Abstract. Molecular mechanisms linking pre-and postsynaptic membranes at the interneuronal synapses are little known. We tested the cadherin adhesion system for its localization in synapses of mouse and chick brains. We found that two classes of cadherin-associated proteins, aN-and 13-catenin, are broadly distributed in adult brains, colocalizing with a synaptic marker, synaptophysin. At the ultrastructural level, these proteins were localized in synaptic junctions of various types, forming a symmetrical adhesion structure. These structures sharply bordered the transmitter release sites associated with synaptic vesicles, although their segregation was less clear in certain types of synapses. N-cadherin was also localized at a similar site of synaptic junctions but in restricted brain nuclei. In developing synapses, the catenin-bearing contacts dominated their junctional structures. These findings demonstrate that interneuronal synaptic junctions comprise two subdomains, transmitter release zone and cateninbased adherens junction. The catenins localized in these junctions are likely associated with certain cadherin molecules including N-cadherin, and the cadherin/catenin complex may play a critical role in the formation or maintenance of synaptic junctions.T HE synapse is a site where the axon terminal of a neuron comes into functional contact with a target cell. To generate the synapses, each neuron selectively contacts and communicates with other particular neurons. Despite the importance of understanding how specific neuronal connections are established, little is known about what kinds of adhesion molecules are essential for the formation and maintenance of interneuronal synaptic junctions. This lack of information is contrasted with a great deal of knowledge about other synaptic structures and functions, including synaptic vesicle traffic, channel/ receptor function, and signal transduction in synapses (Jessell and Kandel, 1993; Stidhof, 1995).Various electron microscopic studies have revealed the ultrastructural features of interneuronal synaptic junctions. Electron-dense cytoplasmic materials accumulate under both pre-and postsynaptic plasma membranes; and the former is, in general, associated with a cluster of synaptic vesicles. These junctional structures are considered to be the sites where synaptic transmission takes place. The interneuronal synaptic junctions exhibit some variations in morphological characteristics of the cytoplasmic density. Gray (1959) and Colonnier (1968) showed that there are two types of synapses, asymmetric (Gray's type I) and symmetric (Gray's type II). The intercellular space beAddress all correspondence to Masatoshi Takeichi, Department of Biophysics, Faculty of Science, Kyoto University, Kitashirakawa, Sakyo-ku, Kyoto 606-01, Japan. Tel.: (81) 75 751 2111. Fax: (81) 75 753 4197. e-maih takeichi@take.biophys.kyoto-u.ac.jp. tween the membrane thickenings, called the synaptic cleft, is occupied by filamentous materials.Molecular analysis of the synaptic junctions has thus far b...

Recent studies have shown that mutations at amino-acid 482 in the ABCG2 gene affect the substrate specificity of the protein. To delineate the effects of these mutations clearly, human embryonic kidney cells (HEK-293) were stably transfected with wild-type 482R or mutant 482G and 482T ABCG2. By flow cytometry, mitoxantrone, BODIPY-prazosin, and Hoechst 33342 were found to be substrates of all ABCG2 proteins, while rhodamine 123, daunorubicin, and LysoTracker Green were transported only by mutant ABCG2. In cytotoxicity assays, all ABCG2 proteins conferred high levels of resistance to mitoxantrone, SN-38, and topotecan, while mutant ABCG2 also exhibited a gain of function for mitoxantrone as they conferred a four-fold greater resistance compared to wild type. Cells transfected with mutant ABCG2 were 13-to 71-fold resistant to the P-glycoprotein substrates doxorubicin, daunorubicin, epirubicin, bisantrene, and rhodamine 123 compared to cells transfected with wild-type ABCG2, which were only three-to four-fold resistant to these compounds. ABCG2 did not confer appreciable resistance to etoposide, taxol or the histone deacetylase inhibitor depsipeptide. None of the transfected cell lines demonstrated resistance to flavopiridol despite our previous observation that ABCG2-overexpressing cell lines are cross-resistant to the drug. Recently reported inhibitors of ABCG2 were evaluated and 50 mM novobiocin was found to reverse wild-type ABCG2 completely, but only reverse mutant ABCG2 partially. The studies presented here serve to underscore the importance of amino-acid 482 in defining the substrate specificity of the ABCG2 protein and raise the possibility that amino-acid 482 mutations in human cancers could affect the clinical application of antagonists for ABCG2.

The fluorescent compounds rhodamine 123, LysoTracker Green DMD-26, mitoxantrone, and BODIPY-prazosin were used with the antagonist fumitremorgin C (FTC) in order to develop functional assays for the half-transporter, MXR/BCRP/ABCP1. A measure of FTC-inhibitable efflux was generated for each compound in a series of MXR-overexpressing drug-selected cell lines and in ten unselected cell lines which were used to determine if the four fluorescent compounds were sensitive enough to detect the low MXR levels found in drug-sensitive cell lines. FTC-inhibitable efflux of mitoxantrone and prazosin was found in four of the ten cell lines, SF295, KM12, NCI-H460, and A549, and low but detectable levels of MXR mRNA were also observed by Northern analysis in these cells. FTC-inhibitable mitoxantrone and prazosin efflux in both selected and unselected cell lines was found to correlate well with MXR levels as determined by Northern blotting, r(2)=0.89 and r(2)=0.70 respectively. In contrast, rhodamine and LysoTracker were not able to reliably detect MXR. Cytotoxicity assays performed on two of the four unselected cell lines confirmed increased sensitivity to mitoxantrone in the presence of FTC. FTC was found to be a specific inhibitor of MXR, with half-maximal inhibition of MXR-associated ATPase activity at 1 microM FTC. Short term selections of the SF295, KM12, NCI-H460 and A549 cell lines in mitoxantrone resulted in a small but measurable increase in MXR by both Northern blot and functional assay. These studies show that flow cytometric measurement of FTC-inhibitable mitoxantrone or prazosin efflux is a sensitive and specific method for measuring the function of the MXR half-transporter in both selected and unselected cell lines.

Single nucleotide polymorphism (SNP) analyses of the ABCG2 gene have revealed three nonsynonymous SNPs resulting in the amino acid changes at V12M, Q141K and D620N. To determine whether the SNPs have an effect on drug transport, human embryonic kidney cells (HEK-293) were stably transfected with full length ABCG2 coding wild-type or SNP variants of ABCG2. In 4-day cytotoxicity assays with mitoxantrone, topotecan, SN-38 or diflomotecan, cells transfected with wild-type R482 ABCG2 showed IC50 values up to 1.2-fold to 5-fold higher than cells expressing comparable levels of Q141K ABCG2, suggesting that the Q141K SNP affects drug transport. FTC-inhibitable mitoxantrone efflux normalized to ABCG2 surface expression as assayed by the anti-ABCG2 antibody 5D3 was significantly lower in cells transfected with Q141K ABCG2 than in those transfected with wild-type R482 ABCG2 (P = 0.0048). Values for V12M and D620N ABCG2 were comparable to those for wild-type R482 ABCG2. The vanadate-sensitive ATPase activity of ABCG2 was assayed in Sf9 insect cells infected with wild-type or SNP variants of ABCG2. Basal ATPase activity in cells transfected with Q141K ABCG2 was 1.8-fold lower than in cells transfected with wild-type ABCG2, but was comparable among cells expressing wild-type, V12M or D620N ABCG2. Confocal studies of ABCG2 localization revealed higher intracellular staining in the Q141K transfectants than in cells transfected with wild-type or V12M ABCG2. Decreased transport of Hoechst 33342 was observed in Sf9 cells expressing V12M ABCG2; however, this was not true in HEK-293 cells expressing V12M ABCG2. These results suggest that the Q141K SNP affects the transport efficiency of ABCG2 and may result in altered pharmacokinetics or drug-resistance profiles in clinical oncology.

No abstract

Golgins are coiled-coil proteins that play a key role in the regulation of Golgi architecture and function. Giantin, the largest golgin in mammals, forms a complex with p115, rab1, GM130, and soluble N-ethylmaleimide-sensitive factor attachment protein receptors (SNAREs), thereby facilitating vesicle tethering and fusion processes around the Golgi apparatus. Treatment with the microtubule destabilizing drug nocodazole transforms the Golgi ribbon into individual Golgi stacks. Here we show that siRNA-mediated depletion of giantin resulted in more dispersed Golgi stacks after nocodazole treatment than by control treatment, without changing the average cisternal length. Furthermore, depletion of giantin caused an increase in cargo transport that was associated with altered cell surface protein glycosylation. Drosophila S2 cells are known to have dispersed Golgi stacks and no giantin homolog. The exogenous expression of mammalian giantin cDNA in S2 cells resulted in clustered Golgi stacks, similar to the Golgi ribbon in mammalian cells. These results suggest that the spatial organization of the Golgi ribbon is mediated by giantin, which also plays a role in cargo transport and sugar modifications.

Variations in the amino acid sequence of ABC transporters have been shown to impact substrate specificity. We identified two acquired mutations in ABCG2, the ABC half-transporter overexpressed in mitoxantrone-resistant cell lines. These mutations confer differences in substrate specificity and suggest that naturally occurring variants could also affect substrate specificity. To search for the existence of single nucleotide polymorphisms (SNPs) in ABCG2, we sequenced 90 ethnically diverse DNAs from the Single Nucleotide Polymorphism Discovery Resource representing the spectrum of human genotypes. We identified 3 noncoding SNPs in the untranslated regions, 3 nonsynonymous and 2 synonymous SNPs in the coding region and 7 SNPs in the intron sequences adjacent to the sixteen ABCG2 exons. Nonsynonymous SNPs at nucleotide 238 (V12M; exon 2) and nucleotide 625 (Q141K; exon 5) showed a greater frequency of heterozygosity (22.2% and 10%) than the SNP at 2062 (D620N; exon 16). Heterozygous changes at nucleotide 238 are in linkage disequilibrium with an SNP observed 36 bases downstream from the end of exon 2. No polymorphism at amino acid 482 was identified to correspond to the R to G or R to T mutations previously found in two drug resistant cell lines. Among 23 drug resistant sublines for which sequence at position 482 was determined, no additional mutations were found. Heterozygosity at amino acid 12 allowed us to identify overexpression of a single allele in a subset of drug resistant cell lines, a feature that could be exploited clinically in evaluating the significance of ABCG2 expression in malignancy. We conclude that ABCG2 is well conserved and that described amino acid polymorphisms seem unlikely to alter transporter stability or function.

Urodele amphibians can regenerate their limbs. During limb regeneration, dermal fibroblasts are transformed into undifferentiated cells called blastema cells. These dermis-blastema cells show multipotency. Such so-called endogenous reprogramming of cell differentiation is one of the main targets of amphibian limb regeneration studies. It is well recognized that nerve presence controls the initiation of limb regeneration. Accordingly, nerve factors have been sought in amphibian limb regeneration. To investigate it, a relatively new study system called the accessory limb model (ALM) was developed. Using ALM, two signaling cascades (Fgf and Gdf5 signaling) came under focus. In the present study, Growth and differentiation factor-5 (Gdf5) application to wounded skin initiated limb regeneration responses and resulted in induction of a blastema-like structure in the absence of a nerve. However, the Gdf5-induced structure showed defects as a regeneration blastema, such as absence of detectable Prrx1 expression by in situ hybridization. The defects could be remedied by additional Fibroblasts growth factor (Fgf) inputs. These two inputs (Gdf5 and Fgfs) were sufficient to substitute for the nerve functions in the induction of limb regeneration. Indeed, Fgf2, Fgf8, and Gdf5 applications with the contralateral skin graft resulted in limb formation without nerve supply. Furthermore, acquisition of cartilage differentiation potential of dermal fibroblasts was tested in an in vivo and in vitro combination assay. Dermal fibroblasts cultured with Gdf5 were difficult to participate in cartilage formation when the cultured cells were grafted into cartilage forming region. In contrast, dermal fibroblasts cultured with Fgf2 and Fgf8 became easier to participate into cartilage formation in the same procedure. These results contribute to our understanding of molecular mechanisms of the early phase of amphibian limb regeneration.