Differential recognition of beta -lactam antibiotics by intestinal and renal peptide transporters, PEPT 1 and PEPT 2.

This study was initiated to determine if there are differences in the recognition of β-lactam antibiotics as substrates between intestinal and renal peptide transporters, PEPT 1 and PEPT 2. Reverse transcription-coupled polymerase chain reaction and/or Northern blot analysis have established that the human intestinal cell line Caco-2 expresses PEPT 1 but not PEPT 2, whereas the rat proximal tubule cell line SKPT expresses PEPT 2 but not PEPT 1. Detailed kinetic analysis has provided unequivocal evidence for participation of PEPT 2 in SKPT cells in the transport of the dipeptide glycylsarcosine and the aminocephalosporin cephalexin. The substrate recognition pattern of PEPT 1 and PEPT 2 was studied with cefadroxil (a cephalosporin) and cyclacillin (a penicillin) as model substrates for the peptide transporters constitutively expressed in Caco-2 cells (PEPT 1) and SKPT cells (PEPT 2). Cyclacillin was 9-fold more potent than cefadroxil in competing with glycylsarcosine for uptake via PEPT 1. In contrast, cefadroxil was 13-fold more potent than cyclacillin in competing with the dipeptide for uptake via PEPT 2. The substrate recognition pattern of PEPT 1 and PEPT 2 was also investigated using cloned human peptide transporters functionally expressed in HeLa cells. Expression of PEPT 1 or PEPT 2 in HeLa cells was found to induce H+-coupled cephalexin uptake in these cells. As was the case with Caco-2 cells and SKPT cells, the uptake of glycylsarcosine induced in HeLa cells by PEPT 1 cDNA and PEPT 2 cDNA was inhibitable by cyclacillin and cefadroxil. Again, the PEPT 1 cDNA-induced dipeptide uptake was inhibited more potently by cyclacillin than by cefadroxil, and the PEPT 2 cDNA-induced dipeptide uptake was inhibited more potently by cefadroxil than by cyclacillin. It is concluded that there are marked differences between the intestinal and renal peptide transporters in the recognition of β-lactam antibiotics as substrates.

Peptide transporters are primarily expressed in the small intestine and kidney. The endogenous substrates for these transporters are small peptides consisting of two or three amino acids (1). These transporters function in the absorption of peptides arising from digestion of dietary proteins (small intestine) and in the reabsorption of peptides present in the glomerular filtrate (2)(3)(4). It became apparent that the peptide transporters can serve as carriers for exogenous compounds which bear structural resemblance to the physiologically occurring peptide substrates, when the transport of cephalexin, a ␤-lactam antibiotic, was shown to be mediated in the kidney (5) and small intestine (6) by the peptide transport system. The peptide substrates of the peptide transport system and cephalexin share certain structural features such as a peptide bond with an ␣-amino group and a terminal carboxylic acid group. This structural similarity is apparently the basis for the molecular mimicry, enabling the peptide transporters to accept cephalexin as a substrate. The pharmacological relevance of the peptide transporters became immediately evident from these studies because of the enormous potential of these transporters to serve as carriers for a variety of peptidomimetic drugs. Subsequent studies have indeed identified a wide spectrum of pharmacologically active compounds that are accepted as substrates by the peptide transporters in the intestine and/or kidney (7)(8)(9).
The two organs, the small intestine and the kidney, in which the peptide transporters are primarily expressed play an important role in the therapeutic efficacy of ␤-lactam antibiotics. The intestinal peptide transport system is responsible for the oral absorption of these drugs. The renal peptide transport system, which functions in the reabsorption of these drugs from the glomerular filtrate, enhances the half-life of these drugs in the circulation. Therefore, detailed studies on the interaction of ␤-lactam antibiotics with the peptide transporters in these two organs are vital to the understanding of the pharmacodynamics of these drugs. Recent molecular cloning studies (10 -13) have shown that the peptide transporters expressed in the small intestine and kidney are structurally different. The human intestinal peptide transporter (PEPT 1) and the human kidney peptide transporter (PEPT 2) exhibit only about 50% homology in amino acid sequence. PEPT 1 is expressed primarily in the small intestine and, to a small extent, in the kidney, whereas PEPT 2 is expressed only in the kidney (12). Nonetheless, both transporters accept small peptides as substrates and are driven by a transmembrane electrochemical H ϩ gradient. The current investigation was undertaken to study in detail the interaction of ␤-lactam antibiotics with PEPT 1 and PEPT 2 and to determine whether there are differences between these two transporters in the recognition of these drugs as substrates. P]dCTP (specific radioactivity, 3000 Ci/mmol) was purchased from Amersham. Cell culture media were purchased from Life Technologies, Inc. (Gaithersburg, MD). Fetal bovine serum, dexamethasone, apotransferrin, nigericin, benzylpenicillin, ampicillin, and cephalothin were obtained from Sigma. Cyclacillin, cephalexin, and cefadroxil were generous gifts from Dr. T. Hoshi, University of Shizioka, Shizioka, Japan. The rat renal proximal tubule cell line SKPT was provided by Dr. Ulrich Hopfer, Case Western Reserve University, Cleveland, OH. The human colon carcinoma cell line Caco-2 was obtained from the American Type Culture Collection. All other chemicals were of analytical grade.

Methods
Cell Culture and Uptake Measurements-SKPT cells and Caco-2 cells were cultured in Dulbecco's modified Eagle's/F-12 (1:1) medium and in minimal essential medium, respectively, as described previously (14,15). Uptake of [ 14 C]glycylsarcosine and [ 3 H]cephalexin in cells was measured with the uptake buffer whose composition was 25 mM Hepes/ Tris (pH 7.5) or 25 mM Mes 1 /Tris (pH 6.0), 140 mM NaCl, 5.4 mM KCl, 1.8 mM CaCl 2 , 0.8 mM MgSO 4 , and 5 mM glucose (14 -16). In experiments where uptake buffers of different pH over the range of 6.0 -9.0 were used, two buffers were prepared, one containing 25 mM Mes/Tris (pH 6.0), 140 mM NaCl, 5.4 mM KCl, 1.8 mM CaCl 2 , 0.8 mM MgSO 4 , and 5 mM glucose, and the other containing 25 mM Tris/Hepes (pH 9.0), 140 mM NaCl, 5.4 mM KCl, 1.8 mM CaCl 2 , 0.8 mM MgSO 4 , and 5 mM glucose, and these buffers were mixed to give the desired pH. The culture medium was first aspirated from the dish and the cell monolayer was washed once with 1 ml of the uptake buffer. Uptake was initiated by the addition of 1 ml of the uptake buffer containing radiolabeled substrate. Incubation was continued for the desired time, after which uptake was terminated by the removal of the medium followed by three times washing with ice-cold uptake buffer. The cells were then solubilized with 1 ml of 0.2 M NaOH, 1% SDS and the contents were transferred to a counting vial for determination of radioactivity.
mRNA Isolation and Northern Blot-Poly(A) ϩ RNA was isolated from Caco-2 and SKPT cells using the FastTrack mRNA isolation kit (Invitrogen). The cDNAs encoding PEPT 1 and PEPT 2 were radiolabeled with [ 32 P]dCTP by random priming. Northern blot analysis was carried out under high stringency conditions as described previously (11). The same blot was used for probing with PEPT 1 cDNA and PEPT 2 cDNA by sequential hybridization.
RT-PCR-RNA samples from human intestine, human kidney, and Caco-2 cells were subjected to RT-PCR to determine whether these tissues contain PEPT 1 mRNA and/or PEPT 2 mRNA. The primers specific for PEPT 1 corresponded to nucleotide positions 342-360 (sense) and 1575-1592 (antisense) of the cDNA (11). The primers specific for PEPT 2 corresponded to nucleotide positions 900 -917 (sense) and 1760 -1777 (antisense) of the cDNA (12). The RT-PCR products were analyzed by agarose gel electrophoresis.
Vaccinia Virus Expression of PEPT 1 and PEPT 2-This was done using the procedure described earlier (11)(12)(13). Subconfluent HeLa cells in 24-well culture plates were first infected with a recombinant vaccinia virus VTF 7-3 which carries the gene for T7 RNA polymerase as a part of its genome. This enables the HeLa cells to express T7 RNA polymerase. Following the infection, the cells were transfected with pBluescript-PEPT 1 cDNA construct or with pBluescript-PEPT 2 cDNA construct. In these constructs, the cDNAs were under control of T7 promotor in the plasmid. Cells transfected with empty plasmid served as control. Transfection was mediated by lipofection (17). The virus-encoded T7 RNA polymerase catalyzes the transcription of the cDNA, allowing transient expression of the PEPT 1 or PEPT 2 protein in the HeLa cell plasma membrane. After 12 h post-infection, transport measurements were made at room temperature with [ 14 C]glycylsarcosine or with [ 3 H]cephalexin. The uptake medium was 25 mM Mes/Tris (pH 6.0) or 25 mM Tris/Hepes (pH 9.0), containing 140 mM NaCl, 5.4 mM KCl, 1.8 mM CaCl 2 , 0.8 mM MgSO 4 , and 5 mM glucose. The time of incubation for uptake measurements was 3 min. At the end of the incubation, the uptake was terminated by removal of the uptake medium by aspiration followed by washing two times with ice-cold uptake medium which did not contain radiolabeled substrate. The cells were then solubilized with 0.5 ml of 1% SDS, transferred to counting vials, and used for determination of radioactivity.
Data Analysis-The kinetic parameters, the Michaelis-Menten con-stant, K t , and the maximal velocity, V max , were calculated by linear regression of the Eadie-Hofstee plot and confirmed by nonlinear regression methods using the Fig. P (18).

RESULTS AND DISCUSSION
Differential Expression of PEPT 1 and PEPT 2 in Caco-2 and SKPT Cells-Mammalian intestine expresses PEPT 1 and not PEPT 2. Mammalian kidney expresses both PEPT 1 and PEPT 2, but the predominant one is PEPT 2. One of the goals of the present study was to investigate the interaction of ␤-lactam antibiotics, a group of peptidomimetic drugs with pharmacological and clinical importance, with PEPT 1 and PEPT 2, using cultured intestinal and renal cell lines. Caco-2 cells, which are of human intestinal origin, are known to possess a low-affinity H ϩ /peptide cotransporter, resembling PEPT 1 (15). SKPT cells, which are of rat kidney origin, have been recently demonstrated to express a high-affinity H ϩ /peptide cotransporter, resembling PEPT 2 (14). Even though the MDCK renal cell line possesses a peptide transport system, available kinetic data suggest that the transport system is most likely PEPT 1 rather than PEPT 2 (19). Therefore, Caco-2 cells and SKPT cells were used in the present study.
To establish unequivocally that Caco-2 cells express PEPT 1 and SKPT cells express PEPT 2, we performed the following experiments. We determined the identity of the peptide transporter present in Caco-2 cells by RT-PCR using PEPT 1-and PEPT 2-specific primers (11,12). The specificity of each pair of primers was established by PCR using respective cDNAs as templates. RNA samples prepared from human intestine, human kidney, and Caco-2 cells were subjected to RT-PCR using these primers and the products were analyzed by agarose gel electrophoresis. The results of these experiments, given in Fig.  1, show that the PEPT 1-specific PCR product of expected size (ϳ1.2 kb) was generated from all three RNA samples. In contrast, the PEPT 2-specific PCR product (ϳ0.9 kb in size) was generated only from kidney RNA. RNA samples from Caco-2 cells and intestine were negative for this product (Fig. 1). These data demonstrate that Caco-2 cells express PEPT 1 and not PEPT 2.
The nucleotide sequences of the rat homologs of PEPT 1 and PEPT 2 have not yet been determined. Therefore, we investigated the expression of PEPT 1/PEPT 2 in the SKPT cell line by Northern blot hybridization using the human PEPT 1 and PEPT 2 cDNAs as probes (Fig. 2). Poly(A) ϩ RNA prepared from SKPT cells and Caco-2 cells was size-fractionated and probed 1 The abbreviations used are: Mes, 4-morpholineethanesulfonic acid; RT-PCR, reverse transcriptase-polymerase chain reaction; kb, kilobase(s). RNA samples isolated from human intestine, human kidney, and Caco-2 cells were subjected to RT-PCR using PEPT 1-and PEPT 2-specific primers. The RT-PCR products were analyzed by agarose gel electrophoresis. The expected size of the product was 1.2 kb in the case of PEPT 1 and 0.9 kb in the case of PEPT 2. with these cDNAs. With the PEPT 1 probe, the presence of a major hybridizing band, 3.1 kb in size, was evident in Caco-2 cells, but this band was absent in SKPT cells. With the PEPT 2 probe, there was a primary hybridizing band, 4.2 kb in size, in SKPT cells. This signal was absent in Caco-2 cells. These results indicate that the peptide transporter expressed in SKPT cells is PEPT 2. PEPT 1, which is expressed in Caco-2 cells, is not present in SKPT cells.
Interaction of ␤-Lactam Antibiotics with PEPT 2 Expressed in SKPT Cells-Having established that the SKPT cells functionally express the kidney-specific PEPT 2, we investigated the interaction of ␤-lactam antibiotics with PEPT 2 using these cells. Initial studies were done by assessing the ability of several unlabeled ␤-lactam antibiotics to inhibit the uptake of radiolabeled glycylsarcosine, a dipeptide substrate for peptide transporters. In these experiments, uptake measurements were made at pH 6.0. Under these conditions, there exists a medium-to-cell H ϩ gradient (pH of the cell cytoplasm is ϳ7.4), which provides the driving force for PEPT 2, a H ϩ -dependent transporter. We have used three cephalosporins (cephalexin, cephalothin, and cefadroxil) and three penicillins (cyclacillin, ampicillin, and benzylpenicillin) (Fig. 3). These experiments have shown that cefadroxil, cyclacillin, and cephalexin interact with PEPT 2 with high affinity. The IC 50 values (i.e. concentration necessary to cause 50% inhibition of glycylsarcosine uptake) for these antibiotics were 3.0 Ϯ 0.2, 42 Ϯ 2, and 73 Ϯ 6 M, respectively. The affinity of PEPT 2 for ampicillin, cephalothin, and benzylpenicillin was comparatively low, the IC 50 values being 1.3 Ϯ 0.2, 7.5 Ϯ 2.0, and Ͼ10 mM, respectively.
We then investigated the kinetics of inhibition of glycylsarcosine uptake by cefadroxil (a cephalosporin) and cyclacillin (a penicillin). The presence of the antibiotics decreased the affinity of PEPT 2 for glycylsarcosine, without affecting the maximal velocity (Fig. 4). The Michaelis-Menten constant (K t ) for glycylsarcosine in the absence of the antibiotics was 48 Ϯ 4 M. This value was increased 3.3-fold to 156 Ϯ 17 M in the presence of 40 M cyclacillin. Similarly, the K t value was increased 2.3-fold to 112 Ϯ 3 M in the presence of 3 M cefadroxil. Therefore, the ␤-lactam antibiotics and the dipeptide substrates apparently compete for the same binding site on PEPT 2.
We also employed radiolabeled cephalexin as a substrate to study the interaction of ␤-lactam antibiotics with PEPT 2 in SKPT cells. Our initial characterization studies have established that the uptake of cephalexin in these cells was stimulated by an inwardly directed H ϩ gradient (Fig. 5). The uptake of the antibiotic (0.1 M) was stimulated severalfold upon acidification of the extracellular medium. However, this stimulation was abolished when the cells were pH-clamped (i.e. intracellular pH ϭ extracellular pH) with nigericin. Kinetic analysis, done over a cephalexin concentration of 25-250 M, showed that the uptake occurred via a single, saturable process (Fig. 6). The Michaelis-Menten constant (K t ) for the uptake process was 49 Ϯ 8 M and the maximal velocity (V max ) was 1.5 Ϯ 0.1 nmol/mg of protein/20 min.
We then performed detailed kinetic studies to establish that the uptake of the dipeptide glycylsarcosine and the uptake of the ␤-lactam antibiotic cephalexin occur via a common transport system (i.e. PEPT 2) in SKPT cells. The experimental approach employed here for this purpose is the so-called "A-B-C test" which is widely used in the transport field (20 -22). We systematically investigated the interaction between glycylsarcosine and cephalexin during uptake in SKPT cells with the  primary aim to determine whether the uptake characteristics of these two compounds meet the criteria of the A-B-C test. The results of these experiments have shown that the uptake of glycylsarcosine was completely inhibitable by cephalexin. Similarly, the uptake of cephalexin was completely inhibitable by glycylsarcosine. The interaction between the two compounds during uptake was strictly competitive. The results of the kinetic experiments are summarized in Table I. The K t value for glycylsarcosine determined from its uptake was 48 Ϯ 4 M which is approximately the same as the K i value (64 Ϯ 4 M) for the inhibition of cephalexin uptake by glycylsarcosine. Similarly, the K t value for cephalexin determined from its uptake was 49 Ϯ 8 M which is very close to the K i value (68 Ϯ 5 M) for the inhibition of glycylsarcosine uptake by cephalexin. Cyclacillin inhibited the uptake of glycylsarcosine and the uptake of cephalexin with similar potency, the K i values being 39 Ϯ 1 M and 37 Ϯ 5 M, respectively. Cefadroxil also inhibited the uptake of these two compounds with similar potency, the K i values being 2.8 Ϯ 0.2 M and 2.5 Ϯ 0.1 M, respectively. The conclusion from these experiments is that the uptake characteristics of glycylsarcosine and cephalexin strictly meet every requirement of the classical A-B-C test, thus strongly indicating that these two compounds are transported by the same transporter (i.e. PEPT 2) in the SKPT renal cell line.
Differential Recognition of ␤-Lactam Antibiotics by PEPT 1 and PEPT 2-Cephalosporins as well as penicillins are ␤-lactam antibiotics, possessing peptide-like chemical structures. The basic structure of these compounds resembles the backbone of a tripeptide in which the C-terminal peptide bond is located in the ␤-lactam ring. The free carboxylic acid group which constitutes the C terminus is present in the dihydrothiazine ring in the case of cephalosporins and in the thiazolidine ring in the case of penicillins. Our results show that some penicillins such as cyclacillin interact with the renal peptide transporter with greater affinity than some cephalosporins such as cephalexin (Fig. 3). At the same time, there are also examples where certain cephalosporins (e.g. cefadroxil) are better substrates for this transporter than cyclacillin (Fig. 3). This indicates that the dihydrothiazine ring in the cephalosporins and the thiazolidine ring in the penicillins are not differentiated to any significant extent by the renal peptide transporter. This conclusion is in contrast to the previously held notion that penicillins in general have much lower affinity than cephalosporins for the transporter (23).
Since it has now become clear from molecular biological studies that the intestinal and renal peptide transporters are distinct proteins with significant differences in their primary structure (10 -13), we initiated studies to see if there are differences in the substrate recognition pattern between these two transporters. In our studies to compare the substrate recognition pattern of PEPT 1 and PEPT 2, we selected cefadroxil (a cephalosporin) and cyclacillin (a penicillin) as model substrates. Initially, we carried out the experiments with Caco-2 cells (PEPT 1) and SKPT cells (PEPT 2) by determining the relative potency of these two peptidomimetic drugs for the inhibition of the uptake of the dipeptide glycylsarcosine. In Caco-2 cells, the dipeptide uptake was inhibited by cefadroxil and by cyclacillin in a dose-dependent manner (Fig. 7A). The respective IC 50 values for the inhibition were 5.4 Ϯ 0.6 and 0.6 Ϯ 0.1 mM. Thus, cyclacillin is severalfold more potent than cefadroxil in competing with glycylsarcosine for uptake via  PEPT 1. Interestingly, even though the uptake of glycylsarcosine in SKPT cells was inhibited by both cyclacillin and cefadroxil as in Caco-2 cells, there were important differences (Fig. 7B). The potency with which these drugs inhibited the uptake in SKPT cells was much greater than in Caco-2 cells.
The IC 50 values were in the micromolar range rather than in millimolar range. In addition, there was a significant difference in the relative inhibitory potency between the two drugs. The IC 50 values for cefadroxil and cyclacillin in SKPT cells were 3.0 Ϯ 0.2 and 41.6 Ϯ 1.5 M, respectively. In other words, cefadroxil is manyfold more potent than cyclacillin in competing with the dipeptide for uptake via PEPT 2. Thus, the relative affinities of PEPT 1 and PEPT 2 for cyclacillin and cefadroxil are reversed. These results show that the substrate recognition pattern is significantly different between PEPT 1 and PEPT 2.
To rule out the possibility that the observed differences in substrate recognition of the peptide transporters between these two cell lines may be due to species differences rather than real differences between PEPT 1 and PEPT 2, we performed similar experiments with the cloned human PEPT 1 and human PEPT 2. These two transporters were functionally expressed in HeLa cells using the vaccinia virus expression system and the uptake of cephalexin was determined using 0.5 M [ 3 H]cephalexin. As shown in Fig. 8A, cephalexin uptake measured at pH 6.0 in HeLa cells expressing PEPT 1 was 1.89 Ϯ 0.05 pmol/10 6 cells/10 min which was 11-fold greater than the uptake in control cells (i.e. HeLa cells transfected with empty vector) under similar conditions. The H ϩ gradient-dependent nature of the uptake process was evident from the findings that the uptake decreased drastically in PEPT 1-expressing cells when measured at pH 9.0 instead of pH 6.0. Similar results were obtained with PEPT 2 (Fig. 8B). These data demonstrate that PEPT 1 and PEPT 2 catalyze the transport of the ␤-lactam antibiotic cephalexin. We then used this experimental system to compare the substrate recognition pattern of PEPT 1 and PEPT 2. These transporters were individually expressed in HeLa cells and the uptake of glycylsarcosine was determined in the presence of increasing concentrations of cyclacillin and cefadroxil. As was the case in Caco-2 cells and SKPT cells, the dipeptide uptake via PEPT 1 and PEPT 2 was inhibited by both cyclacillin and cefadroxil. In the case of PEPT 1, the IC 50 values for cefadroxil and cyclacillin were 0.87 Ϯ 0.12 and 0.35 Ϯ 0.09 mM, respectively (Fig. 9A). In contrast, the corresponding IC 50 values were 66 Ϯ 4 and 610 Ϯ 100 M in the case of PEPT 2 (Fig. 9B). These data with the cloned human PEPT 1 and PEPT 2 show that there are significant differences between the two peptide transporters in substrate selectivity.
Significant differences were noted in relative potency of cefadroxil and cyclacillin as inhibitors of glycylsarcosine uptake in Caco-2 and SKPT cells which express PEPT 1 and PEPT 2 natively and in HeLa cells which express the cloned human PEPT 1 and PEPT 2. It is possible that post-translational

FIG. 7. Differential recognition of cefadroxil and cyclacillin by PEPT 1 in Caco-2 cells and PEPT 2 in SKPT cells. Uptake of [ 14 C]
Gly-Sar was measured at pH 6.0 in Caco-2 cells (A) and in SKPT cells (B) in the absence and presence of increasing concentrations of cefadroxil (q) and cyclacillin (E). Concentration of Gly-Sar was 5 M for Caco-2 cells and 3 M for SKPT cells. Incubation time for uptake measurement was 10 min for both cell types. Uptake of Gly-Sar measured in the absence of inhibitors was taken as 100%. This value was 81.6 Ϯ 8.8 pmol/mg of protein/10 min for Caco-2 cells and 53.7 Ϯ 3.1 pmol/mg of protein/10 min for SKPT cells.
modifications (e.g. N-glycosylation) of PEPT 1 and PEPT 2 expressed in HeLa cells are not identical to those of the native transporters in Caco-2 and SKPT cells. This may contribute to the observed differences. With respect to PEPT 2, species differences may also be a factor because SKPT cells were derived from rat kidney whereas the PEPT 2 cDNA was cloned from human kidney.
In addition to PEPT 1 and PEPT 2, mammalian tissues may express other peptide transporters. A cDNA clone (HPT-1) has been recently isolated from a Caco-2 cell cDNA library and expression of this cDNA in mammalian cells leads to increased uptake of the peptidomimetic drugs cephalexin and bestatin (24). Interestingly, there is no sequence homology between the HPT-1 protein and the peptide transporters PEPT 1 and PEPT 2. It has also been shown that Caco-2 cells express two functionally distinct peptide transporters, one in the apical membrane and the other in the basolateral membrane (25). Among these multiple peptide transporters, PEPT 1 and PEPT 2 have been characterized in detail, both at the functional level and at the molecular level. The present study, which focuses on the handling of peptidomimetic drugs by PEPT 1 and PEPT 2, documents a major functional difference between these two transporters in terms of recognition of ␤-lactam antibiotics as substrates. FIG. 9. Differential recognition of cefadroxil and cyclacillin by human PEPT 1 and human PEPT 2 functionally expressed in HeLa cells. Cells were transfected either with human PEPT 1 cDNA (A) or with human PEPT 2 cDNA (B). The cDNAs were functionally expressed in these cells by the vaccinia virus expression technique. Uptake of [ 14 C]Gly-Sar was measured at pH 6.0 with a 3-min incubation in the absence and presence of increasing concentrations of cefadroxil (q) and cyclacillin (E). Concentration of Gly-Sar was 25 M for cells expressing PEPT 1 and 50 M for cells expressing PEPT 2.