Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Valacyclovir (VACV) is derived from acyclovir (ACV) by esterifying valine to the hydroxyl group of ACV (Beauchamp et al., 1992). ACV is used to treat a variety of viral infections including cytomegalovirus infections, an AIDS opportunistic infection. VACV is completely converted to ACV by first-pass intestinal and hepatic metabolism in rats and humans (Burnette and de Miranda, 1994; de Miranda and Burnette, 1994; Soul-Lawton et al., 1995). Its bioavailability is three to five times higher than that of ACV in humans. It has been demonstrated that the uptake of VACV in primate intestinal brush border membrane vesicles (BBMVs) is concentration dependent, saturable, and inhibited by several dipeptides, suggesting the involvement of the proton-linked intestinal peptide transporter (PEPT1) (Smith et al., 1993). Proton-dependent uptake of VACV, however, was not demonstrated in that study. Studies performed in our laboratory using an in situ perfusion technique in rat intestine and rabbit apical BBMVs have suggested that the transport of VACV may be mediated by multiple membrane transporters, including the intestinal peptide transporter (Hu and Sinko, 1997; Sinko and Balimane, 1998). de Vrueh et al. (1998) and Cook et al. (1997) have demonstrated that the transport kinetics of VACV in Caco-2 cells were concentration dependent and saturable. However, all of these studies were performed in rat and rabbit intestinal tissue or Caco-2 cells, where numerous transporters potentially involved with the transport of VACV are found. Therefore, it is difficult to assess the role of each putative transporter, such as PEPT1. The present study provides direct evidence of the involvement of PEPT1 in the intestinal transport of VACV in a Chinese hamster ovary (CHO) cell system. Furthermore, the functional transport characteristics of VACV (e.g., lack of proton cotransport) in the human PEPT1 (hPEPT1)/CHO cell system were consistent with results obtained in intestinal tissues in vitro.

Active peptide transport has unequivocally been shown to occur in the intestine and the kidney (Fei et al., 1998). The peptide transporter genes PEPT1 and PEPT2 have been cloned from human, rat, and rabbit (Boll et al., 1994, 1996; Fei et al., 1994; Liang et al., 1995; Miyamoto et al., 1996). The PEPT1 and PEPT2 genes were isolated from mammalian intestine and kidney, respectively. Through the expression and characterization of these transporter genes in Xenopus laevis oocytes (XLOs), the intestinal PEPT1 transporter displayed high transport capacity and low substrate affinity. On the other hand, PEPT2 transporters showed a high affinity but low capacity (Döring et al., 1996). Because the PEPT1 transporter has high transport capacity and relatively broad substrate specificity (Amidon and Lee, 1994; Ganapathy and Leibach, 1994; Döring et al., 1996), this transporter could potentially serve as a viable absorption pathway for numerous drugs and prodrugs. Studies characterizing the intestinal absorption mechanisms of VACV suggest the involvement of multiple transporters, including the peptide transporter (Hu and Sinko, 1997; Sinko and Balimane, 1998). Recently, Ganapathy et al. (1998) observed interactions between VACV and the intestinal peptide transporter and also showed interactions of VACV with the renal peptide transporter (PEPT2) in the rat kidney proximal tubular cell line SKPT and in PEPT2-transfected HeLa cells. More recently, there have been several reports that the transport of VACV was concentration dependent in Caco-2 cells transfected with hPEPT1 and XLOs injected with hPEPT1 (Balimane et al., 1998; Han et al., 1998). In the present study, we characterized the interactions and kinetics of VACV with hPEPT1 in a CHO cell system providing further evidence for the potential involvement of this intestinal pathway in the oral absorption of VACV. It is also demonstrated, for the first time, that the mechanism of VACV uptake is not proton dependent but rather substrate structure charge dependent, which is consistent with other results obtained in tissues in vitro.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. VACV was kindly provided by Glaxo Wellcome Inc. (Research Triangle Park, NC). [³H]VACV was synthesized by Moravek Biochemicals (Brea, CA). [¹⁴C]Gly-Sar (specific radioactivity, 110 mCi/mmol) was purchased from Moravek Biochemicals. Medium, nonessential amino acids, and trypsin were purchased from Fisher. FBS and Lipofectamine were purchased from GIBCO BRL (Grand Island, NY). DNA isolation kits were purchased from Qiagen Inc. (Santa Clara, CA). Restriction enzymes were obtained from New England Biolabs (Beverly, MA). The pcDNA3 was purchased from Invitrogen, and CHO cells were supplied by American Type Culture Collection (Rockville, MD). All other chemicals were obtained from Sigma Chemical Co. (St. Louis, MO).

Cell Culture. CHO cells were grown in Dulbecco's modified Eagle's medium containing 90% Dulbecco's modified Eagle's medium, 10% FBS, 1% nonessential amino acids, 100 units/ml penicillin, and 100 µg/ml streptomycin. CHO cells were grown at 37°C in a humidified atmosphere of 5% CO₂. Culture medium was changed every other day, and cells were passed every 3 to 5 days by trypsinizing cells with 0.05% trypsin and 0.53 mM EDTA at 37°C for 2 min.

Construction of hPEPT1 Gene for Expression. The human dipeptide transporter gene hPEPT1 was obtained from Dr. You-Jun Fei (Medical College of Georgia, Atlanta, GA). The hPEPT1 gene was subcloned into the mammalian expression vector (pcDNA3) by digestion of the pBluescript/hPEPT1 with EcoRV and NotI and then ligation into pcDNA3 at EcoRV-NotI sites. The construct was confirmed by restriction enzymes analysis.

Transfection. Cells were transfected with hPEPT1 or pcDNA3 (vector control) by Lipofectamine according to the manufacturer's instructions (GIBCO BRL). Briefly, CHO cells were seeded at a density of 3 × 10⁵ cells/well in 12-well plates and incubated at 37°C for 24 h. For each well, 1 µg of DNA was mixed gently with 200 µl of serum-free medium and 10 µg of Lipofectamine reagent. The mixture was incubated at room temperature for 15 min and transferred to each well. Then, 0.8 ml of serum-free medium was subsequently added to the mixture. After a 5-h incubation at 37°C, the transfection mixture was removed and replaced with 1 ml of complete growth medium containing 10% FBS.

Functional Assay. Transport assays were performed essentially as described by Liang et al. (1995). Cells were washed twice with 25 mM 2-(N-morpholino)ethanesulfonic acid/Tris, pH 6.0, or 25 mM HEPES/Tris, pH 7.5, buffer containing 130 mM NaCl, 5.4 mM KC1, 1.8 mM CaCl₂, 0.8 mM MgSO₄, and 5 mM glucose. Subsequently, the cells were incubated either with 20 µM (1 µCi/ml) [³H]VACV or 20 µM (0.1 µCi/ml) [¹⁴C]Gly-Sar at 37°C for 5 min. The uptake was stopped by washing the cells three times with ice-cold buffer. Nonspecific uptake was measured in parallel experiments with the control pcDNA3 vector-transfected CHO cells. Finally, the cells were solubilized by 0.1% v/v Triton X-100, and 0.6 ml was used for scintillation counting. The remaining volume was saved for protein concentration determination.

Inhibition Studies. Inhibition studies were carried out in triplicate in the coincubation with VACV and competitive substrates acting as inhibitors. [³H]VACV (20 µM, 1 µCi/ml) was used in the control. In the inhibition studies, 20 µM [³H]VACV was coincubated with other inhibitors. Gly-Sar, one of the well known substrate for hPEPT1, and several other dipeptides and -lactam antibiotics were used as inhibitors. Gly-Sar and cefadroxil were used at concentrations of 5 and 10 mM. All others, Gly-Gly, Ala-Ala, Gly-Leu, Leu-Leu, ampicillin, and cephalexin, were used at a concentration of 10 mM. Uptake assays were performed as above.

Protein Assay. After cells were solubilized in Triton X-100, 10 µl of solution was taken from each well for protein concentration determination. Protein concentration was determined using the Bio-Rad reagent according to the manufacturer's instructions (Bowers-Komro et al., 1989). BSA was used as standard.

Data Analysis. Each data set was collected from three or four samples. The kinetic parameters for the Michaelis-Menten studies were calculated using The Scientist (MicroMath, UT) with the equation J = J_max * [C]/(K_m + [C]), where J is the rate of VACV uptake and [C] is the VACV concentration. Results from inhibition studies were plotted using a Lineweaver-Burk analysis. K_i values were obtained by fitting the data using the equation J = J_max * [C]/{[C] + K_m (1 + [I]/K_i)}, where K_i is the dissociation constant for the transporter-inhibitor complex and [I] is the inhibitor concentration. Data were weighted using 1/S.E.M².

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Functional Expression of hPEPT1 in CHO Cells. Transport activity of CHO cells transiently transfected with hPEPT1 was examined with a known peptide transporter substrate, [¹⁴C]Gly-Sar. The initial rate time point (5 min) for the uptake of Gly-Sar was selected because the maximum uptake was observed at 10 min in the time course studies (data not shown). Uptake of Gly-Sar was determined at various concentrations in the range of 0.1 to 5 mM (Fig. 1A). The K_m and J_max values were 3.13 ± 1.09 mM and 2.71 ± 0.48 nmol/mg protein/5 min. The pH effect studies were carried out in CHO cells transfected with hPEPT1 in the presence or absence of 10 mM Gly-Sar (Fig. 1B). The uptake of Gly-Sar was proton dependent with a higher activity at pH 5.5 and 6.0 than that at pH 7.5. 

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Transport of glycylsarcosine in CHO cells transfected with hPEPT1. A, concentration dependence and saturability of Gly-Sar uptake. The uptake was measured over a concentration range of 0.1 to 5 mM at pH 6.0 with 5-min incubation. B, the pH dependence of Gly-Sar in hPEPT1-transfected CHO cells. Transport activity was measured at pH 5.5, 6.0, and 7.5 in the absence () or presence () of 10 mM unlabeled Gly-Sar. Uptake was measured after 5-min incubation.

VACV Transport Activity. The time course for VACV uptake is shown in Fig. 2. The level of VACV uptake in CHO cells was significantly higher than that in the control (i.e., vector only). The accumulation of VACV was linear up to 10 min after incubation with [³H]VACV. For subsequent experiments, an incubation time of 5 min was used for uptake studies. The level of VACV uptake in the control (i.e., pcDNA3-transfected CHO cells), representing the background level of transporter activity, nonmediated diffusion, and nonspecific binding, was very low. When 10 mM unlabeled VACV was added to the uptake solution, the level of VACV transport was almost identical with that in the control, suggesting that all significant transport activity was due to hPEPT1. The experiments were performed in quadruplicate.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Time course of 20 µM [3H]VACV uptake in CHO cells transfected with hPEPT1. The uptake was measured over time in the absence () of 10 mM VACV in hPEPT1-transfected CHO cells. Uptake from controls was measured in CHO cells transfected with pcDNA3 (vector) () or transfected with hPEPT1 in the presence of 10 mM unlabeled VACV ().

Concentration Dependence and Saturability of VACV Uptake. The uptake of VACV in CHO cells transfected with hPEPT1 was concentration dependent and saturable (Fig. 3). The measurements were done at concentrations in the range of 0.2 to 4 mM VACV. Using The Scientist program (MicroMath Utah), the Michaelis-Menten constant and maximum velocity values for VACV uptake were determined as 1.64 ± 0.06 mM and 23.34 ± 0.36 nmol/mg protein/5 min, respectively. Transformation of the data from uptake of VACV resulted in an Eadie-Hofstee plot (r = 0.99) (Fig. 3, inset). The kinetics of VACV uptake matched a single, saturable carrier model.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Concentration dependence and saturability of [3H]VACV uptake. Uptake was measured at 5 min at pH 7.5 over a concentration range of 0.2 to 4 mM. Values represent mean ± S.D. (n = 4). Inset, Eadie-Hofstee linear plot from VACV uptake.

pH Effect on VACV Uptake. The pH effect on the interaction of VACV and the peptide transporter also was characterized. Figure 4 shows VACV uptake in the absence or presence of 10 mM unlabeled VACV measured at different pH. Optimum uptake was observed at pH 7.5. Uptake of VACV at pH 7.5 was almost 2-fold higher than that at pH 5.5, 6.0, and 8.0. It is clear that pH had a different effect on the uptake of VACV compared with the uptake of Gly-Sar. Because the uptake of the prototypical PEPT1 substrate, Gly-Sar, was proton dependent and contrary to the VACV results, the effect of buffer composition and pH was further investigated. To investigate whether the observed pH results were due to a buffer effect, the experiments were repeated four times, with two additional buffers (Moseley et al., 1992; Covitz et al., 1996); however, very similar results were obtained with all three buffers showing maximal VACV uptake at pH 7.5. 

View larger version (21K): [in this window] [in a new window]   Fig. 4.   pH effect on VACV uptake in hPEPT1-transfected CHO cells. Transport activity was measured at pH 5.5, 6.0, 7.5, and pH 8.0 in the absence () or presence () of 10 mM unlabeled VACV. Uptake was measured after 5-min incubation.

Inhibition by Peptides and -Lactam Antibiotics. Dipeptides, tripeptides, and some -lactam antibiotics are known substrates of the peptide transporter, and they have been shown to inhibit the uptake of Gly-Sar (Liang et al., 1995; Covitz et al., 1996). To determine the interaction between VACV and hPEPT1, as well as these compounds, several dipeptides and -lactam antibiotics were studied as putative inhibitors of VACV uptake. As shown in Fig. 5, the uptake in the control, VACV alone, was assigned a value of 100% (739 pmol/mg protein/5 min). VACV uptake was significantly inhibited by all of the inhibitors used in this study. Interestingly, three -lactam antibiotics (ampicillin, cefadroxil, and cephalexin) exhibited very similar levels of inhibition: approximately 50% inhibition compared with the control. All other dipeptides, Gly-Gly, Ala-Ala, Gly-Leu, and Leu-Leu, inhibited VACV uptake by more than 80%.

View larger version (14K): [in this window] [in a new window]   Fig. 5.   Inhibition of [3H]VACV uptake by dipeptides and antibiotics. Uptake was measured at 5 min in the presence of 10 mM concentration of each compound. Data are presented as mean ± S.D. (n = 4).

Inhibitory Effects of Gly-Sar and Cefadroxil on VACV Uptake. To further characterize the interactions between VACV and hPEPT1, a dipeptide and a -lactam antibiotic, Gly-Sar and cefadroxil, respectively, were used to inhibit the uptake of [³H]VACV in hPEPT1-transfected CHO cells. Data were transformed and are shown in two Lineweaver-Burk plots (Fig. 6, A and B). The rates of uptake of VACV were measured in the presence of Gly-Sar or cefadroxil at two fixed concentrations (5 and 10 mM) against various concentrations of VACV. Significant inhibition of VACV uptake was observed in the presence of Gly-Sar at concentrations of 5 and 10 mM (Fig. 6A). The calculated K_i values for Gly-Sar and cefadroxil were 12.8 ± 2.7 and 9.1 ± 1.2 mM, respectively. The ordinate intercepts (1/J_max) from treatments using Gly-Sar as an inhibitor were very similar to that from the control, whereas the Michaelis-Menten constants (K_m) were different for all three treatments. This indicates that inhibition of VACV uptake by Gly-Sar was competitive. Similar results were also observed when cefadroxil was used as the inhibitor, although the inhibition was slightly reduced at 5 mM (Fig. 6B). The J_max value were not significantly different (P < .01) for VACV uptake regardless of the presence or absence of cefadroxil, suggesting a competitive interaction; however, the scatter in the data potentially confounds the interpretation, further suggesting that a mixed-type inhibition model for cefadroxil is possible.

View larger version (13K): [in this window] [in a new window]   Fig. 6.   Inhibition of Gly-Sar and cefadroxil on [3H]VACV uptake. Rates of uptake of VACV were measured at 5 min in the presence of Gly-Sar (A) or cefadroxil (B) at 5 and 10 mM against various concentrations of VACV in CHO cells transfected with hPEPT1. Data were presented in linearized Lineweaver-Burk plots.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

The intestinal peptide transporter has a broad spectrum of substrates. Dipeptides and tripeptides, but not free amino acids, are the primary substrates of these transporters (Matthews, 1991). PEPT1 not only serves to mediate in the absorption of nutrients but also functions in the transport of exogenous compounds that have peptide-like structures. Cefadroxil, cephalexin, and ampicillin are -lactam antibiotics, possessing peptide-like chemical structures (Sinko and Amidon, 1989; Tsuji and Tamai, 1996; Tamai et al., 1997). These antibiotics are known substrates for intestinal PEPT1 and renal PEPT2, although differential recognitions of these drugs by these two transporters were observed (Ganapathy et al., 1995). The wide range of endogenous substrates and peptide-mimicking drugs makes the intestinal transporter PEPT1 an important, potential carrier for drugs and prodrugs. The present results demonstrate that PEPT1 is involved in the apical domain transport of a nonpeptidic prodrug, VACV.

VACV demonstrates an oral bioavailability that is three to five times greater than ACV (Jacobson, 1993; Burnette and de Miranda, 1994). A series of studies characterizing the carrier-mediated intestinal absorption of VACV have been carried out in our laboratory over the past few years. Based on studies using the single-pass intestinal perfusion technique in rats, results show that the uptake of VACV is potentially mediated by several transporters, including the intestinal peptide transporter (Sinko and Balimane, 1998). Studies using rabbit BBMVs, Caco-2 cells, or hPEPT1/Caco-2 cells also demonstrate that the uptake or permeability of VACV is concentration dependent and saturable (Lee et al., 1996; Cook et al., 1997; Hu and Sinko, 1997; Han et al., 1998). These studies were performed in experimental systems where multiple transporters are potentially present. In other words, intact tissues or Caco-2 cells possess multiple transporters that may potentially be involved in the uptake of VACV. Therefore, it is difficult to characterize and evaluate the role of a single transporter such as PEPT1 in the intestinal transport of VACV. Recently, uptake of VACV was also studied in hPEPT1/XLOs, hPEPT1/HeLa cells, or rPEPT2/HeLa cells (Balimane et al., 1998; Ganapathy et al., 1998). To characterize the direct interaction between VACV, the L-valyl ester prodrug of ACV, and hPEPT1, the transport of VACV was studied in CHO cells that were transfected with hPEPT1. The uptake of VACV was concentration dependent and saturable (Fig. 3). The K_m value was 1.64 ± 0.06 mM, which is consistent with the results previously reported in rats (K_m = 1.2 mM), rabbits (K_m = 1.3 mM), monkey (K_m = 3.4 ±1.2 mM), and Caco-2 cells (K_m = 2.0 mM), respectively (Smith et al., 1993; Cook et al., 1997; Hu and Sinko, 1997; Sinko and Balimane, 1998). The present results are slightly higher than that reported by others in Caco-2 cells (K_m = 0.3 mM; Lee et al., 1996; de Vrueh et al., 1998) and lower than that reported by our group in hPEPT1/XLO (K_m = 5.94; Balimane et al., 1998). Uptake of VACV was dramatically inhibited by known substrates of hPEPT1, including all of the peptides and -lactam antibiotics used in the present study. The K_m value from CHO cells with the overexpressed hPEPT1 gene was similar to those obtained from rat and rabbit tissue and Caco-2 cells in vitro. Even though the affinity of VACV was similar in the overexpressed CHO cell system and the in vitro tissue studies, the capacity (J_max) values could not be directly compared due to the differences in study type. Furthermore, previous results from our laboratory in intact rat intestinal segments suggest that nonmediated diffusion of VACV is minimal (Sinko and Balimane, 1998). Therefore, even if there were other transporters involved in the absorption of VACV in vivo, these transporters probably would have an insignificant role. For example, if an organic cation transporter with typical K_m values in the micromolar range was involved in VACV uptake, it would be saturated at low concentrations and therefore would not significantly affect the total uptake of VACV. Given that the operational (i.e., apparent) K_m values are similar in the overexpressed CHO cells and normal intestinal tissues and the high-capacity, low-affinity transport properties of hPEPT1, these results provide direct evidence that hPEPT1 is critical to the intestinal transport of VACV.

Inhibition studies confirmed that the uptake of VACV is mediated by hPEPT1. The uptake of VACV was significantly inhibited by dipeptides and -lactam antibiotics (Figs. 5 and 6). Interestingly, greater inhibition was observed for dipeptides (Gly-Gly, Ala-Ala, Gly-Leu and Leu-Leu) than for -lactam antibiotics (ampicillin, cefadroxil, and cephalexin). The inhibition effect was further characterized using Gly-Sar and cefadroxil as a representative peptide and peptide drug analog, respectively. The calculated K_i values for Gly-Sar and cefadroxil were 12.8 ± 2.7 and 9.1 ± 1.2 mM, respectively. The lower affinity of these inhibitors for PEPT1 is a result of the pH used in the studies and is consistent with other reports (Wenzel et al., 1996; Amasheh et al., 1997). Lineweaver-Burk analysis showed a similar value for the J_max and different K_m values for the uptake of VACV in the presence of 5 and 10 mM Gly-Sar (Fig. 6A). It is apparent that the inhibition of VACV by Gly-Sar fits a competitive inhibition model. However, it is not as clear whether the inhibition of VACV by cefadroxil fits a typical competitive inhibition model given the scatter in Results (Fig. 6B). The results of the analysis indicate that the inhibition appears to be competitive, but scatter in the data also suggests a mixed-type inhibition where the other mechanism is unknown. There are two possible reasons for this behavior. The first is that there might be multiple transporters involved in VACV/cefadroxil transport, which would confound the analysis. Second, the differences in inhibition may be related to a pH-dependent affinity phenomenon, as described in the following section.

Differential effects of pH on the transport of Gly-Sar and VACV in hPEPT1-transfected CHO cells were observed in this study (Figs. 1B and 4). The human intestinal peptide transporter has been reported as a H⁺-dependent transporter (Thwaites et al., 1993a; Liang et al., 1995; Covitz et al., 1996); our studies with the prototypical substrate Gly-Sar confirm this. This is consistent with the observations from hPEPT1 cRNA-microinjected oocytes (Liang et al., 1995), hPEPT1-transfected CHO cells (Covitz et al., 1996), and Caco-2 cells (Thwaites et al., 1993a). However, direct evidence for proton-dependent transport has mainly focused on studies with atypical peptides such as Gly-Sar or Gly-Pro (Thwaites et al., 1993b). Recently, several groups investigated the fundamental question of how the peptide transporter interacts with charged substrates because more than 20% of dipeptides and tripeptides carry a net negative or positive charge at physiological pH (Amasheh et al., 1997). It was found that charged substrates interact with the transporters (Temple et al., 1995, 1996; Wenzel et al., 1996; Amasheh et al., 1997; Lister et al., 1997). Interestingly, it has been shown that the H⁺/peptide coupling ratio depends on the net charge of the peptide substrate, that is, the H⁺/peptide coupling ratio is 1 for zwitterionic peptides, 2 for anionic peptides, and 0 for cationic peptides (Temple et al., 1995). Amasheh et al. (1997) has shown the characteristics by which the peptide transporter differentially transports charged dipeptides in XLOs expressing PEPT1. In their study, they used glycyl-L-glutamine as a zwitterionic substrate, glycyl-L-aspartate as an anionic dipeptide, and glycyl-L-lysine as a cationic dipeptide. They concluded that the dipeptide/PEPT1 binding site is affected by both pH and membrane potential. The binding affinity was reduced at high pH for anionic substrates, whereas cationic substrates have reduced binding affinity at pH values ranging from 5.5 to 8.0, the typical pH range of the intestine (Gray, 1996). Furthermore, Lister et al. (1997) studied the influence of luminal pH on the transport of positively and negatively charged dipeptides using an intact preparation of rat small intestine. They have shown that transport of neutral and negatively charged dipeptides was stimulated by lowering the luminal pH to 6.8, whereas increasing the luminal pH to 8.0 strongly stimulated the transport of positively charged dipeptide Phe-Lys but inhibited that of negatively charged dipeptide Phe-Ala. Gly-Sar is a neutral dipeptide at pH 6.0 to 7.0. Our data show that the transport of Gly-Sar is proton dependent and consistent with the results observed by others for this neutral dipeptide (Thwaites et al., 1993a; Liang et al., 1995; Covitz et al., 1996). On the other hand, in the present study, the optimum pH for VACV uptake is 7.5. The uptake is almost 2-fold higher than that at low pH (5.5 and 6.0) and high pH (8.0) (Fig. 4). This result might be explained by its chemical structure (Fig. 7). VACV has three pK_a values equal to 1.90, 7.47, and 9.43. At low pH conditions (6 or lower), VACV would exist primarily as a cationic moiety. As the pH is increased from 6.0, the net cationic charge present on the drug becomes progressively less, reaching an almost neutral state as the extracellular pH approaches 8.0. In the present study, the optimal uptake of VACV occurs at pH 7.5, presumably because it exists predominantly as a mixture of neutral and cationic species at that pH. Also, the decreased uptake of VACV at pH 8.0 may be explained by the instability of VACV at this pH, even though VACV exists as the favorable neutral moiety. VACV degrades to ACV and valine, both of which are not substrates for PEPT1 (Balimane et al., 1998; Han et al., 1998), thus reducing the amount of VACV available for uptake. Therefore, the present study results are consistent with reports of the interactions of positively charged dipeptides with PEPT1 (Temple et al., 1995, 1996; Amasheh et al., 1997; Lister et al., 1997) and the known pH-dependent stability of VACV (Sinko and Balimane, 1998). Also, the apparent lack of VACV/proton cotransport in hPEPT1/CHO cells is consistent with observations in intact tissues.

View larger version (9K): [in this window] [in a new window]   Fig. 7.   Structure of VACV, which is valyl prodrug of acyclic nucleoside ACV, with pKa values equal to 1.90, 7.47, and 9.43.

In conclusion, the results of the present study provide evidence that a nonpeptidic drug VACV, a prodrug of ACV, is a substrate for the human intestinal peptide transporter hPEPT1 expressed in CHO cells. The uptake of VACV was concentration dependent, saturable, and inhibited by other known hPEPT1 competitive substrates. In some cases, mixed-model-type inhibition cannot be ruled out. We have also shown, for the first time, that the optimal uptake of VACV occurs at pH 7.5 and that the apparent lack of proton cotransport is consistent with results in tissues and Caco-2 cells (Lee et al., 1996; Hu and Sinko, 1997; de Vrueh et al., 1998). The basis for this interaction relates to charge and ionization. Collectively, these results suggest that the hPEPT1/CHO cell system is an appropriate in vitro model for investigating interactions with the human intestinal peptide transporter hPEPT1.