Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Arginine-vasopressin (AVP) fragment 4-9 [AVP_4-9, isoelectric point; (pI) = 9.2; Fig. 1] is a stable major metabolite of AVP in the central nervous system (CNS; Burbach et al., 1983, 1993). It is well documented that AVP_4-9 exerts a more potent memory-facilitative effect than AVP (Gaffori and De Weid, 1986; De Weid et al., 1987).

View larger version (16K): [in this window] [in a new window]   Fig. 1.   The structures of AVP4-9, C-AVP4-9, and B chain. *The 35S-labeled position. **The 125I-labeled position. The structure of the B chain is enclosed in a solid line.

Previous studies indicated that a novel cationic AVP_4-9 analog (C-AVP_4-9, PI = 9.8, Fig. 1) possesses more potent memory-facilitative activity than its parent peptide, AVP_4-9 (Tanabe et al., 1997). C-AVP_4-9 was designed to be converted to the neuroactive parent peptide, AVP_4-9, by the postproline cleaving enzyme (PPCE), which was previously demonstrated to be abundant in the CNS and scarce in the plasma (Yoshimoto et al., 1979).

To act on the CNS, drugs and neuropeptides should be transported through the blood-brain barrier (BBB), which consists of brain capillary endothelial cells possessing tight intercellular junctions. With regard to BBB transport of peptides, previous studies have reported the receptor-mediated endocytosis (RME) of peptides, such as insulin (Duffy and Pardridge, 1987), anti-insulin receptor antibody (Pardridge et al., 1995; Wu et al., 1998), transferrin (Fishman et al., 1987; Skarlatos et al., 1995), antitransferrin receptor antibody (Pardridge et al., 1991; Shin et al., 1995; Walus et al., 1996), and the absorptive-mediated endocytosis (AME) of cationic peptides, such as E-2078 (Terasaki et al., 1989), ebiratide (Shimura et al., 1991), histone (Pardridge et al., 1989), and cationized albumin (Pardridge et al., 1990). In addition, these reports demonstrated that the AME-transported peptides show a far higher maximal internalization capacity into the brain capillary than the RME-transported peptides.

Accordingly, it is of great interest to investigate the BBB transport mechanism of C-AVP_4-9, because C-AVP_4-9 is a cationic peptide at physiological pH.

The purpose of the present study is to evaluate the BBB transport of C-AVP_4-9 based on pharmacokinetic analysis and clarify the BBB transport mechanism of C-AVP_4-9. We also demonstrated the conversion of C-AVP_4-9 to AVP_4-9 by PPCE in the CNS.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Radiolabeling of Peptides. AVP_4-9 ([pGlu⁴,Cyt⁶,Arg⁸]vasopressin fragment 4-9; Sigma Chemical Co., St. Louis, MO) was labeled with ³⁵S. As an intermediate of ³⁵S-labeled AVP_4-9, 3-nitro-2-pyridinesulfenyl-AVP_4-9 was synthesized by a previously demonstrated method (Shimohigashi et al., 1992). [³⁵S]Cysteine (Amersham Pharmacia Biotech, Buckinghamshire, UK) was used immediately after purification to remove dithiothreitol by HPLC (see below). 3-nitro-2-pyridinesulfenyl-AVP_4-9, 16.2 µl (1 mg/ml) was reacted with 74 MBq of [³⁵S]cysteine in 500 µl of distilled water at room temperature for 24 h. C-AVP_4-9 (American Peptide Company, Sunnyvale, CA) was labeled with ¹²⁵I by the chloramine T method (Hunter and Greenwood, 1962; Tanabe et al., 1997). The reaction mixtures of ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 were purified by HPLC as described below. The ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 obtained had specific activities of 22.6 to 46.0 TBq/mmol (purity >94.4%) and 48.3 to 138.5 TBq/mmol (purity >99.0%), respectively.

HPLC Conditions. To quantify the unchanged ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 in in vivo and in vitro uptake studies, HPLC was used. For the separation, the column J'sphere ODS-H80 (YMC Inc., Wilmington, NC) was used. In the studies using radiolabeled peptides, elution was performed at a flow rate of 1.5 ml/min with the following linear gradient: solvent A = 0.1% trifluoroacetic acid; solvent B = acetonitrile; B in A (v/v): from 0 to 20% for 20 min for separation, from 20 to 40% for 2 min, holding 40% for 5 min to wash the column, and from 40% to 0% for 3 min with equilibration for 15 min before subsequent injection. The HPLC eluates were collected automatically (0.75 ml/fraction). In the studies using only ¹²⁵I-labeled peptides, the radioactivity in each eluate was counted using a gamma counter (Autogamma 5550; Packard Instrument Co., Meriden, CT). In the studies using dual nuclides (³⁵S and ³H, or ¹²⁵I and ³H) or only ³⁵S, the radioactivity was counted using a liquid scintillation counter (LSC-5000; Beckman Coulter Inc., Fullerton, CA). In the studies using unlabeled peptides, the HPLC elution was performed with the following linear gradient: solvent B in A (v/v): from 0 to 5% for 15 min, from 5 to 15% for 10 min, from 15 to 40% for 5 min, holding 40% for 5 min, and from 40% to 0% for 5 min with equilibration for 10 min. The eluate was monitored by UV absorbance at 210 nm.

In Vivo Pharmacokinetic Studies. ³⁵S-labeled AVP_4-9 (130 KBq), ¹²⁵I-labeled C-AVP_4-9 (1110 KBq), or 370 KBq of ¹²⁵I-labeled BSA (144 KBq/µg; NEN Life Sciences, Boston, MA) was administered i.v. into the tail vein of normal male ICR mice (6-8 weeks old; CLEA Japan, Tokyo). At 1, 5, 15, 30, and 45 min after the administration, mice were sacrificed, and the plasma, cerebrum, liver, kidney, heart, and lung were obtained. The unchanged ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 in the plasma and tissues were quantified as follows. Ice-cold 2% phosphoric acid (3× volume for plasma, 4× volume for cerebrum, and 10× volume for other tissues) was added, and the mixtures were homogenized with an Ultra-turrax (IKA Labortechnik, Staufen, Germany). The homogenates were then centrifuged (14,000g, 4°C, 30 min), and 500-µl aliquots of the supernatants were analyzed by HPLC. In the ¹²⁵I-labeled BSA studies, plasma and tissues were homogenized with a 5× volume of 30% ice-cold trichloroacetic acid (TCA). After centrifugation (14,000g, 4°C, 30 min), the radioactivity in each pellet was measured as unchanged ¹²⁵I-labeled BSA by a gamma counter. The plasma concentrations [C_p(t)] of unchanged labeled peptides were normalized as a percentage of the injected dose per milliliter. The C_p(t) data were fitted to the function (1) by a nonlinear least-squares regression analysis using a computer program, MULTI (Yamaoka et al., 1980):

(1)

where A and B are the intercepts, and  and  are the rate constants describing the fast and slow compartments of clearance from plasma, respectively. As an index of the distribution of unchanged labeled peptides to the tissues, the apparent tissue-to-plasma concentration ratio (K_p,app), defined by the radioactivity of unchanged radiolabeled peptides per gram of tissue divided by that per milliliter of plasma, was calculated (Terasaki et al., 1984). The BBB permeation clearances (PS/V_br) of labeled peptides were calculated using the function (2):

(2)

where V₀, V_br, and C_br(t) are the rapidly equilibrated distribution volume of labeled peptides, the cerebrum weight, and the cerebral concentration of labeled peptides, respectively. PS/V_br values of labeled peptides were obtained from the initial slopes of C_br(t)/C_p(t) [K_p,app versus the area under the curve [AUC(0  t)] C_p(t) as an integration plot (Smith et al., 1987; Kakee et al., 1996; Yang et al., 1997)].

Capillary Depletion Study. The left and right renal artery and vein of mice were ligated under ether anesthesia before use. ¹²⁵I-labeled BSA (370 KBq), ³⁵S-labeled AVP_4-9 (130 KBq), or ¹²⁵I-labeled C-AVP_4-9 (1110 KBq) was i.v. administered into the tail vein of mice. At 30 min after the administration, plasma and cerebrum were obtained. The cerebra were separated into the parenchyma and the capillary fractions by a method reported previously (Yang et al., 1997). The unchanged ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 in the parenchyma fraction were extracted two times with a 1/10× volume of ice-cold 50% phosphoric acid and a 10× volume of ice-cold acetonitrile. After centrifugation (3500g, 4°C, 1 h), the supernatants were lyophilized and suspended in 1 ml of ice-cold methanol. After further centrifugation (3500g, 4°C, 1 h), the supernatants were again lyophilized and reconstituted with 600 µl of 0.1% trifluoroacetic acid, and 500-µl aliquots were analyzed by HPLC. The unchanged ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 in the capillary fraction were extracted with 600 µl of 5% phosphoric acid, and 500-µl aliquots were analyzed by HPLC. The unchanged ¹²⁵I-labeled BSA in the parenchyma and the capillary fractions were precipitated by ice-cold 30% TCA (5× volume for the parenchyma fraction and 500 µl for the capillary fraction), and the radioactivity in each pellet was measured with a gamma counter.

In Vitro Internalization Studies Using MBEC4. MBEC4 cells (Tatsuta et al., 1992) were used. Monolayers of MBEC4 cells were maintained in Dulbecco's modified Eagle's medium containing 10% FBS in 5% CO₂/95% air at 37°C. The internalization of ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 into MBEC4 was examined by a method reported previously (Terasaki et al., 1992) with minor modifications. Briefly, monolayers of MBEC4 cells cultured in 24-well dishes were washed five times with 1 ml of the incubation buffer [122 mM NaCl, 3 mM KCl, 25 mM NaHCO₃, 1.2 mM MgSO₄, 0.4 mM K₂HPO₄, 1.4 mM CaCl₂, 10 mM D-glucose, 10 mM HEPES, 0.1% BSA, 1 tablet/50 ml Complete (a protease inhibitor cocktail; Boehringer Mannheim, Mannheim, Germany), pH 7.4, 300 mOsM] at 37°C. They were then preincubated in 200 µl of incubation buffer at 37°C for 30 min. In the uptake studies, except for the studies of concentration dependencies, the uptakes were initiated by adding 200 µl of incubation buffer containing ³⁵S-labeled AVP_4-9 (18.5 KBq) or ¹²⁵I-labeled C-AVP_4-9 (18.5 KBq), and 92.5 KBq of [³H]inulin (13.5 GBq/mmol, NEN Life Sciences) to MBEC4. [³H]Inulin was used as the extracellular space marker. The effects of medium osmolarity on the uptakes of ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 were examined with hypertonic buffer (incubation buffer with 1.2 M sucrose, 1600 mOsM) substituted for the incubation buffer. The effects of endocytosis inhibitors, metabolic inhibitor, and several peptides on the uptake of ¹²⁵I-labeled C-AVP_4-9 were examined in the presence of 0.5 mM danylcadaverine (Sigma), 0.1 mM phenylarsine oxide (Aldrich Chemical Co., Milwaukee, WI), 1 mM 2,4-dinitrophenol (Wako Pure Chemical Industries, Osaka, Japan), 0.5 mM V₁ vasopressin receptor antagonist ([(-mercapto-,-cyclopentamethylene propionic acid)¹,O-methyl-Tyr²,Arg⁸]vasopressin; Peninsula Laboratories, Belmont, CA), 0.5 mM V₂ vasopressin receptor antagonist ([adamantaneacetyl¹,O-ethyl-D-Tyr²,Val⁴,aminobutyryl⁶,Arg^8,9]-vasopressin; Sigma), 0.5 mM AVP_4-9, 0.3 mM poly(L-lysine) (Sigma), 0.3 mM protamine (Wako Pure Chemical Industries), or 0.3 mM poly(L-glutamic acid) (Sigma). In the studies of the concentration dependencies, various concentrations of mixtures of ³⁵S-labeled AVP_4-9 and unlabeled AVP_4-9 or ¹²⁵I-labeled C-AVP_4-9 and unlabeled C-AVP_4-9 were added with [³H]inulin (92.5 KBq) to MBEC4 after preincubation. At designated times after incubation, the incubation supernatants (unbound fraction) were recovered, and the cells were washed seven times with 1 ml of ice-cold incubation buffer. An acid wash technique was then used to remove the labeled peptide binding to the cell surface and to evaluate the amounts of labeled peptides internalized into MBEC4. The cells were incubated with 1 ml of ice-cold acetate-barbital buffer (28 mM CH₃COONa, 120 mM NaCl, 20 mM barbital, pH 3.0, 320 mOsM) for 10 min. The buffer was then recovered (acid-soluble fraction), and the cells were subsequently washed three additional times with 1 ml of acetate-barbital buffer. The cells remaining were obtained as the acid-resistant fraction and solubilized with 50 µl of NaOH (0.5 N) and 200 µl of 2% Triton X-100. Aliquots of unbound (100 µl), acid-soluble (500 µl), and solubilized acid-resistant (200 µl) fractions were analyzed by HPLC. The protein assay of the acid-resistant fraction was carried out with BCA Protein Assay Reagent (Pierce, Rockford, IL). In accord with Terasaki et al. (1989, 1992), the data on the acid-resistant binding are expressed as the cell-to-medium concentration ratios (cell/medium ratios), corrected for an extracellular space using [³H]inulin, as follows:

(3)

where ³H, ³⁵S, and ¹²⁵I are the radioactivities of unchanged [³H]inulin, ³⁵S-labeled AVP_4-9, and ¹²⁵I-labeled C-AVP_4-9, respectively; and -R, -S, and -M are the acid-resistant, the acid-soluble, and the unbound fractions, respectively. The maximal internalization capacity (B_max) and the half-saturation constant (K_D) were estimated by MULTI (Yamaoka et al., 1980). The cell/medium ratios were statistically analyzed using the Kruskal-Wallis test and Dunnett's test.

Enzymatic Conversion of C-AVP_4-9 to AVP_4-9 by Mouse Cerebral Homogenate. The mouse cerebra were homogenized with a 4× volume of metabolism assay buffer (the incubation buffer without BSA and Complete). The PPCE activity in the mouse cerebral homogenate was determined by a method reported previously (Yoshimoto et al., 1979). After preincubation at 37°C for 10 min, C-AVP_4-9 was incubated with PPCE derived from Flavobacterium meningosepticum (authentic PPCE, 0.1 U/ml; Seikagaku Corp., Tokyo, Japan) or the mouse cerebral homogenate (20 mg cerebrum/ml, equivalent to 1.5 × 10³ U/ml) in 250 µl of metabolic assay buffer in the presence or absence of the PPCE-specific inhibitor N-benzyloxycarbonyl-prolyl-prolinal (ZPP, 1 mM; Yakult Central Institute for Microbiological Research, Tokyo, Japan) at 37°C for designated times. The metabolic reaction was stopped by adding 250 µl of 2% phosphoric acid. After centrifugation (14,000g, 4°C, 10 min), 400-µl aliquots of supernatants were analyzed by electrospray ionization-liquid chromatography-mass spectrometry (ESI-LC-MS) using a Finnigan TSQ 7000 triple quadrupole mass spectrometer with ESI-LC-MS interface (Thermo Quest Corp., San Jose, CA). The operating conditions were as follows: spray voltage, 4.5 kV; sheath gas pressure, 70 psi (nitrogen); auxiliary gas flow, 5 arbitrary units; heated capillary temperature, 200°C. Mass spectra were acquired every 3 s with a scanning m/z range of 50 to 1500. The other conditions were as described for HPLC conditions. The Michaelis constant (K_m) and maximum velocity (V_max) were also estimated by MULTI (Yamaoka et al., 1980). The intrinsic clearance of the conversion of C-AVP_4-9 to AVP_4-9 (CL_int) was defined as the V_max value divided by the K_m value.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

In Vivo Pharmacokinetic Studies. The plasma concentration of i.v. administered ¹²⁵I-labeled BSA was monoexponentially fitted (Table 1, Fig. 2). In contrast, the plasma concentrations of i.v. administered ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 were characterized by a biexponential function and reached apparent steady states at 15 min after administration (Table 1, Fig. 2). The K_p,app values of both ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 were the highest in the kidney among the tissues examined (Figs. 3 and 4). The K_p,app values of ¹²⁵I-labeled C-AVP4-9 in the liver and the lung were decreased compared with those of ³⁵S-labeled AVP_4-9 (Fig. 3), whereas the K_p,app values of ¹²⁵I-labeled C-AVP_4-9 in the cerebrum were increased compared with those of ³⁵S-labeled AVP_4-9 (Fig. 4). At 45 min after the administration, K_p,app values of ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 in the cerebrum were 0.103 ± 0.019 and 0.330 ± 0.030 ml/g cerebrum, respectively (Fig. 4). The K_p,app values of ¹²⁵I-labeled BSA in the cerebrum were maintained at 0.01 ml/g cerebrum (Fig. 4). By the integration plots, the BBB permeation clearances of ¹²⁵I-labeled BSA, ³⁵S-labeled AVP_4-9, and ¹²⁵I-labeled C-AVP_4-9 were estimated to be 6.26 × 10⁶ ± 1.26 × 10⁵, 1.47 ± 0.09 × 10⁴, and 3.10 ± 0.35 × 10⁴ ml/min/g cerebrum, respectively (Fig. 5).

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Plasma concentration profiles of 125I-labeled BSA, 35S-labeled AVP4-9, and 125I-labeled C-AVP4-9. 125I-labeled BSA (37 pmol/mouse), 35S-labeled AVP4-9 (2.8 pmol/mouse), or 125I-labeled C-AVP4-9 (23.0 pmol/mouse) was administered i.v. into normal mice or mice whose renal vessels were ligated. The unchanged 35S-labeled AVP4-9 and 125I-labeled C-AVP4-9 were extracted from plasma with 2% phosphoric acid and quantified by HPLC. The unchanged 125I-labeled BSA in plasma was precipitated with 30% TCA, and radioactivity was counted. The 125I-labeled BSA () data in the normal mice were fitted to a monoexponential function, whereas 35S-labeled AVP4-9 () and 125I-labeled C-AVP4-9 () data in the normal mice were fitted to a biexponential function. The plasma concentrations of the unchanged 125I-labeled BSA (), 35S-labeled AVP4-9 (), and 125I-labeled C-AVP4-9 () at 30 min after administration into the mice whose renal vessels are also plotted. Each point represents the mean ± S.E. of three or four mice.

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Apparent tissue-to-plasma concentration ratios of 125I-labeled BSA, 35S-labeled AVP4-9, and 125I-labeled C-AVP4-9 after the i.v. administration. 125I-labeled BSA (37 pmol/mouse), 35S-labeled AVP4-9 (2.8 pmol/mouse), or 125I-labeled C-AVP4-9 (23.0 pmol/mouse) was administered i.v. into the normal mice. The unchanged 125I-labeled BSA (), 35S-labeled AVP4-9 (), and 125I-labeled C-AVP4-9 () in the liver (A), kidney (B), heart (C), and lung (D) were analyzed by HPLC or the TCA precipitation method. The tissue distributions are expressed as the apparent tissue-to-plasma concentration ratios (Kp,app). The Kp,app was defined by the concentration of unchanged radio-labeled peptides in the tissues divided by the plasma concentration of each peptide demonstrated in Fig. 2. Each point represents the mean ± S.E. of three or four mice.

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Apparent cerebrum-to-plasma concentration ratios of 125I-labeled BSA, 35S-labeled AVP4-9, and 125I-labeled C-AVP4-9. 125I-labeled BSA (37 pmol/mouse), 35S-labeled AVP4-9 (2.8 pmol/mouse), or 125I-labeled C-AVP4-9 (23.0 pmol/mouse) was administered i.v. into the normal mice. The Kp,app values of 125I-labeled BSA (), 35S-labeled AVP4-9 (), and 125I-labeled C-AVP4-9 () in the cerebrum were determined as described in the legend to Fig. 3. Each point represents the mean ± S.E. of three or four mice.

View larger version (13K): [in this window] [in a new window]   Fig. 5.   BBB permeation clearances of 125I-labeled BSA, 35S-labeled AVP4-9, and 125I-labeled C-AVP4-9. The Kp,app values of 125I-labeled BSA (), 35S-labeled AVP4-9 (), and 125I-labeled C-AVP4-9 () shown in Fig. 4 were plotted against the indicated AUC(0  t)/Cp(t) values to estimate the BBB permeation clearances of each labeled peptide (integration plot). Each point represents the mean ± S.E. of three or four mice.

Capillary Depletion Study. At 30 min after the i.v. administration of ¹²⁵I-labeled BSA, ³⁵S-labeled AVP_4-9, and ¹²⁵I-labeled C-AVP_4-9 to the mice whose renal vessels were ligated, the distributions of labeled peptides in the capillary and the parenchyma fractions of cerebrum were examined. The plasma concentrations of ¹²⁵I-labeled BSA, ³⁵S-labeled AVP_4-9, and ¹²⁵I-labeled C-AVP_4-9 in mice whose renal vessels were ligated were 0.9, 10, and 3 times greater than those in the normal mice, respectively (Fig. 2). The unchanged ³⁵S-labeled AVP_4-9and ¹²⁵I-labeled C-AVP_4-9 in the parenchyma fraction, quantified by HPLC, were 10.4 ± 1.2% and 0.63 ± 0.06% of total recovered radioactivities, respectively. As shown in Table 2, the K_p,app value of ¹²⁵I-labeled BSA in the parenchyma fraction was 1.04 ± 0.13 × 10² ml/g cerebrum. The K_{p, app} values of ¹²⁵I-labeled BSA, ³⁵S-labeled AVP_4-9, and ¹²⁵I-labeled C-AVP_4-9 in the parenchyma fraction were 35, 68, and 22 times higher than those in the capillary fraction, respectively.

Time Courses of Acid-Resistant Binding of ³⁵S-Labeled AVP_4-9 and ¹²⁵I-Labeled C-AVP_4-9 to MBEC4. Unchanged ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 in the acid-resistant fraction at 60 min, quantified by HPLC, were 6.48 ± 0.32 and 68.7 ± 5.6% of total recovered radioactivities respectively. The cell/medium ratios of [³H]inulin coexisting with ³⁵S-labeled AVP_4-9 (Fig. 6A) or ¹²⁵I-labeled C-AVP_4-9 (Fig. 6B) were increased in a slightly time-dependent manner. The acid-resistant binding of ³⁵S-labeled AVP_4-9 increased with time and reached apparent equilibrium at 30 min (Fig. 6A). The acid-resistant binding of ¹²⁵I-labeled C-AVP_4-9 increased time-dependently (Fig. 6B). The acid-resistant binding of ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 at 120 min were 0.93 ± 0.07 and 1.95 ± 0.20 µl/mg protein (as the cell/medium ratios), respectively.

View larger version (21K): [in this window] [in a new window]   Fig. 6.   Time courses of acid-resistant binding of [3H]inulin, 35S-labeled AVP4-9, and 125I-labeled C-AVP4-9 to MBEC4. 35S-labeled AVP4-9 (4.1 nM) or 125I-labeled C-AVP4-9 (0.7 nM) was incubated with MBEC4 in the presence of [3H]inulin (34.3 µM) at 37°C for the indicated time. After the incubation, unbound, acid-soluble, and acid-resistant fractions were recovered, and the unchanged [3H]inulin, 35S-labeled AVP4-9, and 125I-labeled C-AVP4-9 in each fraction was quantified by HPLC. The acid-resistant bindings of 35S-labeled AVP4-9 ( in A), 125I-labeled C-AVP4-9 ( in B), and [3H]inulin () are expressed as the cell/medium ratios. Each point represents the mean ± S.E. of three or four separate experiments.

Concentration Dependencies of Acid-Resistant Binding of ³⁵S-Labeled AVP_4-9 and ¹²⁵I-Labeled C-AVP_4-9. In the studies of concentration dependencies, ³⁵S-labeled AVP_4-9 showed two-phase saturable acid-resistant binding (Fig. 7). The B_max values of ³⁵S-labeled AVP_4-9 for the high- and low-affinity sites of MBEC4 were estimated to be 0.72 ± 0.63 pmol/mg protein and 26.5 ± 6.0 pmol/mg protein, respectively. The K_D values of ³⁵S-labeled AVP_4-9 for the high- and low-affinity sites of MBEC4 were estimated to be 3.81 ± 3.46 nM and 45.7 ± 11.4 µM, respectively. In contrast, ¹²⁵I-labeled C-AVP_4-9 showed single-phase saturable acid-resistant binding (Fig. 7). The B_max and K_D values of ¹²⁵I-labeled C-AVP_4-9 for MBEC4 were estimated to be 14.7 ± 4.1 pmol/mg protein and 16.4 ± 5.0 µM, respectively.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Concentration dependencies of acid-resistant binding of 35S-labeled AVP4-9 and 125I-labeled C-AVP4-9 to MBEC4. Various concentrations of mixtures of 35S-labeled AVP4-9 and unlabeled AVP4-9 or 125I-labeled C-AVP4-9 and unlabeled C-AVP4-9 were incubated with MBEC4 in the presence of [3H]inulin (34.3 µM) at 37°C for 60 min. The acid-resistant bindings of unchanged 35S-labeled AVP4-9 () and 125I-labeled C-AVP4-9 () are expressed as the cell/medium ratios. Each point represents the mean ± S.E. of three or four separate experiments.

Effect of Temperature, Osmolarity, Endocytosis Inhibitors, Metabolic Inhibitor, and Several Peptides on Acid-Resistant Binding of ¹²⁵L-Labeled C-AVP_4-9. As shown in Table 3, the incubation of ¹²⁵I-labeled C-AVP_4-9 with MBEC4 at 4°C for 60 min resulted in a decrease in the acid-resistant binding to 10.4 ± 1.2% of the control value. When the osmolarity of the incubation buffer was increased from 300 to 1600 mOsM, the acid-resistant binding of ¹²⁵I-labeled C-AVP_4-9 decreased significantly (Table 3). The acid-resistant binding of ¹²⁵I-labeled C-AVP_4-9 to MBEC4 was significantly diminished by dancylcadaverine, phenylarsine oxide, and 2,4-dinitrophenol (Table 3); markedly inhibited by AVP_4-9, poly(L-lysine), and protamine (Table 4); and unaffected by poly(L-glutamic acid), V₁ vasopressin receptor antagonist, and V₂ vasopressin receptor antagonist (Table 4).

Enzymatic Conversion of C-AVP_4-9 to AVP_4-9. C-AVP_4-9 was converted by authentic PPCE and the mouse cerebral homogenate to the metabolites eluted at 12.0 and 13.3 min (Fig. 8A) or 13.3 min (Fig. 8C) in the HPLC. The retention times of authentic B chain (arginyl-histidinyl-proline; Fig. 1), AVP_4-9, and C-AVP_4-9 were 12.0, 13.3, and 23.0 min, respectively, in the HPLC conditions used here. The C-AVP_4-9 metabolites produced by the authentic PPCE and the mouse cerebral homogenate eluted at 13.3 min in the HPLC (Fig. 8, A and C) showed the same m/z as protonated authentic AVP_4-9 ([M + H] = 775) and doubly protonated AVP_4-9 ([M + 2H]²⁺ = 338) by ESI-LC-MS. Based on the agreements of retention time in the HPLC and the mass spectrometric behavior, C-AVP_4-9 metabolites eluted at 13.3 min in Fig. 8, A and C, were identified as AVP_4-9. ZZP (1 mM) completely inhibited the conversion of C-AVP_4-9 to AVP_4-9 by the authentic PPCE (Fig. 8B). That concentration of ZPP also significantly (p < .001 by Wilcoxon's test) inhibited 49.3 ± 0.3% (mean ± S.E., n = 4) of C-AVP_4-9 conversion to AVP_4-9 by the cerebral homogenate (Fig. 8D). The authentic PPCE and the mouse cerebral homogenate showed saturable conversion rates of C-AVP_4-9 to AVP_4-9 with the CL_int values 2.47 ± 0.24 (Fig. 9A) and 2.10 ± 0.14 ml/U/min (Fig. 9B), respectively.

View larger version (26K): [in this window] [in a new window]   Fig. 8.   Enzymatic conversion of C-AVP4-9 to AVP4-9. C-AVP4-9 (150 µM) was incubated with the authentic PPCE (0.1 U/ml, A and B) or mouse cerebral homogenate (20 mg cerebrum/ml equivalent to 1.5 × 103 U/ml, C and D) at 37°C for 20 min or 6 h, respectively. The inhibition experiments were carried out in the presence of 1 mM ZPP in the same conditions (B and D). The metabolic reactions were stopped by the addition of 2% phosphoric acid. After the centrifugation the supernatants were analyzed by HPLC.

View larger version (23K): [in this window] [in a new window]   Fig. 9.   Enzymatic conversion rate of C-AVP4-9 to AVP4-9. Various concentrations of C-AVP4-9 were incubated with authentic PPCE (0.1 U/ml, A) or mouse cerebral homogenate (20 mg cerebrum/ml equivalent to 1.5 × 103 U/ml, B) at 37°C for 1 or 30 min, respectively. The metabolic reactions were stopped by the addition of 2% phosphoric acid. After centrifugation, the supernatants were analyzed by HPLC. Each point represents the mean ± S.E. of three separate experiments.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

It has been demonstrated that cationic peptides are effectively transported through the BBB and reach the parenchyma of the CNS via AME (Pardridge et al., 1991; Terasaki et al., 1992). To enhance the BBB permeability of AVP_4-9 (PI = 9.2, Fig. 1), we combined the peptide composed of cationic amino acids, that is, a B-chain (Fig. 1), with the free cysteine residue of AVP_4-9 and obtained C-AVP_4-9 (PI = 9.8, Fig. 1). C-AVP_4-9 was also designed to be converted to AVP_4-9 in the CNS by PPCE, which is abundant in the CNS (Yoshimoto et al., 1979). It was previously demonstrated that C-AVP_4-9 facilitates the memory of rodents more effectively than AVP_4-9 (Tanabe et al., 1997). In the present study, the BBB transport of C-AVP_4-9 and its conversion to AVP_4-9 in the CNS were evaluated.

The K_p,app values of ¹²⁵I-labeled C-AVP_4-9 in the liver and the lung were lower compared with those of ³⁵S-labeled AVP_4-9 (Fig. 3), whereas the K_p,app values of ¹²⁵I-labeled C-AVP_4-9 in the cerebrum were clearly higher compared with those of ³⁵S-labeled AVP_4-9 (Fig. 4). These results suggest that the cationization of AVP_4-9 decreases the peripheral distribution of modified molecule and increases the distribution of the molecule to the whole cerebrum, including the parenchyma, capillary, and plasma. To evaluate the BBB transport of neuroactive peptides, however, it is necessary to evaluate the distribution of peptides to the parenchyma of the cerebrum. Hence, we separated the cerebrum of the mice administered with the labeled peptides into the capillary and the parenchyma fractions by the capillary depletion method (Yang et al., 1997) and evaluated the distributions of labeled peptides to both fractions. We have demonstrated previously that the alkaline phosphatase activity and K_p,app value of i.v. administered [¹²⁵I]-labeled low-density lipoprotein (markers for the capillary) in the parenchyma fraction, which express contamination of the capillary in the parenchyma fraction, were about 10% of those in the whole cerebrum (Yang et al., 1997). In the capillary depletion studies, we ligated the renal vessels of mice before the i.v. administration of ¹²⁵I-labeled BSA, ³⁵S-labeled AVP_4-9, or ¹²⁵I-labeled C-AVP_4-9 to eliminate the distribution of these labeled peptides to the kidney, the tissue where ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 showed the highest K_p,app value among the tissues examined (Figs. 3 and 4), and to elevate the plasma concentrations and the cerebral distributions of the labeled peptides. With the ligation treatments, the plasma concentrations of ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 at 30 min were 10 times and 3 times higher than those in the normal mice, respectively (Fig. 2), and this treatment enabled us to quantify the unchanged ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 in the parenchyma and capillary fractions of the cerebrum by HPLC. Because it is widely accepted that BSA remains in the plasma in the cerebrum (Triguero et al., 1990), the plasma volume of the cerebral parenchyma fraction was estimated to be 1.04 ± 0.13 × 10² ml/g cerebrum (K_p,app of ¹²⁵I-labeled BSA in the parenchyma fraction) (Table 2). The K_p,app values of ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 in the parenchyma fraction were markedly higher than the plasma volume in the parenchyma fraction and K_p,app values of each labeled peptide in the capillary fractions (Table 2), suggesting that almost all of these labeled peptides are distributed to the parenchyma in the whole cerebrum. Therefore, the K_p,app values of ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 in the whole cerebrum demonstrated in Fig. 4 nearly indicate the K_p,app value of these labeled peptides in the cerebral parenchyma, and we evaluated the BBB permeation clearances of ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 with the K_p,app values shown in Fig. 4 by the integration plot method (Smith et al., 1987; Kakee et al., 1996; Yang et al., 1997). With the integration plotting, the BBB permeation clearances of ³⁵S-labeled AVP_4-9 and ¹²⁵I-labeled C-AVP_4-9 were estimated to be 1.47 ± 0.09 × 10⁴ and 3.10 ± 0.35 × 10⁴ ml/min/g cerebrum, respectively (Fig. 5), suggesting that C-AVP_4-9 is transported through the BBB to the cerebral parenchyma more effectively than AVP_4-9.

Monolayers of MBEC4 cells were used in in vitro internalization studies. Because the acid wash treatment removes the labeled peptides binding to the cell surface, the acid-resistant binding of the labeled peptides expresses the internalized labeled peptides into the cells (Terasaki et al., 1989, 1992). The slight time dependence of acid-resistant bindings of extracellular space marker, [³H]inulin, coexisting with ³⁵S-labeled AVP_4-9 or ¹²⁵I-labeled C-AVP_4-9, to MBEC4 (Fig. 6) may be ascribed to the fluid-phase endocytosis of MBEC4 (Terasaki et al., 1992). From 10 min after the incubation, the acid-resistant binding of ¹²⁵I-labeled C-AVP_4-9 was higher than that of ³⁵S-labeled AVP_4-9 (Fig. 6), suggesting that C-AVP_4-9 is internalized into the brain capillary endothelial cells more effectively than AVP_4-9. The K_D values of RME-transported peptides reported previously were as follows: atrial natriuretic factor (400 pM, using cultured bovine brain capillary endothelial cells; Smith et al., 1988), transferrin (5.6 nM, isolated human brain capillaries; Pardridge et al., 1987), insulin (2.9 nM, isolated human brain capillaries; Frank et al., 1986), and leptin (5.1 nM, isolated human brain capillaries; Golden et al., 1997). Although the K_D values of AVP_4-9 for the low-affinity sites (45.7 µM) and of C-AVP_4-9 (16.4 µM) were in good agreement with the values of E-2078 (4.62 µM, using isolated bovine brain capillaries; Terasaki et al., 1989), ebiratide (15.9 µM, cultured bovine brain capillary endothelial cells; Terasaki et al., 1992; 62.1 µM, isolated bovine brain capillaries, Shimura et al., 1991), and histone (15.2 µM, isolated bovine brain capillaries, Pardridge et al., 1989), which are internalized into the BBB via AME. In light of the agreements of K_D values of AVP_4-9 (low affinity) and C-AVP_4-9 with the values of AME-transported peptides in the previous reports and the basicities of both peptides, it is suggested that C-AVP_4-9 and a part of AVP_4-9 are internalized into the BBB via AME.

To more precisely elucidate the mechanism underlying the internalization of C-AVP_4-9 into the BBB, we examined the acid-resistant binding of ¹²⁵I-labeled C-AVP_4-9 in several conditions. The osmolarity dependence of the acid-resistant binding of ¹²⁵I-labeled C-AVP_4-9 to MBEC4 (Table 3) suggests that C-AVP_4-9 is significantly internalized into MBEC4. As shown in Table 3, the acid-resistant binding of ¹²⁵I-labeled C-AVP4-9 to MBEC4 was significantly inhibited by lower temperature, dancylcadaverine (an endocytosis inhibitor and a suppressor of the coated pit formation; Haigler et al., 1980), phenylarsine oxide (an endocytosis inhibitor and a denaturant of the SH group in the cell membrane; Knutson et al., 1983), and 2,4-dinitrophenol (a metabolic inhibitor and an uncoupler of phosphorylation). In addition, as shown in Table 4, the cationic peptides [poly(L-lysine), protamine, and AVP_4-9] significantly inhibited the acid-resistant binding of ¹²⁵I-labeled C-AVP_4-9 to MBEC4, whereas an anionic peptide [poly(L-glutamic acid)] and the V₁ and V₂ vasopressin receptor antagonists had no effect on the acid-resistant binding. These results suggest that C-AVP_4-9 is internalized into the BBB via the energy-dependent AME by its basicity, independently of the vasopressin receptors.

We also examined the conversion of C-AVP_4-9 to AVP_4-9 in the cerebrum in vitro. The conversions of C-AVP_4-9 by the authentic PPCE and the mouse cerebral homogenate were inhibited by the PPCE-specific inhibitor ZPP (Fig. 8), and the CL_int of the C-AVP_4-9 conversion by the mouse cerebral homogenate agreed well with the value by the authentic PPCE (Fig. 9), suggesting that C-AVP_4-9 is converted to AVP_4-9 by PPCE in the cerebrum.

In conclusion, the results of our studies suggest that C-AVP_4-9 is transported through the BBB via AME to the cerebral parenchyma more effectively than its parent peptide, AVP_4-9. The results also suggest that C-AVP_4-9 is converted to AVP_4-9 in the cerebrum, consistent with our molecular design concept of C-AVP_4-9. Chemical modification to obtain a peptide with basicity and convertibility to the active form in the CNS is a promising strategy for the conversion of native physiologically active peptides into neuropharmaceuticals.