AMIGO is expressed in multiple brain cell types and may regulate dendritic growth and neuronal survival

Amphoterin-induced gene and ORF (AMIGO) is a brain-enriched transmembrane immunoglobulin (Ig) superfamily protein with six extracellular leucine-rich repeats (LRR) and a single immunoglobulin-like (Ig) domain. We report here that AMIGO is a glycosylated protein widely expressed in the central nervous system (CNS), and can be found in neurons, astrocytes as well as oligodendrocytes. In morphologically mature primary neurons, endogenous AMIGO, and transfected full length AMIGO (AMIGO-FL) are largely dendritic, while AMIGO with its LRR domain deleted (AMIGO-Ig) is predominantly axonal. In line with AMIGO's dendritic localization, siRNA-mediated silencing of AMIGO resulted in reduced dendritic growth of cortical neurons in culture. SH-SY5Y cells stably over-expressing AMIGO are more resistant to apoptosis induced by staurosporine and H(2) O(2) compared to vector controls. AMIGO therefore likely plays important roles in dendritic outgrowth during development, and could modulate the survival of developing and adult neurons.     

Sequence of human syndecan indicates a novel gene family of integral membrane proteoglycans

The structure of human syndecan, an integral membrane proteoglycan, has been determined by cloning its full-length cDNA, which codes for the entire 310-amino acid-long core protein, including the NH2-terminal signal peptide. Similar to mouse syndecan (Saunders, S., Jalkanen, M., O'Farrell, S., and Bernfield, M. (1989) J. Cell Biol. 108, 1547-1556), the core protein of human syndecan can be divided into three domains: a matrix-interacting ectodomain containing putative glycosaminoglycan attachment sites, a 25-residue hydrophobic membrane-spanning domain, and a 34-residue cytoplasmic domain. Several interesting conserved structures were revealed by comparing the human syndecan sequence to the murine one. (i) Although the ectodomains are only 70% identical, all putative glycosaminoglycan attachment sites are identical (two of them belong to the consensus sequence SGXG and three others to (E/D)GSG(E/D), as are also (ii) the single putative N-glycosylation site and (iii) the proteinase-sensitive dibasic RK site adjacent to the extracellular face of the transmembrane domain. Furthermore, (iv) the transmembrane domain is 96% identical, as the only change in human syndecan was an alteration of an alanine residue to glycine; and finally, (v) the cytoplasmic domain is 100% identical, including 3 identically located tyrosine residues. Comparison of transmembrane and cytoplasmic domains to a third cell-surface proteoglycan, 48K5 from human lung fibroblasts (Marynen, P., Zhang, J., Cassiman, J., Vanden Berghe, H., and David, C. (1989) J. Biol. Chem. 264, 7017-7024), indicates that the transmembrane and cytoplasmic domains are similar also in this molecule regardless of the presence of a totally nonhomologous ectodomain. Thus, the transmembrane and cytoplasmic domains are unique for these cell-surface proteoglycans, which we propose to be members of a novel gene family of syndecans.     

SCAN domain-containing 2 gene (SCAND2) is a novel nuclear protein derived from the zinc finger family by exon shuffling

The SCAN domain is a recently recognized protein domain that characterizes a subfamily of the Krüppel-like zinc finger proteins. We have previously described a novel SCAN domain-containing 2 gene (SCAND2) that does not belong to the zinc finger family. We report structural and sequence analyzes of all known members of the SCAN family and use these data to illustrate a model of gene family evolution. Most of the SCAN containing genes share common gene organization features that support the proposed origin for SCAND2 by disruption of an ancestral SCAN-zinc finger gene by a retroposition event and subsequent exon shuffling.     

The cytoplasmic/transmembrane domain of dipeptidyl peptidase IV, a type II glycoprotein, contains an apical targeting signal that does not specifically interact with lipid rafts

We investigated the signals involved in the apical targeting of dipeptidyl peptidase IV (DPP IV/CD26), an archetypal type II transmembrane glycoprotein. A secretory construct, corresponding to the DPP IV ectodomain, was first stably expressed in both the enterocytic-like cell line Caco-2 and the epithelial kidney MDCK cells. Most of the secretory form of the protein was delivered apically in MDCK cells, whereas secretion was 60% basolateral in Caco-2 cells, indicating that DPP IV ectodomain targeting is cell-type-dependent. A chimera (CTM-GFP) containing only the cytoplasmic and transmembrane domains of mouse DPP IV plus the green fluorescent protein was then studied. In both cell lines, this chimera was preferentially expressed at the apical membrane. By contrast, a secretory form of GFP was randomly secreted, indicating that GFP by itself does not contain cryptic targeting information. Comparison of the sequence of the transmembrane domain of DPP IV and several other apically targeted proteins does not show any consensus, suggesting that the apical targeting signal may be conformational. Neither the DPP IV nor the CTM-GFP chimera was enriched in lipid rafts. Together these results indicate that, besides the well-known raft-dependent apical targeting pathway, the fate of the CTM domain of DPP IV may reveal a new raft-independent apical pathway.     

LIV-1 breast cancer protein belongs to new family of histidine-rich membrane proteins with potential to control intracellular Zn2+ homeostasis

Investigation of the protein product of the oestrogen-regulated gene LIV-1, implicated in metastatic breast cancer, has revealed 10 protein sequences of unknown function that belong to a new family with potential to control intracellular Zn2+ homeostasis. Sequence alignment highlights the similarity in transmembrane domains and extramembrane charged residues, indicating potential ion-transport ability. This family has a novel highly conserved motif of 66 residues, including a transmembrane domain and a catalytic zinc-binding sequence of zinc metalloproteases, containing conserved (indicated in bold type) proline and glutamine residues, HEXPHEXGD. These proteins contain more plentiful histidine-rich repeats than zinc transporters, suggesting an ability to bind or transport zinc across membranes. I propose that these 11 proteins form a new family with the potential to control intracellular Zn2+ homeostasis.     

Internal targeting signal of the BCS1 protein: a novel mechanism of import into mitochondria.

The BCS1 protein is anchored in the mitochondrial inner membrane via a single transmembrane domain and has an N(out)-C(in) topology. Unlike the majority of nuclear encoded mitochondrial preproteins, the BCS1 protein does not contain an N-terminal targeting sequence. A positively charged segment of amino acids which is located immediately C-terminal to the transmembrane domain acts as an internal targeting signal. In order to function, we postulate that this sequence co-operates with the transmembrane domain to form a tight hairpin loop structure. This loop is translocated across the inner membrane via the MIM/mt-Hsp70 machinery in a membrane potential-dependent manner. This novel mechanism of import and sorting of the BCS1 protein is proposed to represent a more general mechanism used by a number of inner membrane proteins.     

The autocrine motility factor receptor gene encodes a novel type of seven transmembrane protein.

Autocrine motility factor receptor (AMFR) is a cell surface glycoprotein of molecular weight 78,000 (gp78), mediating cell motility signaling in vitro and metastasis in vivo. Here, we cloned the full-length cDNAs for both human and mouse AMFR genes. Both genes encode a protein of 643 amino acids containing a seven transmembrane domain, a RING-H2 motif and a leucine zipper motif and showed a 94.7% amino acid sequence identity to each other. Analysis of the amino acid sequence of AMFR with protein databases revealed no significant homology with all known seven transmembrane proteins, but a significant structural similarity to a hypothetical protein of Caenorhabditis elegans, F26E4.11. Thus, AMFR is a highly conserved gene which encodes a novel type of seven transmembrane protein.     

Molecular cloning of Rhombex-40 a transmembrane protein from the ventral medullary surface of the rat brain by differential display

Respiration-related neurons, which detect various chemicals in cerebrospinal fluid, are localized to the ventral medullary surface (VMS). We hypothesized that expression of genes involved in respiratory function is upregulated in the VMS. By differential display, we looked for genes differentially expressed in VMS neurons and cerebral cortical neurons. Seventeen clones of interest were isolated, and sequence analysis revealed that one of these clones encoded a putative transmembrane protein, rhombencephalic expression protein-40 kDa (Rhombex-40). The rat Rhombex-40 was composed of 374 amino acid residues, and the predicted secondary structure displays a signal peptide in the N-terminus and single-pass transmembrane domain in the center of the sequence. An analysis of consensus sequences identified several phosphorylation sites in the intracellular domain. Expression of rat Rhombex-40 mRNA is high in the brain, and low in lung, liver and kidney. No homologous protein sequence was found in database searches. Whereas the biological function of this protein is presently unknown, its structural features and high expression in the brain suggest that Rhombex-40 may function as a novel transmembrane molecule in neural cells of the brain.     

Molecular characterization of a new type of receptor-like kinase (wlrk) gene family in wheat

In plants, several types of receptor-like kinases (RLK) have been isolated and characterized based on the sequence of their extracellular domains. Some of these RLKs have been demonstrated to be involved in plant development or in the reaction to environmental signals. Here, we describe a RLK gene family in wheat (wlrk, wheat leaf rust kinase) with a new type of extracellular domain. A member of this new gene family has previously been shown to cosegregate with the leaf rust resistance gene Lr10. The diversity of the wlrk gene family was studied by cloning the extracellular domain of different members of the family. Sequence comparisons demonstrated that the extracellular domain consists of three very conserved regions interrupted by three variable regions. Linkage analysis indicated that the wlrk genes are specifically located on chromosome group 1 in wheat and on the corresponding chromosomes of other members of the Triticeae family. The wlrk genes are constitutively expressed in the aerial parts of the plant whereas no expression was detected in roots. Protein immunoblots demonstrated that the WLRK protein coded by the Lrk10 gene is an intrinsic plasma membrane protein. This is consistent with the hypothesis that WLRK proteins are receptor protein kinases localized to the cell surface. In addition, we present preliminary evidence that other disease resistance loci in wheat contain genes which are related to wlrk.     

SGR2, a phospholipase-like protein, and ZIG/SGR4, a SNARE, are involved in the shoot gravitropism of Arabidopsis

In higher plants, the shoot and the root generally show negative and positive gravitropism, respectively. To elucidate the molecular mechanisms involved in gravitropism, we have isolated many shoot gravitropism mutants in Arabidopsis. The sgr2 and zig/sgr4 mutants exhibited abnormal gravitropism in both inflorescence stems and hypocotyls. These genes probably are involved in the early step(s) of the gravitropic response. The sgr2 mutants also had misshapen seed and seedlings, whereas the stem of the zig/sgr4 mutants elongated in a zigzag fashion. The SGR2 gene encodes a novel protein that may be part of a gene family represented by bovine phosphatidic acid-preferring phospholipase A1 containing a putative transmembrane domain. This gene family has been reported only in eukaryotes. The ZIG gene was found to encode AtVTI11, a protein that is homologous with yeast VTI1 and is involved in vesicle transport. Our observations suggest that the two genes may be involved in a vacuolar membrane system that affects shoot gravitropism.     

Molecular cloning, tissue distribution, and chromosomal localization of MMEL2, a gene coding for a novel human member of the neutral endopeptidase-24.11 family

Members of the neutral endopeptidase (NEP, also known as MME for membrane metallo-endopeptidase in the Human Gene Nomenclature database) family play significant roles in pain perception, arterial pressure regulation, phosphate metabolism, and homeostasis. In this paper, we report the cloning of a new human member of the NEP family that we named MMEL2 for membrane metallo-endopeptidase-like 2. The MMEL2 protein has the structural characteristics of type II transmembrane proteins, although the presence of a furin-like cleavage site in the ectodomain suggests that it may be released into the medium following proteolytic cleavage. The MMEL2 protein contains the zinc-binding consensus sequence HEXXH and all the residues known to be essential for the enzymatic activity of other members of the family. The MMEL2 mRNA was detected predominantly in testis, but weak expression also was observed in brain, kidney, and heart. The human MMEL2 gene was mapped to 1p36 by fluorescence in situ hybridization. It will be important to test whether MMEL2 defects are associated with diseases such as hereditary motor sensory neuropathy 2A, Schwartz-Jampel-Aberfeld syndrome, or neuroblastoma, which all map to this locus.     

Molecular evolutionary analyses of the Arabidopsis L7 ribosomal protein gene family

Cytoplasmic ribosomal protein (r-protein) genes in Arabidopsis thaliana are encoded by 80 multigene families that contain between two and seven members. Gene family members are typically similar at the protein sequence level, with the most divergent members of any gene family retaining 94% identity, on average. However, three Arabidopsis r-protein families - S15a, L7 and P2 - contain highly divergent family members. Here, we investigated the organization, structure, expression and molecular evolution of the L7 r-protein family. Phylogenetic analyses showed that L7 r-protein gene family members constitute two distinct phylogenetic groups. The first group including RPL7B, RPL7C and RPL7D has homologs in plants, animals and fungi. The second group represented by RPL7A is found in plants but has no orthologs from other fully-sequenced eukaryotic genomes. These two groups may have derived from a duplication event prior to the divergence of animals and plants. All four L7 r-protein genes are expressed and all exhibit a differential expression in inflorescence and flowers. RPL7A and RPL7B are less expressed than the other genes in all tissues analyzed. Molecular characterization of nucleic and protein sequences of L7 r-protein genes and analysis of their codon usage did not indicate any functional divergence. The probable evolution of an extra-ribosomal function of group 2 genes is discussed.     

Human GPR107 and murine Gpr108 are members of the LUSTR family of proteins found in both plants and animals, having similar topology to G-protein coupled receptors.

Two new cDNAs, human GPR107 and murine GPR108, were cloned from mammalian lung that are members of a novel gene family encoding proteins that are predicted to have an amino-terminal hydrophobic signal peptide sequence, a long extracellular domain and a carboxy-terminal seven transmembrane domain (LUSTR domain) similar to GPCRs. The 18-exon human GPR107 gene is located at 9q34.2-3 and spans 86.4 kb and the cDNA encodes a 552 residue protein. The closely related, but not homologous, 17-exon murine Gpr108 gene is located at 17C-D and spans 12.8 kb. The murine Gpr108 cDNA encodes a 562 residue protein that has 49% identity to human GPR107. They are distantly related to two other genes, transmembrane protein 87A and 87B that encode LUSTR domain-containing proteins in the human genome. LUSTR proteins are also found in Drosophila, Saccharomyces and Arabidopsis, but are absent from bacteria, archaea and viruses. This suggests that GPCRs are present in higher plants.     

GlyGly-CTERM and rhombosortase: a C-terminal protein processing signal in a many-to-one pairing with a rhomboid family intramembrane serine protease

The rhomboid family of serine proteases occurs in all domains of life. Its members contain at least six hydrophobic membrane-spanning helices, with an active site serine located deep within the hydrophobic interior of the plasma membrane. The model member GlpG from Escherichia coli is heavily studied through engineered mutant forms, varied model substrates, and multiple X-ray crystal studies, yet its relationship to endogenous substrates is not well understood. Here we describe an apparent membrane anchoring C-terminal homology domain that appears in numerous genera including Shewanella, Vibrio, Acinetobacter, and Ralstonia, but excluding Escherichia and Haemophilus. Individual genomes encode up to thirteen members, usually homologous to each other only in this C-terminal region. The domain's tripartite architecture consists of motif, transmembrane helix, and cluster of basic residues at the protein C-terminus, as also seen with the LPXTG recognition sequence for sortase A and the PEP-CTERM recognition sequence for exosortase. Partial Phylogenetic Profiling identifies a distinctive rhomboid-like protease subfamily almost perfectly co-distributed with this recognition sequence. This protease subfamily and its putative target domain are hereby renamed rhombosortase and GlyGly-CTERM, respectively. The protease and target are encoded by consecutive genes in most genomes with just a single target, but far apart otherwise. The signature motif of the Rhombo-CTERM domain, often SGGS, only partially resembles known cleavage sites of rhomboid protease family model substrates. Some protein families that have several members with C-terminal GlyGly-CTERM domains also have additional members with LPXTG or PEP-CTERM domains instead, suggesting there may be common themes to the post-translational processing of these proteins by three different membrane protein superfamilies.     

The Escherichia coli FtsH protein is a prokaryotic member of a protein family of putative ATPases involved in membrane functions, cell cycle control, and gene expression.

The ftsH gene is essential for cell viability in Escherichia coli. We cloned and sequenced the wild-type ftsH gene and the temperature-sensitive ftsH1(Ts) gene. It was suggested that FtsH protein was an integral membrane protein of 70.7 kDa (644 amino acid residues) with a putative ATP-binding domain. The ftsH1(Ts) gene was found to have two base substitutions within the coding sequence corresponding to the amino acid substitutions Glu-463 by Lys and Pro-587 by Ala. Homology search revealed that an approximately 200-amino-acid domain, including the putative ATP-binding sequence, is highly homologous (35 to 48% identical) to the domain found in members of a novel, eukaryotic family of putative ATPases, e.g., Sec18p, Pas1p, CDC48p, and TBP-1, which function in protein transport pathways, peroxisome assembly, cell division cycle, and gene expression, respectively. Possible implications of these observations are discussed.