Nucleotide sequence analysis of 77.7 kb of the human V beta T-cell receptor gene locus: direct primer-walking using cosmid template DNAs.

The nucleotide sequence of 77.7 kb from the human T-cell receptor beta-chain locus was determined directly from three overlapping cosmid clones using the primer-walking approach. Computer-aided analyses of this sequence reveal the presence of at least 11 genic regions that are closely related to the human T-cell receptor beta variable region (TCRBV) gene family. These include five germline sequences that have previously been determined, V beta 21.2, V beta 8.1, V beta 8.2, V beta 8.3, and V beta 16, and four whose sequences have partially been determined at the mRNA level, V beta 6, V beta 23, V beta 12.2, V beta 24. The two remaining V beta Tcr-related sequences have eluded discovery by cDNA and RT-PCR cloning and genomic blot hybridization methods. These two V beta Tcr-related genes lack > 75% nucleotide sequence identity with any other V beta Tcr gene member and therefore, by convention, are referred to as new subfamily members V beta 25 and V beta 26. This lack of shared identity with other subfamily members explains why they were not detected by hybridization. The promoter regions of these V beta Tcr genes contain the conserved Tcr decamer element located between 80 and 110 bp 5' of the translation start site, generally near a putative TATAA promoter element. Our sequence analysis also reveals that a 3.3-kb duplication unit was involved in the recombination event that produced the closely related V beta 8.1 and 8.2 gene subfamily members. This sequenced region of the V beta locus contains an average number of repetitive DNA elements (21 Alu, three L1, three MER, and three retrovirus-related elements.

Direct sequencing of baculovirus genomic DNA: sequence determination of the engineered respiratory syncytial virus chimeric FG gene.

Primer-directed enzymatic sequencing has proven to be an efficient and effective method for sequencing various size double-stranded DNA templates. We previously developed a primer-directed sequencing procedure for using double-stranded cosmid (50 kb) DNAs as template. We are interested in using this method to directly sequence larger DNA templates. Towards this goal we applied this method to directly sequence an engineered gene that had been transferred and integrated into the 130-kb baculovirus genome. Both crudely prepared and CsCl gradient-banded baculovirus DNAs were tested and reasonable sequencing ladders were obtained for both types of DNA templates. As little as 3 micrograms of gradient-banded baculovirus DNA were found to be sufficient to obtain film exposure times similar to those observed for cosmid size templates, 24 to 48 h. Effectiveness of the described method was demonstrated by obtaining the complete sequence of the engineered respiratory syncytial virus chimeric FG gene (2.5 kb in length) directly from the recombinant baculovirus "Baculo-FG" genome. Thus, our results demonstrate first, that double-stranded DNA templates as large as 130 kb can be sequenced directly and second, that the nucleotide sequence of engineered genes integrated within the baculovirus genome can be determined without the use of any intermediate steps of procedures.

Reexamination of the African hominoid trichotomy with additional sequences from the primate beta-globin gene cluster.

Additional DNA sequence information from a range of primates, including 13.7 kb from pygmy chimpanzee (Pan paniscus), was added to data sets of beta-globin gene cluster sequence alignments that span the gamma 1, gamma 2, and psi eta loci and their flanking and intergenic regions. This enlarged body of data was used to address the issue of whether the ancestral separations of gorilla, chimpanzee, and human lineages resulted from only one trichotomous branching or from two dichotomous branching events. The degree of divergence, corrected for superimposed substitutions, seen in the beta-globin gene cluster between human alleles is about a third to a half that observed between two species of chimpanzee and about a fourth that between human and chimpanzee. The divergence either between chimpanzee and gorilla or between human and gorilla is slightly greater than that between human and chimpanzee, suggesting that the ancestral separations resulted from two closely spaced dichotomous branchings. Maximum parsimony analysis further strengthened the evidence that humans and chimpanzees share the longest common ancestry. Support for this human-chimpanzee clade is statistically significant at P = 0.002 over a human-gorilla clade or a chimpanzee-gorilla clade. An analysis of expected and observed homoplasy revealed that the number of sequence changes uniquely shared by human and chimpanzee lineages is too large to be attributed to homoplasy. Molecular clock calculations that accommodated lineage variations in rates of molecular evolution yielded hominoid branching times that ranged from 17-19 million years ago (MYA) for the separation of gibbon from the other hominoids to 5-7 MYA for the separation of chimpanzees from humans. Based on the relatively late dates and mounting corroborative evidence from unlinked nuclear genes and mitochondrial DNA for the close sister grouping of humans and chimpanzees, a cladistic classification would place all apes and humans in the same family. Within this family, gibbons would be placed in one subfamily and all other extant hominoids in another subfamily. The later subfamily would be divided into a tribe for orangutans and another tribe for gorillas, chimpanzees, and humans. Finally, gorillas would be placed in one subtribe with chimpanzees and humans in another, although this last division is not as strongly supported as the other divisions.

Sequence analysis of linked maize 22 kDa alpha-zein genes.

We have determined the nucleotide sequence of a 7343 bp zein genomic clone (gZ22.8H3) from the maize inbred W64A. Computer-aided analysis of the DNA sequence revealed two contiguous 22 kDa alpha-zein genes. The 5' gene (gZ22.8) encodes a complete polypeptide and contains putative regulatory sequences in both the 5' and 3' flanking regions that are typical of zein genes. In contrast, the 3' gene (psi gZ22.8) appears to be a pseudogene, because it contains numerous insertions and deletions that would prevent translation of the mRNA. Alignment of the 5' and 3' flanking sequences of both genes indicated that they resulted from a 3.3 kb DNA duplication event.

Mitochondrial D-loop sequences are integrated in the rat nuclear genome.

We have cloned two fragments of rat nuclear DNA (nucDNA), 3.3 x 10(3) nucleotide-pairs (knp) and 9.1 knp, that contain a 0.5 knp section sharing 80% sequence identity with the mitochondrial DNA (mtDNA) heavy strand origin of replication (D-loop) nascent strand and 88% identity with each other. The light and heavy strand promoters of the D-loop region are not present in either clone, thus they likely do not function as replication origins in the nuclear genome. The nucDNA sequences surrounding the mtDNA-like sequences are not mitochondrial, thus the mtDNA-like sequences are demonstrably covalently linked in the nuclear genome. Indeed, the surrounding nuclear sequences of each clone also share 88% identity. This sequence arrangement strongly suggests an initial insertion of mtDNA into nucDNA with subsequent amplification of an encompassing region of nucDNA. Divergence calculations suggest that the mtDNA insertion occurred around 13.6 million years ago (MYA) with the subsequent separation occurring around 6.5 MYA. The mtDNA-like sequences of the nuclear clones hybridize strongly to a number of different BamHI-PstI restriction fragments, suggesting either repeated integration and/or frequent mutational events producing new restriction enzyme sites. It is not yet known if one or more of the uncloned D-loop-like sequences are associated with promoters, which would suggest possible function. The 3.3 knp nucDNA fragment is present in low copy number. In contrast, the 9.1 knp nucDNA fragment appears to be moderately repeated. The elements do not appear to be tandemly repeated. The nucDNA clones contain remnants of rat long interspersed repetitive element (LINE) sequences; in addition the 9.1 knp fragment contains sequences with similarity to portions of viral reverse transcriptase and RNaseH genes. Until now, all mtDNA-like sequences found in the nuclear genome have been coding sequences. This is the first confirmation by sequence analysis of a portion of the mtDNA control region in the nuclear genome.

Strategy and methods for directly sequencing cosmid clones.

The primer-directed enzymatic sequencing method for sequencing double-stranded DNA templates has made possible the development of new strategies for directly sequencing large DNA molecules. Toward this goal, we have developed a strategy and the necessary techniques to obtain the complete sequence of cosmid clones (double-stranded DNA molecules in the size range of 50 kb). Our present strategy uses the chemical sequencing method to obtain sequence initiation points internal to a cosmid insert and the primer-directed enzymatic DNA sequencing method to extend these sequence contigs. As part of this development we added a nucleotide "chase" solution to the standard T7 sequencing protocol and included the use of both [alpha-32P]-dATP and -dCTP for labeling. With these modifications our double-stranded cosmid DNA sequencing reactions routinely extend well beyond 1000 bp, and film exposure times are kept to a minimum (24 to 48 h). We can routinely separate sequenced DNA fragments, using a 1-m gel system, which can be accurately read (with less than 0.5% error) to distances of 800 bp or more, from the oligomer primer. The strategy and procedures presented here allow the complete sequence of a cosmid clone to be obtained without subcloning.

Watermelon mosaic virus II and zucchini yellow mosaic virus: cloning of 3'-terminal regions, nucleotide sequences, and phylogenetic comparisons.

The 3'-terminal genomic regions of an isolate of watermelon mosaic virus II (WMVII) and a Florida isolate of zucchini yellow mosaic virus (ZYMV-F) have been cloned. The nucleotide sequence of the WMVII cDNA clone shows the presence of the large nuclear inclusion protein gene, the coat protein gene and 3' untranslated region. The nucleotide sequence of a ZYMV-F cDNA clone shows the presence of the coat protein gene and 3' untranslated region. Comparisons of the nucleotide and deduced amino acid sequences of these clones with those from other potyviruses show that WMVII and the soybean mosaic virus N strain are closely related, thus supporting their classification as different strains of the same virus. Our comparisons also indicate that ZYMV-F is a distinct potyvirus type and that its closest relative is WMVII. Phylogenetic analysis using the most-parsimonious branching arrangement derived from the alignment of coat protein gene sequences suggests the existence of two major potyvirus groupings.

The nucleotide sequences of the 3'-terminal regions of papaya ringspot virus strains W and P.

The sequences of cDNA clones encoding most of the NIb protein, the coat protein and the 3' untranslated region of papaya ringspot virus (PRV) strains W and P have been determined. The open reading frame of P strain PRV was confirmed by amino acid analysis. Nucleotide sequence comparisons of these strains show that they share a 98.2% identity in their NIb gene regions and a 97.7% identity in their coat protein genes. The sequences of these two strains are distinct from other potyvirus types, confirming their classification as two strains of the same virus. The NIb amino acid sequence possesses conserved amino acids characteristic of RNA-dependent RNA polymerases. Comparison of the coat protein amino acid sequence with those of other potyviruses shows perfectly conserved amino acids which may have functional significance.