and the cells were centrifuged for 30 sec through a 0.5 M sucrose cushion. In the second experiment (triangles), an identical mixture was incubated for 10 min, at which time it was diluted by addition of 5 ml of prewarmed buffer and incubated further. One-milliliter aliquots were withdrawn 1, 3, 8, and 18 min after the dilution, and the cells were centrifuged as previously described. Recovery of nonpermeant 14C-sucrose was used to calculate the extracellular contamination of the pellet and the intracellular 3H-bestatin radioactivity (see Materials and Methods).

Scintillation Counting. Samples of the acid-soluble tissue extracts or urines were mixed with 10 ml of scintillation fluid (Liquiscint; National Diagnostics, Atlanta, GA) and counted in separate channels for ³H and ¹⁴C in an LKB 1209 Rackbeta Scintillation counter. Samples containing the same solvents were added to each run with the injected solution, as well as a ¹⁴C-sucrose standard to determine cross-over in the ³H channel.

	Results

Abstract Introduction Results Discussion References

Our primary interest in this study is the uptake of bestatin in liver and muscle, 10 min after its intravenous injection. In this short interval, because of inhibition by bestatin of cytosolic exopeptidases, peptide intermediates accumulate linearly in these tissues and permit an estimate of the rates of cellular protein breakdown (9, 10). In the experiments presented herein, we inject bestatin intravenously to mice, together with trace amounts of ³H-bestatin and ¹⁴C-sucrose. We use ¹⁴C-sucrose as a nonpermeant marker of the extracellular space accessible to ³H-bestatin, and calculate by difference the cellular uptake of bestatin in the liver, kidney, intestine, red cells, and skeletal muscle. We chose labeled sucrose as a marker because its molecular weight (340) is closest to that of bestatin (308). This is significant in tissues, such as skeletal muscle, in which capillaries are lined by a continuous endothelium with tight interendothelial junctions, because in these tissues transcapillary permeability to small solutes are predominantly determined by their molecular size (19). Bestatin is not metabolized; even after 1-2 days, <5% is converted to the bioactive para-hydroxy-derivative (16, 17). We therefore assume that, in the relatively short intervals after its injection, all of the ³H radioactivity represents bestatin at the same specific radioactivity as in the injected solution.

Bestatin and Sucrose Disappear Rapidly from the Circulation and Are Largely Recovered in the Urine. When injected intravenously to mice, bestatin disappears rapidly from the circulation (3). It has not been clear, however, how much of this rapid disappearance is due to its urinary elimination and how much to its cellular uptake. The following experiments indicate that most of the injected bestatin is rapidly eliminated in the urine. A mouse received the intravenous injection of 1 mg of bestatin, 2.4 µCi of ³H-bestatin and 0.75 µCi of ¹⁴C-sucrose. Because of their similar molecular sizes, we expected that sucrose and freely dissolved bestatin will escape the intravascular space at closely similar rates (19). Sucrose does not enter cells, it is not metabolized, and it is eliminated in the urine (20, 21).

³H-Bestatin Is Taken Up Rapidly by the Liver, Kidney, and Intestine, But Slowly by Erythrocytes and Skeletal Muscle. Two mice received an intravenous injection of bestatin (5 mg) with 1 µCi of ³H-bestatin and 0.3 µCi of ¹⁴C-sucrose. The mice were killed 10 min after the injection. The trichloroacetic acid-soluble ¹⁴C and ³H radioactivities were determined in the plasma, red blood cells, skeletal muscles of the abdominal wall, liver, kidneys, and the upper portion of the small intestine. The results are presented in table 1.

Uptake of Bestatin into Erythrocytes Is Slow and Nonsaturable. The experiments previously described show a large difference in the uptake of bestatin, which is high in the liver, kidneys, and intestine, but much lower in red cells or skeletal muscle. This could indicate the entry of bestatin through transporter-mediated processes in some cells, but not others (see Discussion). The slower entry of bestatin into erythrocytes and skeletal muscle cells may depend entirely on passive diffusion across the plasma membrane. Red blood cells, because of their simple structure, the limited number of transporters in their plasma membrane, and the facility with which we can study them in suspension, lent themselves to additional experiments that reproduced in vitro the slow uptake of bestatin and suggested a nonsaturable mechanism for it.

	Discussion

Abstract Introduction Results Discussion References

The immunoenhancing effects of bestatin are attributed to its interaction with enzymes exposed in the surface of some of the cellular components of the immune system (12, 13). Because of this, the primary interest has been to reach effective concentrations of the drug in the extracellular fluid (3). Our purpose is different. We are interested in the uptake of bestatin in the liver and muscle, within 10 min after its intravenous injection. In this short interval, because of inhibition by bestatin of cytosolic exopeptidases, peptide intermediates accumulate linearly in these tissues and permit an estimate of instant rates of cellular protein breakdown (9, 10). This is an important tool in studies of protein metabolism, because there is no other way at present to estimate rates of protein degradation in the tissues of live animals in such a short time.

The experiments reported in this study indicate that, after a single intravenous injection, most of the drug is rapidly eliminated in the urine in the same manner as the impermeant sucrose, with a half-life in the extracellular fluid of ~20 min. Whereas the drug is available to cells, there is a large difference in the uptake of bestatin, which seems to readily enter the liver, kidneys, and intestine, but much less so in red cells or skeletal muscle. This could indicate the entry of bestatin through transporter-mediated processes in some cells but not others, and has implications both in the targeting of the drug to specific tissues and in anticipating the potential toxicity of high doses.

We calculate the cellular uptake by difference (i.e. the total minus the calculated extracellular bestatin). It could be argued that the excess bestatin has not entered the cells, but is bound to extracellular peptidases on their surface. This is the mechanism proposed for the effects of bestatin in some of the cells of the immune system (12, 13). It has been proposed that tissue distribution of bestatin is closely related to the distribution of aminopeptidases, particularly the abundant leucine aminopeptidase (18, 25). This is conceivable in studies (18) in which ³H-bestatin was detected by autoradiography, one or more hours after the injection of a dose of the drug 14-fold smaller than that used in the experiments herein. The intracellular content of bestatin in tables 1 and 2 are too large to represent enzyme-bound bestatin, either on the surface or inside cells. The most abundant aminopeptidase in liver is the cytosolic leucine aminopeptidase. From its concentration in liver (0.1 mg/g), and its hexameric molecular weight (360,000) (26), we calculate that a maximum of six molecules of bestatin bound per hexamer (27) would represent ~0.5 µg of bestatin/g of liver; the concentrations actually found were three orders of magnitude larger. The slow, linear, nonsaturable uptake of bestatin into erythrocytes (fig. 2, table 3) is also consistent with entry into the cells, rather than binding to their surface.

In the liver, previous autoradiographic studies have shown uptake of bestatin primarily in hepatocytes (18). Although we know at this time of no specific transport mechanism that can account for this uptake, the sinusoidal domain of the hepatocyte plasma membrane is known to possess many transporters involved in the uptake of a large variety of substances, many of which are metabolized in the liver or eliminated in the bile (28). The high intracellular concentrations of bestatin found in the liver in this experiment, 10 min after its intravenous injection, explains the rapid inhibition of hepatic cytosolic exopeptidases with one-fifth of the dose used herein (7-9). Hepatic uptake of bestatin is reversible. In another experiment (not shown), a mouse injected and studied in the same manner as those in table 1 was killed 1 hr after the injection. Intracellular bestatin in liver was found to be 16 µg/g, or 20-fold less than in the livers of mice after 10 min.

The abundant uptake of bestatin in kidneys and intestine indicate that the liver is not unique in this respect. Because of the complex and heterogeneous structure of these organs, however, the interpretation of these results is less clear. Autoradiographic measurements, 1 hr after the intravenous injection of bestatin, have shown higher density of silver grains in the brush border of proximal straight renal tubules (18). It suggests that bestatin may enter renal cells from the lumen of the nephron, through which most of the drug is eliminated from the body. Bestatin is a substrate for a H⁺/coupled peptide transporter in the renal brush border (29). This transporter (PEPT2) has been recently cloned and shown to be distinct from the similar transporter in the intestinal brush border (PEPT1) (30). The autoradiographic study previously mentioned did not include intestine. We do not know how much of the drug was in the muscular wall or in the intestinal epithelium. Bestatin has been shown to enter readily the intestinal epithelium from the lumen by H⁺/coupled active transport via the PEPT1 system in the brush border, but not the amino acid transport systems (14, 15). It seems, however, unlikely that within 10 min after its intravenous injection, enough bestatin found its way into the intestinal lumen, from which other small molecules, such as sucrose, are excluded (21).

By contrast with the liver, kidney, or intestine, the slower entry of bestatin into erythrocytes and skeletal muscle cells may depend entirely on passive diffusion across the plasma membrane. The low rate of entry, the lack of saturation, and the paucity of plasma membrane transporters in the erythrocyte membrane make this very likely in red cells. Bestatin is an analog of the dipeptide phenylalanyl-leucine. Because of its hydrophobic side chains, it may enter cells by partition in the lipid bilayer of the plasma membrane. Ten minutes after the intravenous injection of 5 mg of bestatin, we find in the muscle 4.8 µg of intracellular bestatin/g of tissue or 12 µg/ml of intracellular water. This is almost 100-fold less than in liver, but it still is 10-fold higher that the concentrations of bestatin found to produce accumulation of peptide intermediates in reticulocyte lysates (6). This explains why, although with great inefficiency, the intravenous injection of a large dose of bestatin can produce intracellular accumulation of peptide intermediates in skeletal muscle (9).

In addition to its uses as an aminopeptidase inhibitor, bestatin has proven to be valuable as a nonhydrolizable probe for the H⁺/coupled peptide transporters PEPT1 and PEPT2, in the brush border of the intestine and kidney (14, 15, 29). These transporters permit the efficient intestinal absorption of di- and tripeptides resulting from the digestion of dietary proteins, and the reabsorption of peptides from the glomerular filtrate (31, 32). They are also involved in the absorption and retention of many peptidomimetic drugs, including cephalosporins, penicillins, and inhibitors of angiotensin-converting enzyme (32). PEPT1 and PEPT2 are, so far, the only well characterized mammalian peptide transporters, but not the only ones. We have shown that di- and tripeptides resulting from the breakdown of proteins by lysosomal enzymes are released from autophagic organelles to the cytosol, where they are degraded to amino acids by bestatin-sensitive exopeptidases (7, 8). More recent studies have shown the selective release by lysosomes, in vitro, of endocytosed di- and tripeptide probes (33). There is also a transporter in the endoplasmic reticulum that mediates the translocation of oligopeptides for antigen presentation (34). These oligopeptides are, however, larger than the di- and tripeptides discussed herein. The ready uptake of bestatin by the liver, shown in this study, suggests the existence of a novel peptide transporter in the sinusoidal domain of the hepatocyte plasma membrane. It must be different from PEPT1 and PEPT2, because Northern blots using specific cDNA probes failed to reveal mRNAs for either of these transporters in liver (30, 35). Radioactive bestatin should be a useful probe for the identification and study of this transporter in hepatocytes, either freshly isolated or in primary cultures.