Dense Alu clustering and a potential new member of the NF kappa B family within a 90 kilobase HLA class III segment.

We have conducted a detailed structural analysis of 90 kilobases (kb) of the HLA Class III region from the Bat2 gene at the centromeric end to 23 kb beyond TNF. A single contig of 80 kb was sequenced entirely with a group of four smaller contigs covering 10 kb being only partly sequenced. This region contains four known genes and a novel telomeric potential coding region. The genes are bracketed by long, dense clusters of Alu repeats belonging to all the major families. At least six new families of MER repeats and one pseudogene are intercalated within and between the Alu clusters. The most telomeric 3.8 kb contains three potential exons, one of which bears strong homology to the ankyrin domain of the DNA binding factors NF kappa B and I kappa B.

E-box- and MEF-2-independent muscle-specific expression, positive autoregulation, and cross-activation of the chicken MyoD (CMD1) promoter reveal an indirect regulatory pathway.

Members of the MyoD family of gene-regulatory proteins (MyoD, myogenin, myf5, and MRF4) have all been shown not only to regulate the transcription of numerous muscle-specific genes but also to positively autoregulate and cross activate each other's transcription. In the case of muscle-specific genes, this transcriptional regulation can often be correlated with the presence of a DNA consensus in the regulatory region CANNTG, known as an E box. Little is known about the regulatory interactions of the myogenic factors themselves; however, these interactions are thought to be important for the activation and maintenance of the muscle phenotype. We have identified the minimal region in the chicken MyoD (CMD1) promoter necessary for muscle-specific transcription in primary cultures of embryonic chicken skeletal muscle. The CMD1 promoter is silent in primary chick fibroblast cultures and in muscle cell cultures treated with the thymidine analog bromodeoxyuridine. However, CMD1 and chicken myogenin, as well as, to a lesser degree, chicken Myf5 and MRF4, expressed in trans can activate transcription from the minimal CMD1 promoter in these primary fibroblast cultures. Here we show that the CMD1 promoter contains numerous E-box binding sites for CMD1 and the other myogenic factors, as well as a MEF-2 binding site. Surprisingly, neither muscle-specific and the other myogenic factors, as well as a MEF-2 binding site. Surprisingly, neither muscle-specific expression, autoregulation, or cross activation depends upon the presence of of these E-box or MEF-2 binding sites in the CMD1 promoter. These results demonstrate that the autoregulation and cross activation of the chicken MyoD promoter through the putative direct binding of the myogenic basic helix-loop-helix regulatory factors is mediated through an indirect pathway that involves unidentified regulatory elements and/or ancillary factors.

Structure and expression of hcr, a locus rearranged with c-myc in a woodchuck hepatocellular carcinoma.

We have previously described a rearrangement of the proto-oncogene c-myc with a new cellular sequence of unknown function in a woodchuck primary liver tumor. We have now cloned and further analysed the normal woodchuck locus (termed hcr) of the sequence involved in the rearrangement with c-myc. The hcr locus is highly expressed in hepatocytes but not in other cell types examined and is conserved in mammals. Two unspliced hcr transcripts 4.5 and 4.7 kb long accumulate in liver cell nuclei. These transcripts differ only in their 3' extremities, located 180 bases apart, and by additional poly(A) tailing of the longer RNA species. The genomic sequence flanking the transcription start site contains variant elements of a classical eukaryotic promoter. Nucleotide sequence analysis of cDNA clones for the hcr RNA reveals that the 5' end of the hcr transcripts contains a short open reading frame of only 3 gamma codons initiated by an ATG. The biological function of her RNA remains to be determined.

Computer generation and statistical analysis of a data bank of protein sequences translated from GenBank.

We describe PGtrans, a new and freely available protein sequence databank (2625 sequences, 554198 amino-acids). This data bank is routinely produced by automatic computer translation of the nucleotide sequence library GenBank. The information needed for the translation process (transcriptional orientation, location of coding regions, splice sites and pertinent genetic code) is gathered by the translation program through an "intelligent" scanning of the documentary field of each GenBank entry. Inconsistencies resulting in unexpected termination codons are detected and reported thus allowing the correction of data bank errors. PGtrans is intended as a tool for protein similarity searches. Its reasonable overall size (2 Moctets) makes it suitable for micro-computer environments. Up to date amino-acid composition data and relative abundances of di-, tri-, and tetra-peptides in proteins of known sequences are presented and discussed.

Collaborative annotation of genes and proteins between UniProtKB/Swiss-Prot and dictyBase.

UniProtKB/Swiss-Prot, a curated protein database, and dictyBase, the Model Organism Database for Dictyostelium discoideum, have established a collaboration to improve data sharing. One of the major steps in this effort was the 'Dicty annotation marathon', a week-long exercise with 30 annotators aimed at achieving a major increase in the number of D. discoideum proteins represented in UniProtKB/Swiss-Prot. The marathon led to the annotation of over 1000 D. discoideum proteins in UniProtKB/Swiss-Prot. Concomitantly, there were a large number of updates in dictyBase concerning gene symbols, protein names and gene models. This exercise demonstrates how UniProtKB/Swiss-Prot can work in very close cooperation with model organism databases and how the annotation of proteins can be accelerated through those collaborations.

Heuristic informational analysis of sequences.

Nucleotide or amino-acid sequences are interpreted as successions of words of length k (k-tuples) the frequencies of which are highly variable in different statistical populations of genes or proteins. After building k-tuple reference tables from coherent subsets or entire data banks, the local information content profile of individual sequences is drawn. Anomalous regions (peaks or depressions) of such a profile can lead to the discovery and identification of specific sequence patterns. Along the same principle, the simultaneous use of two reference statistical populations and the computation of an index combining the two information profiles lead to a general and powerful discriminant analysis methods. The identification of a "signal" associated with gene conversion, the introns/exons discrimination and the location of function specific patterns in proteins are given as examples of successful applications of this heuristic informational approach.

Relevance of the lariat model to the splicing of the H-2 gene family.

We examined the splicing of the H-2 gene family, taking the H-2Kd as a prototype, in the framework of the lariat model. We mainly investigated the mechanism described by Konarska et al. [Nature (Lond.) 313, 552-557 (1985)] who propose a direct interaction between the 5' splicing site and the branching region. We also checked each of the H-2 introns for the presence of patterns resembling the published consensus for the branching region. The known splicing events in the H-2 gene family are not always consistent with the current models, and our results indicate that slightly different mechanisms govern the splicing of different introns. A tentative explanation of the alternative splicing of the first and last intron, previously described, is given. The removal of the large third intron is not easily rationalized unless new rules for an additional multistep processing are postulated.

Conjugative R plasmids in group C and G streptococci.

Two streptococcal isolates of groups C and G harbored conjugative R plasmids with molecular weights of 17 X 10(6) (pIP646) and 20 X 10(6) (pIP920). These plasmids carried genetic markers for resistance to macrolides and related drugs, as well as to chloramphenicol (pIP920), and have very similar HindIII restriction enzyme patterns.

Objective comparison of exon and intron sequences by means of 2-dimensional data analysis methods.

Here we advocate the use of 2-dimensional data representation in the context of the informational approach of sequence analysis (Claverie & Bougueleret (1986) Nucleic Acids Research 14, 179-196) by applying these methods to the problem of intron/exon discrimination. Two main findings are reported: i) oligonucleotide patterns complementary to the Ul small nuclear RNA are specifically avoided in exon sequences, ii) vertebrate intron sequences, to the exclusion of other eukaryotic phyla, are characterized by a peculiar distribution of CpG containing patterns.

Hepatitis B virus DNA integration in a sequence homologous to v-erb-A and steroid receptor genes in a hepatocellular carcinoma.

Hepatitis B virus (HBV) is clearly involved in the aetiology of human hepatocellular carcinoma (HCC) and the finding of HBV DNA integration into human liver DNA in almost all HCCs studied suggested that these integrated viral sequences may be involved in liver oncogenesis. Several HBV integrations in different HCCs and HCC-derived cell lines have been analysed after molecular cloning without revealing any obvious role for HBV. From a comparison of a HBV integration site present in a particular HCC with the corresponding unoccupied site in the non-tumorous tissue of the same liver, we now report that HBV integration places the viral sequence next to a liver cell sequence which bears a striking resemblance to both an oncogene (v-erb-A) and the supposed DNA-binding domain of the human glucocorticoid receptor and human oestrogen receptor genes. We suggest that this gene, usually silent or transcribed at a very low level in normal hepatocytes, becomes inappropriately expressed as a consequence of HBV integration, thus contributing to the cell transformation.

Rare sequence motifs are common constituents of hypervariable antibody regions.

An analysis of the amino acid sequences of variable regions of human and mouse antibody molecules was performed. It involved comparison of their constituent tetrapeptides with those found in a reference set (the somatic self) built with non-immunological proteins found in a protein data base. It appeared that hypervariable regions, particularly CDR1 and CDR3, are often made up of rare tetrapeptides not present in the reference set. As assessed by simple statistical tests, this bias was significant. We discuss its possible connection with the problem of antibody immunogenicity. This result provides indirect support for the existence of idiopeptides predicted by the "peptidic self model".

Implications of a Fab-like structure for the T-cell receptor.

The antigen-specific receptor of T lymphocytes (TCR) and the Fab moiety of immunoglobulins are expected to fold into similar three-dimensional structures because of their identical protein domain organization, the conservation of key residues and their overall sequence homology. However, T cells mostly appear to recognize short peptide antigens bound to MHC class I or class II presenting molecules. A complete model of the human leucocyte antigen molecule (HLA-A2) reconstructed from the alpha-carbon coordinates was used to investigate the putative organization of a TCR/peptide/HLA-A2 complex. In this article, Jean-Michel Claverie and co-workers show that the respective geometries of a Fab-like TCR structure and of the HLA-A2 antigen binding site suggest a model where the third variable regions of both chains of the TCR mainly interact with the peptide antigen, while the first and/or second less variable regions are in position for making contact with residues pointing up from the alpha 1 and alpha 2 helical regions of the HLA-A2 molecule.

Overproduction of the EcoR V endonuclease and methylase.

Strains overproducing the EcoR V endonuclease and methylase have been obtained by inserting each of the two genes in expression vectors containing the lambda PL promoter. The methylase is overproduced to a level reaching 5-10% of the total cellular proteins, which represents a 50-100 fold increase. A 30 fold overproduction of endonuclease was achieved by randomly positioning the EndRV gene downstream of the lambda PL promoter. The situation in the endonuclease overproducing clone resembles that encountered in maxi-cells. The strains described here allowed a quick purification of both enzymes in sufficient amounts for crystallisation attempts.

Characterization of the genes coding for the Eco RV restriction and modification system of Escherichia coli.

A plasmid encoding the recently described Eco RV restriction and modification system has been isolated and characterized. This plasmid, pLB1 , is 6.2 kb long and carries only the Eco RV genes. A subclone of 3 kb has been inserted in pBR322. The relative positions of the endonuclease and the methylase genes were determined by the construction of a set of overlapping deletions generated by Bal31 resection. The DNA sequence of a 2.2 kb fragment containing the two genes was determined. The two genes are transcribed divergently from a 310 bp region and the assignment of the coding region has been confirmed by direct aminoacid sequence analysis. Possible mechanisms of regulation of the endonuclease gene expression at the translational level are proposed and discussed.

Sequence comparison of five polymerases (L proteins) of unsegmented negative-strand RNA viruses: theoretical assignment of functional domains.

The large (L) protein subunit of unsegmented negative-strand RNA virus polymerases is thought to be responsible for the majority of enzymic activities involved in viral transcription and replication. In order to gain insight into this multifunctional role we compared the deduced amino acid sequences of five L proteins of rhabdoviruses (vesicular stomatitis virus and rabies virus) or paramyxoviruses (Sendai virus, Newcastle disease virus and measles virus). Statistical analysis showed that they share an atypical amino acid usage, outlining the uniqueness of the negative-strand virus life style. Similarity studies between L proteins traced evolutionary relationships in partial disagreement with the present taxonomic arrangement of this group of viruses. The five L proteins exhibit a high degree of homology along most of their length, with strongly invariant amino acids embedded in conserved blocks separated by variable regions, suggesting a structure of concatenated functional domains. The most highly conserved central block contains the probable active site for RNA synthesis. We tentatively identified some other functional sites, distributed around this central core, that would naturally work together to assure the polymerase activity. This provides detailed guidelines for the future study of L proteins by site-directed mutagenesis.

Nucleotide sequence and evolution of ETn elements.

The ETn (for "early transposon") family of long repeated sequences in abundantly transcribed in early mouse embryos from retroviral-like long terminal repeats. Nucleotide sequencing of two elements does not reveal any long open reading frame nor significant homology to retroviral proteins. The genetic polymorphism, monitored by Southern blotting within and across mouse species, reflects a concerted mode of evolution for the ETn sequences.

High-level, plasmid-borne resistance to gentamicin in Streptococcus faecalis subsp. zymogenes.

Each of three isolates of Streptococcus faecalis subsp. zymogenes harbored three R plasmids and a hemolysin-bacteriocin plasmid. The plasmids carried by one of these strains were physically characterized after their conjugative transfer. In each strain one of the plasmids carried genetic markers for resistance to gentamicin, kanamycin, sisomicin, netilmicin, and tobramycin.

Characterization and chromosomal assignment of a human cDNA encoding a protein related to the murine 102-kDa cadherin-associated protein (alpha-catenin).

We report the characterization of a human cDNA encompassing the complete coding region of a 945-residue putative protein (CAP-R) 80% identical to the recently described murine 102-kDa alpha-catenin (CAP102). The CAP-R protein mostly differs from CAP102 by the presence of a 48-residue insert. This insert exhibits similarity with a segment of the type 1 neurofibromatosis gene product. The analysis of a publicly available human "expressed sequence tag" collection revealed the existence of another human cDNA more closely related (89% identical) to CAP102. This strongly suggests that CAP-R is not the human homologue of the murine 102-kDa alpha-catenin but a new closely related gene of the vinculin family. This is further supported by the computed mutation rates falling outside the range observed for mammalian orthologous genes. Using in situ hybridization, the CAP-R gene could be mapped to the p11.1-p12 region of human chromosome 2 and to the homologous B3-D region of mouse chromosome 6.