A screen for fast evolving genes from Drosophila

In an attempt to quantify the rates of protein sequence divergence in Drosophila, we have devised a screen to differentiate between slow and fast evolving genes. We find that over one-third of randomly drawn cDNAs from a Drosophila melanogaster library do not cross-hybridize with Drosophila virilis DNA, indicating that they evolve with a very high rate. To determine the evolutionary characteristics of such protein sequences, we sequenced their homologs from a more closely related species (Drosophila yakuba). The amino acid substitution rates among these cDNAs are among the fastest known and several are only about 2-fold lower than the corresponding values for silent substitutions. An analysis of within-species polymorphisms for one of these sequences reveals an exceptionally high number of polymorphic amino acid positions, indicating that the protein is not under strong negative selection. We conclude that the Drosophila genome harbors a substantial proportion of genes with a very high divergence rate.

Chromosomal aberrations in PARP−/− mice: Genome stabilization in immortalized cells by reintroduction of poly(ADP-ribose) polymerase cDNA

Depletion of poly(ADP-ribose) polymerase (PARP) increases the frequency of recombination, gene amplification, sister chromatid exchanges, and micronuclei formation in cells exposed to genotoxic agents, implicating PARP in the maintenance of genomic stability. Flow cytometric analysis now has revealed an unstable tetraploid population in immortalized fibroblasts derived from PARP−/− mice. Comparative genomic hybridization detected partial chromosomal gains in 4C5-ter, 5F-ter, and 14A1-C1 in PARP−/−mice and immortalized PARP−/−fibroblasts. Neither the chromosomal gains nor the tetraploid population were apparent in PARP−/− cells stably transfected with PARP cDNA [PARP−/−(+PARP)], indicating negative selection of cells with these genetic aberrations after reintroduction of PARP cDNA. Although the tumor suppressor p53 was not detectable in PARP−/− cells, p53 expression was partially restored in PARP−/− (+PARP) cells. Loss of 14D3-ter that encompasses the tumor suppressor gene Rb-1 in PARP−/− mice was associated with a reduction in retinoblastoma(Rb) expression; increased expression of the oncogene Jun was correlated with a gain in 4C5-ter that harbors this oncogene. These results further implicate PARP in the maintenance of genomic stability and suggest that altered expression of p53, Rb, and Jun, as well as undoubtedly many other proteins may be a result of genomic instability associated with PARP deficiency.

Inhibition of intrathymic T cell development by expression of a transgenic antagonist peptide

The mature T cell receptor (TCR) repertoire is shaped by positive- and negative-selection events taking place during T cell development. These events are regulated by interactions between the TCR and major histocompatibility complex molecules presenting self-peptides. It has been shown that many antagonist peptides are efficient at mediating positive selection. In this study we analyzed the effects of a transgene encoding an antagonist peptide (influenza NP34) that is presented by H-2Db in a Tap-1-independent fashion in mice expressing the influenza NP68-specific TCR F5. We find that the transgenic peptide does not mediate positive or negative selection in F5+Tap-1−/− mice, but inhibits maturation of CD8+ single positive thymocytes in F5+Tap-1+ mice without inducing signs of negative selection. We conclude that antagonism of antigen recognition occurs not only at the level of mature T cells but also in T cell development.

The insect endosymbiont Sodalis glossinidius utilizes a type III secretion system for cell invasion

Sodalis glossinidius is a maternally transmitted secondary endosymbiont residing intracellularly in tissues of the tsetse flies, Glossina spp. In this study, we have used Tn5 mutagenesis and a negative selection procedure to derive a S. glossinidius mutant that is incapable of invading insect cells in vitro and is aposymbiotic when microinjected into tsetse. This mutant strain harbors Tn5 integrated into a chromosomal gene sharing high sequence identity with a type III secretion system invasion gene (invC) previously identified in Salmonella enterica. With the use of degenerate PCR, we have amplified a further six Sodalis inv/spa genes sharing high sequence identity with type III secretion system genes encoded by Salmonella pathogenicity island 1. Phylogenetic reconstructions based on the inv/spa genes of Sodalis and other members of the family Enterobacteriaceae have consistently identified a well-supported clade containing Sodalis and the enteric pathogens Shigella and Salmonella. These results suggest that Sodalis may have evolved from an ancestor with a parasitic intracellular lifestyle, possibly a latter-day entomopathogen. These observations lend credence to a hypothesis suggesting that vertically transmitted mutualistic endosymbionts evolve from horizontally transmitted parasites through a parasitism–mutualism continuum.

A new method for generating point mutations in bacterial artificial chromosomes by homologous recombination in Escherichia coli

Bacterial artificial chromosomes (BACs) and P1 artificial chromosomes (PACs), which contain large fragments of genomic DNA, have been successfully used as transgenes to create mouse models of dose-dependent diseases. They are also potentially valuable as transgenes for dominant diseases given that point mutations and/or small rearrangements can be accurately introduced. Here, we describe a new method to introduce small alterations in BACs, which results in the generation of point mutations with high frequency. The method involves homologous recombination between the original BAC and a shuttle vector providing the mutation. Each recombination step is monitored using positive and negative selection markers, which are the Kanamycin-resistance gene, the sacB gene and temperature-sensitive replication, all conferred by the shuttle plasmid. We have used this method to introduce four different point mutations and the insertion of the β-galactosidase gene in a BAC, which has subsequently been used for transgenic animal production.

Mutations Affecting Induction of Glycolytic and Fermentative Genes during Germination and Environmental Stresses in Arabidopsis1

Expression of the alcohol dehydrogenase gene (ADH) of Arabidopsis is known to be induced by environmental stresses and regulated developmentally. We used a negative-selection approach to isolate mutants that were defective in regulating the expression of the ADH gene during seed germination; we then characterized three recessive mutants, aar1–1, aar1–2, and aar2–1, which belong to two complementation groups. In addition to their defects during seed germination, mutations in the AAR1 and AAR2 genes also affected anoxic and hypoxic induction of ADH and other glycolytic genes in mature plants. The aar1 and aar2 mutants were also defective in responding to cold and osmotic stress. The two allelic mutants aar1–1and aar1–2 exhibited different phenotypes under cold and osmotic stresses. Based on our results we propose that these mutants are defective in a late step of the signaling pathways that lead to increased expression of the ADH gene and glycolytic genes.

Top-down free-energy minimization on protein potential energy landscapes

The hierarchical properties of potential energy landscapes have been used to gain insight into thermodynamic and kinetic properties of protein ensembles. It also may be possible to use them to direct computational searches for thermodynamically stable macroscopic states, i.e., computational protein folding. To this end, we have developed a top-down search procedure in which conformation space is recursively dissected according to the intrinsic hierarchical structure of a landscape's effective-energy barriers. This procedure generates an inverted tree similar to the disconnectivity graphs generated by local minima-clustering methods, but it fundamentally differs in the manner in which the portion of the tree that is to be computationally explored is selected. A key ingredient is a branch-selection algorithm that takes advantage of statistically predictive properties of the landscape to guide searches down the tree branches that are most likely to lead to the physically relevant macroscopic states. Using the computational folding of a β-hairpin-forming peptide as an example, we show that such predictive properties indeed exist and can be used for structure prediction by free-energy global minimization.

Long term trends in the evolution of H(3) HA1 human influenza type A

We have studied the HA1 domain of 254 human influenza A(H3N2) virus genes for clues that might help identify characteristics of hemagglutinins (HAs) of circulating strains that are predictive of that strain’s epidemic potential. Our preliminary findings include the following. (i) The most parsimonious tree found requires 1,260 substitutions of which 712 are silent and 548 are replacement substitutions. (ii) The HA1 portion of the HA gene is evolving at a rate of 5.7 nucleotide substitutions/year or 5.7 × 10−3 substitutions/site per year. (iii) The replacement substitutions are distributed randomly across the three positions of the codon when allowance is made for the number of ways each codon can change the encoded amino acid. (iv) The replacement substitutions are not distributed randomly over the branches of the tree, there being 2.2 times more changes per tip branch than for non-tip branches. This result is independent of how the virus was amplified (egg grown or kidney cell grown) prior to sequencing or if sequencing was carried out directly on the original clinical specimen by PCR. (v) These excess changes on the tip branches are probably the result of a bias in the choice of strains to sequence and the detection of deleterious mutations that had not yet been removed by negative selection. (vi) There are six hypervariable codons accumulating replacement substitutions at an average rate that is 7.2 times that of the other varied codons. (vii) The number of variable codons in the trunk branches (the winners of the competitive race against the immune system) is 47 ± 5, significantly fewer than in the twigs (90 ± 7), which in turn is significantly fewer variable codons than in tip branches (175 ± 8). (viii) A minimum of one of every 12 branches has nodes at opposite ends representing viruses that reside on different continents. This is, however, no more than would be expected if one were to randomly reassign the continent of origin of the isolates. (ix) Of 99 codons with at least four mutations, 31 have ratios of non-silent to silent changes with probabilities less than 0.05 of occurring by chance, and 14 of those have probabilities <0.005. These observations strongly support positive Darwinian selection. We suggest that the small number of variable positions along the successful trunk lineage, together with knowledge of the codons that have shown positive selection, may provide clues that permit an improved prediction of which strains will cause epidemics and therefore should be used for vaccine production.

Genetic characterization of a mammalian protein-protein interaction domain by using a yeast reverse two-hybrid system.

Many biological processes rely upon protein-protein interactions. Hence, detailed analysis of these interactions is critical for their understanding. Due to the complexities involved, genetic approaches are often needed. In yeast and phage, genetic characterizations of protein complexes are possible. However, in multicellular organisms, such characterizations are limited by the lack of powerful selection systems. Herein we describe genetic selections that allow single amino acid changes that disrupt protein-protein interactions to be selected from large libraries of randomly generated mutant alleles. The strategy, based on a yeast reverse two-hybrid system, involves a first-step negative selection for mutations that affect interaction, followed by a second-step positive selection for a subset of these mutations that maintain expression of full-length protein (two-step selection). We have selected such mutations in the transcription factor E2F1 that affect its ability to heterodimerize with DP1. The mutations obtained identified a putative helix in the marked box, a region conserved among E2F family members, as an important determinant for interaction. This two-step selection procedure can be used to characterize any interaction domain that can be tested in the two-hybrid system.

Pathogenic autoantibodies are routinely generated during the response to foreign antigen: a paradigm for autoimmune disease.

The immune system's ability to distinguish self and nonself is essential for both host defense against foreign agents and protection of self-antigens from autoimmune destruction. Such discrimination is complicated by extensive structural homology shared between foreign and self antigens. One hypothesis to explain the development of an autoimmune response is that some B cells activated by foreign antigen acquire, through somatic mutation, specificity for both the eliciting foreign antigen and self antigen. If such clones arise frequently, there must be a mechanism for their elimination. We have analyzed the extent of autoreactivity arising in a nonautoimmune host during the response to a foreign antigen. To overcome the process of apoptosis in primary B cells that might routinely eliminate autoreactive clones, we generated B-cell hybridomas from spleen cells of immunized mice by using a fusion partner constitutively expressing bcl-2. Multiple lines were obtained that recognize simultaneously the hapten phosphorylcholine and the self antigen double-stranded DNA. This dual specificity was not present early but was detected by day 10 after immunization. Some of these cross-reactive antibodies deposit in kidneys in a pattern similar to what is seen in autoimmune disease. These results demonstrate that autoantibodies arise at a high frequency as part of a response to foreign antigen. It has previously been shown that autoreactivity is regulated by central deletion; these data demonstrate a need for negative selection in peripheral lymphoid organs also, to regulate autoantibodies acquiring their self-specificity by somatic mutation.

Enhanced and accelerated lymphoproliferation in Fas-null mice.

Fas is a 45-kDa membrane protein that transduces an apoptotic signal. The mouse lymphoproliferation (lpr) mutation is a leaky mutation of Fas. In this study, we examined lymphocyte development in Fas-null mice generated by gene targeting. The Fas-/- mice progressively accumulated abnormal T cells (Thy1+, B220+, CD4-, and CD8-) and developed lymphadenopathy and splenomegaly, which were much more accelerated and pronounced than those in lpr mice. In addition, the Fas-null mice showed lymphocytosis, accompanied by lymphocytic infiltration in the lungs and liver. The number of apparently normal B cells also increased, and large amounts of immunoglobulins, including anti-DNA antibodies, were produced. Thymic clonal deletion, assessed by deletion of T cells reactive to mouse endogenous superantigens, was apparently normal in the Fas-/- mice, whereas the peripheral clonal deletion of mature T cells against a bacterial superantigen was impaired. These results suggested that Fas plays a decisive role in peripheral clonal deletion but not in negative selection in the thymus.

Deletion of high-avidity T cells by thymic epithelium.

Tolerance induction by thymic epithelium induces a state of so-called "split tolerance," characterized in vivo by tolerance and in vitro by reactivity to a given thymically expressed antigen. Using a model major histocompatibility complex class I antigen, H-2Kb (Kb), three mechanisms of thymic epithelium-induced tolerance were tested: induction of tolerance of tissue-specific antigens exclusively, selective inactivation of T helper cell-independent cytotoxic T lymphocytes, and deletion of high-avidity T cells. To this end, thymic anlagen from Kb-transgenic embryonic day 10 mouse embryos, taken before colonization by cells of hemopoietic origin, were grafted to nude mice. Tolerance by thymic epithelium was not tissue-specific, since Kb-bearing skin and spleen grafts were maintained indefinitely. Only strong priming in vivo could partially overcome the tolerant state and induce rejection of some skin grafts overexpressing transgenic Kb. Furthermore, the hypothesis that thymic epithelium selectively inactivates those T cells that reject skin grafts in a T helper-independent fashion could not be supported. Thus, when T-cell help was provided by a second skin graft bearing an additional major histocompatibility complex class II disparity, tolerance to the Kb skin graft was not broken. Finally, direct evidence could be obtained for the avidity model of thymic epithelium-induced negative selection, using Kb-specific T-cell receptor (TCR) transgenic mice. Thymic epithelium-grafted TCR transgenic mice showed a selective deletion of those CD8+ T cells with the highest density of the clonotypic TCR. These cells presumably represent the T cells with the highest avidity for Kb. We conclude that split tolerance induced by thymic epithelium was mediated by the deletion of those CD8+ T lymphocytes that have the highest avidity for antigen.

Quantitative assessment of sense and antisense transcripts from genes involved in antigenic variation (vsp genes) and encystation (cwp 1 gene) of Giardia lamblia clone GS/M-83-H7.

Antigenic variation of the intestinal protozoan parasite Giardia lamblia is caused by an exchange of the parasite's variant surface protein (VSP) coat. Many investigations on antigenic variation were performed with G. lamblia clone GS/M-83-H7 which produces surface antigen VSP H7. To generate novel information on giardial vsp gene transcription, vsp RNA levels were assessed by quantitative reverse transcription-(RT)-PCR in both axenic VSP H7-type trophozoites and subvariants obtained after negative selection of GS/M-83-H7 trophozoites by treatment with a cytotoxic, VSP H7-specific monoclonal antibody. Our investigation was not restricted to the assessment of the sense vsp transcript levels but also included an approach aimed at the detection of complementary antisense vsp transcripts within the two trophozoite populations. We found that sense vsp H7 RNA predominated in VSP H7-type trophozoites while sense RNA from only one (vsp IVg) of 8 subvariant vsp genes totally analysed predominated in subvariant-type trophozoites. Interestingly, the two trophozoite populations exhibited a similar relative distribution regarding the vsp H7 and vsp IVg antisense RNA molecules. An analogous sense versus antisense RNA pattern was also observed when the transcripts of gene cwp 1 (encoding cyst wall protein 1) were investigated. Here, both types of RNA molecules only appeared after cwp 1 had been induced through in vitro encystation of the parasite. These findings for the first time demonstrated that giardial antisense RNA production did not occur in a constitutive manner but was directly linked to complementary sense RNA production after activation of the respective gene systems.

The 2dF-SDSS LRG and QSO (2SLAQ) survey: the z < 2.1 quasar luminosity function from 5645 quasars to g=21.85

We have used the Two-Degree Field (2dF) instrument on the Anglo-Australian Telescope (AAT) to obtain redshifts of a sample of z < 3 and 18.0 < g < 21.85 quasars selected from Sloan Digital Sky Survey (SDSS) imaging. These data are part of a larger joint programme between the SDSS and 2dF communities to obtain spectra of faint quasars and luminous red galaxies, namely the 2dF-SDSS LRG and QSO (2SLAQ) Survey. We describe the quasar selection algorithm and present the resulting number counts and luminosity function of 5645 quasars in 105.7 deg(2). The bright-end number counts and luminosity functions agree well with determinations from the 2dF QSO Redshift Survey (2QZ) data to g similar to 20.2. However, at the faint end, the 2SLAQ number counts and luminosity functions are steeper (i.e. require more faint quasars) than the final 2QZ results from Croom et al., but are consistent with the preliminary 2QZ results from Boyle et al. Using the functional form adopted for the 2QZ analysis ( a double power law with pure luminosity evolution characterized by a second-order polynomial in redshift), we find a faint-end slope of beta =-1.78 +/- 0.03 if we allow all of the parameters to vary, and beta =-1.45 +/- 0.03 if we allow only the faint-end slope and normalization to vary (holding all other parameters equal to the final 2QZ values). Over the magnitude range covered by the 2SLAQ survey, our maximum-likelihood fit to the data yields 32 per cent more quasars than the final 2QZ parametrization, but is not inconsistent with other g > 21 deep surveys for quasars. The 2SLAQ data exhibit no well-defined 'break' in the number counts or luminosity function, but do clearly flatten with increasing magnitude. Finally, we find that the shape of the quasar luminosity function derived from 2SLAQ is in good agreement with that derived from Type I quasars found in hard X-ray surveys.

Binary sparse nonnegative matrix factorization

This paper presents a fast part-based subspace selection algorithm, termed the binary sparse nonnegative matrix factorization (B-SNMF). Both the training process and the testing process of B-SNMF are much faster than those of binary principal component analysis (B-PCA). Besides, B-SNMF is more robust to occlusions in images. Experimental results on face images demonstrate the effectiveness and the efficiency of the proposed B-SNMF.