Molecular and cellular permeability control at the blood-brain barrier.

The blood-brain barrier (BBB) is formed by brain capillary endothelial cells. These cells have at least three properties which distinguish them from their peripheral counterparts: (1) tight junctions (TJs) of extremely low permeability; (2) low rates of fluid-phase endocytosis; (3) specific transport and carrier molecules. In combination, these features restrict the nonspecific flux of ions, proteins, and other substances into the central nervous system (CNS) environment. The restriction protects neurons from harmful compositional fluctuations occurring in the blood and allows uptake of essential molecules. Breakdown of the BBB is associated with a variety of CNS disorders and results in aggravation of the condition. Restoration of the BBB is thus one strategy during therapy of CNS diseases. Its success depends on a precise knowledge of the structural and functional principles underlying BBB functionality. In this review we have tried to summarise the current knowledge of TJs, including information gained from non-neuronal systems, and describe selected mechanisms involved in permeability regulation.

Down-regulation of occludin expression in astrocytes by tumour necrosis factor (TNF) is mediated via TNF type-1 receptor and nuclear factor-kappaB activation.

Tight junctions form the diffusion barrier of brain microcapillary endothelial cells and support cell polarity. Also astrocytes express tight junction components such as occludin, claudin-1, ZO-1 and ZO-2, but do not establish a permeability barrier. However, little is known about the function and regulation of these molecules in astrocytes. We studied the impact of tumour necrosis factor (TNF) on occludin and ZO-1 expression in astrocytes. TNF decreased occludin, but not ZO-1 expression. In brain microcapillary endothelial cells, as well as in epithelial cells, occludin expression was not influenced by TNF. Removal of TNF from astrocytes restored the basal level of occludin. Down-regulation was inhibited by caffeic acid phenethyl ester, a specific inhibitor of nuclear factor-kappaB (NF-kappaB) activation. Exposure of astrocytes isolated from mice deficient in either TNF type-1 receptor (TNFR1), TNF type-2 receptor (TNFR2), or both, respectively, revealed that down-regulation was mediated entirely by TNFR1. ZO-1, which can interact with occludin, was found to co-precipitate connexin43, but not occludin. These findings demonstrate that TNF selectively down-regulates occludin in astrocytes, but not in cells forming established tight junctions, through TNFR1 and suggest that NF-kappaB is involved as a negative regulator.

Identification of a novel neuroligin in humans which binds to PSD-95 and has a widespread expression.

Neuroligins, first discovered in rat brain, form a family of three synaptically enriched membrane proteins. Using reverse transcription-PCR of human brain polyadenylated RNA and extensive database searches, we identified the human homologues of the three rat neuroligins and a cDNA encoding a fourth member, which we named neuroligin 4. Neuroligin 4 has 63-73% amino acid identity with the other members of the human neuroligin family, and the same predicted domain structure. DNA database analyses, furthermore, indicated that a possible fifth neuroligin gene may be present in the human genome. Northern-blot analysis revealed expression of neuroligin 4 in heart, liver, skeletal muscle and pancreas, but barely at all in brain. Overexpression of neuroligin 4 cDNA in COS-7 cells led to the production of a 110 kDa protein. Immunofluorescence analysis demonstrated that the protein was integrated into the plasma membrane. Overexpression of cDNAs encoding neuroligin 4 and the PDZ-domain protein, PSD-95, in COS-7 cells resulted in the formation of detergent-resistant complexes. Neuroligin 4 did not bind to ZO-1, another PDZ-domain protein. Together, our data show that the human neuroligin family is composed of at least one additional member, and suggest that neuroligin 4 may also be produced outside the central nervous system.

Structure elucidation of beta-mannanase: from the electron-density map to the DNA sequence.

The crystal structure of affinity-purified Thermomonospora fusca beta-mannanase has been solved despite the lack of the major part of the amino-acid sequence. A high-quality electron-density map allowed the identification of a stretch of eight amino acids close to the C-terminus which was used to design a degenerate downstream PCR primer. Together with a specific primer previously derived from the N-terminus, 95.7% of the mannanase gene sequence was obtained from genomic T. fusca DNA by PCR. The structure-derived sequence was then compared with the DNA-derived sequence and corrected when necessary. Applying the presented protocol, there was no need to manually build a model at an early stage of structure determination, an erroneous and tedious process, especially in the absence of the amino-acid sequence. Using the DNA sequence information and the current version of ARP/wARP, 281 residues, or 93% of the polypeptide chain (including side chains), were built and refined to an R factor of 16.5% without any manual intervention.

sICAM-1 and TNF-alpha induce MIP-2 with distinct kinetics in astrocytes and brain microvascular endothelial cells.

The dysfunction of the blood-brain barrier (BBB) occurring after traumatic brain injury (TBI) is mediated by intracerebral neutrophil accumulation, chemokine release (e.g., interleukin (IL)-8) and upregulation of adhesion molecules (e.g., intercellular adhesion molecule (ICAM)-1). In patients with severe TBI, we previously found that elevated cerebrospinal fluid (CSF) IL-8 and soluble (s)ICAM-1 correlate with BBB dysfunction, and this prompted us to concomitantly monitor IL-8, sICAM-1 and their stimulator tumor necrosis factor (TNF)-alpha in CSF. Potential mechanisms for upregulation of the IL-8 analogue, murine macrophage inflammatory protein (MIP)-2, and sICAM-1 at the BBB were studied using cultured mouse astrocytes and brain microvascular endothelial cells (MVEC). In CSF of seven patients, IL-8 and sICAM-1 were elevated for 19 days after severe TBI, whereas TNF-alpha exceeded normal values on 9 days. Stimulation of MVEC and astrocytes with TNF-alpha simultaneously induced the release of MIP-2 reaching saturation by 4-8 hr and of sICAM-1 increasing continuously from 2-4 hr to 12 hr. Augmented sICAM-1 production correlated with enhanced membrane-bound (m)ICAM-1 expression in both cell types (r(s) = 0.96 and 0.90, P < 0.0001), but was markedly higher in astrocytes. The release of sICAM-1 was not influenced by IL-8 or MIP-2, although astrocytes and MVEC expressed the IL-8/MIP-2 receptor (CXCR-2) as determined by FACS analysis. Instead, we found that sICAM-1 strongly induced MIP-2 secretion by both cell types with kinetics differing from those evoked by TNF-alpha. If added together, sICAM-1 and TNF-alpha synergistically induced MIP-2 production suggesting the involvement of two different pathways for MIP-2 regulation.

Species differences in the sites of cleavage of pro-lactase to lactase supports lack of selective pressure.

The pro-sequences in pro-lactase-phlorizin hydrolase (LPH) are needed for lactase to proceed past the ER, but are irrelevant as to the enzymatic activities. Hence, in all species removal of the pro- sequences (or most of them) must take place after the ER. Contrary to this, the details of the removal of these pro-sequences are to be expected to differ in the various species, since they are not subjected to selective pressure. Using site-directed mutagenesis we investigated processing in rabbit. The first cleavage occurs by furin (or furin-like PCs) and takes place at R-A-A-R(349) in the pro-sequence, generating the known 180 kDa intermediate. Replacing R(349) by Q results in a mutant which is not cleaved but nevertheless transported to the cell surface as demonstrated by immunofluorescence. Further processing of either the 180 kDa intermediate or the mutant is not directly mediated by furin-like PCs, but involves (also) other proteases. These results demonstrate that formation of the 180 kDa intermediate, consistently found only in rabbits, but not in man, is not essential for lactase transport: in all likelihood lack of selective pressure has led to species-specific processing of pro-LPH.

Occludin proteolysis and increased permeability in endothelial cells through tyrosine phosphatase inhibition.

Regulation of epithelial and endothelial permeability is essential for proper function of compartmentalized organisms, and tyrosine phosphorylation plays an important role in this process. We analyzed the impact of protein tyrosine phosphatase (PTP) inhibition on the structure of endothelial junctional proteins. In human umbilical vein endothelial cells (HUVECs) the PTP inhibitors phenylarsine oxide (PAO) and pervanadate induced proteolysis of the tight junction protein occludin. Occludin proteolysis was inhibited by the metalloproteinase inhibitor 1,10-phenanthroline (PHEN), but not by inhibitors against other types of proteases. The junctional proteins ZO-1, cadherin and beta-catenin were not cleaved. Under conditions of occludin proteolysis, PAO treatment elevated permeability for FITC-dextran. Simultaneous incubation of HUVECs with PAO and PHEN inhibited the rise in permeability by more than 60%. PAO treatment lead to progressive disappearance of occludin from the cell periphery. In contrast, ZO-1, cadherin and beta-catenin retained their positions at the sites of intercellular contact. Simultaneous administration of PAO and PHEN greatly prevented the redistribution of occludin. These results demonstrate a selective cleavage of occludin by a metalloproteinase and suggest that this process can contribute to the control of paracellular permeability in endothelial cells.

Inhibition of Na,K-ATPase activity by cGMP is isoform-specific in brain endothelial cells.

cGMP has been shown to either activate or inhibit Na,K-ATPase activity. Using mouse brain endothelial cells which express both ouabain-resistant alpha1 and ouabain-sensitive alpha2 and alpha3 isoforms, we show that cGMP reduces total Na,KATPase activity to about 58%. The inhibition is prevented by the protein kinase G (PKG)-specific inhibitor KT5823, indicating that cGMP-mediated activation of PKG leads to inhibition of the pump. A similar extent of inhibition is obtained with nitric oxide. cGMP-induced inhibition acts mainly on alpha1 isoforms but hardly affects alpha2/alpha3 isoforms. These data suggest that inhibition of Na,K-ATPase activity by cGMP occurs in an isoform-selective manner in brain endothelial cells.

High-resolution native and complex structures of thermostable beta-mannanase from Thermomonospora fusca - substrate specificity in glycosyl hydrolase family 5.

BACKGROUND: . beta-Mannanases hydrolyse the O-glycosidic bonds in mannan, a hemicellulose constituent of plants. These enzymes have potential use in pulp and paper production and are of significant biotechnological interest. Thermostable beta-mannanases would be particularly useful due to their high temperature optimum and broad pH tolerance. The thermophilic actinomycete Thermomonospora fusca secretes at least one beta-mannanase (molecular mass 38 kDa) with a temperature optimum of 80 degreesC. No three-dimensional structure of a mannan-degrading enzyme has been reported until now. RESULTS: . The crystal structure of the thermostable beta-mannanase from T. fusca has been determined by the multiple isomorphous replacement method and refined to 1.5 A resolution. In addition to the native enzyme, the structures of the mannotriose- and mannohexaose-bound forms of the enzyme have been determined to resolutions of 1.9 A and 1.6 A, respectively. CONCLUSIONS: . Analysis of the -1 subsite of T. fusca mannanase reveals neither a favourable interaction towards the axial HO-C(2) nor a discrimination against the equatorial hydroxyl group of gluco-configurated substrates. We propose that selectivity arises from two possible mechanisms: a hydrophobic interaction of the substrate with Val263, conserved in family 5 bacterial mannanases, which discriminates between the different conformations of the hydroxymethyl group in native mannan and cellulose; and/or a specific interaction between Asp259 and the axial hydroxyl group at the C(2) of the substrate in the -2 subsite. Compared with the catalytic clefts of family 5 cellulases, the groove of T. fusca mannanase has a strongly reduced number of aromatic residues providing platforms for stacking with the substrate. This deletion of every second platform is in good agreement with the orientation of the axial hydroxyl groups in mannan.

Processing of human intestinal prolactase to an intermediate form by furin or by a furin-like proprotein convertase.

Human lactase-phlorizin hydrolase (human-LPH) is synthesized as a large precursor (prepro-LPH), then cleaved to a pro-LPH of 220 kDa which is further cut to a "mature-like LPH" of a size close to that of mature LPH, i.e. about 150 kDa (in the processing of rabbit pro-LPH the intermediate has a mass of approximately 180 kDa). By coexpression of human prepro-LPH with furin in COS-7 cells we show that furin generates a mature-like LPH. Radioactive amino acid sequence analysis reveals that furin recognizes the motif R-T-P-R832, a protein convertase consensus, to generate a NH2 terminus located 36 amino acids upstream of the NH2 terminal found in vivo at Ala869. This intermediate is ultimately cleaved to the mature LPH form by other proteases including the pancreatic ones. These data demonstrate that human pro-LPH, like the rabbit enzyme, is processed to the mature enzyme by furin or furin-like enzymes through at least an intermediate form that has, however, an apparent mass close to that of the mature enzyme.

Intestinal lactase-phlorizin hydrolase (LPH): the two catalytic sites; the role of the pancreas in pro-LPH maturation.

Brush border lactase-phlorizin hydrolase carries two catalytic sites. In the human enzyme lactase comprises Glu-1749, phlorizin hydrolase Glu-1273. The proteolytic processing of pro-lactase-phlorizin hydrolase by (rat) enterocytes stops two amino acid residues short of the N-terminus of 'mature' final, brush border lactase-phlorizin hydrolase. Only these two amino acid residues are removed by luminal pancreatic protease(s), probably trypsin.

Interleukin-1 modulates protein tyrosine phosphatase activity and permeability of brain endothelial cells.

Interleukin-1 alpha (IL-1 alpha) and interleukin-6 (IL-6), both known to be able to open the blood-brain barrier (BBB), downregulated plasma membrane-associated tyrosine phosphatase activity in primary porcine brain endothelial cells (PBEC). In contrast, transforming growth factor beta (TGF-beta) upregulated PTP activity and tumor necrosis factor alpha (TNF-alpha) had no effect. Plasma membrane-associated PTP activity of PBEC was upregulated at contact inhibited growth arrest. Tightly confluent cells reduced 3H-inulin permeability by 34% compared with just confluent cells indicating the formation of barrier properties. The decrease in permeability temporally correlated with the elevated PTP activity of the cells at growth arrest and was reversed to control by IL-1 alpha. Vanadate, a broad-specificity PTP inhibitor, also enhanced 3H-inulin permeability. These data suggest that IL-1 alpha-induced endothelial permeability could be controlled through lowering PTP activity.

Relevance of Na,K-ATPase to local extracellular potassium homeostasis and modulation of synaptic transmission.

The ion gradients generated by the Na,K-ATPase are essential for Na+-coupled transport systems, osmoregulation and restoration of ion concentrations in excitable tissues. Indirectly, the sodium pump controls intracellular Ca2+ concentration through the Na/Ca exchanger. In the nervous system various neurotransmitters can modulate Na,K-ATPase activity. The great diversity of Na,K-ATPase subunit isoforms, their complex spatial and temporal regulation of expression and their cellular localisation imply a functional role of the sodium pump in different regulatory pathways. Among these, potassium homeostasis and modulation of synaptic transmission are discussed here.

Lactase persistence versus decline in human adults: multifactorial events are involved in down-regulation after weaning.

BACKGROUND & AIMS: In nonhuman mammals, lactase activity declines during or after weaning. In contrast, about one half of the human species maintains high lactase activity even in adulthood. To clarify this difference, this study examined some parameters for which contrasting observations have been reported in connection with lactase decline. METHODS: Lactase activity, lactase messenger RNA (mRNA) levels, and in vitro lactase biosynthesis were determined in normal jejunal samples from a large group of white adults, all born in or near Naples. RESULTS: Of 44 individuals, 10 were lactase persistent and 34 were hypolactasic. Biosynthesis of prolactase correlated well with lactase mRNA levels, indicating transcriptional control; it did less so with steady-state lactase activity. Examination of lactase mRNA levels and lactase activity/lactase mRNA ratios revealed a heterogeneous pattern of lactase mRNA level, lactase synthesis, and activity in both lactase persistent and hypolactasic subjects. CONCLUSIONS: Both transcriptional and posttranscriptional factors cause the decline of intestinal lactase. This probably explains the multifarious observations that most studies on adult-type hypolactasia have reported. The single overriding factor distinguishing lactase-persistent subjects from hypolactasic subjects is the high rate of lactase biosynthesis.

Isoform-specific interactions of Na,K-ATPase subunits are mediated via extracellular domains and carbohydrates.

The functional unit of the Na,K-ATPase consists of a catalytic alpha subunit noncovalently linked with a glycoprotein subunit, beta. Using ouabain binding assays and immunoprecipitation of rodent alpha/beta complexes, we show here that all six possible isozymes between three alpha and two beta isoforms can be formed in Xenopus oocytes. Two isoform-specific differences in alpha/beta interactions are observed: (i) alpha1/beta1 and alpha2/beta2 complexes, in contrast to alpha1/beta2 complexes, are stable against Triton X-100-mediated dissociation, and (ii) beta2 subunits must carry N-glycans to combine with alpha1 but not with alpha2. The interacting surfaces are mainly exposed to the extracellular side because coexpression of a truncated beta1 subunit comprising the ectodomain results in assembly with alpha1 and alpha2, but not with alpha3; the beta2 ectodomain combines with alpha2 only. A chimera consisting of 81% and 19% of the alpha1 N terminus and alpha2 C terminus, respectively, behaves like alpha2 and coprecipitates with the beta2 ectodomain. In contrast, the reciprocal chimera does not coprecipitate with the beta2 ectodomain. These results provide evidence for a selective interaction of Na,K-ATPase alpha and beta subunits.

The extracellular domain of the sodium pump beta isoforms determines complex stability with alpha 1.

Both subunits of the Na,K-ATPase are encoded by several genes giving rise to at least six isozymes. To examine whether beta isoforms assemble with alpha 1 in a selective manner, we have overexpressed wild-type and chimeric beta subunits in L929 cells and examined assembly as a function of resistance towards detergent-mediated dissociation. In the presence of digitonin all beta chimeras coimmunoprecipitate the endogenous alpha 1 subunit. Only beta proteins with the ectodomain of beta 1 coimmunoprecipitate alpha 1 in the presence of Triton X-100. All beta chimeras stimulate Na,K-ATPase activity in L929 cells. These data indicate that the beta subunit ectodomains mediate interactions with alpha 1 and influence the stability of this complex.

Co-culture blood-brain barrier models and their use for pharmatoxicological screening.

The availability of an in vitro blood-brain barrier model would represent a powerful alternative to experimental animals in pharmacological and toxicological research. This overview collects the various current approaches to build an in vitro model of the blood-brain barrier for these purposes. Purified bovine, porcine and human brain microcapillary endothelial cells as well as several immortalized cell lines have been used to model the blood-brain barrier in vitro, partly in co-culture with astrocytes of various species, or various cell lines such as C6 glioma or N2a neuroblastoma cells. The collected data indicate that functional parameters often can be induced by soluble and membrane-bound factors in such cell systems. Relevant barrier-specific parameters are reviewed: electrical resistance, and structure and function of the multidrug resistance P-glycoprotein and the y-glutamyl transpeptidase. Both P-glycoprotein and gamma-glutamyl transpeptidase have great influence on the pharmacodynamics, toxicology and metabolic capacity of the blood-brain barrier (drug efflux, oxidative damage, detoxification of endotoxins, etc.). Several available in vitro models appear to be suited for pharmacotoxicological screening, if the functional parameters gamma-glutamyl transpeptidase, P-glycoprotein as well as transendothelial resistance are monitored.

cDNA sequence coding for a rat glia-derived nexin and its homology to members of the serpin superfamily.

Rat glial cells release a neurite-promoting factor with serine protease inhibitory activity. By using a rat glioma cDNA clone as a probe, it was possible to isolate rat cDNAs containing the entire sequence coding for this neurite-promoting factor. The largest rat cDNA (approximately 2100 bp) was characterized by DNA sequencing. It contained the entire coding region, 135 bp of the 5' nontranslated region, and about 750 bp of the 3' nontranslated region. The open reading frame coded for 397 amino acids including a putative signal peptide of 19 amino acids. The correct identity of the coding sequence was substantiated by the fact that the sequence of tryptic peptides, derived from the purified rat factor, matched exactly with the deduced amino acid sequence. The rat protein sequence had 84% homology with the corresponding protein from human glioma cells. Both amino acid sequences indicated that the proteins belong to the protease nexins [Baker, B.J., Low, D. A., Simmer, R. L., & Cunningham, D.D. (1980) Cell (Cambridge, Mass.) 21, 37-45] and therefore can be defined as glia-derived nexins (GDNs). Further analysis showed that both rat and human GDN belong to the serpin superfamily and share 41%, 32%, and 25% homology with human endothelial-cell-type plasminogen activator inhibitor, antithrombin III, and alpha-1 proteinase inhibitor, respectively.