Address correspondence and reprints requests to Wolfgang Sattler, University of Graz, Institute of Medical Biochemistry and Molecular Biology, Harrachgasse 21, 8010 Graz, Austria. E-mail: firstname.lastname@example.org
It is clearly established that an efficient supply to the brain of α-tocopherol (αTocH), the most biologically active member of the vitamin E family, is of the utmost importance for proper neurological functioning. Although the mechanism of uptake of αTocH into cells constituting the blood–brain barrier (BBB) is obscure, we previously demonstrated that high-density lipoprotein (HDL) plays a major role in the supply of αTocH to porcine brain capillary endothelial cells (pBCECs). Here we studied whether a porcine analogue of human and rodent scavenger receptor class B, type I mediates selective (without concomitant lipoprotein particle internalization) uptake of HDL-associated αTocH in a similar manner to that described for HDL-associated cholesteryl esters (CEs). In agreement with this hypothesis we observed that a major proportion of αTocH uptake by pBCECs occurred by selective uptake, exceeding HDL3 holoparticle uptake by up to 13-fold. The observation that selective uptake of HDL-associated CE exceeded HDL3 holoparticle up to fourfold suggested that a porcine analogue of SR-BI (pSR-BI) may be involved in lipid uptake at the BBB. In line with the observation of selective lipid uptake, RT-PCR and northern and western blot analyses revealed the presence of pSR-BI in cells constituting the BBB. Adenovirus-mediated overexpression of the human analogue of SR-BI (hSR-BI) in pBCECs resulted in a fourfold increase in selective HDL-associated αTocH uptake. In accordance with the proposed function of SR-BI, selective HDL–CE uptake was increased sixfold in Chinese hamster ovary cells stably transfected with murine SR-BI (mSR-BI). Most importantly stable mSR-BI overexpression mediated a twofold increase in HDL-associated [14C]αTocH selective uptake in comparison with control cells. In line with tracer experiments, mass transfer studies with unlabelled lipoproteins revealed that mSR-BI overexpression resulted in a twofold increase in endogenous HDL3-associated αTocH uptake. The results of this study indicate that SR-BI promotes the uptake of HDL-associated αTocH into cells constituting the BBB and plays an important role during the supply of the CNS with this indispensable micronutrient.
Vitamin E is a generic term for a group of compounds known as tocopherols and tocotrienols. α-Tocopherol (αTocH) is the major constituent of the vitamin E family found in mammalian tissues and has the highest biological activity of all the tocopherols and tocotrienols. Plasma αTocH levels are controlled by oral intake, absorption and transfer to newly synthesized lipoproteins, the latter step being under the tight control of the α-tocopherol transfer protein (αTTP) (Arita et al. 1995). In rats, the major site of rat αTTP expression is the liver, however, additional αTTPs have also been identified in lung, heart and brain (Murphy and Mavis 1981), where it is predominantly localized to Bergmann glial cells (Hosomi et al. 1998). Mutations in the gene for αTTP result in a neurologic syndrome of spinocerebellar ataxia (termed ataxia with associated vitamin E deficiency; AVED), characterized by progressive ataxia, areflexia and sensory loss (Ouahchi et al. 1995). In addition to AVED, secondary vitamin E deficiency due to lipid malabsorption (e.g. cholestatic liver disease) or abetalipoproteinemia may be associated with neurological syndromes identical to ataxia (Muller 1995). From the aforementioned findings it is evident that αTocH must be effectively absorbed, transported and delivered to organs and tissues following assembly with newly synthesized lipoproteins and exchange to other circulating lipoprotein classes (Kayden and Traber 1993).
As the hydrophobic αTocH molecule is transported exclusively in association with circulating lipoproteins, the major pathways involved in the delivery of lipoprotein-associated lipids may also be involved in αTocH uptake by organs and tissues. In healthy, fasting human subjects low-density lipoprotein (LDL) carries between 50 and 70% of the circulating αTocH, with the remainder being transported by high-density lipoproteins (HDL) (Sattler et al. 1996). In comparison with humans, the total antioxidant content is approximately fourfold lower in pigs and the αTocH content of porcine LDL is 2–3-fold lower than in humans (Knipping et al. 1990). Uptake of LDL-associated αTocH is facilitated by LDL-receptor mediated pathways (Kayden and Traber 1993), although studies in rabbits lacking functional intact LDL receptors suggest that alternative uptake pathways for lipoprotein-associated αTocH must exist (Cohn et al. 1992). In line with that study it appears that, on a quantitative basis, HDL is a more potent αTocH donor system than LDL (Goti et al. 1998, 2000). Along this line, it is important to note that HDL-associated lipids are metabolized during a process that is fundamentally different from the classical LDL receptor-mediated endocytotic pathway (Goldstein and Brown 1977), namely by selective uptake of HDL lipids (reviewed in Trigatti et al. 2000) without concomitant lipoprotein particle internalization. Selective uptake of HDL-associated lipids, particularly cholesteryl esters (CEs), in rodents is mediated by scavenger receptor class B, type I (SR-BI). Major sites of SR-BI expression are the liver and steroidogenic tissues, however, low level expression in brain has been reported for murine SR-BI (mSR-BI) (Acton et al. 1996). In liver mSR-BI mediates selective uptake of HDL-associated CE and facilitates removal of peripheral, excess cholesterol. Selective uptake of HDL-associated CE by adrenals provides cholesterol that is available for steroidogenesis (Trigatti et al. 2000).
The following observations have prompted us to study whether SR-BI mediates the uptake of HDL-associated αTocH by cells constituting the blood–brain barrier (BBB): (i) HDL is a more effective αTocH donor than LDL (Goti et al. 1998, 2000); (ii) the majority of HDL-associated αTocH is taken up independent of lipoprotein particles by brain endothelial cells (Goti et al. 2000); (iii) the majority of αTocH in mice is transported in the HDL fraction (Sattler et al. 1996); (iv) the adrenal gland, the organ with the highest expression of SR-BI (Acton et al. 1996), has the highest αTocH content on a nmol/g organ basis (Bjornboe et al. 1986); and (v) female mice containing a targeted null mutation for the SR-BI gene are infertile (Trigatti et al. 1999), a major manifestation of vitamin E deficiency in rodents (Evans and Bishop 1922). For these reasons we analysed the expression of SR-BI in a porcine model of the BBB (porcine brain capillary endothelial cells; pBCECs) using RT-PCR, northern and western blot analysis. Furthermore, we assessed the effect of SR-BI overexpression in pBCECs (transient hSR-BI overexpressors generated by adenovirus-mediated transfection) and Chinese hamster ovary (CHO) cells (stable SR-BI overexpressors) on the uptake efficacy of HDL-associated αTocH.
Materials and methods
Earle's medium M199, Dulbecco's modified Eagle's medium (DMEM)/Ham's F12 (1 : 1, v/v) medium, penicillin/streptomycin, gentamicin, l-glutamine and trypsin were obtained from Biochrom (Berlin, Germany). Ox serum was from PAA Laboratories (Linz, Austria). Plastic ware for cell cultures was from Costar (Vienna, Austria). 125INa (specific activity 629 GBq/mg) and [cholesteryl-1,2,6,7-3H(N)]-palmitate ([3H]CE; specific activity 3.7 TBq/mmol) was from NEN (Vienna, Austria), [14C]all-rac-αTocH (termed [14C]αTocH throughout this paper; specific activity 2.07 GBq/mmol) was a gift from Henkel. Rabbit anti-rat SR-BI polyclonal antibody was from Abcam (Cambridge, UK). PD10 size-exclusion columns were obtained from Pharmacia (Uppsala, Sweden). All solvents were the highest available quality from Sigma and all other chemicals were from Boehringer Mannheim (Vienna). LC-18 reversed-phase HPLC columns were obtained from Supelco (250 × 4.6 mm i.d.; Vienna) or Chrompack (200 × 2 mm i.d.; Vienna). Ready Safe scintillation cocktail was from Beckman (Vienna). RNeasy kit was from Qiagen (Vienna), Poly(A+) RNA from human placenta was provided by Clontech (Heidelberg, Germany), RQ1 RNase-free Dnase I was from Promega (Mannheim, Germany), dNTPs, RNAguard and random hexamer primers were from Pharmacia Biotech Inc. (Vienna), First Strand Buffer and Moloney murine leukaemia virus reverse transcriptase were from Life Technologies Inc. (Vienna), PCR buffer and DyNAzyme II DNA polymerase were from Finnzymes Oy (Vienna), primers were from MWG Biotech (Ebersberg, Germany), and the nylon membrane for northern blotting was from Biodyne B, Pall (Vienna).
Isolation and culture of pBCECs
pBCECs were isolated by several enzymatic digestion and centrifugation steps according to Tewes et al. (1997) and characterized as described by Goti et al. (2000). After 2 days in vitro (DIV) the cells were seeded on collagen-coated multiwell cell culture clusters or 10 cm diameter Petri dishes for the experiment at a minimum density of 30 000 cells/cm2. Prior to the experiments, which were performed after 4 or 5 DIV with confluent dishes, cells were pre-incubated with serum-free medium [DMEM/Ham's F12 medium (1 : 1, v/v), containing 2 mm glutamine, 50 units/mL penicillin and 50 µg/mL streptomycin].
Isolation and labelling of HDL3
Human apolipoprotein E (ApoE)-free HDL3 was isolated from the blood of healthy volunteers by ultracentrifugation as described previously (Sattler et al. 1994). Sodium dodecyl sulfate (SDS)–PAGE analysis revealed the presence of ApoA-I as the major apolipoprotein in all preparations. Molar HDL3 concentrations were calculated from total lipoprotein mass using a molecular mass of 175 kDa (Goti et al. 2000). The protein content was measured according to Lowry et al. (1951).
HDL3[1.5 mg protein/2 mL phosphate-buffered saline (PBS)] was labelled by addition of an ethanolic [14C]αTocH solution (max. 25 µL) and incubation at 37°C (3 h under argon in a shaking water bath), and resulted in specific activities of 5–10 dpm/ng HDL3–protein. Non-lipoprotein-associated [14C]αTocH was removed by size-exclusion chromatography on PD-10 columns.
HDL3 was labelled with [3H]CE by cholesteryl ester transfer protein catalysed transfer from donor liposomes and re-isolated by ultracentrifugation in a TLX120 bench-top ultracentrifuge using the TLA100.4 rotor (Sattler and Stocker 1993). This labelling procedure resulted in specific activities of 5–9 c.p.m./ng HDL3–protein.
HDL3 was iodinated using N-Br-succinimide as the oxidizing agent (Goti et al. 1998). Routinely, 1 mCi of 125INa was used to label 5 mg of HDL3–protein. This procedure resulted in specific activities of between 300 and 450 d.p.m./ng HDL3–protein. Lipid-associated radioactivity was always < 3% total radioactivity. No cross-linking or fragmentation of ApoA-I due to the iodination procedure was detected by SDS–PAGE and subsequent autoradiography.
Cellular lipids were extracted with hexane/propan-2-ol (3 : 2, v/v) at 25°C (30 min on a plateau shaker). Lipid extracts were dried under nitrogen, redissolved in 200 µL of methanol and analysed by reversed-phase HPLC on a LC-18 column with methanol as the mobile phase (1 mL/min) and fluorescence detection (excitation and emission wavelengths 292 and 335 nm, respectively). Quantitation was performed by peak area comparison with external standards of known concentrations. The concentration of freshly prepared αTocH standards in hexane was estimated photometrically (ε292 = 3467 m−1.cm−1).
SDS–PAGE and immunoblots
SDS–PAGE of detergent-extracted pBCEC membrane proteins was performed on 8% gels with electrophoresis at 150 V for 90 min in a BioRad Miniprotean chamber. Samples for SDS–PAGE (50 µg protein/lane) were treated with sample buffer (0.1 m Tris/HCl, pH 6.8, 4% SDS, 15% glycerol and 1% mercaptoethanol) at a ratio of 1 : 1 (v/v). For western blots, proteins were electrophoretically transferred to nitrocellulose membranes (150 mA; 4°C, 90 min). Immunochemical detection of pSR-BI was performed with a sequence specific polyclonal rabbit anti-rat SR-BI IgG as primary antibody (dilution 1 : 1000). Immunoreactive bands were visualized with peroxidase-conjugated goat-anti rabbit IgG (dilution 1 : 2000) and subsequent enhanced chemiluminescence development.
Reverse-transcriptase polymerase chain reaction
Total RNA from pBCECs was isolated by using RNeasy kit. Poly(A+) RNA (3 µg) from human placenta was used as a positive control. Total RNA (3 µg) was treated with RQ1 RNase-free DNase I for 15 min at 37°C and subsequently used as a template for first-strand cDNA synthesis in a 30-µL reaction. The reaction mix contained 0.5 mm dNTPs, 15 units RNAguard, 3.3 µm random hexamer primers, 10 mm dithiothreitol, 1× First Strand Buffer and 200 units Moloney murine leukemia virus reverse transcriptase. Reactions were incubated for 1 h at 37°C, heated to 75°C for 10 min and 2.5 µL of the completed reactions were used as template for PCR. Fifty-microlitre PCR reactions contained 0.2 mm dNTPs, appropriate oligonucleotide primers at 10 µm, 1× PCR buffer and 1 unit of Finnzyme DyNAzyme II DNA polymerase. The reaction mix was heated at 94°C for 4 min, and subsequent amplification was carried out for 35 cycles (denaturation, 30 s at 94°C; annealing, 30 s at 58°C; extension 1 min at 72°C). Oligonucleotide primers used for amplification of pSR-BI cDNA were: forward primer 5′-TCGCTCATCAAGCAGCAGGT-3′ (nucleotides 169–188) and reverse primer 5′-GCCCAGAGTCGGAGTTGTTG-3′ (nucleotides 721–602). A 553-bp fragment was subsequently cloned into TA-cloning vector and sequenced revealing 91% homology to human SR-BI (data not shown).
Total RNA was prepared from pBCECs and human placenta (used as a positive control for SR-BI expression; Cao et al. 1997) by using RNeasy kit. For Northern blot analysis, 15 µg of total RNA was separated by 1% formaldehyde–agarose gel electrophoresis and subsequently blotted onto a nylon membrane. The blot was prehybridized for 6 h at 65°C and hybridized overnight at 65°C in a buffer containing 0.15 m sodium phosphate (pH 7.2), 1 mm EDTA, 7% SDS and 1% BSA. To detect pSR-BI mRNA, a 553-bp RT-PCR fragment from the 5′-end of the hSR-BI cDNA was used as a probe.
Cell culture studies
Mass transfer of HDL3-associated αTocH
pBCECs were plated on 10 cm Petri dishes and cultured as described above. Freshly isolated HDL3 was added to the medium at a final concentration of 190 ng αTocH/mL. After the indicated incubation times, cells were washed twice with PBS containing BSA (10 mg/mL), followed by two washes with PBS prior to lipid extraction and αTocH analysis.
Radioactive tracer uptake
pBCECs were cultured on six-well cluster plates. [14C]αTocH- or [3H]CE-labelled HDL3 was added to the medium at the indicated final concentrations. At the end of the incubation cells were washed as described above. Cells were lysed in NaOH (0.3 n) and an aliquot of the cell lysate was used to determine the cellular protein concentration, the remaining lysate was mixed with scintillation cocktail to determine the radioactivity. [14C]αTocH and [3H]CE–lipoprotein uptake is expressed in terms of apparent particle uptake. Based on the specific activity of the labelled lipoprotein preparations the amount of lipoprotein that would be necessary to deliver the observed amount of tracer was calculated. This is necessary to allow quantitative comparison of the data and is a result of lipoprotein particle-independent (selective) lipid uptake by nonendocytotic mechanisms (Pittman et al. 1987).
For 125I-HDL3 uptake experiments, cells were plated onto 12-well cluster plates. 125I-HDL3 was added to DMEM/Ham's F12 medium at the concentrations indicated. After incubation, cells were washed and lysed as described above. All incubations were performed in the absence or presence of a 20-fold excess of the corresponding competitor to differentiate between total and nonspecific uptake. Cell degradation of 125I-HDL3 was estimated by measuring the nontrichloroacetic acid (TCA)-precipitable radioactivity in the medium after precipitation of free iodine with AgNO3 (Goti et al. 2000).
Construction of recombinant hSR-BI adenovirus
The adenoviral plasmid shuttle vector (pAvCvSv) and pJM17 vector were kindly supplied by L. Chan (Baylor College of Medicine, Houston, TX, USA). Human SR-BI cDNA (kindly supplied by H. Hauser, ETH, Zürich, Switzerland), which was originally inserted into pcEXV-3 vector (Miller and Germain 1986), was partially restricted with EcoRI and the 2.5 kb band was eluted from gel. This band was subcloned into pBluescript using the EcoRI site, amplified, restricted with KpnI and this fragment was finally partially restricted with BamHI. Plasmid shuttle vector was opened using KpnI and BglII and KpnI/BamHI restricted hSR-BI cDNA was inserted. These modifications were necessary to enable insertion of hSR-BI cDNA under the control of the cytomegalovirus promoter in the plasmid shuttle vector. Recombinant vector (pAvCvSv/hSR-BI) was used to transform E. coli DH5-α competent cells in order to amplify recombinant plasmid. Positive clones were confirmed by restriction analysis and DNA sequencing. The resulting recombinant shuttle plasmid (5 µg) was cotransfected with 5 µg supercoiled pJM17 into 293 cells by calcium–phosphate coprecipitation. Two weeks after transfection, infectious recombinant adenoviral vector plaques were picked, propagated and screened for hSR-BI sequences using PCR. Adenoviral vectors containing hSR-BI were further expanded in 293 cells and the expression was determined by Western blotting. Large-scale production of high-titre recombinant adenovirus was performed as described previously (Teng et al. 1994). The virus was purified twice by caesium chloride density gradient centrifugation and dialysed for 14 h at 4°C against a buffer containing 10 mm Tris–HCl, pH 7.5, 1 mm MgCl2, 10% glycerol and stored at −70°C. Virus titre was determined by plaque-assay on 293 cells and was found to be 2 × 1010 p.f.u./mL. Control β-galactosidase (β-gal) and apolipoprotein(a) [Apo(a)] viruses were amplified and purified as described above.
Adenovirus infection of brain endothelial cells
Brain endothelial cells were cultivated in six-well culture dishes. At a density of 5 × 104 cells/cm2 they were rinsed once with PBS and infected with recombinant adenoviruses (MOI = 1000 p.f.u./mL) in 1 mL of infection media (DMEM medium containing 2% fetal calf serum, 50 units/mL penicillin and 50 µg/mL streptomycin) for 16 h. During infection six-well plates were rocked slowly and gently in an incubator (37°C, 5% CO2, humidified atmosphere). After removing infection media, the cells were supplied with 2 mL of S12K medium (containing 5% fetal bovine serum, 2 mm glutamine, 50 units/mL penicillin and 50 µg/mL streptomycin) and the incubation was continued for 24 h without rocking. Control cells were infected with β-gal or Apo(a) virus as described for hSR-BI-infected cells. The expression rate of hSR-BI was determined by Western blotting.
SR-BI assays in transfected CHO cells
ldlA cells (clone 7), an LDL receptor-deficient CHO line were cultured in Ham's F12 medium containing 5% (v/v) fetal bovine serum, 2 mm glutamine, 50 units/mL penicillin and 50 µg/mL streptomycin (Webb et al. 1998). Stable transfectants expressing mSR-BI were prepared as described previously using the expression vector pCMV5 (Webb et al. 1998).
The scavenger receptor class B, type I was originally identified in mice and is referred to as murine (m)SR-BI. The human homologue of SR-BI is a member of the CD36/LIMPII gene family (also termed CLA-1) and fulfils functions comparable with those described for mSR-BI (Calvo et al. 1997; Murao et al. 1997). CLA-1 is referred to as human (h)SR-BI throughout. Porcine SR-BI is referred to as pSR-BI.
Uptake of HDL3-associated αTocH by pBCECs
The following experiments were performed to clarify the uptake mechanisms of HDL3-associated αTocH and CE and to compare lipid uptake with HDL3 holoparticle uptake. For these experiments, HDL3 was labelled in the protein (125I) and in the lipid ([14C]αTocH, [3H]CE) moiety. Data from these experiments are shown in Fig. 1 and are expressed in terms of apparent particle uptake, calculated as described in Materials and methods. 125I-HDL3 holoparticle uptake was obtained by measuring the amount of degraded (i.e. TCA precipitable) and cell-associated HDL3 as described in Materials and methods. While HDL3 uptake revealed saturability, [3H]CE and [14C]αTocH uptake showed a near linear increase during the 6 h time course. At the 6 h time point both lipid tracers were taken up in excess of holoparticles, indicating efficient selective holoparticle-independent uptake. After 6 h, [3H]CE uptake exceeded 125I-HDL3 holoparticle uptake by 4.2-fold (6.7 vs. 1.6 µg/mg cell protein), an effect even more pronounced for [14C]αTocH uptake (13-fold in excess of holoparticles, 20 vs. 1.6 µg/mg cell protein). These findings clearly indicate that the uptake of HDL3-associated αTocH and HDL3-CE occurs principally via the selective uptake pathway without internalization of the lipoprotein particle, these data are in agreement with one of our previous reports (Goti et al. 2000).
PSR-BI expression by pBCECs
In order to verify SR-BI expression by pBCECs we performed RT-PCR, northern and western blot analysis. We detected significant levels of both pSR-BI mRNA and protein in pBCEC from 3 DIV until at least 10 DIV (data not shown). RT-PCR analysis (Fig. 2a) revealed that cDNA from pBCECs could be amplified using primers corresponding to the 5′-terminus of hSR-BI cDNA. Interestingly, we were not able to identify a RT-PCR product when primers corresponding to the 5′-terminus of mSR-BI (data not shown) were used. It is evident that the RT-PCR product from pBCECs (lane 2) co-migrated with the product obtained after amplification of RNA isolated from human placenta (lane 3), a tissue with relatively high levels of SR-BI expression (Cao et al. 1997). During northern blot analysis a major band of 2.8 kb was detected (Fig. 2b, lane 1), also co-migrating with the corresponding product from human placenta (lane 2) which was used as a positive control.
pSR-BI expression was also tested by immunoblot analysis (Fig. 3). Detergent-solubilized membrane proteins were isolated from CHO cells (ldlA7 and mSR-BI overexpressors) and pBCECs. In ldlA7 cells a major immunoreactive band with an apparent molecular mass of 84 kDa was detected (Fig. 3, lane 1). A second, very faint band was detectable at 81 kDa. As observed for ldlA7 cells, membrane proteins isolated from pBCECs revealed similar immunoreactivity with a major band detected at 84 kDa and a faint band at 81 kDa (Fig. 3, lane 2). In protein extracts obtained from mSR-BI overexpressing CHO cells, the major band was detected at 84 kDa (Fig. 3, lane 3) and the 81 kDa band was not resolved on this gels.
SR-BI contributes to selective uptake of HDL-associated αTocH in transiently transfected pBCECs
To determine whether SR-BI is responsible for the high rates of selective HDL–αTocH uptake by pBCECs, we transiently transfected pBCECs with the hSR-BI gene using adenoviral infection. In the same experiments we also tested whether pSR-BI expression in wild-type pBCECs is regulated by exogenous cholesterol. In Fig. 4 immunoblot analyses of wild-type and adenovirus-infected pBCECs are shown. In the first two lanes western blots of detergent-solubilized protein extracts of pBCECs precultured in lipoprotein-depleted serum (LPDS) (lane 1) or fetal calf serum (lane 2) are displayed. In line with the data shown in Fig. 3, we observed two immunoreactive bands with an apparent molecular mass of 84 and 81 kDa, respectively. Interestingly pSR-BI expressed at the BBB is regulated by the cholesterol content of the cellular supernatant. Following 14 h pre-incubation under cholesterol-depleted conditions pSR-BI expression by pBCECs was induced approximately threefold [Fig. 4, lane 1 (LPDS) vs. lane 2 (fetal calf serum), densitometric evaluation]. Adenoviral transfection of pBCECs under the conditions described in Materials and methods resulted in a high level of overexpression of hSR-BI 36 h post infection (Fig. 4, lane 3). The gene product was localized predominantly at the plasma membrane as determined by immunoblot analysis of plasma membrane preparations (data not shown).
In order to test the functionality of the adenoviral hSR-BI construct, pBCECs were first incubated in the presence of [3H]CE-labelled HDL3. In line with the role of SR-BI during selective uptake of HDL-associated CEs, we observed a twofold increase in selective CE uptake by SR-BI overexpressors (4.6 vs. 9.4 µg HDL3 protein/mg cell protein; Fig. 5). Most importantly, adenoviral overexpression of hSR-BI in pBCECs induced αTocH accumulation that was ≈ 4 times higher than in wild-type and mock, constructs containing a β-gal or Apo(a) insert, transfected cells; apparent particle uptake of 24.7 (hSR-BI), 6.6 (wild-type), 7.3 (β-gal) and 6.4 Apo(a) µg/mg cell protein (Fig. 5). From these results it is evident that adenovirus-mediated transient overexpression of hSR-BI greatly enhances the efficacy of pBCECs for selective uptake of HDL-associated αTocH.
SR-BI mediates selective uptake of HDL-associated αTocH in permanently transfected CHO cells
To further corroborate the results obtained with adenoviral hSR-BI transient transfection of pBCECs, we used stably transfected CHO cells expressing high levels of mSR-BI. A LDL receptor-deficient CHO clone (ldlA7) was used as the control. In order to confirm the expression of functional SR-BI, uptake experiments with [3H]CE-labelled HDL3 were performed. In line with another report (Fluiter et al. 1999) using the same cell lines, the stable transfectants revealed sixfold higher (at 100 µg HDL-protein) selective HDL–CE uptake compared with control cells (Fig. 6a). In agreement with data shown in Fig. 1αTocH uptake in control cells exceeded CE uptake substantially (up to 20-fold). Most importantly, stable overexpression of SR-BI led to a further twofold increase in αTocH uptake by the SR-BI overexpressing cells (Fig. 6a). Holoparticle uptake was increased slightly in the mSR-BI overexpressing CHO cells (at the highest concentration approximately twofold higher; 75 vs. 35 ng HDL3 protein/mg cell protein; Fig. 6a, inset). However, these values represented only 1% of selective uptake of αTocH and cannot account for the increase in selective αTocH uptake by the overexpressors.
A major question arising from the results shown in Fig. 6(a) is whether uptake of [14C]αTocH reflected true mass transfer or merely tracer exchange. To clarify this issue, control and SR-BI overexpressing CHO cells were incubated in 10 cm Petri dishes and received 125 µg HDL3 protein corresponding to 190 ng endogenous αTocH. Lipoproteins were analysed for their αTocH content immediately after isolation and revealed an average content of 1.8 nmol αTocH/mg HDL3 (total lipoprotein mass). This corresponds to a molar ratio of 0.3 mol αTocH/mol HDL3. At the indicated time points, cells were washed and analysed for their αTocH content by HPLC analysis of the cellular lipid extracts as described in Materials and methods. From the data shown in Fig. 6(b), it is evident that overexpression of SR-BI led to a pronounced increase (twofold) in αTocH mass transfer from HDL3 particles, closely reflecting results obtained with [14C] αTocH-labelled HDL3 (Fig. 6a). Over the 6 h time course investigated the αTocH content increased from 13 ± 9 (endogenous, cellular αTocH content at time zero) to 42 ± 3.1 (controls) and 85 ± 4 (SR-BI overexpressors) ng αTocH/mg cell protein indicating enhanced uptake rates by the overexpressors vs. wild-type cells (12.0 vs. 4.8 ng/h/mg cell protein). These data clearly demonstrate that SR-BI plays an important role during selective uptake of HDL3-associated αTocH uptake by cells constituting the BBB.
The presence of tight junctions at the BBB provides a gate-keeping function and reduces opportunities for the movement of molecules, and thus require the existence of specific transport/receptor systems for the passage of certain critical compounds into the brain (Rubin and Staddon 1999). In light of the entry of αTocH into the brain it is important to note that the LDL receptor (Dehouck et al. 1997), HB2, a HDL-binding protein (Matsumoto et al. 1997), and scavenger receptors (de Vries et al. 1993) are expressed at the BBB and in the brain. In addition, class A scavenger receptors contributing to the uptake of chemically modified LDL were identified in bovine brain microvessels (Lucarelli et al. 1997) and bovine brain endothelial cells (de Vries et al. 1993). Although lipoprotein receptors are mainly thought of as necessary systems contributing to cholesterol homeostasis, they also deliver vitamins that exert critical functions in the regulation of cellular processes and lipoprotein binding and uptake may be of particular importance for vitamin transport across the BBB.
In a previous report (Goti et al. 2000) we were able to demonstrate that αTocH delivery to the BBB is more effective from HDL particles than from LDL particles. In addition, HDL-associated αTocH is taken up in a holoparticle-independent manner (Goti et al. 1998, 2000), a fact compatible with SR-BI-mediated selective uptake. The physiological significance of SR-BI expression in liver and steroidogenic tissues during HDL metabolism and steroidogenesis was clarified in mice in which the levels of SR-BI expression were manipulated by transgenic approaches (Trigatti et al. 2000). In this study we were able to demonstrate for the first time SR-BI expression in cells constituting the BBB. We identified two immunoreactive pSR-BI bands with apparent molecular masses of 84 and 81 kDa, findings in line with another report (Murao et al. 1997). Whether this is a result of a differential splicing process is not clear. In line with sterol regulatory element binding protein-dependent transcription of SR-BI (Lopez and McLean 1999), we observed cholesterol-dependent regulation of SR-BI expression in our in vitro BBB model.
An important question arising from this study is the physiological relevance of SR-BI expression at the BBB. Using two different SR-BI overexpressing cell systems (adenoviral transfection of pBCECs with hSR-BI and CHO cells stably overexpressing mSR-BI) we obtained clear evidence that the capacity for selective HDL–CE and HDL–αTocH uptake was dependent on SR-BI protein levels. While selective HDL–CE uptake is of utmost importance in liver and steroidogenic tissues (see above), the contribution of this pathway to cholesterol uptake across the BBB might be less significant. The major supply for growth and differentiation of the brain and CNS results from endogenous cholesterol biosynthesis (Turley et al. 1998). This might also explain the major abnormalities during brain development in patients suffering from the Smith-Lemli-Opitz syndrome, which is characterized by Δ7-reductase deficiency (Tint et al. 1994).
Along this line, it is noteworthy that SR-BI mediates bi-directional cholesterol flux, i.e. SR-BI is also involved in cellular cholesterol efflux (Rothblat et al. 1999). In peripheral tissues excess cholesterol stores are mobilized by an ApoA-I and SR-BI dependent pathway and subsequently esterified by the lecithin-cholesterol-acyl transferase reaction (Rothblat et al. 1999). The brain appears to use a different strategy to maintain cholesterol homeostasis, namely conversion of cerebral cholesterol to 24-hydroxycholesterol and subsequent release into the peripheral circulation (Björkhem et al. 1998). As 24-hydroxycholesterol is neurotoxic (Kölsch et al. 1999), it is conceivable that this compound has to follow an efficient elimination route. Whether SR-BI at the BBB contributes to this process in a fashion similar to that described for reverse cholesterol transport is currently under investigation.
The findings of this study, and indirect evidence from other studies (Bjornboe et al. 1986; Sattler et al. 1996; Goti et al. 1998; Trigatti et al. 1999; Orso et al. 2000), strongly support a key function of SR-BI facilitating selective uptake of HDL-associated αTocH at the BBB (and possibly other organs). Selective uptake of αTocH at the BBB is most probably the first step facilitating the supply of the brain with this micronutrient generating a pool of nonlipoprotein-associated αTocH in brain endothelial cells. However, the subsequent steps of paracellular αTocH transport are obscure and we are currently investigating whether specialized lipid domains, such as rafts or caveolae, are involved in the transport of αTocH across the BBB. Another question is the physiological relevance of αTTP expression in brain where it is localized in Purkinje cells (Copp et al. 1999) and surrounding Bergmann glia cells (Hosomi et al. 1998). The authors of the latter study suggested that αTocH uptake by Bergmann glia could occur via the long processes from the basement membrane of the blood vessels. Subsequently, αTTP could facilitate the assembly of αTocH with newly synthesized ApoE-containing lipoprotein-like particles that are secreted by Bergmann glia thus supplying Purkinje cells (and possibly other cerebellar cells) with αTocH. This concept could provide a basis for intercellular αTocH transport in the cerebellum. In contrast, neuronal transport mechanisms of αTocH from the cell body to the axons are largely unknown. However, the expression of several members of the LDL receptor superfamily by neurons (Posse de Caves et al. 2000) may facilitate the uptake (and probably anterograde neuronal transport) of αTocH in association with ApoE-containing lipoproteins. Along this line it is important to note that SR-BI expression was not detectable in cultured rat sympathetic neurons (Posse de Caves et al. 2000). The possibility that cubilin (predominantly expressed in polarized cells; Hammad et al. 1999) and/or megalin could contribute to the neuronal uptake of HDL-associated αTocH warrants further investigation.
Free radicals have been implicated in the pathogenesis of neurodegenerative diseases such as amyothropic lateral sclerosis, Alzheimer's disease and Parkinson's disease (reviewed in Sun and Chen 1998), and it is possible that αTocH which is transported across the BBB to the deeper regions of the CNS beneficially interferes with these oxidative processes. Another prominent feature of neurodegenerative diseases is the presence of activated microglial cells (Aschner et al. 1999). Although microglia may play a role during normal development and repair in the brain, chronic activation of microglia may be injurious. This is most probably a result of the production of pro-inflammatory cytokines and induction of inflammation-related enzymes such as the NADPH–oxidase complex, inducible nitric oxide synthase or cyclooxygenase. Reactive oxygen intermediates that are produced by these enzymatic systems are thought to contribute to neuronal injury although the exact underlying mechanisms are not clear. As microglia are defined as CNS-resident macrophages, some of the microglial enzymatic systems implicated in neuronal injury, such as the NADPH–oxidase complex, are activated by pathways similar to those described for monocytes (Della Bianca et al. 1999), and it seems conceivable that some of the modulatory properties of αTocH during monocyte activation could be extrapolated to microglia activation. Therefore, in analogy to activated monocytes, αTocH could interfere with microglia activation: for example, inhibition of PKC activity by αTocH (Ricciarelli et al. 1998) results in attenuated phosphorylation and translocation of p47phox, one of the cytosolic subunits of the NADPH–oxidase complex in human monocytes (Cachia et al. 1998), thereby decreasing superoxide production. Whether αTocH is able to modulate chemokine production by glial cells as described for monocytes (Deveraj et al. 1996) remains to be elucidated. αTocH-mediated modulation of adhesion molecule expression at the BBB, as observed in activated human monocytes obtained from αTocH-supplemented donors (Deveraj et al. 1996), could affect the entry of leukocytes across the BBB into the brain (Miller 1999). Finally, Heppner et al. (1998) were able to demonstrate a direct effect of αTocH on microglia ramification, a process indicative for immunological deactivation. Whether αTocH is able to induce differential gene expression in the CNS is currently an open question.
In summary, SR-BI-mediated selective uptake of HDL-associated αTocH into the BBB and subsequent redistribution to deeper brain regions by as yet undefined mechanisms could provide an important pathway facilitating a continuous supply of this indispensable micronutrient to the CNS.
We thank R. Parks for technical assistance. This work was supported by grants of the Austrian Research Foundation (P12000 Med and P14109Gen to WS; P14186 to EM), the Austrian National Bank (8778 to EM and 8127 to WS) and NIH grant R01 HL59376 (to DvdW).