Search

Background: There is tremendous potential for genome sequencing to improve clinical diagnosis and care once it becomes routinely accessible, but this will require formalizing research methods into clinical best practices in the areas of sequence data generation, analysis, interpretation and reporting. The CLARITY Challenge was designed to spur convergence in methods for diagnosing genetic disease starting from clinical case history and genome sequencing data. DNA samples were obtained from three families with heritable genetic disorders and genomic sequence data were donated by sequencing platform vendors. The challenge was to analyze and interpret these data with the goals of identifying disease-causing variants and reporting the findings in a clinically useful format. Participating contestant groups were solicited broadly, and an independent panel of judges evaluated their performance., Results: A total of 30 international groups were engaged. The entries reveal a general convergence of practices on most elements of the analysis and interpretation process. However, even given this commonality of approach, only two groups identified the consensus candidate variants in all disease cases, demonstrating a need for consistent fine-tuning of the generally accepted methods. There was greater diversity of the final clinical report content and in the patient consenting process, demonstrating that these areas require additional exploration and standardization., Conclusions: The CLARITY Challenge provides a comprehensive assessment of current practices for using genome sequencing to diagnose and report genetic diseases. There is remarkable convergence in bioinformatic techniques, but medical interpretation and reporting are areas that require further development by many groups.

Tissue regeneration and scarless healing involves the complete replacement and functional restoration of damaged organs and tissues. In this study of the "scarless healing" MRL mouse model, we demonstrate that 2-mm diameter through-and-through holes made in the cartilaginous part of previously injured MRL mouse ears are closed more efficiently, and that the regenerative repair response is significantly accelerated compared with unprimed MRL and control "nonhealer" strains of mice. Accelerated healing was detected both locally and distally from the original site of injury indicating the involvement of systemic components such as circulating cell types or soluble factors. Histologically, we observed early differences during the wound repair process (before Day 4 post injury) with accelerated formation of blastema-like structures, epidermal downgrowths, and enhanced epithelium thickening in wound border zones in primed MRL mice versus unprimed MRL mice. Although the mechanism of tissue regeneration remains unclear, the results from this study justify the use of the MRL model for further experimentation directed toward the identification of proteins and cell types capable of stimulating scarless tissue regeneration.

Loss-of-function mutations in the cystatin B (Cstb) gene cause a neurological disorder known as Unverricht-Lundborg disease (EPM1) in human patients. Mice that lack Cstb provide a mammalian model for EPM1 by displaying progressive ataxia and myoclonic seizures. We analyzed RNAs from brains of Cstb-deficient mice by using modified differential display, oligonucleotide microarray hybridization and quantitative reverse transcriptase polymerase chain reaction to examine the molecular consequences of the lack of Cstb. We identified seven genes that have consistently increased transcript levels in neurological tissues from the knockout mice. These genes are cathepsin S, C1q B-chain of complement (C1qB), beta2-microglobulin, glial fibrillary acidic protein (Gfap), apolipoprotein D, fibronectin 1 and metallothionein II, which are expected to be involved in increased proteolysis, apoptosis and glial activation. The molecular changes in Cstb-deficient mice are consistent with the pathology found in the mouse model and may provide clues towards the identification of therapeutic points of intervention for EPM1 patients.

The development of high-throughput methods for gene discovery has paved the way for the design of new strategies for genome-scale protein analysis. Lawrence Livermore National Laboratory and Onyx Pharmaceuticals, Inc., have produced an automatable system for the expression and purification of large numbers of proteins encoded by cDNA clones from the IMAGE (Integrated Molecular Analysis of Genomes and Their Expression) collection. This high-throughput protein expression system has been developed for the analysis of the human proteome, the protein equivalent of the human genome, comprising the translated products of all expressed genes. Functional and structural analysis of novel genes identified by EST (Expressed Sequence Tag) sequencing and the Human Genome Project will be greatly advanced by the application of this high-throughput expression system for protein production. A prototype was designed to demonstrate the feasibility of our approach. Using a PCR-based strategy, 72 unique IMAGE cDNA clones have been used to create an array of recombinant baculoviruses in a 96-well microtiter plate format. Forty-two percent of these cDNAs successfully produced soluble, recombinant protein. All of the steps in this process, from PCR to protein production, were performed in 96-well microtiter plates, and are thus amenable to automation. Each recombinant protein was engineered to incorporate an epitope tag at the amino terminal end to allow for immunoaffinity purification. Proteins expressed from this system are currently being analyzed for functional and biochemical properties., (Copyright 2000 Wiley-Liss, Inc.)

The ability to rapidly survey and compare gene expression levels between reference and test samples is moving the drug discovery process towards a more genomic orientation. The success of the Human Genome Project and related private genomics initiatives, combined with new technologies to probe, image and access expression data, are responsible for this transformation. This article reviews the history, status and future direction of high-throughput gene expression analysis. It describes classical approaches, explains the development of methods such as differential display for discovering novel genes, and discusses how microarray technology is exploiting collections of known sequences to pinpoint drug targets.

Neurocan is a chondroitin sulfate proteoglycan thought to be involved in the modulation of cell adhesion and migration. Its sequence has been determined previously in rat and mouse (Rauch et al., 1992. Cloning and primary structure of neurocan, a developmentally regulated, aggregating, chondroitin sulfate proteoglycan of the brain. J. Biol. Chem. 267, 19536-19547; Rauch et al., 1995. Structure and chromosomal location of the mouse neurocan gene. Genomics 28, 405-410). We describe here the complete coding sequence of the human neurocan mRNA, known as CSPG3, as well as mapping data, expression analysis, and genomic structure. A cDNA known as CP-1 was initially sequenced as part of a gene discovery project focused on characterizing chromosome 19-specific cDNAs. Sequence homology searches indicated close homology to the mouse and rat proteoglycan, neurocan (GenBank accession Nos X84727 and M97161). Northern analysis identified a brain-specific transcript of approx. 7.5kb. A longer cDNA clone, GT-5, was obtained, fine-mapped to the physical map of chromosome 19 by hybridization to a chromosome-specific cosmid library, and sequenced. Full coding sequence of the mRNA indicates a 3963bp open reading frame corresponding to a 1321 amino acid protein, similar to the protein length found in mouse and rat. The amino acid sequence of human neurocan shows 63% identity with both the mouse and rat sequences. Finally, genomic sequencing of a cosmid containing the complete neurocan gene was performed to determine the genomic structure of the gene, which spans approx. 41kb, and is transcribed in the telomere to centromere orientation.

The yeast and human RAD51 genes encode strand-transfer proteins that are thought to be involved in both recombinational repair of DNA damage and meiotic recombination. In yeast, the Rad51 family of related proteins also includes Rad55, Rad57 and Dmc1. In mammalian cells, five genes in this family have been identified (HsRAD51, XRCC2, XRCC3, RAD51B/hREC2 and HsDMC1), and here we report the isolation of the sixth member, RAD51C. RAD51C was originally identified by a computer screen of the EST database. A full-length approximately 1.3 kb cDNA clone has been isolated that encodes a protein of 376 aa, having a 18-26% aa identity with other human Rad51 family members. RAD51C includes a previously mapped sequenced-tagged site location near the end of chromosome 17q. The RAD51C transcript is expressed in various human tissues, with highest level of expression in testis, followed by heart muscle, spleen and prostate. Yeast two-hybrid experiments indicate that the Rad51C protein binds to two other members of the Rad51 protein family (Xrcc3 and Rad51B) but not to itself. These findings suggest that Rad51C may function similarly to the yeast Rad55 or Rad57 proteins, rather than as a Rad51 functional homolog.

The highly conserved Saccharomyces cerevisiae RAD51 protein functions in both mitotic and meiotic homologous recombination and in double-strand break repair. Screening of the public cDNA sequence database for RAD51-like genes led to the identification of a partial sequence from a breast tissue library present in the I.M.A.G.E. (Integrated Molecular Analysis of Genes and their Expression) collection. An extended 1764-bp cDNA clone encoding an open reading frame of 350 amino acids was isolated. This clone showed significant amino acid identity with other human RAD51 homologs. The new homolog, named RAD51B, was mapped to human chromosome 14q23-q24.2 using a panel of human-hamster somatic cell hybrids and fluorescence in situ hybridization. Northern blot analysis demonstrated that RAD51B mRNA is widely expressed and most abundant in tissues active in recombination. Functions associated with known RAD51 homologs suggest a role for RAD51B in meiotic recombination and/or recombinational repair.

We have used exon amplification to identify putative transcribed sequences from an 823-kb contig consisting of 28 cosmids that form a minimum tiling path from the interval 19p12-p13.1. This region contains the genes responsible for multiple epiphyseal dysplasia (MED) and pseudoachondroplasia (PSACH). We have trapped 66 exons (an average of 2.4 exons per cosmid) from pools of 2 or 3 cosmids. The majority of exons (51.5%) show only weak similarity or no similarity (36.3%) to sequences in current databases. Six of 8 exons examined from these groups, however, show cross-species sequence conservation, indicating that many of them probably represent authentic exons. Eight exons show identity or significant similarity to ESTs or known genes, including the human TNF receptor 3 '-flanking region gene, human epoxide hydrolase (EPHX), human growth/differentiation factor (GOF-1), human myocyte-specific enhancer factor 2, the rat neurocan gene, and the human cartilage oligomeric matrix protein gene (COMP). Mutations in this latter gene have recently been shown to be responsible for MED and PSACH.

Using cDNA hybridization selection techniques, we identified seven new genes in a 280 kilobase YAC covering the HLA-F locus. The new genes were mapped back to the YAC by a combination of optical restriction mapping and pulse field gel electrophoresis. Northern analysis of individual clones demonstrated the presence of either different mRNA sizes or different expression patterns. Two of the cDNA clones were expressed only in lymphoid cell lines: one in Jurkat cells (T cell) and another in JY cells (B cell). All the genes lacked sequence similarity to any known classical and non-classical major histocompatibility complex (MHC) class I genes, indicating that the MHC class I region has more functions than anticipated. Of the seven new genes, one is highly similar (97%) to mouse 60S ribosomal protein, and another is homologous to diubiquitin proteins. Of the two G-coupled receptor-like cDNAs, one was fully sequenced and found to be an olfactory receptor-like gene. The study strengthens evidence that the MHC complex not only plays a key role in the immune system, but also contributes to non-immunological functions.

We have developed a method for direct selection of cDNAs using whole chromosomes as target DNA. Double-strand cDNAs were synthesized from human fetal brain polyadenylated mRNAs. Flow-sorted chromosomes 17 and 19 were amplified by degenerate oligonucleotide primed polymerase chain reaction (DOP-PCR) and used to capture ds cDNAs by an improved magnetic bead capture protocol. To demonstrate the capabilities of this method, the selected cDNAs were used as probes in FISH experiments. The selected cDNA populations specifically painted chromosomes 17 or 19 on metaphase spreads. These results demonstrate that it is possible to do chromosome painting using cDNA probes and that this method is a means to rapidly select expressed sequences encoded by any portion of the genome.

In the course of isolation and sequence analysis of microsatellite repeat containing human cDNAs, we have isolated the human homologue of the rat visinin-like peptide gene. The human gene shows a high degree of conservation at both the amino acid and the DNA sequence level. The (CA)n microsatellite repeat embedded in the 3' untranslated region of the gene is conserved between rat and human, along with the flanking DNA sequences. We have mapped the VSNL1 gene to the short arm of chromosome 2.

Using a high-speed flow cytometer/sorter, we constructed spot-blot "stamps" measuring 3.5 x 2.0 cm containing 21 separate human chromosome fractions. Through hybridization to these stamps, 20 randomly selected cDNAs were assigned to specific chromosomes. Sequencing and BLAST database screening confirmed the location of one gene (UCHL1) and allowed the assignment of two other previously identified genes (LRP130 and cDNA IB871.)

Pseudoachondroplasia (PSACH) is a well characterized dwarfing condition mapping to chromosome 19p12-13.1. Cartilage oligomeric matrix protein (COMP), a cartilage specific protein, maps to the same location within a contig that spans the PSACH locus. Using single strand conformation polymorphism (SSCP) analysis and nucleotide sequencing we have identified COMP mutations in eight familial and isolated PSACH cases. All mutations involve either a single base-pair change or a three base-pair deletion in exon 17B. Six mutations delete or change a well conserved aspartic acid residue within the calcium-binding type 3 repeats. These results demonstrate that mutations in the COMP gene cause pseudochondroplasia.

The human genome contains multiple copies of sequences related to the HERV-K family of endogenous retroviruses, homologous to the B-type mouse mammary tumour virus. A DNA fragment closely resembling an HERV-K long tandem repeat (LTR) was detected in a library of hncDNA clones enriched for sequences from human chromosome 19. Sites showing homology to the sequence of this fragment have been identified on human chromosome 19 by hybridization to previously mapped chromosome 19 cosmids. Thus the distribution of LTR sequences on a specific human chromosome has been mapped for the first time. We estimate the total number of such sites on human chromosome 19 to be at least 110. Many of these sites are located in the vicinity of known genes. The precise localizations (to specific cosmids) of LTR-homologous sequences on chromosome 19 can serve as a reference source and will automatically provide further insight into LTR-gene relationships as new genes are mapped onto the chromosome.

We have identified and studied the chromosomal location of the human Rieske Fe-S protein-encoding gene UQCRFS1. Mapping by hybridization to a panel of monochromosomal hybrid cell lines indicated that a UQCRFS1 partial cDNA was derived from either chromosome 19 or 22. By screening a human chromosome 19 specific genomic cosmid library with a probe from this cDNA sequence, we identified a corresponding cosmid. Portions of this cosmid were sequenced directly. The exon, exon:intron junction and flanking sequences verified that this cosmid contains the genomic locus. Fluorescent in situ hybridization (FISH) was performed to localize this cosmid to chromosome band 19q12.

We have used the human H blood group alpha(1,2)fucosyltransferase (FUT1) cDNA to screen chromosome 19 cosmid libraries in a search for the human Secretor (Se) blood group gene (FUT2). One cosmid has been isolated that contains two distinct segments that cross-hybridize with FUT1. We have assembled a 100-kilobase (kb) cosmid contig, localized to 19q13.3, encompassing FUT1 and the two FUT1-related sequences, termed Sec1 and Sec2, for Secretor candidate 1 and 2. Sec1 and Sec2 are separated by 12 kb and are 65.5 kb and 35 kb apart, respectively, from the FUT1 gene. We used a cosmid-dependent direct cDNA selection method to clone a cDNA corresponding to a transcript that emanates from Sec2. This cDNA detects a 3.35-kb transcript in human tissues known to express the Se locus. Together with sequence and expression data reported in the accompanying article (Kelly, R. J., Rouquier, S., Giorgi, D., Lennon, G. G., and Lowe, J. B. (1995) J. Biol. Chem. 270, 4640-4649), these data demonstrate that Sec2 corresponds to the human Se blood group locus (FUT2). Our results furthermore define the physical relationship between the H and Se loci and confirm a hypothesis that these two loci represent distinct but closely linked alpha(1,2)fucosyltransferase genes.

Synthesis of soluble A, B, H, and Lewis b blood group antigens in humans is determined by the Secretor (Se) (FUT2) blood group locus. Genetic, biochemical, and molecular analyses indicate that this locus corresponds to an alpha(1,2)fucosyltransferase gene distinct from the genetically-linked H blood group alpha(1,2)fucosyltransferase locus. The accompanying paper (Rouquier, S., Lowe, J. B., Kelly, R. J., Fertitta, A. L., Lennon, G. G., and Giorgi, D. (1995) J. Biol. Chem. 270, 4632-4639) describes the molecular cloning and mapping of two human DNA segments that are physically linked to, and cross-hybridize with, the H locus. We present here an analysis of these two new DNA segments. One of these, termed Sec1, is a pseudogene, because translational frameshifts and termination codons interrupt potential open reading frames that would otherwise share primary sequence similarity with the H alpha(1,2)fucosyltransferase. The other DNA segment, termed Sec2, predicts a 332-amino acid-long polypeptide, and a longer isoform, that share 68% sequence identity with the COOH-terminal 292 residues of the human H blood group alpha(1,2)fucosyltransferase. Sec2 encodes an alpha(1,2)fucosyltransferase with catalytic properties that mirror those ascribed to the Secretor locus-encoded alpha(1,2)fucosyltransferase. Approximately 20% of randomly-selected individuals were found to be apparently homozygous for an enzyme-inactivating nonsense allele (Trp143-->ter) at this locus, in correspondence to the frequency of the non-secretor phenotype in most human populations. Furthermore, each of six unrelated non-secretor individuals are also apparently homozygous for this null allele. These results indicate that Sec2 corresponds to the human Secretor blood group locus (FUT2) and indicate that homozygosity for a common nonsense allele is responsible for the nonsecretor phenotype in many non-secretor individuals.

We have cloned a series of overlapping cDNA clones encoding a 5194 bp transcript for human DNA methyltransferase (DNA MTase). This sequence potentially codes for a protein of 1495 amino acids with a predicted molecular weight of 169 kDa. The human DNA MTase cDNA has eighty percent homology at the nucleotide level, and the predicted protein has seventy-four percent identity at the amino acid level, to the DNA MTase cDNA cloned from mouse cells. Like the murine DNA MTase, the amino terminal two-thirds of the human protein contains a cysteine-rich region suggestive of a metal-binding domain. The carboxy terminal one-third of the protein shows considerable similarity to prokaryotic (cytosine-5)-methyltransferases. The arrangement of multiple motifs conserved in the prokaryotic genes is preserved in the human DNA MTase, including the relative position of a proline-cysteine dipeptide thought to be an essential catalytic site in all (cytosine-5)-methyltransferases. A single 5.2 kb transcript was detected in all human tissues tested, with the highest levels of expression observed in RNA from placenta, brain, heart and lung. DNA MTase cDNA clones were used to screen a chromosome 19 genomic cosmid library. The DNA MTase-positive cosmids which are estimated to span a genomic distance of 93 kb have been localized to 19p13.2-p13.3 by fluorescence in situ hybridization. Isolation of the cDNA for human DNA MTase will allow further study of the regulation of DNA MTase expression, and of the role of this enzyme in establishing DNA methylation patterns in both normal and neoplastic cells.

As an aid to the fission yeast genome project, we describe a database for Schizosaccharomyces pombe consisting of both genetic and physical information. As presented, it is therefore both an updated gene list of all the nuclear genes of the fission yeast, and provides an estimate of the physical distance between two mapped genes. Additionally, a field indicates whether the sequence of the gene is available. Currently, sequence information is available for 135 of the 501 known genes.

As part of an integrated mapping and sequencing analysis of genomes, we have developed an approach allowing the characterization of large numbers of cDNA library clones with a minimal number of experiments. Three basic elements used in the analysis of cDNA libraries are responsible for the high efficiency of this new approach: (1) high-density library arrays allowing thousands of clones to be screened simultaneously; (2) hybridization fingerprinting techniques to identify clones abundantly expressed in specific tissues (by hybridizations with labeled tissue cDNA pools) and to avoid the repeated selection of identical clones and of clones containing noncoding inserts; and (3) a computerized system for the evaluation of hybridization data. To demonstrate the feasibility of this approach, we hybridized high-density cDNA library arrays of human fetal brain and embryonal Drosophila with radiolabeled cDNA pools derived from whole mouse tissues. Fingerprints of the library arrays were generated, localizing clones containing cDNA sequences from mRNAs expressed at middle to high abundance (> 0.1-0.15%) in the respective tissue. Partial sequencing data from a number of clones abundantly expressed in several tissues were generated to demonstrate the value of the approach, especially for the selection of cDNA clones for the analyses of genomes based on expressed sequence tagged sites. Data obtained by the technique described will ultimately be correlated with additional transcriptional and sequence information for the same library clones and with genomic mapping information in a relational database.

A limitation to the use of oligonucleotide probes as tools for genetic and physical mapping has been the low hybridization positive frequency obtained by oligonucleotides of sufficient length to hybridize preferentially to cloned insert DNA (and not host E. coli genomic DNA). Both computer and experimental results now indicate that oligonucleotide probes composed of W (A or T) sequence are preferentially found in eukaryotic DNA, and can be used to provide high frequency, discriminative hybridization. Such W sequences may be useful as either probes or PCR primers in molecular diagnostic applications as well as in genetic and physical mapping.

Three differently made, primary Drosophila cosmid libraries of 16-fold genome coverage have been generated. Also, a jumping library has been created by a new method that takes advantage of methylation differences between genomic DNA and vector. Thirdly, two cDNA libraries have been picked. All these libraries have been arrayed on high-density in situ filters, each containing 9216 clones. As a reference system, such filters are distributed and identified clones are provided. Single-copy probes have identified on average 1.4 cosmids per genome equivalent. Together with cytogenetically mapped yeast artificial chromosomes, the libraries are also being used for physically mapping the genome, mainly by oligonucleotide fingerprinting and pool hybridizations. cDNA clones are further examined by a partial sequencing analysis by oligomer hybridization.

A direct hybridization protocol is described for screening cosmid and yeast artificial chromosome libraries with pools of Alu-PCR products from somatic cell or irradiation hybrids. This method eliminates purification, cloning and analysis of each individual Alu-PCR product before library screening. A series of human X chromosome irradiation hybrids were mapped by this method, using a cosmid reference library for comparisons between overlapping hybrids to identify interesting clones for further analysis.

The developmental time course of RNA transcribed from unrearranged (germ-line) and rearranged Ig genes in murine fetal liver was determined by a quantitative Northern blot analysis. Sterile Cmu transcripts and germ-line VH transcripts are detectable as early as day 14, whereas significant amounts of rearranged VDJCmu H chain transcripts do not appear until days 16 to 17. The sterile Cmu transcripts continuously increase in abundance throughout fetal development, in contrast to the germ-line VH transcripts, which decrease abruptly after day 16. Transcripts of germ-line and rearranged CK genes are detectable on day 17, and continue to increase in abundance on day 18. Transcripts from a pre-B cell specific gene, lambda 5, first appear on day 15, and reach maximum abundance on day 17. The order of these events is consistent with the known order of gene rearrangements and with the idea that transcriptional activation of germ-line loci is a prerequisite for Ig gene rearrangement. The lag between the onsets of germ-line Cmu and VH transcription and the appearance of VDJCmu transcripts suggests that additional developmentally regulated events may be necessary to achieve efficient expression of completely rearranged H chain genes.

Nucleotide sequences were converted into purine (R)-pyrimidine (Y) series and divided into several groups, embracing higher and lower organisms. The frequencies of R-Y doublets, triplets and quartets in each were calculated. Whereas eukaryotes uniformly show RR + YY greater than RY + YR, in bacteria and phage no such relationship is observed. The triplet and quartet patterns in higher organisms differ from those seen in prokaryotes. In the higher organisms a correlation is observed between the frequencies of triplets and quartets and some DNA structural parameters. Specifically, the most frequent triplets are those with minimal torsion angle deviations from a regular B-DNA. The most frequent quartets are those with minimal roll angle deviations. No such correlations are observed in prokaryotes. We therefore propose that in eukaryotic DNA, tight, smooth packaging imposes sequence constraints.

The scid mouse mutant is severely deficient in lymphocytes; cells of the B or T lymphocyte lineage cannot be detected by either serological or functional assays. However, as shown here, germ-line transcripts of B cell immunoglobulin (Ig) constant and variable region genes and of T cell receptor (TCR) genes are detectable in lymphopoietic tissues of scid mice, as well as B and T lineage-specific lambda 5 and T3 delta transcripts. We conclude that B and T lineage-committed cells do arise in scid mice and that their Ig and TCR genes are accessible to enzymes involved in their recombination. This suggests that scid impairs lymphopoiesis at the stage at which antigen receptor genes normally undergo rearrangement.

Transcriptional competence of the immunoglobulin heavy-chain locus (IgH) is established at an early stage of lymphoid cell development, leading to the appearance of RNA components, previously called C mu RNA1 or sterile-mu RNA2, which contain constant-region sequences but lack variable-region sequences. These components are of two types: those which initiate in the D region of alleles that have undergone DJH (diversity-joining region) rearrangement (D mu transcripts) and those which initiate within the JH-C mu intron (hereafter termed I mu transcripts). In pre-B and early B cells, D mu and I mu transcripts are nearly as abundant as the messenger RNA encoding mu heavy chain. The D mu transcripts are spliced into RNAs containing D, JH and C mu sequences, and in some, but not all, cases these RNAs are translated into D mu proteins. To establish whether the I mu transcripts have any translational potential and to elucidate the structure of their promoter region, we have determined their transcription initiation sites and their mode of splicing. As reported here, by using sequence analysis of cloned I mu complementary DNAs, primer extension and S1 nuclease mapping, we have found that these transcripts have remarkable 5' heterogeneity: there are more than five distinct start sites spanning a region of 44 nucleotides that is located downstream of an octanucleotide found in all variable-region promoters. Such imprecise initiation may result from the lack of a well-defined TATAA motif and the unusual proximity of the octanucleotide to the enhancer region. Approximately 700 nucleotides downstream from these initiation sites, a cryptic splice site is used to create a nontranslatable exon ('nontron') which is joined to the C mu 1 domain. The properties of the nontron may be important for the mechanism of allelic exclusion.

Three well-defined E. coli operator regions were examined for recurring conformational deviation from a regular B-DNA helix. All three, the lac, trp, and gal, show repeats of the same set of neighboring helical twist angles. These angles recur with a periodicity equal to the helix periodicity on one side of the operator's axes of symmetry. The probability that their occurrence is random was found to be extremely small. Therefore, we propose that in addition to specific bases, repeating twist angles patterns are likely to be among the local parameters involved in repressor-operator recognition.

Pelham has shown that the Drosophila hsp 70 gene is not transcribed under heat shock conditions unless a given upstream region is present. Davidson et al. have recently compiled a list of sequences homologous to this region in other Drosophila heat shock genes. They proposed that a set of unlinked genes, such as the heat shock genes, could be coordinately induced through an interaction in cis with a common regulatory molecule. That this interaction involves structural elements is suggested by the fact that these upstream regions share inverted repeats as well as areas of Z-DNA potential. Furthermore, using the Calladine-Dickerson rules for local helical parameters, we show that these regions share structural homology. This is significant because the presence of regions homologous to a derived consensus sequence does not necessarily imply structural similarity. Therefore, we suggest that these structural features are at least as important as the sequence homologies in enabling the heat shock response.

The relationship between the deficiency of CpG dinucleotides and the coding-noncoding segments of DNA has been examined. Analysis of five human alpha-like globin DNA sequences and five human beta-like globin DNA sequences reveal that there is no apparent difference between protein coding and non-coding portions of DNA. Rather CpG deficiency appears to be a property of long contiguous segments of DNA consisting of several genes and their intergenic regions. Thus we propose that CpG deficiency is not involved with translation or transcription but rather is related to chromosomal constraints.

Calladine (1982) proposed that steric repulsion between adjacent purines on opposite helix backbones accounts for the local variation seen in four helical parameters. By defining simple sum functions based on Calladine 's proposal, Dickerson (1983) has taken what he terms "the first step in establishing a sequence/structure vocabulary". In this letter we analyze the implications of the Calladine - Dickerson model with regard to " homonyms ", "synonyms" and mutations. Specifically, we (1) show that because of the number of adjacent helical positions affected by one transversion, two purine-pyrimidine sequences may be similar at the sequence level yet be very different structurally ( homonyms ); (2) list all sequences which, though different at the sequence level, share adjacent structural parameter values (synonyms); and (3) use two simple statistical measures to show that transversion mutations occurring between a purine and a pyrimidine (5'----3' on the same strand) are in general less disruptive of local helical structure than transversions occurring between a pyrimidine and purine. On the assumption that they are not inconsistent with experimental findings, we discuss the significance of these implications.