Hideki Chiba

Tight junctions and human diseases

Tight junctions are intercellular junctions adjacent to the apical end of the lateral membrane surface. They have two functions, the barrier (or gate) function and the fence function. The barrier function of tight junctions regulates the passage of ions, water, and various macromolecules, even of cancer cells, through paracellular spaces. The barrier function is thus relevant to edema, jaundice, diarrhea, and blood-borne metastasis. On the other hand, the fence function maintains cell polarity. In other words, tight junctions work as a fence to prevent intermixing of molecules in the apical membrane with those in the lateral membrane. This function is deeply involved in cancer cell biology, in terms of loss of cell polarity. Of the proteins comprising tight junctions, integral membrane proteins occludin, claudins, and JAMs have been recently discovered. Of these molecules, claudins are exclusively responsible for the formation of tight-junction strands and are connected with the actin cytoskeleton mediated by ZO-1. Thus, both functions of tight junctions are dependent on the integrity of the actin cytoskeleton as well as ATP. Mutations in the claudin14 and the claudin16 genes result in hereditary deafness and hereditary hypomagnesemia, respectively. Some pathogenic bacteria and viruses target and affect the tight-junction function, leading to diseases. In this review, the relationship between tight junctions and human diseases is summarized.

Tight Junction Proteins Claudin-2 and -12 are Critical for Vitamin D-dependent Ca2+ Absorption between Enterocytes

Ca(2+) is absorbed across intestinal epithelial monolayers via transcellular and paracellular pathways, and an active form of vitamin D(3), 1alpha,25-dihydroxyvitamin D(3) [1alpha,25(OH)(2)D(3)], is known to promote intestinal Ca(2+) absorption. However, the molecules driving the paracellular Ca(2+) absorption and its vitamin D dependency remain obscure. Because the tight junction proteins claudins are suggested to form paracellular channels for selective ions between neighboring cells, we hypothesized that specific intestinal claudins might facilitate paracellular Ca(2+) transport and that expression of these claudins could be induced by 1alpha,25(OH)(2)D(3). Herein, we show, by using RNA interference and overexpression strategies, that claudin-2 and claudin-12 contribute to Ca(2+) absorption in intestinal epithelial cells. We also provide evidence showing that expression of claudins-2 and -12 is up-regulated in enterocytes in vitro and in vivo by 1alpha,25(OH)(2)D(3) through the vitamin D receptor. These findings strongly suggest that claudin-2- and/or claudin-12-based tight junctions form paracellular Ca(2+) channels in intestinal epithelia, and they highlight a novel mechanism behind vitamin D-dependent calcium homeostasis.

Possible involvement of gap junctions in the barrier function of tight junctions of brain and lung endothelial cells.

Gap‐junction plaques are often observed with tight‐junction strands of vascular endothelial cells but the molecular interaction and functional relationships between these two junctions remain obscure. We herein show that gap‐junction proteins connexin40 (Cx40) and Cx43 are colocalized and coprecipitated with tight‐junction molecules occludin, claudin‐5, and ZO‐1 in porcine blood–brain barrier (BBB) endothelial cells. Gap junction blockers 18β‐glycyrrhetinic acid (18β‐GA) and oleamide (OA) did not influence expression of Cx40, Cx43, occludin, claudin‐5, junctional adhesion molecule (JAM)‐A, JAM‐B, JAM‐C, or ZO‐1, or their subcellular localization in the porcine BBB endothelial cells. In contrast, these gap‐junction blocking agents inhibited the barrier function of tight junctions in cells, determined by measurement of transendothelial electrical resistance and paracellular flux of mannitol and inulin. 18β‐GA also significantly reduced the barrier property in rat lung endothelial (RLE) cells expressing doxycycline‐induced claudin‐1, but did not change the interaction between Cx43 and either claudin‐1 or ZO‐1, nor their expression levels or subcellular distribution. These findings suggest that Cx40‐ and/or Cx43‐based gap junctions might be required to maintain the endothelial barrier function without altering the expression and localization of the tight‐junction components analyzed.

Evidence for auto/paracrine actions of vitamin D in bone: 1α-hydroxylase expression and activity in human bone cells

Vitamin D is an important regulator of mineral homeostasis and bone metabolism. 1α‐Hy‐droxylation of 25‐(OH)D3 to form the bioactive vitamin D hormone, 1α,25‐(OH)2D3, is classically considered to take place in the kidney. However, 1α‐hydroxylase has been reported at extrarenal sites. Whether bone is a 1α,25‐(OH)2D3 synthesizing tissue is not univocal. The aim of this study was to investigate an autocrine/ paracrine function for 1α,25‐(OH)2D3 in bone. We show that 1α‐hydroxlase is expressed in human osteoblasts, as well as the vitamin D binding protein receptors megalin and cubilin. Functional analyses demonstrate that after incubation with the 1α‐hydoxylase substrate 25‐(OH)D3, the osteoblasts can produce sufficient 1α,25‐(OH)2D3 to modulate osteoblast activity, resulting in induced alkaline phosphatase (ALP) activity, osteocalcin (OC) and CYP24 mRNA expression, and mineralization. The classical renal regulators of 1α‐hydroxylase, parathyroid hormone, and ambient calcium do not regulate 1α‐hydroxylase in osteoblasts. In contrast, interleukin (IL)‐1β strongly induces 1α‐hydroxylase. Besides the bone‐forming cells, we demonstrate 1α‐hydroxylase activity in the bone resorbing cells, the osteoclasts. This is strongly dependent on osteoclast inducer RANKL. This study showing expression, activity, and functionality of 1α‐hydoxylase unequivocally demonstrates that vitamin D can act in an auto/paracrine manner in bone.—van Driel, M., Koedam, M., Buurman, C. J., Hewison, M., Chiba, H., Uitterlinden, A. G., Pols, H. A. P., van Leeuwen, J. P. T. M. Evidence for auto/paracrine actions of vitamin D in bone: 1α‐hydroxylase expression and activity in human bone cells. FASEB J. 20, E1811–E1819 (2006)

Differential Expression and Subcellular Localization of Claudin-7, −8, −12, −13, and −15 Along the Mouse Intestine

Among tight-junction proteins, claudins, which play a key role in paracellular transport across epithelia, claudins 1 to 5 are expressed in the intestine, and changes in their abundance and/or distribution are considered to contribute to various gastrointestinal diseases. We investigated, by reverse transcription-PCR, immunoblot, and immunofluorescence analyses, which other claudin species were expressed in the mouse intestine, and whether they showed unique expression profiles. Rabbit polyclonal antibodies against mouse claudin-8, claudin-12, and claudin-15 were generated, and their specificity was verified by immunoblotting using COS-7 cells transfected with individual claudin cDNAs. Claudin-7, −8, −12, −13, and −15 appeared to be expressed in the duodenum, jejunum, ileum, and/or colon with remarkable variations in the expression levels along the intestinal tract, and had distinct subcellular localization in the intestinal epithelium. In addition, claudin-13 and −15 exhibited gradients along the crypt-surface axis of the colon. By contrast, claudin-6, −9, −10, −11, −14, −16, −18, and −19 were not observed in the intestine. Our results indicate that five additional species of claudins have very complex expression patterns along and within the intestine, and that this may reflect differences in paracellular permeable properties, providing valuable resources for studying the significance of these claudins in gastrointestinal disorders. This manuscript contains online supplemental material available at http://www.jhc.org. Please visit this article online to view these materials.

Cyclic AMP induces phosphorylation of claudin-5 immunoprecipitates and expression of claudin-5 gene in blood-brain-barrier endothelial cells via protein kinase A-dependent and -independent pathways.

Cyclic AMP (cAMP) promotes functions of tight junctions in endothelial cells, although its target remains unknown. We showed here that cAMP increased gene expression of claudin-5 and decreased that of claudin-1 in porcine blood-brain-barrier endothelial cells via protein kinase A (PKA)-independent and -dependent pathways, respectively. cAMP also enhanced immunoreactivity of claudin-5 along cell borders and in the cytoplasm, reorganized actin filaments, and altered signals of claudin-5, occludin, ZO-1, and ZO-2 along cell boundaries from zipperlike to linear patterns. In contrast, claudin-1 was detected only in the cytoplasm in a dotlike pattern, and its immunolabeling was reduced by cAMP. Interestingly, 31- and 62-kDa claudin-5 immunoprecipitates in the NP-40-soluble and -insoluble fractions, respectively, were highly phosphorylated on threonine residue(s) upon cAMP treatment. All these changes induced by cAMP, except for claudin-5 expression and its signals in the cytoplasm, were reversed by an inhibitor of PKA, H-89. We also demonstrated that cAMP elevated the barrier function of tight junctions in porcine blood-brain-barrier endothelial cells in PKA-dependent and -independent manners. These findings indicate that both PKA-induced phosphorylation of claudin-5 immunoprecipitates and cAMP-dependent but PKA-independent induction of claudin-5 expression could be involved in promotion of tight-junction function in endothelial cells.

RXRalpha-null F9 embryonal carcinoma cells are resistant to the differentiation, anti-proliferative and apoptotic effects of retinoids.

The F9 murine embryonal carcinoma (EC) cell line, a well established model system for the study of retinoic acid (RA)‐induced differentiation, differentiates into cells resembling three types of extra‐embryonic endoderm (primitive, parietal and visceral), depending on the culture conditions and RA concentration used. A number of previously identified genes are differentially expressed during this process and serve as markers for the different endodermal cell types. Differentiation is also accompanied by a decreased rate of proliferation and an apoptotic response. Using homologous recombination, we have disrupted both alleles of the retinoid X receptor (RXR) alpha gene in F9 cells to investigate its role in mediating these responses. The loss of RXRalpha expression impaired the morphological differentiation of F9 EC cells into primitive and parietal endoderm, but has little effect on visceral endodermal differentiation. Concomitantly the inducibility of most primitive and parietal endoderm differentiation‐specific genes was impaired, while several genes upregulated during visceral endodermal differentiation were induced normally. We also demonstrate that RXRalpha is required for both the anti‐proliferative and apoptotic responses in RA‐treated F9 cells. Additionally, we provide further evidence that retinoic acid receptor (RAR)‐RXR heterodimers are the functional units transducing the effects of retinoids in F9 cells.

Distinct retinoid X receptor-retinoic acid receptor heterodimers are differentially involved in the control of expression of retinoid target genes in F9 embryonal carcinoma cells.

The F9 murine embryonal carcinoma cell line represents a well-established system for the study of retinoid signaling in vivo. We have investigated the functional specificity of different retinoid X receptor (RXR)-retinoic acid (RA) receptor (RAR) isotype pairs for the control of expression of endogenous RA-responsive genes, by using wild-type (WT), RXR alpha(-/-), RAR alpha(-/-), RAR gamma(-/-), RXR alpha(-/-)-RAR alpha(-/-), and RXR alpha(-/-)-RAR gamma(-/-) F9 cells, as well as panRXR and RAR isotype (alpha, beta, and gamma)-selective retinoids. We show that in these cells the control of expression of different sets of RA-responsive genes is preferentially mediated by distinct RXR-RAR isotype combinations. Our data support the conclusion that RXR-RAR heterodimers are the functional units transducing the retinoid signal and indicate in addition that these heterodimers exert both specific and redundant functions on the expression of particular sets of RA-responsive genes. We also show that the presence of a given receptor isotype can hinder the activity of another isotype and therefore that functional redundancy between retinoid receptor isotypes can be artifactually generated by gene knockouts.

Evidence that both 1α,25‐dihydroxyvitamin D3 and 24‐hydroxylated D3 enhance human osteoblast differentiation and mineralization

Vitamin D plays a major role in the regulation of mineral homeostasis and affects bone metabolism. So far, detailed knowledge on the vitamin D endocrine system in human bone cells is limited. Here we investigated the direct effects of 1α,25‐(OH)2D3 on osteoblast differentiation and mineralization. Also, we studied the impact of 24‐hydroxylation, generally considered as the first step in the degradation pathway of vitamin D, as well as the role of the nuclear and presumed membrane vitamin D receptor (VDR). For this we used a human osteoblast cell line (SV‐HFO) that has the potency to differentiate during culture forming a mineralized extracellular matrix in a 3‐week period. Transcriptional analyses demonstrated that both 1α,25‐(OH)2D3 and the 24‐hydroxylated metabolites 24R,25‐(OH)2D3 and 1α,24R,25‐(OH)3D3 induced gene transcription. All metabolites dose‐dependently increased alkaline phosphatase (ALP) activity and osteocalcin (OC) production (protein and RNA), and directly enhanced mineralization. 1α,24R,25‐(OH)3D3 stimulated ALP activity and OC production most potently, while for mineralization it was equipotent to 1α,25‐(OH)2D3. The nuclear VDR antagonist ZK159222 almost completely blocked the effects of all metabolites. Interestingly, 1β,25‐(OH)2D3, an inhibitor of membrane effects of 1α,25‐(OH)2D3 in the intestine, induced gene transcription and increased ALP activity, OC expression and mineralization. In conclusion, not only 1α,25‐(OH)2D3, but also the presumed 24‐hydroxylated “degradation” products stimulate differentiation of human osteoblasts. 1α,25‐(OH)2D3 as well as the 24‐hydroxylated metabolites directly enhance mineralization, with the nuclear VDR playing a central role. The intestinal antagonist 1β,25‐(OH)2D3 acts in bone as an agonist and directly stimulates mineralization in a nuclear VDR‐dependent way. J. Cell. Biochem. 99: 922–935, 2006.

Regulation of the blood-biliary barrier: interaction between gap and tight junctions in hepatocytes.

Hepatocytes tightly connect with each other by intercellular junctions to form liver cell plates. The junctions composed of gap, tight, and adherens junctions and desmosomes concentrate around bile canaliculi. In particular, tight junctions serve as a barrier to keep bile in bile canaliculi away from the blood circulation. Thus, it is very reasonable to call tight junctions of hepatocytes the blood–biliary barrier. On the other hand, gap junctions of hepatocytes are considered to enable ordered contraction of bile canaculi from centrizonal to periportal hepatocytes by their function of intercellular communication. Gap and tight junctions may thus play a crucial role in bile secretion, one of the most differentiated functions of the liver. In intrahepatic cholestasis, a common pathological condition of the liver, downregulation of gap and tight junctional functions is seen, which results in impaired intercellular communication and in leaky tight junctions. Although the changes in gap and tight junctions had been considered to be independent of each other, recent findings that the tight junction-associated proteins ZO-1 and occludin bind to connexins indicate the possibility of either coordinate or reciprocal regulation of macromolecular complexes containing gap- and tight-junction proteins. In this review, we introduce the interaction and regulation between gap and tight junctions of hepatocytes in vitro and discuss the regulatory mechanisms of the “blood–biliary barrier” to study the molecular pathogenesis of cholestasis.