Search tips
Search criteria 


Logo of jcbfmJournal of Cerebral Blood Flow & Metabolism
J Cereb Blood Flow Metab. 2011 January; 31(1): 250–261.
Published online 2010 June 23. doi:  10.1038/jcbfm.2010.84
PMCID: PMC2965812

Distribution of glycylsarcosine and cefadroxil among cerebrospinal fluid, choroid plexus, and brain parenchyma after intracerebroventricular injection is markedly different between wild-type and Pept2 null mice


The purpose of this study was to define the cerebrospinal fluid (CSF) clearance kinetics, choroid plexus uptake, and parenchymal penetration of PEPT2 substrates in different regions of the brain after intracerebroventricular administration. To accomplish these objectives, we performed biodistribution studies using [14C]glycylsarcosine (GlySar) and [3H]cefadroxil, along with quantitative autoradiography of [14C]GlySar, in wild-type and Pept2 null mice. We found that PEPT2 deletion markedly reduced the uptake of GlySar and cefadroxil in choroid plexuses at 60 mins by 94% and 82% (P<0.001), respectively, and lowered their CSF clearances by about fourfold. Autoradiography showed that GlySar concentrations in the lateral, third, and fourth ventricle choroid plexuses were higher in wild-type as compared with Pept2 null mice (P<0.01). Uptake of GlySar by the ependymal–subependymal layer and septal region was higher in wild-type than in null mice, but the half-distance of penetration into parenchyma was significantly less in wild-type mice. The latter is probably because of the clearance of GlySar from interstitial fluid by brain cells expressing PEPT2, which stops further penetration. These studies show that PEPT2 knockout can significantly modify the spatial distribution of GlySar and cefadroxil (and presumably other peptides/mimetics and peptide-like drugs) in brain.

Keywords: autoradiography, blood–CSF barrier, disposition, ependyma, mannitol, mice


The successful development of drugs to treat diseases of the central nervous system will require adequate transport of these agents across the blood–brain and/or blood–cerebrospinal fluid barriers (i.e., BBB and BCSFB) (Taylor, 2002; Begley, 2004; Ohtsuki and Terasaki, 2007). These barrier systems probably arose from the necessity to maintain strict homeostasis of the extracellular fluid and CSF for normal brain function. Although originally conceived as a physical barrier formed by the linking tight junctions of brain capillary endothelial cells and epithelial cells of the choroid plexus and arachnoid membrane, the BBB and BCSFB also include degradative enzymes as well as efflux transporters that can affect the delivery of these agents to specific regions in brain (Smith et al, 2004). This is important not only for drug delivery but also for endogenous neuropeptides that are released within the brain. Beyond the BBB and BCSFB, the pharmacology of drugs or the physiology of neuromodulating peptides depends on their movement through the interstitial fluid into adjacent brain tissue, their uptake and binding by brain cells, their exchange between brain and CSF at the ependyma and glia limitans, and their distribution through the ventricles and subarachnoid space.

PEPT2 (SLC15A2), a high-affinity, low-capacity carrier of the proton-coupled oligopeptide transporter family, facilitates the movement of substrates across biologic membranes through an electrochemical membrane gradient (Brandsch et al, 2008; Rubio-Aliaga and Daniel, 2008). PEPT2 is the most studied protein of the proton-coupled oligopeptide transporter family found in brain where it is expressed at the apical surface of choroid plexus epithelia (i.e., CSF-facing), as well as in astrocytes (newborn rats) and neuronal cells (newborn and adult rats) (Shen et al, 2004). Residing at the blood–CSF interface, PEPT2 has a significant role in limiting the exposure of peptides/mimetics and peptide-like drugs in CSF by transporting these substrates into choroid plexus tissue (Shu et al, 2002; Teuscher et al, 2004; Shen et al, 2005). In accordance with this, the synthetic dipeptide glycylsarcosine (GlySar) (Ocheltree et al, 2005), the aminocephalosporin antibiotic cefadroxil (Shen et al, 2007), and the naturally occurring dipeptide carnosine (Kamal et al, 2009) were found to have reduced CSF levels in vivo after intravenous bolus administration, but increased choroid plexus levels in wild-type mice as compared with Pept2 null animals. Moreover, pharmacologic differences linked to PEPT2 expression in brain have been observed. For instance, this protein protects against 5-aminolevulinic acid (5-ALA) neurotoxicity after subcutaneous dosing in wild-type mice (Hu et al, 2007) and produces dissimilar antinociceptive responses between wild-type and Pept2 null mice after intracerebroventricular (icv) administration of kyotorphin (Jiang et al, 2009). Such findings show that PEPT2 expression in brain can have profound effects on both the disposition and dynamics of peptide/mimetic substrates and peptide-like drugs.

In a study of convectional and diffusional movement in the CSF–brain system, the intracranial distribution of [14C]sucrose after icv injection was examined in rats by quantitative autoradiographic analysis (Ghersi-Egea et al, 1996). The resulting data indicated that the circulatory and delivery functions of CSF were complex and varied among CSF compartments. Rapid convective (bulk) flow of CSF and CSF-entrained [14C]sucrose were observed throughout the ventricular system and through the velum interpositum and superior medullary velum into subarachnoid cisterns, but such flow was rather slow and marginal to and over the cerebral and cerebellar cortices. Results also indicated a CSF–brain barrier to sucrose diffusion in some areas of the glia limitans but not ependyma. Sucrose was also found to accumulate in and/or around the blood vessels within the subarachnoid cisterns, suggesting that the pia-arachnoid investing such vessels is different than that at cerebral and cerebellar cortical surfaces.

The preceding pharmacologic and physiologic findings are limited and indicate how little is known about the effects of specific transporters on the circulation and distribution of drugs or endogenous peptide/mimetic substrates in CSF and brain parenchyma. This is not surprising because the first mammalian ion-coupled drug transporters were cloned around the early mid-1990s (Fei et al, 1994; Gründemann et al, 1994) and genetically modified mice were developed even more recently (Schinkel et al, 1994; Jonker et al, 2002; Shen et al, 2003). In view of this, the objectives of this study were to (1) define the CSF clearance kinetics of PEPT2 substrates (i.e., GlySar and cefadroxil) after icv administration, (2) determine if regional differences exist in the choroid plexus uptake of GlySar across the lateral, third, and fourth ventricles, and (3) evaluate how PEPT2 might alter the penetration of GlySar from CSF into parenchymal tissue in different regions of the brain. To address these objectives, we performed experiments in wild-type and Pept2 null mice, a unique resource to evaluate the role of a single gene on drug disposition and dynamics (Kamal et al, 2008); in addition, quantitative autoradiography (QAR) was applied for the accurate localization of radioactive drug within parts of the CSF system and specific regions of brain tissue.

Materials and methods


For the CSF clearance studies and autoradiography of dipeptide, [14C]GlySar (106 mCi/mmol; purity 97.3%) was obtained from Amersham Life Science (Piscataway, NJ, USA) and [3H]mannitol (20 mCi/mmol; purity 99.0%) from American Radiolabeled Chemicals (St Louis, MO, USA). For the CSF clearance studies of peptide-like drug, [3H]cefadroxil (1 Ci/mmol; purity 98.7%) and [14C]mannitol (53 mCi/mmol; purity 99.8%) were obtained from Moravek Biochemicals (Brea, CA, USA). Unlabeled GlySar and cefadroxil were obtained from Sigma-Aldrich (St Louis, MO, USA), and hyamine hydroxide from MP Biomedicals (Irvine, CA, USA). All other chemicals were obtained from standard sources.

The proposed studies were conducted in wild-type (Pept2+/+) and null (Pept2−/−) mice that were generated in-house (Shen et al, 2003). The mice were gender- and weight-matched (8 to 10 weeks of age), and congenic (>99% C57BL/6 genetic background). The animals were kept in a temperature-controlled environment of 12-h light and dark cycles, with access to a standard diet and water ad libitum. Studies were performed in accordance with guidelines from the National Institutes of Health for the care and use of animals, and were approved by the Institutional Animal Care and Use Committee of the University of Michigan, Ann Arbor.

CSF Clearance Studies

After the mice were anesthetized with sodium pentobarbital (50 mg/kg intraperitoneally), their heads were positioned in a stereotaxic device. Using a 1.0 μL syringe (SGE 26 gauge (0.47 mm), 70 mm length; World Precision Instruments, Sarasota, FL, USA), 30-sec icv infusions were made into the right lateral ventricles through a hole drilled in the skull (2.5 mm in depth, 1.2 mm lateral from midline, 0.4 mm caudal from bregma). The infusate contained substrate in 1 μL of artificial CSF buffer (127 mmol/L NaCl, 20 mmol/L NaHCO3, 2.4 mmol/L KCl, 0.5 mmol/L KH2PO4, 1.1 mmol/L CaCl2, 0.85 mmol/L MgCl2, 0.5 mmol/L Na2SO4, and 5.0 mmol/L glucose (pH 7.4)). After completing the infusion, the needle remained in place until sampling and termination of the experiment. [14C]GlySar (0.02 μCi) was dosed at 190 pmol, in the absence or presence of excess unlabeled GlySar. The infusate also contained [3H]mannitol (0.1 μCi). In other experiments, [3H]cefadroxil (0.08 μCi) was dosed at 80 pmol along with [14C]mannitol (0.02 μCi).

At designated times, a single sample of CSF (approximately 5 μL) was obtained from the cisterna magna of each mouse. The animal was decapitated immediately thereafter, and the lateral and fourth ventricle choroid plexuses harvested. The tissue samples were blotted dry, weighed, and digested in 0.5 mL of 1 mol/L hyamine hydroxide (a tissue solubilizer) for 24 h at 37°C. Ecolite(+) liquid scintillation cocktail (ICN, Irvine, CA, USA) was added to the solubilized tissue and CSF samples. The radioactivity of each sample was measured by a dual-channel liquid scintillation counter (Beckman LS 3801; Beckman Coulter, Fullerton, CA, USA). The initial dose of GlySar (or cefadroxil) remaining in CSF (%) was calculated as

equation image

where INJ is injectate. Mannitol was used to adjust the GlySar and cefadroxil data for variations among experiments in the amount infused, passive diffusion, and CSF turnover. The uptake of GlySar (or cefadroxil) in choroid plexus (μL/mg) was calculated as

equation image

where CP is choroid plexus. The mannitol data in equation (2) corrects for the extracellular content of GlySar in the choroid plexus sample.

Autoradiography Studies and Image Analysis

Thirty-sec icv infusions of [14C]GlySar (0.02 μCi), without mannitol, were made as described above. The mice were decapitated 60 mins after the infusion of dipeptide, and the heads immediately frozen in 2-methylbutane cooled to −40°C using dry ice (Ghersi-Egea et al, 1996; Nagaraja et al, 2005). The brains frozen in situ were carefully dissected in a chest freezer without allowing them to thaw. Such handling has been shown to preserve intact the intracranial fluid compartments and to minimize the postmortem movement of injected radiotracer (Bereczki et al, 1992; Ghersi-Egea et al, 1996; Nagaraja et al, 2005). The dissected brains were immediately covered with an embedding matrix (M1 matrix; Lipshaw, Pittsburgh, PA, USA) and stored frozen in individual plastic bags until sectioning.

Each brain was affixed onto a holder using the embedding matrix on its caudal/cerebellar end. The holder was placed in a cryostat (Microm, Walldorf, Germany) set to −18°C and allowed to equilibrate for 30 mins. The brain was then cut in blocks of five 20-μm-thick coronal sections beginning from the rostral end (Knight et al, 2005). Of the five sections in every block, the first and fifth were placed on serially numbered slides and instantly dried on a hotplate at 60°C; the middle three were placed on serially numbered individual coverslips and dried. This sectioning pattern continued through most of the forebrain (n=4). After initial analysis, sectioning was subsequently extended until the end of the cerebellum (end of fourth ventricle) was reached (n=4). Sections placed on slides underwent de- and rehydration in an alcohol and xylene series and were then Nissl stained. The coverslips bearing brain sections were glued in numerical order to art-mounting boards (Crescent, Wheeling, IL, USA) and, along with a strip of 14C standards, exposed to a sheet of Kodak MR2000 X-ray film (Eastman Kodak, Rochester, NY, USA). The films were developed after approximately 25 days of exposure to obtain autoradiograms (ARGs) of radiotracer disposition.

Autoradiograms were digitized using an image analysis system (Model AIS; Imaging Research, St Catharines, Ontario, Canada) (Knight et al, 2005). A standard curve of radioactivity versus optical density was constructed for each film using the 14C standards. Images of brain sections were analyzed with respect to the injection site in each experiment. Radioactivity was quantified in several brain regions of interest by measuring their optical density and converting those values into radioactivity units (nCi/g) using the standard curve. Values obtained were averaged across the three sections representing one tissue block and then summed over the distance a given structure traversed along the rostrocaudal axis to get total radioactivity for that particular structure or region. After this method, radioactivity was quantified for choroid plexuses in different ventricles, ependyma (which also included the subependyma) and the brain midline structure of triangular septal nucleus (hereafter called the septal region) that lies between the two lateral ventricles. In addition, using the image analysis system, a horizontal bar (transect) was drawn across the coronal image representing the icv injection site, and the radioactivity along its entire length was plotted to evaluate the spread (diffusion or penetration) of radioactivity from the ventricles across the ependyma into brain parenchyma. The distance for tissue radioactivity to decline to 50% of peak ependymal–subependymal levels was determined as was the distance for a further 50% decline. These values were averaged (for both ipsi- and contralateral hemispheres) to give an empirical measure of isotope penetration (D1/2). Of note, the ependymal–subependymal layer is approximately 25 μm thick, extends to the neuropil, and includes ependymal, neuronal, and glial cell bodies, neuronal and glial processes, and numerous capillaries. Because the spatial resolving power of QAR with 14C-radiation is around 25 μm, the measurements of concentration in ependymal–subependymal layer in this study are at the limit of this resolution. In the following text, figures, and their legends, the ependymal–subependymal layer will simply be referred to as the ependyma.


Data are reported as mean±s.e. Statistical comparisons between wild-type and Pept2 null mice were performed by a two sample t-test (Prism v5.0; GraphPad Software, San Diego, CA, USA) in which a probability of P≤0.05 was considered significant. Statistical differences between multiple treatment groups were determined by analysis of variance and pairwise comparisons were made using the Tukey test. Linear regression analyses were also performed with Prism software.


CSF Clearance and Choroid Plexus Accumulation of GlySar in Wild-Type and Pept2 Null Mice

A comparison of results from the two mouse genotypes indicates that PEPT2 ablation had a dramatic effect on the persistence of [14C]GlySar in CSF after icv injections. The clearance of GlySar from CSF was 3.6 times faster in wild-type mice than in Pept2 null animals (Figure 1A; half-lives of 23.6 and 84.5 mins, respectively). The change in half-life was specific for PEPT2-mediated clearance because the coinfused mannitol data, as described in equation (1), corrects for any potential differences in passive diffusion and CSF turnover between the two genotypes. There was also a time-dependent accumulation of [14C]GlySar in the choroid plexuses of wild-type mice after icv injections (Figure 1B). In contrast, dipeptide concentrations were relatively low in the choroid plexuses of Pept2 null mice, declined slowly over time, and were significantly less than the values observed in wild-type animals.

Figure 1
Cerebrospinal fluid (CSF) clearance (A) and choroid plexus uptake (B) of [14C]glycylsarcosine (GlySar) in wild-type and Pept2 null mice after a 30-sec intracerebroventricular (icv) infusion of 0.02 μCi (190 pmol) dipeptide. For ...

To further examine the role of PEPT2 on the CSF efflux of [14C]GlySar and its uptake into choroid plexus tissue, we examined the effect of coinjecting different concentrations of unlabeled GlySar (0 to 20 mmol/L). In these self-inhibition experiments, the CSF and choroid plexuses were sampled 15 mins after an icv injection. As shown in Figure 2A, there was a dose-dependent and significant increase in percent of [14C]GlySar remaining in the CSF of wild-type mice; at the highest concentration of unlabeled GlySar (20 mmol/L), radiotracer retention in the CSF was 57% higher than that observed at 0 mmol/L unlabeled dipeptide. In contrast, the percent of [14C]GlySar remaining in CSF of Pept2 null mice was not consistently different among the concentration groups (Figure 2A); the observed 16% increase in CSF retention of radiotracer at 15 mins between the unlabeled GlySar treatments at 0 versus 20 mmol/L was not statistically significant. At each concentration, the percent of [14C]GlySar in CSF was different between genotypes with the Pept2 null mice having higher values. Of note, the retention of [14C]GlySar in CSF appeared to reach a plateau at approximately 70% for the wild-type mice and at approximately 80% for the Pept2 null mice. This could be the result of the expression of another dipeptide transporter in the choroid plexus–CSF–brain system. For the choroid plexus, a dose-dependent inhibition of [14C]GlySar uptake by unlabeled dipeptide was found in wild-type, but not in Pept2 null, mice (Figure 2B). In contrast with the CSF retention results, the choroidal uptake of [14C]GlySar does not appear to reach a limiting value at 20 mmol/L unlabeled GlySar concentration (in the infusate) of the wild-type mice.

Figure 2
Cerebrospinal fluid (CSF) retention (A) and choroid plexus uptake (B) of [14C]glycylsarcosine (GlySar; ±0 to 20 mmol/L unlabeled GlySar) in wild-type and Pept2 null mice, 15 mins after a 30-sec intracerebroventricular (icv) infusion ...

CSF Clearance of Cefadroxil in Wild-Type and Pept2 Null Mice

Cefadroxil was also chosen for study because, as a therapeutic agent and PEPT2 substrate, the influence of PEPT2 ablation on its transport at the CSF–choroid plexus would have more clinical relevance. The percent of [3H]cefadroxil remaining in CSF was less in wild-type than in Pept2 null mice at 15, 30, and 60 mins after icv infusion (Figure 3A), and its half-life was fourfold shorter in the former (46 mins) than in the latter (188 mins). Accordingly, as with GlySar, [3H]cefadroxil was cleared from CSF more rapidly in wild-type than in Pept2 null mice. Because PEPT2 mediates dipeptide transport from CSF to choroid plexus, the choroidal concentrations of cefadroxil were significantly lower (three- to fourfold) in Pept2 null mice than in wild-type animals at all sampling times (Figure 3B).

Figure 3
Cerebrospinal fluid (CSF) clearance (A) and choroid plexus uptake (B) of [3H]cefadroxil in wild-type and Pept2 null mice after a 30-sec intracerebroventricular (icv) infusion of 0.08 μCi (80 pmol) β-lactam antibiotic. For ...

Distribution of GlySar in the Brain Parenchyma of Wild-Type and Pept2 Null Mice

The GlySar data, along with those of cefadroxil, showed that the expression of PEPT2 by the choroid plexus had a profound effect on substrate concentration and availability in circulating CSF. The possible impact of different CSF concentrations on GlySar uptake by brain parenchyma after icv infusions was examined in both wild-type and Pept2 mice by QAR. Autoradiograms of [14C]GlySar distribution were made at a single time, 60 mins after a 30-sec icv infusion. Representative ARGs and their associated Nissl histologies from wild-type and Pept2 null mice are shown in Figure 4. In this figure, the coronal ARGs were at the level of the injection site (Figures 4A-1 and 4B-1) and at the level of the fourth ventricle (cerebellum/brain stem; Figures 4D-1 and 4E-1); the corresponding histologies are shown in Figures 4A-2, 4B-2, 4D-2, and 4E-2, respectively. Such ARGs were used to examine various parameters of the spatial distribution of GlySar including (1) the concentration of [14C]GlySar (nCi/g) in the lateral, third, and fourth ventricle choroid plexuses; (2) the concentration of [14C]GlySar in the ipsilateral and contralateral ependyma at the level of the injection site; (3) the concentration of [14C]GlySar in the septal region, an area identified as having high radioactivity in wild-type mouse ARGs; (4) the lateral spread of [14C]GlySar into the brain parenchyma along the transect (Figures 4A-1 and 4B-1) as demarcated by the distance for the concentration to fall by one-half (Figure 4C); and (5) the rostrocaudal distribution of [14C]GlySar, which was assessed by determining the total radioactivity in each coronal section excluding that in the choroid plexuses.

Figure 4
Examples of coronal autoradiograms at the level of the injection site (A-1 and B-1) and the fourth ventricle (D-1 and E-1) 60 mins after a 30-sec icv infusion of 0.02 μCi [14C]glycylsarcosine (GlySar) in wild-type (A-1 and D-1 ...

As expected from the previous experimental data (Figure 1B), the ARGs showed that the lateral ventricle choroid plexus concentrations of [14C]GlySar were significantly higher in wild-type than in Pept2 null mice (Figure 5A). This relationship also held true for the third and fourth ventricle choroid plexuses (Figure 5A). The concentrations of [14C]GlySar in the ependyma were higher on the ipsilateral side of wild-type than of Pept2 null mice (Figure 5B) but were equal on the contralateral side in the two genotypes. As for brain parenchyma, the concentrations of [14C]GlySar were significantly higher in the septal region of wild-type than of Pept2 null mice (Figure 5B). The same genotypic difference in [14C]GlySar concentration was also visible in other brain areas around the lateral and the dorsal part of the third ventricle. As another indicator of genotypic differences in distribution, the distance along the tissue transect over which [14C]GlySar concentrations decreased by one-half was 56% larger in Pept2 null mice (Figure 5C). Finally, analysis of the rostral-to-caudal series of ARGs (choroid plexus excluded) and the resulting plots of radioactivity showed that the brains of Pept2 null mice contained more [14C]GlySar (nCi) in the forebrain around the site of infusion (see arrow in Figure 6A) and at the caudal end of the forebrain–midbrain region (see distance 4.0 to 4.5 mm in Figure 6B) than those of wild-type animals.

Figure 5
Autoradiographic analysis of [14C]glycylsarcosine (GlySar) distribution after a 30-sec intracerebroventricular (icv) infusion of 0.02 μCi dipeptide in wild-type (WT) and Pept2 null mice. (A) Concentrations of [14C]GlySar in the lateral ...
Figure 6
Autoradiographic analysis of the rostrocaudal distribution of [14C]glycylsarcosine (GlySar) after a 30-sec intracerebroventricular (icv) infusion of 0.02 μCi dipeptide in wild-type (WT) and Pept2 null mice. The amount of [14C]GlySar per ...


This study showed that PEPT2 deletion markedly reduced the uptake of GlySar (a model dipeptide) and cefadroxil (a peptide-mimetic antibiotic) by the choroid plexus and greatly diminished the clearance of these peptides from CSF after an icv infusion. The resulting elevation of [14C]GlySar concentration in CSF of the Pept2 null mice led to somewhat farther penetration into brain parenchyma than in the wild-type mice.

Effect of PEPT2 Deletion on [14C]GlySar Distribution at the CSF–Choroid Plexus Interface

If a compound gains access to a part of the CSF system (e.g., synthesis and release by periventricular tissue or entry through the choroid plexus), further movement may involve a number of processes including (1) distribution within the CSF system by convection/bulk flow; (2) loss from the CSF system at the arachnoid villi and other CSF absorption sites; (3) choroid plexus transporter-mediated uptake; and (4) uptake (diffusional and/or transport mediated) into brain parenchyma and, if permeable, loss from tissue into blood through parenchymal capillaries. In the first part of this study, a coinjection of [14C]GlySar and [3H]mannitol was used to examine the involvement of PEPT2-mediated transport in the clearance of GlySar from the CSF system. Mannitol and GlySar have similar molecular weights (182 and 146 Da) and loss through diffusion (aqueous diffusion coefficients at 37°C=9 × 10−6 cm2/sec for mannitol (Enna and Schanker, 1972) and 11 × 10−6 cm2/sec for GlySar, calculated by the Hayduk–Laudie expression (Reid et al, 1977)) into brain parenchyma and movement by bulk CSF flow (solvent drag) are likely to be nearly identical. Unlike GlySar, there are no brain transporters for mannitol and, therefore, a comparison of the ratio of [14C]GlySar to [3H]mannitol in CSF to the ratio in initial injectate gives a measure of transporter-mediated loss of GlySar from CSF. This technique has been used previously to examine the clearance of imipenem and cefodizime (Suzuki et al, 1989; Matsushita et al, 1991), and kyotorphin (Jiang et al, 2009). Because this technique compares the isotope ratio in CSF with that in injectate, it also has the advantage of compensating for small variations in injection volume. Using this technique, we found a substantial decline in the GlySar to mannitol ratio in CSF of wild-type mice. The rate of that decline was three to four times slower in Pept2 null mice, and, thus, PEPT2 appears to be the dominant transporter involved in clearing GlySar from CSF. It is unclear why the GlySar slope decreases in Pept2 null mice (Figure 1A); it may, however, be because of the presence of other peptide transporters (i.e., PHT1 and PHT2) in brain (Yamashita et al, 1997; Sakata et al, 2001).

PEPT2 is expressed on the apical surface of choroid plexus epithelium of primary cell culture and whole animal models (Shu et al, 2002; Shen et al, 2004). In vitro experiments indicated that PEPT2 was involved in clearing [14C]GlySar from CSF into choroid plexus (Shu et al, 2002) and in vivo experiments confirmed that PEPT2 deletion enhanced the CSF exposure of GlySar (Ocheltree et al, 2005). In this study, a brief icv infusion of [14C]GlySar resulted in a time-dependent accumulation of isotope by the choroid plexus of wild-type but not of Pept2 null mice. The 60-mins choroid plexus uptake of [14C]GlySar in null mice was 6% of that in wild-type mice (Figure 1B). These findings are consistent with our prior in vitro and in vivo results, indicating that PEPT2 is the predominant mechanism for apical uptake of GlySar in choroid plexus. These results also suggest that the absence of PEPT2 in choroid plexus is a major factor contributing to the lower transport-mediated removal of [14C]GlySar from the CSF in null mice.

Effect of PEPT2 Deletion on [3H]Cefadroxil Distribution at the CSF–Choroid Plexus Interface

Similar findings to that of GlySar were observed in vivo with the β-lactam antibiotic and PEPT2 substrate, cefadroxil (Shen et al, 2007). In isolated choroid plexus, PEPT2 was the predominant mechanism for [3H]cefadroxil uptake, although there was a small contribution by organic acid transporter(s) (Ocheltree et al, 2004). In this study, the 60-mins choroid plexus uptake of [3H]cefadroxil after brief icv infusion was 82% lower in Pept2 null mice compared with wild-type animals (Figure 3B). Moreover, the transporter-mediated efflux of [3H]cefadroxil from CSF was four times slower in PEPT2 ablation animals than in controls (Figure 3A). Some, but not all, β-lactam antibiotics are PEPT2 substrates (Brandsch et al, 2008), and those that are currently used to treat meningitis, that is, cefotaxime, ceftriaxone, and cefepime, are not PEPT2 substrates (Smith et al, 2004; Brandsch et al, 2008). As a result, these antibiotics avoid PEPT2-mediated clearance from the CSF and may achieve higher CSF concentrations. In addition, Shen et al (2007) found that CSF levels of [3H]cefadroxil (adjusted for differences in blood levels) were increased 6.5-fold in Pept2 null mice compared with wild-type animals.

A difference between the GlySar and cefadroxil results was that where [14C]GlySar values progressively accumulated in the choroid plexus of wild-type mice with time (Figure 1B), [3H]cefadroxil values actually decreased from 5 to 60 mins after infusion (Figure 3B). The cause of this difference is uncertain but could be because of dissimilarities in the affinity of these two substrates for PEPT2, their clearances from CSF, and/or their fate upon entering the choroid plexus epithelial cells. Little is known about oligopeptide and peptidomimetic transport at the basolateral membrane of epithelia (Shu et al, 2002). It may be that the rate of transport from epithelium to blood is higher for cefadroxil than for GlySar, resulting in lower epithelial cell concentrations.

Effect of PEPT2 Deletion on [14C]GlySar Distribution in Brain Assessed by Autoradiography

The autoradiographic images were examined in three different ways. First, [14C]GlySar radioactivity was assayed in several regions of interest, the choroid plexuses (lateral, third, and fourth), ependyma, and septal region. Radiolabeled dipeptide concentrations were significantly higher in each of the choroid plexuses of wild-type mice than of Pept2 null animals (Figure 5A). These differences in [14C]GlySar uptake were not as great as when measurements were made with the coinjection of [14C]GlySar and [3H]mannitol because there was no correction for adherent extracellular [14C]GlySar. The ependyma also expresses PEPT2, and this expression may vary from region to region (Shen et al, 2004). In this study, [14C]GlySar concentration in ipsilateral ependyma was higher in wild-type than in Pept2 null mice; the concentration of [14C]GlySar in contralateral ependyma was, however, not significantly elevated. This difference probably reflects the lower [14C]GlySar levels in CSF perfusing the contralateral lateral ventricle. Inspection of the ARGs revealed that the levels of [14C]GlySar were higher in the septal region of wild-type than of Pept2 null mice (Figures 4A-1, 4B-1, and and5B).5B). The septum is surrounded on three sides by the lateral and third ventricles. The appreciable uptake of [14C]GlySar by the septum may, thus, be the result of 60 mins of exposure to CSF radioactivity on three of its four sides and PEPT2-mediated uptake of dipeptide by septal parenchymal cells.

Second, QAR was used to examine the spread of [14C]GlySar from the lateral ventricle toward the cortical surface and, in particular, to determine the distance for a 50% decrease in [14C]GlySar concentrations from the CSF–brain interface into tissue. In Pept2 null mice, the penetration was significantly greater than in wild-type animals (Figure 5C), and this difference is most likely the result of more cellular uptake and trapping of [14C]GlySar in wild-type mice and less free, diffusible [14C]GlySar in interstitial fluid (Figure 7). Such an overall reduction in the apparent rate of diffusion has been reported in studies with various antitumor drugs in normal brain (Blasberg et al, 1975).

Figure 7
Schematic of how the proton-coupled oligopeptide transporter SLC15A2 (PEPT2 shown as An external file that holds a picture, illustration, etc.
Object name is jcbfm201084e3.jpg) affects the distribution of GlySar (GS) in different regions of the brain after an icv injection. Interstitial fluid (ISF) concentrations of GS are reduced in wild-type ...

The distances for a 50% decrease in [14C]GlySar concentration were ~0.5 mm in the null mice and ~0.3 mm in the wild-type animals. This limited penetration in both genotypes over 60 mins suggests the distribution of peptides/mimetics in normal human with their much larger brains would also be minimal. It should be borne in mind, however, that the [14C]GlySar was injected over 30 secs and that concentration in most CSF compartments decreased continuously from 5 to 60 mins. In contrast, a sustained infusion would lead to a fairly high and constant CSF concentration and greater parenchymal penetration of dipeptide. Of note, the ependymal cells that line the ventricles have gap junctions and do not appreciably hinder the movement of many drugs (Blasberg et al, 1975) and even large compounds such as proteins (Smith et al, 2004) between CSF and brain.

The third parameter obtained from ARGs was the rostrocaudal distribution of radioactivity at 60 mins across the entire mouse brain, excluding the choroid plexuses (Figure 6). In the tissue sections at and around the level of the injection site and at the caudal end of the forebrain/midbrain region, there was greater [14C]GlySar content in Pept2 null mice compared with wild-type animals. These differences in brain levels of [14C]GlySar between Pept2 null and wild-type mice were, however, relatively small (<35% near injection site and <50% in caudal forebrain/midbrain region) and were likely to be the result, in part, of the dissimilarities in the time course of [14C]GlySar concentration in CSF between Pept2 null and wild-type mice.

Autoradiography offers several other advantages for examining [14C]GlySar disposition. It allowed confirmation of the correct placement of the injection needle and the proper delivery of icv administered dipeptide. Correct placement was observed in all the experiments. The injected radiotracer was distributed bilaterally in both mouse strains indicating normal (or near-normal) mixing within the various CSF compartments by 60 mins after icv infusion. The serial sectioning plan spanned the complete brain and, thus, brain regions such as the septal region with unexpected [14C]GlySar uptake could be identified. As mentioned previously, QAR has linear spatial resolution for 14C-emissions on the order of 25 μm and can fairly accurately assess the radioactivity of two sites separated by that distance and very accurately for separations of 50 μm or more. This feature aided in measuring the submillimeter spread of GlySar into brain parenchyma (Figure 4C).

Physiologic and Pharmacologic Effects of Cerebral PEPT2 on Substrate Disposition-Dynamics

This study showed that PEPT2 is the major transporter impacting CSF and choroid plexus concentrations of GlySar and cefadroxil after icv injection. In addition, PEPT2 deletion had some effects on brain distribution. These findings raise the question of whether these changes in brain distribution with PEPT2 deletion will affect the physiologic or pharmacologic effects of the substrates of this transporter. In a prior study (Jiang et al, 2009), we found that PEPT2 deletion in mice increased the analgesic effects induced by icv injection of kyotorphin, a PEPT2 substrate, and that this deletion reduced the uptake of kyotorphin in choroid plexus and its loss from CSF. This indicates that PEPT2 can affect the distribution of kyotorphin at its site of action. The current findings of limited but significant penetration of GlySar into brain after a single injection and the short duration of kyotorphin-induced analgesia suggest that kyotorphin-induced analgesic effects are mediated by receptors close to the ventricular system. Hu et al (2007) have previously reported that PEPT2 deletion also affected the neurotoxicity induced by systemic delivery of 5-ALA, greatly enhanced 5-ALA exposure in CSF, but had no effect on entry into cortex distant from CSF (it should be noted that PEPT2 is not present at the BBB; Shen et al, 2004). This implies that 5-ALA neurotoxicity is mediated by receptors close to the CSF system, just as with kyotorphin.

Choroid plexus tissue can synthesize and secrete a variety of neuropeptides in the process of forming CSF, and PEPT2 may have an important role in maintaining homeostatic control of these endogenous compounds and their ‘oligo' fragments (Smith et al, 2004). Choroid plexus PEPT2 can also influence the in vivo disposition of peptides/mimetics and peptide-like drugs at the BCSFB when administered exogenously (Ocheltree et al, 2005; Shen et al, 2007; Hu et al, 2007; Kamal et al, 2009). Two distinct common haplotypes of human PEPT2 (i.e., hPEPT2*1 and hPEPT2*2) were found in a genetically diverse group of subjects of Asian, African, Caucasian, Mexican, and Pacific Island origin (Pinsonneault et al, 2004; Kroetz et al, 2010). The variants, expressed in Chinese hamster ovary cells, displayed similar Vmax values for GlySar uptake but a threefold higher Km value in hPEPT2*2 (i.e., 233 versus 83 μmol/L in hPEPT2*1). Children homozygous for the hPEPT2*2 polymorphism had higher blood levels of lead than children with the hPEPT2*1 variant (Sobin et al, 2009), suggesting an increased risk of 5-ALA-induced neurotoxicity for this population. Finally, another variant, Arg57His, had a complete loss of transport function when transfected in HEK293 cells and Xenopus oocytes (Terada et al, 2004), although the frequency of this single-nucleotide polymorphism was considered rare. These studies show that PEPT2 variants may serve as potential biomarkers for toxicity, and that information on pharmacogenetic variability may be useful for drug design, drug avoidance, and a priori predictions of drug–drug (or drug–peptide) interactions.

In conclusion, these data show that PEPT2 has a major effect on the distribution of GlySar at the CSF–choroid plexus interface, the ependyma, and the septal region. The observation of a difference in [14C]GlySar penetration into periventricular brain tissue between wild-type and null mice (Figure 5C; D1/2 in WT<Null) is significant and shows the effects of PEPT2 deletion on parenchymal distribution. This was the first time that the uptake and retention of a PEPT2 substrate was characterized in choroid plexuses of all four ventricles, the third ventricle being very difficult to access using more conventional drug disposition methods. Autoradiography combined with gene knockout, as shown in this study, may be valuable in studying the effects of the BBB and BCSFB on regional distribution of other drugs with high spatial resolution.


The authors are grateful to Kelly Keenan and Jun Xu for their expert technical assistance.


The authors declare no conflict of interest.


This work was supported by the NIH Grants R01 GM035498 (DES) and R01 NS034709 (RFK), and by the American Heart Association Grant 0635403N (TNN and JDF).


  • Begley DJ. Delivery of therapeutic agents to the central nervous system: the problems and the possibilities. Pharmacol Ther. 2004;104:29–45. [PubMed]
  • Bereczki D, Wei L, Acuff V, Gruber K, Tajima A, Patlak C, Fenstermacher J. Technique-dependent variations in cerebral microvessel blood volumes and hematocrits in the rat. J Appl Physiol. 1992;73:918–924. [PubMed]
  • Blasberg RG, Patlak C, Fenstermacher JD. Intrathecal chemotherapy: brain tissue profiles after ventriculocisternal perfusion. J Pharmacol Exp Ther. 1975;195:73–83. [PubMed]
  • Brandsch M, Knütter I, Bosse-Doenecke E. Pharmaceutical and pharmacological importance of peptide transporters. J Pharm Pharmacol. 2008;60:543–585. [PubMed]
  • Enna SJ, Schanker LS. Absorption of saccharides and urea from the rat lung. Am J Physiol. 1972;222:409–414. [PubMed]
  • Fei Y-J, Kanai Y, Nussberger S, Ganapathy V, Leibach FH, Romero MF, Singh SK, Boron WF, Hediger MA. Expression cloning of a mammalian proton-coupled oligopeptide transporter. Nature. 1994;368:563–566. [PubMed]
  • Ghersi-Egea JF, Finnegan W, Chen JL, Fenstermacher JD. Rapid distribution of intraventricularly administered sucrose into cerebrospinal fluid cisterns via subarachnoid velae in rat. Neuroscience. 1996;75:1271–1288. [PubMed]
  • Gründemann D, Gorboulev V, Gambaryan S, Veyhl M, Koepsell H. Drug excretion mediated by a new prototype of polyspecific transporter. Nature. 1994;372:549–552. [PubMed]
  • Hu Y, Shen H, Keep RF, Smith DE. Peptide transporter 2 (PEPT2) expression in brain protects against 5-aminolevulinic acid neurotoxicity. J Neurochem. 2007;103:2058–2065. [PubMed]
  • Jiang H, Hu Y, Keep RF, Smith DE. Enhanced antinociceptive response to intracerebroventricular kyotorphin in Pept2 null mice. J Neurochem. 2009;109:1536–1543. [PMC free article] [PubMed]
  • Jonker JW, Buitelaar M, Wagenaar E, Van Der Valk MA, Scheffer GL, Scheper RJ, Plosch T, Kuipers F, Elferink RP, Rosing H, Beijnen JH, Schinkel AH. The breast cancer resistance protein protects against a major chlorophyll-derived dietary phototoxin and protoporphyria. Proc Natl Acad Sci USA. 2002;99:15649–15654. [PubMed]
  • Kamal MA, Jiang H, Hu Y, Keep RF, Smith DE. Influence of genetic knockout of Pept2 on the in vivo disposition of endogenous and exogenous carnosine in wild-type and Pept2 null mice. Am J Physiol Regul Integr Comp Physiol. 2009;296:R986–R991. [PubMed]
  • Kamal MA, Keep RF, Smith DE. Role and relevance of PEPT2 in drug disposition, dynamics, and toxicity. Drug Metab Pharmacokinet. 2008;23:236–242. [PMC free article] [PubMed]
  • Knight RA, Nagaraja TN, Ewing JR, Nagesh V, Whitton PA, Bershad E, Fagan SC, Fenstermacher JD. Quantitation and localization of blood-to-brain influx by magnetic resonance imaging and quantitative autoradiography in a model of transient focal ischemia. Magn Reson Med. 2005;54:813–821. [PubMed]
  • Kroetz DL, Yee SW, Giacomini KM. The pharmacogenomics of membrane transporters project: research at the interface of genomics and transporter pharmacology. Clin Pharmacol Ther. 2010;87:109–116. [PMC free article] [PubMed]
  • Matsushita H, Suzuki H, Sugiyama Y, Sawada Y, Iga T, Kawaguchi Y, Hanano M. Facilitated transport of cefodizime into the rat central nervous system. J Pharmacol Exp Ther. 1991;259:620–625. [PubMed]
  • Nagaraja TN, Patel P, Gorski M, Gorevic PD, Patlak CS, Fenstermacher JD. In normal rat, intraventricularly administered insulin-like growth factor-1 is rapidly cleared from CSF with limited distribution into brain. Cerebrospinal Fluid Res. 2005;26:2:5. [PMC free article] [PubMed]
  • Ocheltree SM, Shen H, Hu Y, Keep RF, Smith DE. Role and relevance of PEPT2 in the kidney and choroid plexus: in vivo studies with glycylsarcosine in wild-type and PEPT2 knockout mice. J Pharmacol Exp Ther. 2005;315:240–247. [PubMed]
  • Ocheltree SM, Shen H, Hu Y, Xiang J, Keep RF, Smith DE. Mechanisms of cefadroxil uptake in the choroid plexus: studies in wild type and PEPT2 knockout mice. J Pharmacol Exp Ther. 2004;308:462–467. [PubMed]
  • Ohtsuki S, Terasaki T. Contribution of carrier-mediated transport systems to the blood-brain barrier as a supporting and protecting interface for the brain; importance for CNS drug discovery and development. Pharm Res. 2007;24:1745–1758. [PubMed]
  • Pinsonneault J, Nielsen CU, Sadée W. Genetic variants of the human H+/dipeptide transporter PEPT2: analysis of haplotype functions. J Pharmacol Exp Ther. 2004;311:1088–1096. [PubMed]
  • Reid RC, Prausnitz JM, Sherwood TK.(eds) (1977. The Properties of Gases and Liquids3rd ed.New York, NY: McGraw-Hill; p 573
  • Rubio-Aliaga I, Daniel H. Peptide transporters and their roles in physiological processes and drug disposition. Xenobiotica. 2008;38:1022–1042. [PubMed]
  • Sakata K, Yamashita T, Maeda M, Moriyama Y, Shimada S. Cloning of a lymphatic peptide/histidine transporter. Biochem J. 2001;356:53–60. [PubMed]
  • Schinkel AH, Smit JJ, van Tellingen O, Beijnen JH, Wagenaar E, van Deemter L, Mol CA, van der Valk MA, Robanus-Maandag EC, te Riele HP, Berns AJ, Borst P. Disruption of the mouse mdr1a P-glycoprotein gene leads to a deficiency in the blood-brain barrier and to increased sensitivity to drugs. Cell. 1994;77:491–502. [PubMed]
  • Shen H, Keep RF, Hu Y, Smith DE. PEPT2 (Slc15a2)-mediated unidirectional transport of cefadroxil from CSF into choroid plexus. J Pharmacol Exp Ther. 2005;315:1101–1108. [PubMed]
  • Shen H, Ocheltree SM, Hu Y, Keep RF, Smith DE. Impact of genetic knockout of PEPT2 on cefadroxil pharmacokinetics, renal tubular reabsorption and brain penetration in mice. Drug Metab Dispos. 2007;35:1209–1216. [PubMed]
  • Shen H, Smith DE, Keep RF, Brosius FC., III Immunolocalization of the proton-coupled oligopeptide transporter PEPT2 in developing rat brain. Mol Pharm. 2004;1:248–256. [PubMed]
  • Shen H, Smith DE, Keep RF, Xiang J, Brosius FC., III Targeted disruption of the PEPT2 gene markedly reduces dipeptide uptake in choroid plexus. J Biol Chem. 2003;278:4786–4791. [PubMed]
  • Shu C, Shen H, Teuscher N, Lorenzi P, Keep RF, Smith DE. Role of PEPT2 in peptide/mimetic trafficking at the blood-CSF barrier: studies in rat choroid plexus epithelial cells in primary culture. J Pharmacol Exp Ther. 2002;301:820–829. [PubMed]
  • Smith DE, Johanson CE, Keep RF. Peptide and peptide analog transport systems at the blood-CSF barrier. Adv Drug Deliv Rev. 2004;56:1765–1791. [PubMed]
  • Sobin C, Gutierrez M, Alterio H. Polymorphisms of delta-aminolevulinic acid dehydratase (ALAD) and peptide transporter 2 (PEPT2) genes in children with low-level lead exposure. Neurotoxicology. 2009;30:881–887. [PMC free article] [PubMed]
  • Suzuki H, Sawada Y, Sugiyama Y, Iga T, Hanano M, Spector R. Transport of imipenem, a novel carbapenem antibiotic, in the rat central nervous system. J Pharmacol Exp Ther. 1989;250:979–984. [PubMed]
  • Taylor EM. The impact of efflux transporters in the brain on the development of drugs for CNS disorders. Clin Pharmacokinet. 2002;41:81–92. [PubMed]
  • Terada T, Irie M, Okuda M, Inui K. Genetic variant Arg57His in human H+/peptide cotransporter 2 causes a complete loss of transport function. Biochem Biophys Res Commun. 2004;316:416–420. [PubMed]
  • Teuscher NS, Shen H, Shu C, Xiang J, Keep RF, Smith DE. Carnosine uptake in rat choroid plexus primary cell cultures and choroid plexus whole tissue from PEPT2 null mice. J Neurochem. 2004;89:375–382. [PubMed]
  • Yamashita T, Shimada S, Guo W, Sato K, Kohmura E, Hayakawa T, Takagi T, Tohyama M. Cloning and functional expression of a brain peptide/histidine transporter. J Biol Chem. 1997;272:10205–10211. [PubMed]

Articles from Journal of Cerebral Blood Flow & Metabolism are provided here courtesy of SAGE Publications