Measurement of the anteroposterior translation of the humeral head using MRI

We studied the effect of recombinant murine interferons (rMuIFNs) on the growth of Orientia (formerly Rickettsia) tsutsugamushi Gilliam in mouse L929 cells. Rickettsial growth was measured by flow cytometry. rMUIFN-gamma inhibited the growth of O. tsutsugamushi at the concentrations of 100 i.u./ml and 1,000 i.u./ml in accord with previous reports. Relatively low concentrations (10 i.u./ml and 100 i.u./ml) of rMUIFN-beta also inhibited the growth of O. tsutsugamushi. On the other hand, high concentrations (1,000 i.u./ml and 10,000 i.u./ml) of rMuIFN-beta enhanced the growth of the rickettsia. This enhancement of rickettsial growth was blocked by anti-murine IFN-beta monoclonal antibody (MoAb). rMuIFN-beta also enhanced the growth of Rickettsia sibirica 246 in L929 cells to some extent.     

High-level expression of a 56-kilodalton protein gene (bor56) of Rickettsia tsutsugamushi Boryong and its application to enzyme-linked immunosorbent assays

The 56-kDa protein of Rickettsia tsutsugamushi, which is located on the rickettsial surface, has been shown to be an immunodominant antigen. The gene that encodes the 56-kDa protein of R. tsutsugamushi Boryong (bor56) was cloned. Sequencing revealed an open reading frame of 1,602 bp encoding 534 amino acids with a molecular weight of 56,803. The 56-kDa protein of R. tsutsugamushi Boryong (Bor56) was expressed as a fusion protein with the maltose-binding protein of Escherichia coli by deleting 252 bp from the 5' end of the open reading frame and subcloning it into the StuI site of pIH821. The recombinant fusion protein was purified by amylose column chromatography for application in an enzyme-linked immunosorbent assay to evaluate the ability of the method to detect the antibody to R. tsutsugamushi in human patient sera. By using sera from 100 patients with scrub typhus and 70 patients with other febrile diseases, a high diagnostic sensitivity (95%) and a high diagnostic specificity (100%) were demonstrated, suggesting the suitability of the recombinant antigen for use as an immunodiagnostic tool.     

A radiomicroassay for cytotoxic antibody to human spermatozoa. Quantification by tritiated actinomycin d

A radiomicroassay for titration of spermocytotoxic antibody is described. The assay used [3H]AACTINOMYCIN D ([3H]Act D) to label damaged spermatozoa in a fashion analogous to penetration by vital dye. Optimal conditions for and some kinetics of the assay are presented. The assay is sensitive, reliable, simple to perform and uses only small amounts of serum and spermatozoa. Applied to sperm antibody positive human postvasectomy sera, the assay compared favourably in sensitivity eith vital dye microscopic observations and with parallel titration by the Isojima's immobilization tests.     

Induction of homologous immune response to Rickettsia tsutsugamushi Boryong with a partial 56-kilodalton recombinant antigen fused with the maltose-binding protein MBP-Bor56

Although the 56-kDa protein of Rickettsia tsutsugamushi has been presumed to play important roles in generating protective immunity against scrub typhus, studies of this protein have been impeded. We used the recombinant 56-kDa protein of R. tsutsugamushi Boryong fused with the maltose-binding protein of Escherichia coli (MBP-Bor56) to analyze its ability to induce protective immunity in a C3H/HeDub murine model. Intraperitoneal immunization of mice with MBP-Bor56 resulted in an increase in the 50% minimal lethal dose of more than 160 times compared with that for the control mice. Splenic mononuclear cells from the mice immunized with MBP-Bor56 showed a dose-dependent pattern of lymphocyte proliferation response and secreted gamma interferon and interleukin-2 when stimulated with irradiated R. tsutsugamushi Boryong, which is a cytokine profile of Th1 cells. High titers of antibody to R. tsutsugamushi were also demonstrated by indirect immunofluorescent-antibody testing. These findings suggest that the 56-kDa protein of R. tsutsugamushi is one of the candidates for a vaccine against scrub typhus.     

Thymidine metabolism and DNA synthesis in Newcastle disease virus-infected cells

The inhibition of thymidine incorporation into DNA in Newcastle disease virus-infected cells has been studied. At 6 h after infection of L-929 cells at high multiplicity, transport of exogenous thymidine across the cell membrane was inhibited. The kinetics of this inhibition, decreased Vmax with no change in Km, suggest that there are fewer sites available for transport in infected cells. The conversion of thymidine to dTTP was not inhibited. Equilibrium of exogenous thymidine with the acid-soluble pool occurred more slowly and at a lower level of radioactivity than in uninfected cells, and there was a reduction in the rate of incorporation of exogenous thymidine into DNA. The reduction of incorporation into the pool and into DNA was proportionate. The size of total cellular dTTP pools was changed very little in infected cells. DNA synthesized in infected cells in the presence of [3H]BrdUrd had reduced incorporation of tritium but similar buoyant density to that from uninfected cells. The results show that Newcastle disease virus inhibits DNA synthesis directly and, in addition, decreases thymidine transport. Together these account for the overall decrease in thymidine incorporation into DNA of infected cells.     

Geranylgeraniol overcomes the block of cell proliferation by lovastatin in C6 glioma cells

It is well documented that 3-hydroxy-3-methylglutaryl-CoA reductase inhibitors prevent cultured mammalian cells from progressing through the cell cycle, suggesting a critical role for a mevalonate-derived product. Recently, it has been shown that free geranylgeraniol (GG-OH) and farnesol (F-OH) can be utilized by C6 glioma cells for protein isoprenylation. The ability of GG-OH and F-OH to restore protein geranylgeranylation or farnesylation selectively has enabled us to examine the possibility that mevalonate is essential for cell proliferation because it is a precursor of farnesyl pyrophosphate or geranylgeranyl pyrophosphate, the isoprenyl donors involved in the posttranslational modification of key regulatory proteins. In this study we report that GG-OH, as well as mevalonate, overcomes the arrest of cell proliferation of C6 glioma cells treated with lovastatin, as assessed by increased cell numbers and a stimulation in [3H]thymidine incorporation. The increase in cell number and [3H]thymidine incorporation were significantly lower when F-OH was added. Under these conditions [3H]mevalonate and [3H]GG-OH are actively incorporated into a set of isoprenylated proteins in the size range of small, GTP-binding proteins (19-27 kDa) and a polypeptide with the molecular size (46 kDa) of the smaller isoform of 2 ',3'-cyclic nucleotide 3'-phosphodiesterase. Analysis of the proteins metabolically labeled by [3H]mevalonate and [3H]GG-OH reveals the presence of labeled proteins containing geranylgeranylated cysteinyl residues. Consistent with geranylgeranylated proteins playing a critical role in the entry of C6 cells into the cell cycle, a (phosphonoacetamido)oxy derivative of GG-OH, a drug previously shown to interfere with protein geranylgeranylation, prevented the increase in cell number when mevalonate or GG-OH was added to lovastatin-treated cells. These results strongly suggest that geranylgeranylated proteins are essential for progression of C6 cells into the S phase of the cell cycle and provide the first evidence that the "salvage" pathway for the utilization of the free isoprenols is physiologically significant in the CNS.     

Diversity of the blocking effects of antisperm antibodies on fertilization in human and mouse

The blocking effects of complement-dependent sperm immobilizing antibodies in the sera of infertile women and monoclonal antisperm antibodies against humans and mice on fertilization were investigated. The hemizona assay (HZA) and sperm penetration assay (SPA) were used to study the inhibitory effects of sera from 22 infertile patients positive for sperm immobilizing antibodies. Use of these tests allowed us to differentiate whether the antibody blocked sperm-zona pellucida tight binding and/or sperm penetration into the ooplasm. The zona pellucida penetration assay (ZPA) was also used to study the effects of four monoclonal antibodies (mAbs) on human sperm penetration into the zona pellucida. Seven mAbs against murine spermatozoa were tested for their inhibitory effects on in-vitro fertilization (IVF) and HZA in mice. Of 22 patient sera with sperm immobilizing antibodies, 21 (95.5%) inhibited HZA attachment and penetration, whereas this did not occur in any of 13 patient sera without these antibodies. However, 19 of 22 (86.4%) patient sera with sperm immobilizing antibodies and eight of 13 (61.5%) patient sera without these antibodies inhibited the SPA. Two (2C6, 1G12) of four mAbs against human spermatozoa showed strong inhibitory effects in all the assays (HZA, ZPA and SPA). One mAb (3B10) did not inhibit HZA but blocked ZPA and SPA. Another mAb (H6-3C4) seemed to have no inhibitory effects on fertilization. Two (Vx 5 and Vx 8) of seven mAbs against murine spermatozoa inhibited IVF in mice but did not block mouse HZA. These findings suggest that antisperm antibodies block fertilization at specific stages. Some of them may inhibit sperm capacitation and thus prevent all processes of fertilization that follow. Some other antibodies may not affect capacitation and sperm binding to zona pellucida but inhibit the acrosome reaction, followed by the blocking of sperm penetration through zona pellucida and ooplasm.     

Endotoxin-stimulated spleen cells: dissociation between DNA synthesis and IgM production at the cellular level

LPS stimulation of mouse spleen cells in the presence of Hu resulted in almost total suppression of [3H]thymidine incorporation without affecting the percentage of cells induced to produce IgM. Utilizing a method which permitted the simultaneous measurement of IgM production and [3H]thymidine incorporation in individual cells, it was demonstrated directly that LPS stimulation of IgM production can occur in the absence of DNA synthesis.     

Down-regulation of the noradrenaline transporter by interferon-alpha in cultured bovine adrenal medullary cells

The aim of this study was to investigate the effect of long-term treatment with interferon (IFN)-alpha on the noradrenaline transporter of bovine adrenal medullary cells. Treatment of cultured adrenal medullary cells with IFN-alpha caused a decrease in uptake of [3H]noradrenaline by the cells in time (4-48 h)- and concentration (300-1,000 U/ml)-dependent manners. IFN-beta also inhibited [3H]noradrenaline uptake to a lesser extent than did IFN-alpha, whereas IFN-gamma had little effect. An anti-IFN-alpha antibody reduced the effect of IFN-alpha on [3H]noradrenaline uptake. Saturation analysis of [3H]noradrenaline uptake showed that the inhibitory effect of IFN-alpha was due to a reduction in the maximal uptake velocity (Vmax) values without altering apparent Michaelis constant (Km) values. Incubation of cells with IFN-alpha caused a translocation of protein kinase C from the soluble to the particulate fraction in the cells. The effect of IFN-alpha on [3H]noradrenaline uptake was diminished in protein kinase C-down-regulated cells. Incubation of cells with IFN-alpha for 48 h significantly reduced the specific binding of [3H]desipramine to crude plasma membranes isolated from cells. Scatchard analysis of [3H]desipramine binding revealed that IFN-alpha decreased the maximal binding (Bmax) values without any change in the dissociation constant (K(D)) values. These findings suggest that IFN-alpha suppresses the function of noradrenaline transporter by reducing the density of the transporter in cell membranes through, at least in part, a protein kinase C pathway.     

Heterogeneous effects of G-CSF and GM-CSF on cell growth and ara-C cytotoxicity in childhood leukemias which express myeloid markers

It is uncertain if acute lymphoblastic leukemia (ALL) cells expressing myeloid makers can respond to granulocyte colony-stimulating factor (G-CSF) and granulocyte macrophage colony-stimulating factor (GM-CSF). We investigated the effects of G-CSF (0.01 microgram/ml) and GM-CSF (0.01 microgram/ml) on [3H]thymidine (TdR) uptake, and the cytotoxicity of 1-beta-D-arabinofuranosylcytosine (ara-C) in leukemia cells from 17 pediatric patients. ALL cells without myeloid markers did not respond to G-CSF or GM-CSF. On the other hand, these cytokines enhanced the [3H]TdR uptake and cell growth, not only of AML cells but also of ALL cells expressing myeloid antigens. However, G-CSF and GM-CSF did not always enhance the growth inhibitory effect of the cell cycle specific drug ara-C when the cells were co-cultured with the drug. There was no relationship between cell growth and the amount of [3H]TdR incorporation or the intracellular ara-CTP level. These results indicate the heterogeneous effects of G-CSF and GM-CSF on cell growth and ara-C sensitivity in childhood leukemia cells.     

Glucuronidation of retinoids by rat recombinant UDP: glucuronosyltransferase 1.1 (bilirubin UGT)

Rat liver recombinant BR1UGT1.1 was found to have significant activity toward retinoid substrates. UGT1.1 glucuronidation activity was 91 +/- 18 pmol/mg x min for atRA and 113 +/- 19 pmol/mg x min for 5,6-epoxy-atRA. The apparent K(M) and V(max) of atRA acid glucuronidation by UGT1.1 were 59.1 +/- 5.4 microM and 158 +/- 43 pmol/mg x min, respectively. SDS-PAGE and Western blot analysis of UGT1.1-transfected HK293 membrane proteins photolabeled with [11,12-3H]atRA revealed a protein of approximately 56 kDa that was labeled by [3H]atRA, detected by anti-pNP UGT antibody and not present in membranes from nontransfected HK293 cells. Liver microsomes from Gunn rats, which lack UGT1.1, had significant activity toward atRA (111 +/- 28 pmol/mg x min).     

Nucleocytoplasmic shuttling by protein nuclear import factors

Anti-Neospora caninum antibody was detected in anti-Toxoplasma gondii positive and negative human sera by ELISA, western blot and immunofluorescence assay (IFA). Twelve cases out of 172 (6.7%) Toxoplasma-positive sera cross-reacted with both T. gondii and N. caninum antigens, and one out of 110 Toxoplasma-negative sera reacted with N. caninum antigen by ELISA. By western blot, all 12 sera reacted with T. gondii antigens with various banding patterns but specifically at 30 kDa (SAG1) and 22 kDa (SAG2) bands. With N. caninum antigen, the number of reactive bands was reduced, however a 43 kDa band reacted in three cases in Toxoplasma-positive sera in addition to one in Toxoplasma-negative control sera. All sera of the Toxoplasma-positive group labeled surface membrane of T. gondii, but reacted differently with N. caninum. Fluorescence was detected in surface membrane, subcellular organelles, or both in N. caninum. And one case in the Toxoplasma-negative group also reacted with N. caninum strongly in subcellular organelles. This suggested that the antibody against N. caninum may be present in human sera although the positive rate was very low in this study. The possibility of human infection with N. caninum remains to be evaluated further.     

Reactivities of human sera with human herpesvirus-8-infected BCBL-1 cells and identification of HHV-8-specific proteins and glycoproteins and the encoding cDNAs

The reactivates of human sera with uninduced and phorbol ester (TPA)-induced human herpesvirus-8 (HHV-8)-infected BCBL-1 cells were examined by immunofluorescence assay (IFA) and by radioimmunoprecipitation reactions (RIP). The seroprevalence of HHV-8 infections is low in the United States general population and only low levels of HHV-8 antibodies were detected in the seropositive sera. In contrast, high levels of antibodies against HHV-8 lytic and latent antigens were detected by IFA in the sera from HIV+ Kaposi's sarcoma (KS)-positive individuals. These sera recognized several proteins and glycoproteins from BCBL-1 cells in RIP reactions. Two types of antibody responses were detected in the sera from HIV+ KS- homosexual men. In majority of the sera with and without detectable HHV-8 DNA in the peripheral blood mononuclear cells (PBMC), significantly low levels of HHV-8 antibodies were detected by IFA. These sera recognized only a subset of HHV-8 proteins and glycoproteins in RIP reactions. In contrast, in a subgroup of sera from HIV+ KS- homosexual men, higher levels of IFA antibodies against HHV-8 lytic and latent antigens were detected. These sera also recognized several viral proteins and glycoproteins in RIP reactions. These results suggest that antibody response profiles to HHV-8 infection vary significantly and serologic assays to detect antibody responses to a panel of both lytic and latent antibodies may be required for maximum sensitivity. Screening of a cDNA library from TPA-induced BCBL-1 cells with an HIV+ KS+ serum identified cDNAs encoding 12 HHV-8 proteins. Further characterization of these HHV-8 proteins would define the HHV-8 antigens useful for seroepidemiological studies and in discriminating lytic, latent, past, and/or reactivation infections.     

Src family tyrosine kinase Lyn binds several proteins including paxillin in rat basophilic leukemia cells

Aggregation of the high affinity IgE receptors on rat basophilic leukemia (RBL-2H3) cells results in protein tyrosine phosphorylation although the receptor has no intrinsic enzymatic activity. The Src related protein tyrosine kinase p53/56lyn present in RBL-2H3 cells could play a role in this reaction. Here we have isolated the cDNA for rat Lyn and found it to be very homologous at the amino acid level to both the human and mouse proteins. A bacterially expressed maltose binding protein-Lyn (MBP-Lyn) fusion protein was already tyrosine phosphorylated and had tyrosine kinase activity. In a filter-binding assay, MBP-Lyn fusion protein (at 0.1 microM) specifically bound to several proteins of RBL-2H3 cells. In lysates of IgE receptor-activated cells, there was increased binding of MBP-Lyn to 65, 72, 78 and 110 kDa tyrosine phosphorylated proteins. The 72, 78 and 110 kDa tyrosine phosphorylated proteins were precipitated by a fusion protein containing the Lyn Src Homology 2 (SH2) domain. The 72 kDa Lyn binding protein was different from p72syk. Furthermore, paxillin, a cytoskeletal protein, was identified as one of the Lyn binding proteins. Thus Fc epsilon RI mediated signal transduction in RBL-2H3 cells may result from the interaction of p53/56lyn with paxillin, pp72, pp110 and other proteins.     

Involvement of a single periplasmic hydrogenase for both hydrogen uptake and production in some Desulfovibrio species

Various sulphate-reducing bacteria differing in the number of genes encoding hydrogenase were shown to ferment lactate in coculture with Methanospirillum hungatei, in the absence of sulphate. The efficiency of interspecies H2 transfer carried out by these species of sulphate-reducing bacteria does not appear to correlate with the distribution of genes coding for hydrogenase. Desulfovibrio vulgaris Groningen, which possesses only the gene for [NiFe] hydrogenase, oxidizes hydrogen in the presence of sulphate and produces some hydrogen during fermentation of pyruvate without electron acceptor. The hydrogenase of D. vulgaris was purified and characterized. It exhibits a molecular mass of 87 kDa and is composed of two different subunits (60 and 28 kDa). D. vulgaris hydrogenase contains 10.6 iron atoms, 0.9 nickel atom and 12 acid-labile sulphur atoms/molecule, and the absorption spectrum of the enzyme is characteristic of an iron-sulphur protein. Maximal H2 uptake and H2 evolution activities were 332 and 230 units/mg protein, respectively. D. vulgaris cells contain exclusively the [NiFe] hydrogenase, whatever the growth conditions, as shown by biochemical and immunological studies. Immunocytolocalization in ultrathin frozen sections of cells grown on lactate and sulphate, on H2 and sulphate and on pyruvate showed that the [NiFe] hydrogenase was located in the periplasmic space. Labelling was enhanced in cells grown on H2 and sulphate and on pyruvate. The results enable us to conclude that D. vulgaris Groningen contains a single hydrogenase of the [NiFe] type, located in the periplasmic space like that described for D. gigas. This enzyme appears to be involved in both H2 uptake and H2 production, depending on the growth conditions.     

Cytotoxic activity of BCG-activated macrophages against L929 tumor cells is nitric oxide-dependent

The tumoricidal activity of activated macrophages has been attributed largely to the release of tumor necrosis factor (TNF), or to the production of reactive oxygen or nitrogen intermediates. The L929 tumor cell line (a murine fibroblast-like cell) when treated with actinomycin D (ActD) has been used to measure TNF alpha cytotoxicity. In the present study, we determined the cytotoxic activity of BCG-activated peritoneal macrophages against ActD-untreated L929 tumor cells. Furthermore, we measured the production of hydrogen peroxide (H2O2), nitric oxide (NO) and TNF by macrophages cultured in the presence or absence of L929 cells. As expected, BCG-activated macrophages produced significant amounts of H2O2 (16.0 +/- 3.0 microM), TNF (512 U/ml) and NO (71.5 +/- 3.2 microM). TNF (256 U/ml) and NO (78.9 +/- 9.7 microM) production was unchanged in co-cultures of L929 cells with BCG-activated macrophages but H2O2 production was totally inhibited. The cytotoxic activity was dependent on NO release since L-NAME (2.5, 5.0 and 10 mM), which blocks NO synthase, inhibited the killing of L929 cells. Addition of anti-TNF (20 micrograms/ml) antibodies to the cultures did not affect the tumoricidal activity of macrophages. Our results indicate that macrophage-mediated killing of L929 cells is largely dependent on NO production but independent of H2O2 or TNF release.     

Down-regulation of the human norepinephrine transporter in intact 293-hNET cells exposed to desipramine

The effects of continuous exposure of cultured cells expressing the human norepinephrine transporter (hNET) to the hNET inhibitor desipramine on hNET expression and function were studied. Exposure of HEK-293 cells transfected stably with the hNET cDNA (293-hNET cells) to desipramine for 3 days reduced the specific binding of [3H]nisoxetine in membrane homogenates in a concentration-dependent manner. The magnitude of the reductions in [3H]nisoxetine binding to hNET was dependent on the length of time of the exposure to desipramine, reaching 77% after a 21-day exposure. The reduction of [3H]nisoxetine binding returned to control levels within 72 h after a 3-day exposure to desipramine. Reductions in [3H]nisoxetine binding to hNET were accompanied by time-dependent and exposure concentration-dependent reductions in hNET protein levels as determined by western blotting. Similar to binding, hNET protein levels returned to control levels 72 h after cessation of desipramine exposure. Northern blotting indicated that exposure of 293-hNET cells to desipramine did not significantly alter hNET mRNA levels. Uptake of [3H]norepinephrine by 293-hNET cells was markedly reduced after a 3-day exposure to desipramine. However, desipramine exposure had no effect on uptake of [3H]glutamate or [3H]alanine. The present findings imply that down-regulation of the hNET in 293-hNET cells induced by desipramine results from a selective reduction in hNET protein levels, presumably a consequence of either a reduction in the translation of hNET mRNA or from an enhanced degradation of hNET protein.     

Interruption of a transforming growth factor alpha autocrine loop in Caco-2 cells by antisense oligodeoxynucleotides

BACKGROUND & AIMS: We have previously shown that Caco-2 cell proliferation is driven by basolateral membrane epidermal growth factor receptors. The aim of this study was to investigate whether autocrine production of transforming growth factor alpha (TGF-alpha) activates these receptors and stimulates proliferation using antisense oligodeoxynucleotides. METHODS: Caco-2 cells grown on microporous membranes or Jurkat cells were exposed to conventional or 5' cholesterol-modified oligodeoxynucleotides synthesized with random, antisense, or missense base sequences. Indices of proliferation were measured, including [3H]thymidine or [3H]uridine uptake for studies of short-term stimulation and the methylthiotetrazole assay as an index of cell number increase over longer periods. Secretion of TGF-alpha by cells was detected using a soft agar bioassay. RESULTS: Incubation with antisense oligodeoxynucleotides inhibited TGF-alpha secretion compared with controls. Random and missense oligodeoxynucleotides had no effect on proliferation. The TGF-alpha antisense oligodeoxynucleotides markedly inhibited proliferation, an effect that was abolished by adding TGF-alpha to the medium. Oligonucleotides had no effect on Jurkat cells, a lymphocytic cell line lacking epidermal growth factor receptors. Cholesterol-modified oligodeoxynucleotides were more effective and specific than unmodified oligodeoxynucleotides. CONCLUSIONS: Caco-2 cell proliferation is driven by autocrine stimulation of epidermal growth factor receptors by TGF-alpha. This mechanism may be effectively inhibited by antisense oligodeoxynucleotides, particularly those modified by the 5' attachment of cholesterol.     

Identification of a saturable uptake system for deoxyribonucleosides at the blood-brain and blood-cerebrospinal fluid barriers

Substances can enter the brain either directly across the blood-brain barrier or indirectly across the choroid plexuses and arachnoid membrane (blood-CSF barrier) into the CSF and then by diffusion into the brain. Earlier studies have demonstrated a saturable thymidine uptake across the blood-CSF barrier, but not across the blood-brain barrier. In this study transport of [3H]thymidine across both barriers was measured in vivo by means of a bilateral vascular brain perfusion technique in the anaesthetised guinea-pig. This method allows simultaneous and quantitative measurement of slowly penetrating solutes into both brain and CSF, under controlled conditions of arterial inflow. The results of the present study carried out over perfusion periods of up to 30 min indicated a progressive uptake of [3H]thymidine into brain and CSF, which was found to be significantly greater than the transport of D-[14C]mannitol (a plasma space marker). Furthermore, the addition of 1 mM unlabelled thymidine in the perfusate caused saturation of [3H]thymidine uptake into both brain and CSF. In conclusion, these findings suggest that thymidine can cross both the blood-brain and blood-CSF barriers in the guinea-pig by carrier-mediated transport systems.