Relocation of genes generates non-conserved chromosomal segments in Fusarium graminearum that show distinct and co-regulated gene expression patterns

Background

Genome comparisons between closely related species often show non-conserved regions across chromosomes. Some of them are located in specific regions of chromosomes and some are even confined to one or more entire chromosomes. The origin and biological relevance of these non-conserved regions are still largely unknown. Here we used the genome of Fusarium graminearum to elucidate the significance of non-conserved regions.

Results

The genome of F. graminearum harbours thirteen non-conserved regions dispersed over all of the four chromosomes. Using RNA-Seq data from the mycelium of F. graminearum, we found weakly expressed regions on all of the four chromosomes that exactly matched with non-conserved regions. Comparison of gene expression between two different developmental stages (conidia and mycelium) showed that the expression of genes in conserved regions is stable, while gene expression in non-conserved regions is much more influenced by developmental stage. In addition, genes involved in the production of secondary metabolites and secreted proteins are enriched in non-conserved regions, suggesting that these regions could also be important for adaptations to new environments, including adaptation to new hosts. Finally, we found evidence that non-conserved regions are generated by sequestration of genes from multiple locations. Gene relocations may lead to clustering of genes with similar expression patterns or similar biological functions, which was clearly exemplified by the PKS2 gene cluster.

Conclusions

Our results showed that chromosomes can be functionally divided into conserved and non-conserved regions, and both could have specific and distinct roles in genome evolution and regulation of gene expression.

Electronic supplementary material

The online version of this article (doi:10.1186/1471-2164-15-191) contains supplementary material, which is available to authorized users.     

Identification of three extra-chromosomal replicons in Leptospira pathogenic strain and development of new shuttle vectors

Background

The genome of pathogenic Leptospira interrogans contains two chromosomes. Plasmids and prophages are known to play specific roles in gene transfer in bacteria and can potentially serve as efficient genetic tools in these organisms. Although plasmids and prophage remnants have recently been reported in Leptospira species, their characteristics and potential applications in leptospiral genetic transformation systems have not been fully evaluated.

Results

Three extrachromosomal replicons designated lcp1 (65,732 bp), lcp2 (56,757 bp), and lcp3 (54,986 bp) in the L. interrogans serovar Linhai strain 56609 were identified through whole genome sequencing. All three replicons were stable outside of the bacterial chromosomes. Phage particles were observed in the culture supernatant of 56609 after mitomycin C induction, and lcp3, which contained phage-related genes, was considered to be an inducible prophage. L. interrogans–Escherichia coli shuttle vectors, constructed with the predicted replication elements of single rep or rep combined with parAB loci from the three plasmids were shown to successfully transform into both saprophytic and pathogenic Leptospira species, suggesting an essential function for rep genes in supporting auto-replication of the plasmids. Additionally, a wide distribution of homologs of the three rep genes was identified in L. interrogans isolates, and correlation tests showed that the transformability of the shuttle vectors in L. interrogans isolates depended, to certain extent, on genetic compatibility between the rep sequences of both plasmid and host.

Conclusions

Three extrachromosomal replicons co-exist in L. interrogans, one of which we consider to be an inducible prophage. The vectors constructed with the rep genes of the three replicons successfully transformed into saprophytic and pathogenic Leptospira species alike, but this was partly dependent on genetic compatibility between the rep sequences of both plasmid and host.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-1321-y) contains supplementary material, which is available to authorized users.     

Genome-wide and molecular evolution analysis of the subtilase gene family in Vitis vinifera

Background

Vitis vinifera (grape) is one of the most economically significant fruit crops in the world. The availability of the recently released grape genome sequence offers an opportunity to identify and analyze some important gene families in this species. Subtilases are a group of subtilisin-like serine proteases that are involved in many biological processes in plants. However, no comprehensive study incorporating phylogeny, chromosomal location and gene duplication, gene organization, functional divergence, selective pressure and expression profiling has been reported so far for the grape.

Results

In the present study, a comprehensive analysis of the subtilase gene family in V. vinifera was performed. Eighty subtilase genes were identified. Phylogenetic analyses indicated that these subtilase genes comprised eight groups. The gene organization is considerably conserved among the groups. Distribution of the subtilase genes is non-random across the chromosomes. A high proportion of these genes are preferentially clustered, indicating that tandem duplications may have contributed significantly to the expansion of the subtilase gene family. Analyses of divergence and adaptive evolution show that while purifying selection may have been the main force driving the evolution of grape subtilases, some of the critical sites responsible for the divergence may have been under positive selection. Further analyses of real-time PCR data suggested that many subtilase genes might be important in the stress response and functional development of plants.

Conclusions

Tandem duplications as well as purifying and positive selections have contributed to the functional divergence of subtilase genes in V. vinifera. The data may contribute to a better understanding of the grape subtilase gene family.

Electronic supplementary material

The online version of this article (doi:10.1186/1471-2164-15-1116) contains supplementary material, which is available to authorized users.     

Whole-genome sequencing of Mesorhizobium huakuii 7653R provides molecular insights into host specificity and symbiosis island dynamics

Background

Evidence based on genomic sequences is urgently needed to confirm the phylogenetic relationship between Mesorhizobium strain MAFF303099 and M. huakuii. To define underlying causes for the rather striking difference in host specificity between M. huakuii strain 7653R and MAFF303099, several probable determinants also require comparison at the genomic level. An improved understanding of mobile genetic elements that can be integrated into the main chromosomes of Mesorhizobium to form genomic islands would enrich our knowledge of how genome dynamics may contribute to Mesorhizobium evolution in general.

Results

In this study, we sequenced the complete genome of 7653R and compared it with five other Mesorhizobium genomes. Genomes of 7653R and MAFF303099 were found to share a large set of orthologs and, most importantly, a conserved chromosomal backbone and even larger perfectly conserved synteny blocks. We also identified candidate molecular differences responsible for the different host specificities of these two strains. Finally, we reconstructed an ancestral Mesorhizobium genomic island that has evolved into diverse forms in different Mesorhizobium species.

Conclusions

Our ortholog and synteny analyses firmly establish MAFF303099 as a strain of M. huakuii. Differences in nodulation factors and secretion systems T3SS, T4SS, and T6SS may be responsible for the unique host specificities of 7653R and MAFF303099 strains. The plasmids of 7653R may have arisen by excision of the original genomic island from the 7653R chromosome.

Electronic supplementary material

The online version of this article (doi: 10.1186/1471-2164-15-440) contains supplementary material, which is available to authorized users.     

Clostridium sordellii genome analysis reveals plasmid localized toxin genes encoded within pathogenicity loci

Background

Clostridium sordellii can cause severe infections in animals and humans, the latter associated with trauma, toxic shock and often-fatal gynaecological infections. Strains can produce two large clostridial cytotoxins (LCCs), TcsL and TcsH, related to those produced by Clostridium difficile, Clostridium novyi and Clostridium perfringens, but the genetic basis of toxin production remains uncharacterised.

Results

Phylogenetic analysis of the genome sequences of 44 strains isolated from human and animal infections in the UK, US and Australia placed the species into four clades. Although all strains originated from animal or clinical disease, only 5 strains contained LCC genes: 4 strains contain tcsL alone and one strain contains tcsL and tcsH. Four toxin-positive strains were found within one clade. Where present, tcsL and tcsH were localised in a pathogenicity locus, similar to but distinct from that present in C. difficile. In contrast to C. difficile, where the LCCs are chromosomally localised, the C. sordellii tcsL and tcsH genes are localised on plasmids. Our data suggest gain and loss of entire toxigenic plasmids in addition to horizontal transfer of the pathogenicity locus. A high quality, annotated sequence of ATCC9714 reveals many putative virulence factors including neuraminidase, phospholipase C and the cholesterol-dependent cytolysin sordellilysin that are highly conserved between all strains studied.

Conclusions

Genome analysis of C. sordellii reveals that the LCCs, the major virulence factors, are localised on plasmids. Many strains do not contain the LCC genes; it is probable that in several of these cases the plasmid has been lost upon laboratory subculture. Our data are consistent with LCCs being the primary virulence factors in the majority of infections, but LCC-negative strains may precipitate certain categories of infection. A high quality genome sequence reveals putative virulence factors whose role in virulence can be investigated.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-1613-2) contains supplementary material, which is available to authorized users.     

Characterization of chromosomal and megaplasmid partitioning loci in Thermus thermophilus HB27

Background

In low-copy-number plasmids, the partitioning loci (par) act to ensure proper plasmid segregation and copy number maintenance in the daughter cells. In many bacterial species, par gene homologues are encoded on the chromosome, but their function is much less understood. In the two-replicon, polyploid genome of the hyperthermophilic bacterium Thermus thermophilus, both the chromosome and the megaplasmid encode par gene homologues (parABc and parABm, respectively). The mode of partitioning of the two replicons and the role of the two Par systems in the replication, segregation and maintenance of the genome copies are completely unknown in this organism.

Results

We generated a series of chromosomal and megaplasmid par mutants and sGFP reporter strains and analyzed them with respect to DNA segregation defects, genome copy number and replication origin localization. We show that the two ParB proteins specifically bind their cognate centromere-like sequences parS, and that both ParB-parS complexes localize at the cell poles. Deletion of the chromosomal parAB genes did not apparently affect the cell growth, the frequency of cells with aberrant nucleoids, or the chromosome and megaplasmid replication. In contrast, deletion of the megaplasmid parAB operon or of the parB gene was not possible, indicating essentiality of the megaplasmid-encoded Par system. A mutant expressing lower amounts of ParABm showed growth defects, a high frequency of cells with irregular nucleoids and a loss of a large portion of the megaplasmid. The truncated megaplasmid could not be partitioned appropriately, as interlinked megaplasmid molecules (catenenes) could be detected, and the ParBm-parSm complexes in this mutant lost their polar localization.

Conclusions

We show that in T. thermophilus the chromosomal par locus is not required for either the chromosomal or megaplasmid bulk DNA replication and segregation. In contrast, the megaplasmid Par system of T. thermophilus is needed for the proper replication and segregation of the megaplasmid, and is essential for its maintenance. The two Par sets in T. thermophilus appear to function in a replicon-specific manner. To our knowledge, this is the first analysis of Par systems in a polyploid bacterium.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-1523-3) contains supplementary material, which is available to authorized users.     

Unusual genome complexity in Lactobacillus salivarius JCM1046

Background

Lactobacillus salivarius strains are increasingly being exploited for their probiotic properties in humans and animals. Dissemination of antibiotic resistance genes among species with food or probiotic-association is undesirable and is often mediated by plasmids or integrative and conjugative elements. L. salivarius strains typically have multireplicon genomes including circular megaplasmids that encode strain-specific traits for intestinal survival and probiotic activity. Linear plasmids are less common in lactobacilli and show a very limited distribution in L. salivarius. Here we present experimental evidence that supports an unusually complex multireplicon genome structure in the porcine isolate L. salivarius JCM1046.

Results

JCM1046 harbours a 1.83 Mb chromosome, and four plasmids which constitute 20% of the genome. In addition to the known 219 kb repA-type megaplasmid pMP1046A, we identified and experimentally validated the topology of three additional replicons, the circular pMP1046B (129 kb), a linear plasmid pLMP1046 (101 kb) and pCTN1046 (33 kb) harbouring a conjugative transposon. pMP1046B harbours both plasmid-associated replication genes and paralogues of chromosomally encoded housekeeping and information-processing related genes, thus qualifying it as a putative chromid. pLMP1046 shares limited sequence homology or gene synteny with other L. salivarius plasmids, and its putative replication-associated protein is homologous to the RepA/E proteins found in the large circular megaplasmids of L. salivarius. Plasmid pCTN1046 harbours a single copy of an integrated conjugative transposon (Tn6224) which appears to be functionally intact and includes the tetracycline resistance gene tetM.

Conclusion

Experimental validation of sequence assemblies and plasmid topology resolved the complex genome architecture of L. salivarius JCM1046. A high-coverage draft genome sequence would not have elucidated the genome complexity in this strain. Given the expanding use of L. salivarius as a probiotic, it is important to determine the genotypic and phenotypic organization of L. salivarius strains. The identification of Tn6224-like elements in this species has implications for strain selection for probiotic applications.

Electronic supplementary material

The online version of this article (doi:10.1186/1471-2164-15-771) contains supplementary material, which is available to authorized users.     

Evolution of multipartite mitochondrial genomes in the booklice of the genus Liposcelis (Psocoptera)

Background

The genus Liposcelis (Psocoptera: Troctomorpha) has more than 120 species with a worldwide distribution and they pose a risk for global food security. The organization of mitochondrial (mt) genomes varies between the two species of booklice investigated in the genus Liposcelis. Liposcelis decolor has its mt genes on a single chromosome, like most other insects; L. bostrychophila, however, has a multipartite mt genome with genes on two chromosomes.

Results

To understand how multipartite mt genome organization evolved in the genus Liposcelis, we sequenced the mt genomes of L. entomophila and L. paeta in this study. We found that these two species of booklice also have multipartite mt genomes, like L. bostrychophila, with the mt genes we identified on two chromosomes. Numerous pseudo mt genes and non-coding regions were found in the mt genomes of these two booklice, and account for 30% and 10% respectively of the entire length we sequenced. In L. bostrychophila, the mt genes are distributed approximately equally between the two chromosomes. In L. entomophila and L. paeta, however, one mt chromosome has most of the genes we identified whereas the other chromosome has largely pseudogenes and non-coding regions. L. entomophila and L. paeta differ substantially from each other and from L. bostrychophila in gene content and gene arrangement in their mt chromosomes.

Conclusions

Our results indicate unusually fast evolution in mt genome organization in the booklice of the genus Liposcelis, and reveal different patterns of mt genome fragmentation among L. bostrychophila, L. entomophila and L. paeta.

Electronic supplementary material

The online version of this article (doi:10.1186/1471-2164-15-861) contains supplementary material, which is available to authorized users.     

Genome sequencing of two Neorhizobium galegae strains reveals a noeT gene responsible for the unusual acetylation of the nodulation factors

Background

The species Neorhizobium galegae comprises two symbiovars that induce nodules on Galega plants. Strains of both symbiovars, orientalis and officinalis, induce nodules on the same plant species, but fix nitrogen only in their own host species. The mechanism behind this strict host specificity is not yet known. In this study, genome sequences of representatives of the two symbiovars were produced, providing new material for studying properties of N. galegae, with a special interest in genomic differences that may play a role in host specificity.

Results

The genome sequences confirmed that the two representative strains are much alike at a whole-genome level. Analysis of orthologous genes showed that N. galegae has a higher number of orthologs shared with Rhizobium than with Agrobacterium. The symbiosis plasmid of strain HAMBI 1141 was shown to transfer by conjugation under optimal conditions. In addition, both sequenced strains have an acetyltransferase gene which was shown to modify the Nod factor on the residue adjacent to the non-reducing-terminal residue. The working hypothesis that this gene is of major importance in directing host specificity of N. galegae could not, however, be confirmed.

Conclusions

Strains of N. galegae have many genes differentiating them from strains of Agrobacterium, Rhizobium and Sinorhizobium. However, the mechanism behind their ecological difference is not evident. Although the final determinant for the strict host specificity of N. galegae remains to be identified, the gene responsible for the species-specific acetylation of the Nod factors was identified in this study. We propose the name noeT for this gene to reflect its role in symbiosis.

Electronic supplementary material

The online version of this article (doi:10.1186/1471-2164-15-500) contains supplementary material, which is available to authorized users.     

Genome-wide exploration of the molecular evolution and regulatory network of mitogen-activated protein kinase cascades upon multiple stresses in Brachypodium distachyon

Background

Brachypodium distachyon is emerging as a widely recognized model plant that has very close relations with several economically important Poaceae species. MAPK cascade is known to be an evolutionarily conserved signaling module involved in multiple stresses. Although the gene sequences of MAPK and MAPKK family have been fully identified in B. distachyon, the information related to the upstream MAPKKK gene family especially the regulatory network among MAPKs, MAPKKs and MAPKKKs upon multiple stresses remains to be understood.

Results

In this study, we have identified MAPKKKs which belong to the biggest gene family of MAPK cascade kinases. We have systematically investigated the evolution of whole MAPK cascade kinase gene family in terms of gene structures, protein structural organization, chromosomal localization, orthologs construction and gene duplication analysis. Our results showed that most BdMAPK cascade kinases were located at the low-CpG-density region, and the clustered members in each group shared similar structures of the genes and proteins. Synteny analysis showed that 62 or 21 pairs of duplicated orthologs were present between B. distachyon and Oryza sativa, or between B. distachyon and Arabidopsis thaliana respectively. Gene expression data revealed that BdMAPK cascade kinases were rapidly regulated by stresses and phytohormones. Importantly, we have constructed a regulation network based on co-expression patterns of the expression profiles upon multiple stresses performed in this study.

Conclusions

BdMAPK cascade kinases were involved in the signaling pathways of multiple stresses in B. distachyon. The network of co-expression regulation showed the most of duplicated BdMAPK cascade kinase gene orthologs demonstrated their convergent function, whereas few of them developed divergent function in the evolutionary process. The molecular evolution analysis of identified MAPK family genes and the constructed MAPK cascade regulation network under multiple stresses provide valuable information for further investigation of the functions of BdMAPK cascade kinase genes.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-1452-1) contains supplementary material, which is available to authorized users.     

Unveiling nonessential gene deletions that confer significant morphological phenotypes beyond natural yeast strains

Background

Phenotypes are variable within species, with high phenotypic variation in the fitness and cell morphology of natural yeast strains due to genetic variation. A gene deletion collection of yeast laboratory strains also contains phenotypic variations, demonstrating the involvement of each gene and its specific function. However, to date, no study has compared the phenotypic variations between natural strains and gene deletion mutants in yeast.

Results

The morphological variance was compared between 110 most distinct gene deletion strains and 36 typical natural yeast strains using a generalized linear model. The gene deletion strains had higher morphological variance than the natural strains. Thirty-six gene deletion mutants conferred significant morphological changes beyond that of the natural strains, revealing the importance of the genes with high genetic interaction and specific cellular functions for species conservation.

Conclusion

Based on the morphological analysis, we discovered gene deletion mutants whose morphologies were not seen in nature. Our multivariate approach to the morphological diversity provided a new insight into the evolution and species conservation of yeast.

Electronic supplementary material

The online version of this article (doi:10.1186/1471-2164-15-932) contains supplementary material, which is available to authorized users.     

Seed colour loci, homoeology and linkage groups of the C genome chromosomes revealed in Brassica rapa-B. oleracea monosomic alien addition lines

Background and Aims

Brassica rapa and B. oleracea are the progenitors of oilseed rape B. napus. The addition of each chromosome of B. oleracea to the chromosome complement of B. rapa results in a series of monosomic alien addition lines (MAALs). Analysis of MAALs determines which B. oleracea chromosomes carry genes controlling specific phenotypic traits, such as seed colour. Yellow-seeded oilseed rape is a desirable breeding goal both for food and livestock feed end-uses that relate to oil, protein and fibre contents. The aims of this study included developing a missing MAAL to complement an available series, for studies on seed colour control, chromosome homoeology and assignment of linkage groups to B. oleracea chromosomes.

Methods

A new batch of B. rapa–B. oleracea aneuploids was produced to generate the missing MAAL. Seed colour and other plant morphological features relevant to differentiation of MAALs were recorded. For chromosome characterization, Snow''s carmine, fluorescence in situ hybridization (FISH) and genomic in situ hybridization (GISH) were used.

Key Results

The final MAAL was developed. Morphological traits that differentiated the MAALs comprised cotyledon number, leaf morphology, flower colour and seed colour. Seed colour was controlled by major genes on two B. oleracea chromosomes and minor genes on five other chromosomes of this species. Homoeologous pairing was largely between chromosomes with similar centromeric positions. FISH, GISH and a parallel microsatellite marker analysis defined the chromosomes in terms of their linkage groups.

Conclusions

A complete set of MAALs is now available for genetic, genomic, evolutionary and breeding perspectives. Defining chromosomes that carry specific genes, physical localization of DNA markers and access to established genetic linkage maps contribute to the integration of these approaches, manifested in the confirmed correspondence of linkage groups with specific chromosomes. Applications include marker-assisted selection and breeding for yellow seeds.     

Virulence related sequences; insights provided by comparative genomics of Streptococcus uberis of differing virulence

Background

Streptococcus uberis, a Gram-positive, catalase-negative member of the family Streptococcaceae is an important environmental pathogen responsible for a significant proportion of subclinical and clinical bovine intramammary infections. Currently, the genome of only a single reference strain (0140J) has been described. Here we present a comparative analysis of complete draft genome sequences of an additional twelve S. uberis strains.

Results

Pan and core genome analysis revealed the core genome common to all strains to be 1,550 genes in 1,509 orthologous clusters, complemented by 115-246 accessory genes present in one or more S. uberis strains but absent in the reference strain 0140J. Most of the previously predicted virulent genes were present in the core genome of all 13 strains but gene gain/loss was observed between the isolates in CDS associated with clustered regularly interspaced short palindromic repeats (CRISPRs), prophage and bacteriocin production. Experimental challenge experiments confirmed strain EF20 as non-virulent; only able to infect in a transient manner that did not result in clinical mastitis. Comparison of the genome sequence of EF20 with the validated virulent strain 0140J identified genes associated with virulence, however these did not relate clearly with clinical/non-clinical status of infection.

Conclusion

The gain/loss of mobile genetic elements such as CRISPRs and prophage are a potential driving force for evolutionary change. This first “whole-genome” comparison of strains isolated from clinical vs non-clinical intramammary infections including the type virulent vs non-virulent strains did not identify simple gene gain/loss rules that readily explain, or be confidently associated with, differences in virulence. This suggests that a more complex dynamic determines infection potential and clinical outcome not simply gene content.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-1512-6) contains supplementary material, which is available to authorized users.     

SPoRE: a mathematical model to predict double strand breaks and axis protein sites in meiosis

Background

Meiotic recombination between homologous chromosomes provides natural combinations of genetic variations and is a main driving force of evolution. It is initiated via programmed DNA double-strand breaks (DSB) and involves a specific axial chromosomal structure. So far, recombination regions have been mainly determined by experiments, both expensive and time-consuming.

Results

SPoRE is a mathematical model that describes the non-uniform localisation of DSB and axis proteins sites, and distinguishes high versus low protein density. It is based on a combination of genomic signals, based on what is known from wet-lab experiments, whose contribution is precisely quantified. It models axis proteins accumulation at gene 5’-ends with a discrete approximation of their diffusion and convection along genes. It models DSB accumulation at approximated gene promoter positions with intergenic region length and GC-content. SPoRE can be used for prediction and it is parameterised in an obvious way that makes it easy to understand from a biological viewpoint.

Conclusions

When compared to Saccharomyces cerevisiae experimental data, SPoRE predicts axis protein and DSB positions with high sensitivity and precision, axis protein density with an average local correlation r=0.63 and DSB density with an average local correlation r=0.62. SPoRE outbreaks previous DSB predictors, which are based on nucleotide patterning, and it reaches 85% of success rate in DSB prediction compared to 54% obtained by available tools on a benchmarked dataset.SPoRE is available at the address http://www.lcqb.upmc.fr/SPoRE/.

Electronic supplementary material

The online version of this article (doi:10.1186/s12859-014-0391-1) contains supplementary material, which is available to authorized users.     

A genetic approach of wine yeast fermentation capacity in nitrogen-starvation reveals the key role of nitrogen signaling

Background

In conditions of nitrogen limitation, Saccharomyces cerevisiae strains differ in their fermentation capacities, due to differences in their nitrogen requirements. The mechanisms ensuring the maintenance of glycolytic flux in these conditions are unknown. We investigated the genetic basis of these differences, by studying quantitative trait loci (QTL) in a population of 133 individuals from the F2 segregant population generated from a cross between two strains with different nitrogen requirements for efficient fermentation.

Results

By comparing two bulks of segregants with low and high nitrogen requirements, we detected four regions making a quantitative contribution to these traits. We identified four polymorphic genes, in three of these four regions, for which involvement in the phenotype was validated by hemizygote comparison. The functions of the four validated genes, GCN1, MDS3, ARG81 and BIO3, relate to key roles in nitrogen metabolism and signaling, helping to maintain fermentation performance.

Conclusions

This study reveals that differences in nitrogen requirement between yeast strains results from a complex allelic combination. The identification of three genes involved in sensing and signaling nitrogen and specially one from the TOR pathway as affecting nitrogen requirements suggests a role for this pathway in regulating the fermentation rate in starvation through unknown mechanisms linking nitrogen signaling to glycolytic flux.

Electronic supplementary material

The online version of this article (doi: 10.1186/1471-2164-15-495) contains supplementary material, which is available to authorized users.