In vitro models for the blood-brain barrier.

The aim of the present study was to identify a model for the blood-brain barrier based on the use of a continuous cell line, and to investigate the specificity of this model. A set of test compounds, reflecting different transport mechanisms and different degrees of permeability, as well as different physiochemical properties was selected. In vivo data for transport across the blood-brain barrier of this set of test compounds was generated as part of the study using two different in vivo models. A computational prediction model was also developed, based on 74 proprietary Pharmacia compounds, previously tested in one of the in vivo models. Molsurf descriptors were calculated and 21 descriptors were correlated with log(Brain(conc.)/Plasma(conc.)) using partial least squares projection to latent structures (PLS). However, the correlation between predicted and measured values was found to be rather low and differed between one and two log units for several of the compounds. The test compounds were analyzed in vitro using primary bovine and human brain endothelial cells co-cultured with astrocytes, and also using two different immortalized brain endothelial cell lines, one originating from rat and one from mouse. Cell models using cells not derived from the blood-brain barrier, ECV/C6, MDCK and Caco-2 cell lines, were also used. No linear correlation between in vivo and in vitro permeability was found for any of the in vitro models when all compounds were included in the analysis. The highest r2 values were seen in the bovine brain endothelial cells (r2=0.43) and MDCKwt (r2=0.46) cell models. Higher correlations were seen when only passively transported compounds were included in the analysis, bovine brain endothelial cells (r2=0.74), MDCKwt (r2=0.65) and Caco-2 (r2=0.86). By plotting in vivo Papp values against logDpH7.4 it was possible to classify compounds into four different classes: (1) compounds crossing the blood-brain barrier by passive diffusion, (2) compounds crossing the blood-brain barrier by blood-flow limited passive diffusion, (3) compounds crossing the blood-brain barrier by carrier mediated influx, and (4) compounds being actively excreted from the brain by active efflux. Papp and Pe values obtained using the different in vitro models were also plotted against logDpH7.4 and compared to the plot obtained when in vivo Papp values were used. Several of the in vitro models could distinguish between passively distributed compounds and efflux substrates. Of the cell lines included in the present study, the MDCKmdr-1 cell line gave the best separation of passively and effluxed compounds. Ratios between AUC in brain and AUC in blood were also calculated for six of the compounds and compared to ratios between Pe or Papp for transport in the apical to basolateral and basolateral to apical direction. Again the MDCKmdr-1 cell line gave the best correlation with only one compound (AZT) giving large discrepancy between in vitro and in vivo data. None of the in vitro models could identify compounds known to be substrates for carrier mediated influxed as such, and the results indicate that a tighter in vitro blood-brain barrier model probably is needed in order to facilitate studies on carrier mediated influx. The findings presented also indicate that identification of "batteries" of in vitro tests are likely to be necessary in order to improve in vitro-in vivo correlations and to make it possible to perform acceptable predictions of in vivo brain distributions from in vitro data.

Stable suppression of MDR1 gene expression and function by RNAi in Caco-2 cells.

Vector-based RNAi was used to establish a stable Caco-2 cell line with a persistent knockdown of multidrug resistant gene 1 (MDR1) and P-glycoprotein (P-gp). Several positive clones were collected, many of which showed significantly reduced levels of MDR1 mRNA and P-gp compared to wt Caco-2 cells. Selected clones were sub-cultivated for six passages and real-time PCR showed that MDR1 expression remained significantly reduced (up to 96%) over this period of time. RNAi-MDR1 clones frozen long term also kept their low MDR1 expression levels when re-cultured. Permeability studies were performed across RNAi-MDR1 clone cell monolayers, and the efflux of cyclosporine A, digoxin, vinblastine, and vincristine showed 58%, 61%, 91%, and 78% decrease in active transport, respectively, compared to wt Caco-2 cells. This stably modified Caco-2 cell line provides a novel tool for studies on MDR1 and other ABC transporter protein gene cellular functions.

Correlation of gene expression of ten drug efflux proteins of the ATP-binding cassette transporter family in normal human jejunum and in human intestinal epithelial Caco-2 cell monolayers.

This investigation describes the expression and interindividual variability in transcript levels of multiple drug efflux systems in the human jejunum and compares the expression profiles in these cells with that of the commonly used Caco-2 cell drug absorption model. Transcript levels of ten-drug efflux proteins of the ATP-binding cassette (ABC) transporter family [MDR1, MDR3, ABCB5, MRP1-6, and breast cancer resistance protein (BCRP)], lung resistance-related protein (LRP), and CYP3A4 were determined using quantitative polymerase chain reaction in jejunal biopsies from 13 healthy human subjects and in Caco-2 cells. All genes except ABCB5 were expressed, and transcript levels varied between individuals only by a factor of 2 to 3. Surprisingly, BCRP and MRP2 transcripts were more abundant in jejunum than MDR1 transcripts. Jejunal transcript levels of the different ABC transporters spanned a range of three log units with the rank order: BCRP approximately MRP2 > MDR1 approximately MRP3 approximately MRP6 approximately MRP5 approximately MRP1 > MRP4 > MDR3. Furthermore, transcript levels of 9 of 10 ABC transporters correlated well between jejunum and Caco-2 cells (r2 = 0.90). However, BCRP exhibited a 100-fold lower transcript level in Caco-2 cells compared with jejunum. Thus, the expression of a number of efflux protein transcripts in jejunum are equal to, or even higher than, that of MDR1, suggesting that the roles of these proteins (in particular BCRP and MRP2) in intestinal drug efflux have been underestimated. Also, we tentatively conclude that the Caco-2 cell line is a useful model of jejunal drug efflux, if the low expression of BCRP is taken into account.

Automated absorption assessment using Caco-2 cells cultured on both sides of polycarbonate membranes.

PURPOSE: To increase the capacity of in vitro absorption assessment and to decrease the amount of substance needed to perform early mechanistic investigations. METHODS: A liquid handling system, combined with a shaker and heating plates, was used to automate the Caco-2 cell based in vitro absorption assessment assay. In order to decrease the amount of substance needed for early mechanistic studies, a method for culturing Caco-2 cells on the lower side of polycarbonate membranes was also developed. RESULTS: Similar results were obtained with the automated assay as compared to manually performed assays. Data presented suggest that active transport and efflux were decreased in cells cultured on the lower side of the membranes as compared to ordinary seeded cells. CONCLUSIONS: Implementation of a liquid handling system for in vitro absorption assessment as reported here decrease the manual workload and increases the capacity of this in vitro assay substantially. Caco-2 cells cultured on the lower side of polycarbonate membranes, as described in this article, can not be used for analysis of transport mechanisms.

Binding of tellurium to hepatocellular selenoproteins during incubation with inorganic tellurite: consequences for the activity of selenium-dependent glutathione peroxidase.

The metallic group XVIa elements selenium and tellurium possess remarkably similar chemical properties. However, unlike selenium, tellurium is not an essential micronutrient and, indeed, induces both acute and chronic toxicity in a variety of species. Despite this, very little is known of the molecular mechanisms of toxicity of tellurium, particularly with respect to potential chemical interactions with selenium-containing components in the cell. In this work we describe a novel interaction of inorganic tellurite with hepatocellular selenoproteins, particularly with selenium-dependent glutathione peroxidase. The accumulation of (121Te)-tellurite into cultured primary rat liver hepatocytes was shown to be much more rapid than that of (75Se)-selenite on a molar basis. Neither the uptake of (121Te)-tellurite nor of (75Se)-selenite was affected by a large molar excess of the unlabelled counterpart, respectively. Interestingly, separation of the hepatocellular proteins on continuous pH denaturing gels demonstrated clear binding of radiolabelled tellurium to a number of protein bands, including one at 23 and one at 58 kDa, which corresponded to proteins readily labelled in cells treated with (75Se)-selenite. The binding of (121Te) to these proteins was insensitive to reduction with mercaptoethanol and not affected by pre-treatment of the cells with cycloheximide. When purified selenium-dependent glutathione peroxidase was treated directly with (121Te)-tellurite, the protein became labelled in an analogous manner to that achieved in intact cells. This was not affected by coincubation of the enzyme with (121Te)-tellurite and one or both of its substrates. Additionally, incubation of the peroxidase with tellurite effectively inhibited its ability to catalyse glutathione-dependent reduction of hydrogen peroxide. These data suggest that inorganic tellurite delivers tellurium to the intracellular milieu in a form capable of binding to some intracellular selenoproteins and at least in the case of glutathione peroxidase, cause inhibition of catalytic activity. The nature of the binding seems not to be due to the insertion of tellurocysteine into the protein and the insensitivity to reductive cleavage with mercaptoethanol seems to preclude the formation of stable telluro-selenides in the proteins. These data may offer alternative explanations for the established toxicity of tellurium via disruption of selenoprotein function, particularly by the induction of intracellular oxidative stress by the inhibition of Se-dependent glutathione peroxidase.

Changes in liver fatty acid-binding protein in rat enzyme-altered foci.

The level of liver fatty acid-binding protein (L-FABP) was analyzed in enzyme-altered foci (EAF) positive for GST-P, or after classification of foci into different subclasses by haematoxylin and eosin staining. Rats were treated with either an initiating single dose of diethylnitrosamine (DEN) followed by no treatment, treatment with phenobarbital, PCB, nafenopin or repeated injections of DEN, or alternatively non-treated or treated with nafenopin alone. Changes in the level of L-FABP were detected in the majority of EAF and both L-FABP-positive and -negative foci were seen. However, in rats initiated with DEN, EAF were almost exclusively L-FABP-negative. The fraction of L-FABP-negative foci increased with increasing foci size, while the time of treatment or the dose of the promoter did not seem to have any effect. It was also found that treatment with DEN gave a higher fraction of L-FABP-negative foci as compared to treatment with phenobarbital or PCB, indicating a specific effect of DEN. These data together with previously published findings suggest that L-FABP expression in EAF is determined by the initiating carcinogenic regimen and that it might be possible to use the expression of L-FABP in tumours to differentiate initiating chemicals.

In vitro models of the blood-brain barrier.

In this literature review on in vitro models of the blood-brain barrier (BBB), it is concluded that there is a need to identify a unified in vitro model for the BBB. The best evaluated model at present is based on the use of primary cultures of bovine brain endothelial cells. Primary cell cultures are usually shown to retain several BBB characteristics, but are time-consuming and difficult to establish. To make a unified in vitro model for the BBB more generally available, it is strongly suggested that such a model should be based on the use of an established cell line. To identify the best in vitro model, an evaluation of the most promising immortalised BBB-derived endothelial cell lines, as well as other established cell lines presently used as BBB models, is highly recommended. An evaluation of possible species variation is also important, in order to establish the most relevant species to be used. Furthermore, it is also suggested that the specific properties of in vitro BBB models, as compared to models for the "intestinal barrier", for example, should be evaluated. Finally, it is recommended that an evaluation of available computer models is performed, to further improve early predictions for drug candidates with regard to BBB permeability.

Decreased glutathione peroxidase activity in mice in response to nafenopin is caused by changes in selenium metabolism.

The activity of selenium-dependent glutathione peroxidase is known to be reduced in the liver of both rats and mice after exposure to nafenopin, as well as other peroxisome proliferators. The mechanism for this down-regulation is not known, but might involve changes in incorporation of selenium into selenoproteins. In this paper we show that both incorporation of selenium into selenoproteins and the level of selenium in liver is reduced in mice treated with nafenopin. The activity of selenium dependent glutathione peroxidase (GPx), as well as incorporation of selenium into its 23 kD subunit were found to be decreased. Contrary to what might have been expected, the decreased GPx activity was detected concomitantly with a slight increase in mRNA levels after 10 days of treatment, while a small decrease in mRNA levels was detected in treated animals after 26 weeks, together with the decrease in GPx-activity. Incorporation of selenium into liver fatty acid binding protein (L-FABP) was also decreased, even though large increases in protein and mRNA levels were detected. Taken together these data suggest that the decrease in GPx-activity in response to nafenopin is due to post-transcriptional mechanisms, involving changes in selenium metabolism.

The involvement of selenium in peroxisome proliferation caused by dietary administration of clofibrate to rats.

The effects of dietary treatment with clofibrate (0.5% w/w for 10 days) on the livers of selenium-deficient male rats were examined. The peroxisome proliferation (as determined by electron microscopy) in the livers of selenium-deficient animals was much less pronounced than in the case of selenium-adequate rats and no increase in peroxisomal fatty acid beta-oxidation (assayed both as antimycin-insensitive palmitoyl-CoA oxidation and lauroyl-CoA oxidase activity) was observed in the deficient animals. On the other hand, in selenium-deficient rats clofibrate caused increases in the specific activity of microsomal lauric acid omega- and omega-1-hydroxylation and an apparent change in mitochondrial size, seen as a redistribution of mitochondria from the 600 x g(av) pellet to the 10,000 x g(av) pellet, which were approximately 50% as great as the corresponding effects on control animals. Obviously, then, these three different effects of clofibrate are not strictly coupled and may involve at least partially distinct underlying mechanisms. Initial experiments demonstrated that peroxisome proliferation could be obtained by exposing primary hepatocyte cultures derived from selenium-deficient rats to clofibric acid (an in vivo hydrolysis product of clofibrate which is the proximate peroxisome proliferator), nafenopin or mono(2-ethylhexyl)phthalate. This finding suggests that selenium deficiency does not have a direct influence on the basic process(es) underlying peroxisome proliferation, but rather has indirect effects, influencing, for example, the pharmacokinetics of clofibrate and/or hormonal factors.

Peroxisome proliferation and resistance to hydrogen peroxide in rat hepatocytes: is development of resistance an adaptation to cytotoxicity?

In this work the resistance of peroxisome-proliferated hepatocytes to hydrogen peroxide (H2O2) has been studied. The question has been raised as to whether this resistance is a response to cytotoxicity. In an initial series of experiments, hepatocytes were isolated from rats that had been treated with nafenopin (NAF-hepatocytes). Isolated cells were exposed to a H2O2-generating system or to H2O2 in pulses. The ability to attach to collagen was used as a toxicological endpoint. Loss of attachment was found to be correlated to glutathione (GSH) depletion, and NAF-hepatocytes were more resistant to GSH depletion and to loss of attachment induced by H2O2 than were control hepatocytes. NAF-hepatocytes were not resistant to hydroquinone or to adriamycin. It was also indicated that this resistance was related to an altered metabolism of H2O2, less dependent on GSH. In a second series of experiments, hepatocytes from altered hepatic foci-bearing rats, treated with nafenopin or di(2-ethylhexyl)phthalate (DEHP), were used. This model was used in an attempt to monitor the development of resistance in different subpopulations of hepatocytes. It was found that the majority of hepatocytes developed resistance towards H2O2, and that, for example, foci marker-positive hepatocytes were as resistant as marker-negative cells. In control experiments with this model, it was found that marker-positive cells were more resistant towards diethyl maleate (DEM) or phorone than were marker-negative cells. In addition to demonstrating the validity of the model, these control experiments indicate an increased steady-state level of H2O2 in cells from peroxisome proliferator-treated rats. Other control experiments suggested that a low GSH-peroxidase activity protected from, rather than aggravated, the effect of peroxisome proliferation on marker-negative and GSH-depleted cells. It is concluded that H2O2 metabolism may affect the function of collagen receptors, but that a shift in H2O2 metabolism, so that it becomes less dependent on GSH, conferred resistance to this effect. The apparent non-focal induction of resistance to peroxisome proliferators, as opposed to the focal induction of resistance induced by most liver carcinogens, may explain the lack of development of gamma-glutamyltranspeptidase-positive foci in peroxisome proliferator-treated rats.

Studies on Se incorporation in selenoproteins; effects of peroxisome proliferators and hydrogen peroxide generating system.

The objective of this study was to characterize the influence of peroxisome proliferation on the metabolism of physiological concentrations of Se. In an initial series of experiments hepatocytes in primary cultures and isolated from ordinary-fed rats, were used. The cells were exposed to 75Se-selenite (30 nM) and after 24 h the labelling of selenoproteins was analysed with SDS-PAGE. Treatments with mono(2-ethylhexyl)phthalate (MEHP; a metabolite of di(2-ethylhexyl)phthalate (DEHP)), nafenopin, decreased oxygen tension and a H2O2 generating system decreased the labelling of a 23-kDa and a 15-kDa protein. The decreased labelling of the 23- and the 15-kDa proteins was usually accompanied by an increased labelling of a 58-kDa protein. Increased oxygen tension induced uncertain effects, possibly due to toxicity. In order to further evaluate the validity of the model, the labelling was also studied in hepatocytes isolated from Se-deficient and torula yeast-fed rats. In these cells there was a decreased labelling of the 23-kDa protein as compared to cells from Se-supplemented controls when 100 nM selenite was used. In in vivo experiments it was found that a DEHP-induced decrease in glutathione peroxidase (GSH-Px) activity was potentiated by high doses of selenite. To a large extent, the labelling data are compatible with enzyme activity data and in vivo data. For example, the decreased labelling of the 23-kDa protein may reflect the decreased GSH-Px activity. It is concluded that the effects induced by MEHP on Se-labelling can be explained by an increase in the steady state level of H2O2.

Selenium metabolism in isolated hepatocytes: inhibition of incorporation in proteins by mono(2-ethylhexyl)phthalate, a metabolite of the peroxisome proliferator di(2-ethylhexyl)phthalate.

Di(2-ethylhexyl)phthalate (DEHP), a rat liver carcinogen, induces peroxisomal proliferation and a concomitant oxidative stress, but decreases liver glutathione peroxidase (GSH-Px) activity. This enzyme is a selenoprotein and we have investigated the influence of mono(2-ethylhexyl)phthalate (MEHP), a major metabolite of DEHP, on selenium incorporation in hepatocellular proteins. [75Se]Selenious acid (6 nM) was added to primary cultures of rat hepatocytes and protein incorporation was assessed by SDS-PAGE and autoradiography. High concentrations of MEHP (1.0-3.0 mM) inhibited selenium labeling of all major selenoproteins in 3-24 h experiments, but also inhibited protein synthesis as assessed by leucine incorporation. The protein synthesis inhibition was reversible. Lower concentrations of MEHP (0.3-0.5 mM) did not decrease the 75Se-labeling in 24 h experiments and did not inhibit leucine incorporation. However, conditions that significantly induced peroxisomal proliferation also affected the 75Se-labeling. Thus in 72 h experiments, 0.05-0.25 mM MEHP increased the labeling of a 58 kd protein, decreased the labeling of a 23 kd protein (with the same mol. wt as GSH-Px), had no effect on a 20 kd protein and decreased the labeling of a 15 kd protein (as compared to MEHP-free control plates). The pattern of changes associated with peroxisomal proliferation mimicked that seen in livers from selenium-deficient animals, as reported by others. These data indicate that the bioavailability of selenium is decreased by DEHP. This effect may relate to a transient inhibition of protein synthesis, but also to the DEHP-induced peroxisomal proliferation.

Studies of the role of DNA fragmentation in selenium toxicity.

The role of DNA damage in selenite cytotoxicity was studied in isolated hepatocyte model systems. An initial series of experiments, with hepatocytes in suspension, indicated that selenite-induced DNA fragmentation was oxygen dependent and could be inhibited by cyanide, HgCl2 and CuDIPS. These findings were interpreted to imply that selenite-induced redox cycles were involved in this effect. In a second series of experiments, the effect of inhibitors of poly(ADP-ribose)polymerase (3-aminobenzamide and theophylline) and DNA alkylating agents on selenite-induced cellular lysis was studied. These experiments were performed with hepatocytes in primary culture and 20-30 microM selenite lysed the cultured cells after about 20 hr exposure. It was found that alkylators added 20 hr before selenite acted synergistically with selenite, and that inhibitors of poly(ADP-ribose)polymerase antagonized lysis. Further studies also indicated NAD degradation before lysis. These data indicate a modulating role for DNA damage in selenite cytotoxicity mediated by poly(ADP-ribose)polymerase. Taken together with previously published data on, for example, potentially lethal oxidation of NADPH (Anundi et al., Chem. Biol. Interact. 50, 277, 1984) they also suggest that cell death resulted from interactions between several events that may deplete energy supplies. The results are compatible with a selective killing of DNA-damaged hepatocytes by low doses of selenite.

Evaluation of a genotoxicity test measuring DNA-strand breaks in mouse lymphoma cells by alkaline unwinding and hydroxyapatite elution.

A rapid genotoxicity test, based on the measurement of the proportion of single- to double-stranded DNA by alkaline unwinding and hydroxyapatite elution in mouse lymphoma cells treated in vitro with various chemicals, was evaluated. Seventy-eight compounds from diverse chemical groups, including commonly tested mutagens, toxic compounds not usually tested for genotoxicity and non-toxic compounds not thought to be genotoxic were tested. The results obtained were compared with those from the mouse lymphoma TK locus forward-mutation assay, providing a basis for assessing the relative sensitivity of the 2 assays using the same cells exposed to chemicals under similar conditions. Clear evidence of DNA-damaging activity was obtained with 43 of the compounds, while 4 gave equivocal results. Of the remaining 31 compounds, 14 were toxic without inducing DNA damage while the rest were non-toxic and did not induce any DNA damage. Results were available from both the alkaline unwinding assay and the mouse lymphoma assay for 61 compounds; they showed a concordance between the 2 assays of 77%. Of the 47 compounds that were positive or equivocal in the alkaline unwinding assay, only carbon tetrachloride and prednisolone were negative in the mouse lymphoma assay, while 12 of the 19 compounds that were negative in the alkaline unwinding assay were positive in the mouse lymphoma assay. These included 3 compounds that interfere with nucleic acid metabolism, and 3 crosslinking agents, which would be expected to produce mutations to a greater extent than strand breaks. The other 6 compounds were anthranilic acid, benzoquinone, p-chloroaniline, diethylmaleate, glucose and procarbazine HCl. Of these only the last is a known carcinogen. It is concluded from the present study that there was good overall agreement between the results of the DNA alkaline unwinding and mouse lymphoma TK locus assays, but that the sensitivity of the alkaline unwinding assay is lower for some classes of compounds. Bearing this in mind, the alkaline unwinding assay is considered suitable as a rapid screen for genotoxic activity in eukaryotic cells.

Recovery of malondialdehyde in urine as a 2,4-dinitrophenylhydrazine derivative analyzed with high-performance liquid chromatography.

Malondialdehyde (MDA) in urine was measured as a 2,4-dinitrophenylhydrazine (DNPH) derivative using high-performance liquid chromatography (HPLC) for the analysis. MDA standard coeluted with a peak obtained from rat urine after i.p. administration of MDA standard. This peak was also the only peak containing 14C after injection of a [14C]MDA standard, and was shown by mass spectrometry to contain 1-(2,4-dinitrophenyl)pyrazole, the derivative formed when MDA is treated with DNPH. Depending on the amount given (0.3-5.5 mumol), the recovery (after 24 h sampling period) in urine was 0.7-2.6%. This apparent non-linear kinetics may relate to several factors, such as dose-dependent metabolism. However, the peak urinary concentration approached the expected plasma concentration and reproducible recovery data were obtained, suggesting that MDA was passively excreted in a reasonably stable form. These data indicate that monitoring MDA excretion in urine can give useful information about lipid peroxidation in vivo.