Origin of a novel protein-coding gene family with similar signal sequence in Schistosoma japonicum
- First Online:
- Cite this article as:
- Mbanefo, E.C., Chuanxin, Y., Kikuchi, M. et al. BMC Genomics (2012) 13: 260. doi:10.1186/1471-2164-13-260
Evolution of novel protein-coding genes is the bedrock of adaptive evolution. Recently, we identified six protein-coding genes with similar signal sequence from Schistosoma japonicum egg stage mRNA using signal sequence trap (SST). To find the mechanism underlying the origination of these genes with similar core promoter regions and signal sequence, we adopted an integrated approach utilizing whole genome, transcriptome and proteome database BLAST queries, other bioinformatics tools, and molecular analyses.
Our data, in combination with database analyses showed evidences of expression of these genes both at the mRNA and protein levels exclusively in all developmental stages of S. japonicum. The signal sequence motif was identified in 27 distinct S. japonicum UniGene entries with multiple mRNA transcripts, and in 34 genome contigs distributed within 18 scaffolds with evidence of genome-wide dispersion. No homolog of these genes or similar domain was found in deposited data from any other organism. We observed preponderance of flanking repetitive elements (REs), albeit partial copies, especially of the RTE-like and Perere class at either side of the duplication source locus. The role of REs as major mediators of DNA-level recombination leading to dispersive duplication is discussed with evidence from our analyses. We also identified a stepwise pathway towards functional selection in evolving genes by alternative splicing. Equally, the possible transcription models of some protein-coding representatives of the duplicons are presented with evidence of expression in vitro.
Our findings contribute to the accumulating evidence of the role of REs in the generation of evolutionary novelties in organisms’ genomes.
KeywordsSignal sequence trap Schistosoma japonicum Repetitive elements Gene duplication Secreted proteins Non-allelic homologous recombination
Signal sequence trap
Non-allelic homologous recombination
DNA level recombination
Non-homologous end joining
Open reading frame
Whole genome shotgun.
Evolutionary novelties generated as an upshot of the “nascence” of new protein-coding genes are the bedrock of adaptive evolution and acquisition of novel molecular functions. The ever-growing vast and diverse protein repertoire in organisms can be ascribed to these events, and may explain the increasing heterogeneity among organisms of otherwise common ancestry [1, 2, 3, 4, 5]. Since the pioneering definitive treatise on gene duplication by Ohno about four decades ago , geneticists and evolutionary biologists have advanced this traditional notion; creating remarkable insights into the composite patterns and underlying mechanisms of genetic innovations. Some of these mechanisms are illustrated in a supplementary figure (Additional file 1). The advent of the genomics era has most importantly armed scientists with a valuable tool to enhance discovery of the rather intriguing mechanisms underlying the “birth” of new genes .
Apart from the canonical gene duplication model as proposed by Ohno ; extensive studies in various organisms have not only elucidated other models of gene duplication, including “dispersed” duplication in addition to the more definitive “tandem” duplication [7, 8, 9, 10, 11, 12, 13]; but has also revealed multiple mechanisms leading to the emergence of new functional genes. These include but not limited to: recombination by exon shuffling or exon “scrambling” [4, 14, 15, 16, 17, 18]; retrotransposition by retrotransposons yielding intronless chimeric genes [18, 19, 20, 21, 22, 23, 24, 25]; transduction of genomic segments by transposable elements by skipping the characteristic weak polyadenylation signal in retrotransposons leading to the mobilization of adjacent genomic sequence; or may involve a repetitive element (RE) mediated DNA level recombination (DLR) by a non-allelic homologous recombination (NAHR) mechanism, in which the REs provide the requisite homologous sequences for the recombination of genomic sequences in a non-allelic manner [7, 20, 26, 27, 28, 29, 30]. Horizontal gene transfer between organisms although infrequent, can give rise to new genes in the recipient organism [31, 32, 33]. De novo origination of protein coding genes from previously non-coding genomic sequences is a very important mechanism previously underrated, but accumulating data in many organism show that this event occur more often than previously thought [2, 3, 34, 35, 36, 37, 38, 39, 40]. Equally, a new gene can arise from the fusion of two genes [1, 3, 22] or fission of a “parent” gene . These mechanisms seldom operate singly as they frequently overlap, collaborating in the creation of nascent genes as depicted in the famous origins of Jingwei and Sphinx in Drosophila species [14, 19].
Schistosoma japonicum along with S. mansoni and S. haematobium are the principal schistosome species causing human schistosomiasis. Uncharacteristic of other human invading schistosomes, S. japonicum is also able to infect several non-human mammalian hosts. While S. japonicum and S. mansoni inhabit the periportal veins and cause an intestinal form of the disease, characterized by liver granulomatous fibrosis as a consequence of host immune response to the eggs lodged in the hepatic sinusoids [42, 43]; S. haematobium causes urinary schistosomiasis at the vesical bladder plexus. Although S. japonicum produces similar lesions like S. mansoni, the fibrotic lesions and hepatosplenomegaly, the most severe outcome of schistosomiasis, is relatively more frequent and severe in S. japonicum. Also, in contrast to S. mansoni and S. haematobium, acute disease due to S. japonicum is common in endemic foci and is associated with severe and persistent manifestations that may rapidly progress the host mediated immunopathogenesis, terminating in a network of fibrotic lesions . Secreted proteins from the parasite ova embolized in the liver of the host are accessible to the host immune cells being located at the host-parasite interface and thus constantly exposed to the host liver tissues. Such interactions play critical role in the initiation and progression of granuloma and fibrosis formation by mediating inflammation [42, 43, 44, 45]. Secreted protein candidates thus, possess great potentials for application in interventions aimed at preventing severe hepatic pathogenesis [46, 47] among other applications.
Nascent genes confer extra functional capacities for the organisms to confront the challenges of the ever dynamic environment, and may equally, albeit rarely, inflict some functional constraints. In any case, recently evolved characteristics could best be attributed to either: protein family or domains expansion, gene loss events , or more likely, evolution of new genes. S. japonicum relatively exhibit a higher degree of parasitism and dependence on host derived molecules and signals as inferred from genomic and transcriptomic studies [49, 50, 51]; it is able to infect a wide range of hosts, and produces relatively more severe pathogenesis . While these could be attributed to a number of other factors including: selective pressure of parasite-host interactions, the extensive gene loss and protein domain elimination or expansion events observed in its genome and transcriptome ; the evolution of novel functional protein coding genes before and after the divergence from other members of the genus Schistosoma could account for these extra characteristics.
Here, we report putative evolutionary novel gene family of Asian schistosomes, S. japonicum on the premise that no homologs of the genes were found in the genome of its evolutionary close relatives in the genus Schistosoma, or in any other organism with a complete sequenced genome. The genes first caught our attention as genes bearing similar or same signal sequence from our previous work that identified some secreted protein coding genes from the eggs of S. japonicum using a signal sequence trap (SST) . Given the available tools prior to the publication of the S. japonicum genome sequence, we had attributed this observation to some alternative or trans-splicing models. The present analysis was inspired by the availability of the invaluable tool presented by the recently published partially assembled genome of this parasite . We adopted an integrated approach utilizing extensive BLAST queries and other bioinformatics tools, transcription and expression analyses, southern hybridization of genomic DNA and evolutionary analyses. We describe evidence of “genome-wide” dispersed duplication of a protein coding gene locus, which may have arisen recently from previously non-coding genomic sequence. The role of repetitive elements as major mediators of the dispersive duplication is analyzed and discussed. Detailed evidence of the potential transcription models of some protein-coding representatives of the duplicons with similar signal sequence is presented and supported by our observations. Finally, based on the identification of non-coding mRNA transcripts as alternatively spliced variants of protein coding mRNAs, we propose that the new genes could be under significant functional selection.
Results and discussion
Sequence characteristics of a novel protein-coding gene family with similar signal sequences in S. japonicum
*SST isolatedS. japonicumegg cDNAs with similar signal peptide
GenBank cDNA Accession
GenBank Protein Accession
AY570737 (1027 bp)
AY570744 (983 bp)
AY570753 (1038 bp)
AY570748 (854 bp)
AY570756 (848 bp)
AY570742 (1037 bp)
*UniGene entries forS. japonicummRNAs and ESTs bearing the similar signal sequence (n = 195)
UniGene ID (UID)
Set of likely mRNA transcripts (GenBank)
Gene Products (Database annotation)
Egg protein SjCP3611
AY570742, FN320556, FN320555, FN320553, FN320552, FN320551, FN320550, FN320549
Egg protein SjCP1531
AY570748SST, AY223245, AY222916, AY813542, EF127834, EF140742, FN323799, FN323800, FN323801, FN323803, FN323793, FN323792, FN323791, FN323790, FN323788, FN323785, FN323782, FN323781, FN323779, FN323778, FN323777, FN323776, FN323773, FN323772, FN323771, FN323770, FN323769, FN323768, FN323767, FN323766, FN323765, FN323764, FN323763, FN323762, BU772060est, BU766145est, CX862012est
Egg protein SjCP3842
Egg protein SjCP1084
AY570753SST, AY570744SST, AY814685, FN327232, FN327137, FN318042, FN321065, FN321060, FN321059, FN321058, FN321057, FN321056, FN321055, FN329815nc, BU768978est, BU780021est
Egg protein SjCP3611, Egg protein SjCP501, Hypothetical proteins
Egg protein SjCP1731
AY570756SST, FN327121, FN327254, FN327253, FN327241, FN327233, FN327229, FN327224, FN327222, FN327216, FN327196, FN327185, FN327163, FN327158, FN327154, FN327129, FN327125, FN327115, FN327089, FN327083, FN327073, FN327057, FN327050, FN327049, FN327045, FN327042, FN327035, FN327022, FN327018, FN327014, FN327000, FN326998, FN326978, FN326973, FN326961, FN326960, FN326959, FN326930, FN326905, FN326883, FN326882, FN326881, FN326859, FN326857, FN326852, FN326851, FN326841, FN326831, FN326829, FN326822, FN326808, FN326801, FN326790, FN326770, FN326740, FN330540nc
Egg protein SjCP400, Somula protein
AY915467, FN327219, FN327063, FN326828, FN326826, FN323794, FN323797, FN323798, FN323802, FN323789, FN323787, FN323786, FN323784, FN323783, FN323780, FN323774, FN323761, FN323760, FN323759, FN323758, FN323757, FN320521, FN320520, FN320519, FN320518, FN320517, FN320516, FN320515, FN320513
Egg protein SjCP3842, Hypothetical protein
AY813755, FN320057, FN320056, FN320514, FN329566nc, BU768160est, BU774105est, BU770186est, BU779051est
Egg protein SjCP3842, Hypothetical protein
FN327242, FN327131, FN327087, FN326854, BU776301est
AY813975, FN329814nc, BU769048est
Egg protein SjCP1084
Egg protein SjCP1084
FN327139, FN323795, FN323796, FN323775
Egg protein SjCP3842
FN327130, FN326955, FN326901
Egg protein SjCP1084
Egg protein SjCP1084
FN326786, FN318043, CX861530est
Schistosoma japonicumgenome contigs containing the similar signal sequence (n = 34)
*Contig [GenBank Accession No.]
Signal Sequence coordinates
4848 – 4779
276 – 205
999 – 928
2413 – 2342
3856 – 3785
3284 – 3217
19023 – 19094
6502 – 6431
42511 – 42440
9335 – 9264
493 – 564
10498 – 10427
7860 – 7789
433 – 504
5768 – 5697
12367 – 12438
383 – 322
4663 – 4602
2484 – 2544
337 – 408
4411 – 4342
388 – 459
925 – 854
271 – 200
1646 – 1717
446 – 517
2389 – 2319
5231 – 5160
1295 – 1366
656 – 585
1094 – 1165
1918 – 1847
1131 – 1060
4249 – 4320
Schistosoma japonicumScaffolds containing the similar signal sequence (n = 18)
Scaffolds [GenBank Accession]
Contigs within the scaffolds
CABF01002611, CABF01002612, CABF01002622, CABF01002623, CABF01002628, CABF01002630
CABF01027854, CABF01027861, CABF01027866
To assess whether some homologs or at least some similar domains exist in other species, BLASTN and BLASTP searches using both the signal sequence and the entire coding sequences of the mRNAs and protein sequences as queries showed that these genes have no homologs in any other organism, but their expression in S. japonicum is supported by evidence from transcriptome and proteomic data. A search on several protein domain databases showed that although our candidates were classified in the same protein family with similar domains and assigned to a domain ID (ProDom:PD884968), no related domain or protein family was found in any other organism. The absence of these genes in the genome of S. mansoni, S. haematobium and other published genomes cannot possibly be attributed to sequencing gaps or annotation errors since the WGS sequencing approach is considerably reliable , and the fact that we adopted a multiple species approach covering the entire available sequenced genomes of all species makes this even more improbable [37, 40]. Given the accumulating evidences of de novo origin of new genes from previously non-coding DNA sequences [2, 34, 35, 36, 37, 38, 39, 40], we propose that the coding sequence of these genes may have recently originated de novo from previously non-coding DNA sequences in the ancestral forms, and subsequently duplicated and dispersed in the genome. This represents a more plausible interpretation than the improbable alternative hypothesis of concurrent gene deletion or inactivation in multiple ancestral lineages.
Species and strain specific expression
Nevertheless, while it is completely normal to verify this exclusive evolution and dispersed duplication hypothesis by confirming the physical localization of the gene loci in the genome and chromosomes by performing synteny analysis, we are unable to achieve this because we do not have access to a fully mapped chromosome information of the genome of S. japonicum. However, the distribution of the contigs and scaffolds bearing the similar signal sequence apparently suggests a dispersed distribution. To confirm this hypothesis and to exclude the possibility of overlapping among the loci, we generated the restriction map of six of the genome scaffolds bearing duplicated loci based on the information on the genome map, performed southern hybridization using restriction endonuclease digested genomic DNA from S. japonicum species and strains; and were able to match the expected probe binding fragment sizes with the observed bands on the hybridization blots (Additional file 4). Also an ancestral homolog is required for synteny analysis, however, we could not find a homolog in S. mansoni, another member of the genus with sequenced genome; and the genome of other more closely related species like S. mekongi and S. malayensis are not yet sequenced. Unless new evidences emerge from future updates in the sequenced genomes, we hold true that these genes have newly evolved, probably from modifications on previously non-coding ancestral DNA sequences and subsequently disperse duplicated. As opined in previous studies, the short length of our identified genes is an expected property of nascent genes because of improbability of evolution of long open reading frames (ORFs) and the complexity of intron splicing signal . We expect these novel genes to be of functional significance since new genes tend to display accelerated sequence and structural changes towards neo-functionalization , and most newly characterized genes from other species have been shown to be characteristically functional [35, 56]. Other workers showed that the common pathway for de novo protein-coding gene evolution involves a piece of DNA sequence to be transcribed via recruitment of all transcription core promoters, other elements and machines; followed by the acquisition of a translatable ORF through mutations or other sequence alteration mechanisms [2, 35]. Together, our findings support the presence of these intrinsic features of novel genes in the identified candidates, including the gradual model of novel protein coding gene origination.
Evidence of dispersed duplication from a source gene locus
Additional file 5: Simulations using our raw data to show DNA-Level recombination mediated by REs by NAHR mechanism. The movie created from a Powerpoint presentation (Additional file 10) represents the basic approach we utilized in our analysis to show evidence of DNA level recombination by a non-allelic homologous recombination mechanism. The raw data obtained from BLAST searches and RepBase repetitive element prediction report was used to present a simulation that demonstrates that the duplicated locus is flanked on 5` and 3` ends by retrotransposons of the classes RTE_SJ and Perere respectively. We proposed that these repetitive elements could have provided the requisite homologous stretch of DNA that is required for such DNA level recombination. NAHR can be inter-chromosomal, intra-chromosomal, inter-sister chromatid, or intra-chromatid to give rise to disperse duplicates of the intervening genomic locus. This movie was created from an original Powerpoint presentation (Additional file 10) (MOV 6 MB)
The role of repetitive elements (REs) in dispersed duplication of genomic sequences is fairly documented from previous studies in model organisms [15, 20, 27, 28, 30, 58, 59]. The precise mechanism of this retrotransposon mediated dispersed duplication is not clear but may likely involve RE-mediated DNA level recombination, most likely by non-allelic homologous recombination (NAHR), alternatively called ectopic recombination (see illustration in Additional file 6). Due to their extremely high copy numbers, REs create structural modifications in the genome by providing the requisite highly similar DNA sequences, initiating recombination between non-allelic elements [20, 25, 60], the result of which could be deletion, shuffling, duplication or transduction of a genomic DNA segment. Structural modifications introduced in the genome by NAHR mechanism can progress between non-homologous chromosomes (inter-chromosomal), between homologous chromosomes (inter-homologous or intra-chromosomal), between sister chromatids (inter-sister chromatid) or within a chromatid (intra-chromatid); giving rise to dispersed duplication of genomic segments, several forms of deletions or may create isodicentric chromosome by forming a mirrored segment in the chromosome by inversion. See detailed cartoon in Additional file 6. .
Many studies in other organisms have elucidated the role of REs in mediating sequence duplication, transduction and other structural variations by ectopic recombination mechanism. Notable among these is the human Alu element for which several reports suggest a role in mediating NAHR and other structural modifications in the human genome [7, 20, 61]. Yang et al found an excess of repetitive sequences proximate to the breakpoints of duplicated gene loci in the genome of the fruit fly Drosophyla melanogaster, and have suggested that a NAHR mechanism, mediated by REs accounted for the birth of the new duplicons [1, 27]. Another study performed on human individuals concluded that NAHR accounted for over 40 % of detected genomic sequence duplications in the human genome . Illegitimate recombination (IR), incomplete crossing over and non-homologous end joining (NHEJ) are other possible mechanisms of gene duplication by DNA-level recombination, but NAHR play a more significant role in producing typical dispersed duplications  while the other mechanisms in addition to NAHR are more likely to produce tandem duplicates. Although we could not clearly identify the exact breakpoints of the duplications at both ends still for lack of a reference ancestral homolog and partly due to sequencing gaps, the fact that homology among all the scaffolds examined uniformly terminated at the same point with Perere on the 3` end (Figure 3 and Additional file 5), and traces of the observed predominant retrotransposons (SjR2) was found at the exact positions as they occur in the putative source locus (Figure 4) confirm that these gene loci could be products of dispersed duplication from a single genomic source locus.
In addition to RE-mediated DNA-level recombination by NAHR, gene duplication events are also attributable to RE-mediated retrotransduction mechanism either on the 5` or 3` directions . Xing et al and other groups have demonstrated the role of retrotransposons in the duplication of entire genes and creation of previously un-described genes by analyzing SVA (SINE, VNTR and Alu)-mediated retrotransduction events in the human genome [20, 29]. However, we did not specifically identify any chimeric duplicon originating via a retrotransduction mechanism among our datasets. Furthermore, retrotransposons including SjR2 characteristically encode reverse transcriptase and endonuclease, and can therefore transcribe and ‘paste’ a gene sequence into new locations in the genome [3, 22, 62]. However, retrotransposed genes are characteristically intronless since the introns are usually spliced out during the process of retrotransposition. Our duplicons retained their introns, although in some case some portion of the introns may have either degenerated or deleted during duplication and subsequent sequence modifications [3, 22, 63]. A further evidence that a retrotransposition mechanism is unlikely in our observed cases was that while retrotransposons would not duplicate the promoter regions of duplicated gene based on the process of transcription and insertion of retrocopies [1, 57] which leaves the newly retrotransposed sequences to acquire new regulatory sequences from adjacent genes or through mutations in order to be functional [14, 19, 24]; the protein coding duplicons observed among our duplicated gene loci retained the same or similar core regulatory region and signal sequence as the source locus, suggesting that they may not have been products of retroposition and may equally explain the parallel assumption of coding potential at their new duplication loci without the need to form chimeric structures with adjacent genes.
Evolution of translatable ORF and evidence of expression of duplicated genes
Some of the duplicons appear degenerative in homology and are relatively shorter than the source locus (Figure 3, also see Additional file 5) thus are consequently redundant and non-coding at the new locations as opined in the canonical view on the fate of new duplicons [6, 9]; which assumes a tendency to be lost because of genetic drift under natural evolution [29, 57]. However, our data provide evidence that some of the duplicons have evolved into protein coding genes with distinct products at their new loci, the fate of which could tend to either sub-functionalization to the source gene [8, 64] or neo-functionalization by acquiring new distinct functions [9, 65]. In addition to the two duplicons with alternative splicing variants, which we further explored in the next section, some representatives of the protein coding duplicons were depicted in a supplementary figure (Additional file 7). The nucleotide sequences of these genes are still appreciably similar but accumulation of mutations and other sequence modifications have given rise to novel protein coding ORFs, encoding putatively distinct products. We identified and mapped each cDNA sequence to the genomic contigs using information we generated from GeneMark and GeneQuest gene predictions  and confirmed by alignment of the cDNAs to the genomic sequences using NCBI Spling program. This approach was necessitated because the fully mapped and annotated genome of S. japonicum is not presently available in the public databases. Intriguingly, our results corroborate the available UniGene and GenBank entries. Nevertheless, it is notable that we only assessed the duplicated copies on the basis of possessing the similar signal sequence. There is possibility that some other duplicons from this source locus could be involved in initiating other forms of structural modifications at other loci when incorporated into the coding region of other genes, but this was not investigated here.
Functional selection by alternative splicing
The precise recognition of exon-intron junction in a precursor mRNA (pre-mRNA) by the splicing machinery is central for the production of functional translatable mRNAs. However, there is often uncertainties in the choice of recognizable splice signals, resulting in a process termed alternative splicing , which enables the origination of multiple mRNA transcript variants from a single gene locus [67, 68, 69]. Alternative splicing mechanism could result in ‘intronization’ of an exon or ‘exonization’ of an intronic sequence. Ideally, the creation of an intron from a previously exonic sequence could lead to the loss of an ORF in coding genes. In evolving genes however, functional selection possibly by mutations may evolve the required splice signals and induce the intronization of an exon in a transcribed but non-coding mRNA gene sequence to create a translatable ORF encoding a functional protein. Conversely, while exonization of an intron could disrupt a translatable ORF in a coding gene, selective pressure may also evolve new splice signals within an intron to yield exons that could create a translatable ORF from a previously non-coding gene locus or a chimeric ORF from a protein-coding gene.
On the other hand, exons 5 and 6 of a coding mRNA variant [GenBank:AY570742] predictably transcribed from one of the progeny loci [GenBank:CABF01023364] were skipped in a non-coding shorter variant [GenBank:FN329677] without a translatable ORF (Figure 6 and Additional file 9). We observed that the sequences of exons 5 and 6 were similar and was repeated five times in tandem within this locus, but only two copies of the tandemly duplicated potential exons were incorporated into the coding sequence of the mRNA to create exons 5 and 6 of a protein-coding ORF of 274 codons (SjCP1531). These results represent typical models of alternative splicing by intronization and exonization respectively.
Although in evolutionary perspective, intron retention that creates a translatable ORF is considered more plausible than the reverse process; our data show that both mechanisms are potentially possible. Other groups have also identified intron gains recently in mammalian and rodent retrogenes [68, 69]. The identification of non-coding mRNA variant alternatively transcribed from a single gene locus with a protein coding mRNA (Figure 6) is evidence that a novel protein-coding gene can originate from previously transcribed regions that contain the necessary transcription elements and provide RNA material for a protein translation machine [2, 39, 68]. Exon repetition has also been observed from our data to exist in this organism and could be instrumental in expanding the organism`s coding potential. The ‘parallel’ expression of the non-coding variant alongside the protein-coding transcripts is of significance and could suggest further that the gene may have been recently evolved. Non-coding RNAs have also been shown to perform some regulatory roles at various levels during gene expression [2, 68, 70]. This could be further explored with our data set. In the two described cases in our analyses, we have treated the non-coding isoforms as evolutionally preceding the coding variants; nevertheless, the reverse could also be the case. In addition to these two cases, we also identified a two-nucleotide insertion into a non-coding mRNA sequence [GenBank: FN330540] that yielded the coding mRNA of schistosomula protein with the similar signal peptide, with many similar transcripts in the database. However, this last observation could be an artifact from sequencing error since the existence of the non-coding transcript was not traceable to the genomic sequence.
We have passably delineated the possible mechanism that led to the identification of several protein coding genes with similar signal sequence, following lead from our work that isolated secreted proteins candidate genes using SST. A trend was described in the genome of S. japonicum whereby a ‘newly evolved’ gene served as a source locus for dispersed duplication events leading to the formation of several expressed genes with similar transcription core promoter region and signal sequence. We further found that the duplicated gene locus was flanked by non-long terminal repetitive elements (REs), especially of the RTE- like and Perere class. We therefore inferred that REs may have played an important role in this dispersed gene duplication by creating the requisite homologous DNA sequence that mediate a DNA-level recombination, most probably by a non-allelic homologous recombination (NAHR) mechanism. Our findings also provide evidence of logical sequential process of novel gene origination by evolution of transcription core elements followed by translatable ORF. While similar RE mediated phenomena had been observed in other organisms, unlike our dataset, most analyses have centered on the model organisms. Our data contribute to the accumulating evidence that REs mediate diverse recombination events leading to novel gene origination and other evolutionary novelties.
We had earlier identified a particular 81 nucleotides (27 amino acids) sequence, which was commonly utilized as signal sequence by several of our signal sequence trap (SST) isolated S. japonicum cDNAs (Table 1) . The sequence of this signal sequence was employed as query to search for matches in the GenBank non-redundant nucleotide sequence database and expressed sequence tags (ESTs) database for all organisms using BLASTN program in National Center for Biotechnology Information (NCBI) Basic Local Alignment Search Tool (BLAST) . A search on the NCBI UniGene database  that provides information on sets of transcript sequences that appear to come from the same transcription locus was performed to ascertain redundancy and group identified transcripts. For genome-wide searches, the same query sequence and program were used to search the WGS reads from all organisms with sequenced genomes deposited in the NCBI genome databases. In a similar search in the protein database, the amino acid and nucleotide sequence of the same signal sequence was used as query for BLASTP and BLASTX searches respectively. Conserved domain architecture searches on all translation products of the SST identified candidate genes were performed using the conserved domain architecture retrieval tool on NCBI website  and compared with same analyses on the ProDom database of protein domain families available online at .
All multiple sequence alignments of DNA and protein sequences were performed in parallel with ClustalW on MegAlign program in Lasergene 7 DNASTAR software, NCBI bl2seq, COBALT multiple alignment programs, and Multialin interface software . cDNA-to-genome sequence alignments were computed using the free NCBI Spling program . The latest update of the S. japonicum genome map is accessible at . Phylogenetic and molecular evolutionary analyses were conducted using MEGA version 5 .
Gene predictions were performed using the GeneQuest program (Lasergene 7 DNASTAR) to predict potential coding regions, starts, stops, acceptors and donor sites using Borodovsky matrix files for Caenorhabditis elegans; and the results compared with that of the Eukaryotic GeneMark.hmm  gene prediction server provided freely on the website of Georgia Institute of Technology, Atlanta, USA.
The whole sequences of all the genome contigs bearing the similar signal sequence were screened against a reference collection of repetitive DNA elements in the RepBase database available at the Genomic Information Research Institute website, using the CENSOR repeat masking software . Sequence analysis figures were generated using real DNA sequences on Vector NTI Advanced 11.0 (Invitrogen).
Designation of putative duplication source locus and probable breakpoint
Reference to a parent gene is required for accurate determination of duplication breakpoint. However, in absence of a reference homolog, we putatively selected the most prominent contig [GenBank:CABF01020060], the longest among the identified dataset (43.7 kb), which significantly covered the length of the other contigs (Figure 3, also see Additional file 5) as the putative duplication source locus and utilized it as such for most of the analyses performed in this study. When the contigs were aligned with the putative source locus, homology was not lost till the 3` end of the aligned sequences. We therefore recruited two contigs [GenBank:CABF01020061 and GenBank:CABF01020062] downstream of the source locus based on genome assembly information, thereby generating at least 5 kb flanking sequences on either side of the duplication source locus. This sequence was then aligned with the genome contigs and scaffolds to identify the exact point where sequence identity disappeared. This point was arguably chosen as the possible duplication breakpoint and utilized as such in our discussions. We further attempted to identify a recurrent consensus sequence at the breakpoints but this was hampered by several sequencing gaps in the partially assembled scaffolds.
Parasites, genomic DNA and developmental stage mRNA samples
Chinese strain of S. japonicum (hereafter abbreviated as Sj) was obtained from Jiangsu Provincial Institute of Parasitic Diseases Wuxi, Jiangsu Province, PR China, while the Philippine and Japanese strains of S. japonicum in addition to S. mekongi (Smk) samples, were maintained in the Laboratory of Tropical Medicine and Parasitology, Dokkyo Medical University, Tochigi, Japan. S. mansoni (Sm) adult worms were maintained by, and kindly provided by the Department of Parasitology, Institute of Tropical Medicine, Nagasaki University, Japan. S. haematobium (Sh) sample was from Department of Immunology and Parasitology, University of Occupational and Environmental Health, Kitakyushu, Japan. Total genomic DNA was purified from cut tissues of mixed sex adult worms from different species of Schistosoma using QIAamp DNA Mini Kit (QIAGEN) according to the manufacturer’s instructions. Qualification and quantification of genomic DNA extract was assessed by gel electrophoresis and ND-1000 spectrophotometer (NanoDrop, USA). To obtain sufficient amount of genomic DNA for southern hybridization experiments, the whole genome of each sample was amplified using the GenomePhi DNA Amplification Kit (GE Healthcare) according to the manufacturer’s instructions. Equally, total RNA was extracted from parasite eggs, cercariae, 24 h cultured schistosomulae and adult worms of S. japonicum according to the instruction manual of PureLink Micro-to-midi total RNA Purification System Kit (Invitrogen).
Reverse transcription polymerase chain reaction (RT-PCR)
mRNA from eggs, cercariae, 24 h culture schistosomulae and adult worms of the Chinese strain of S. japonicum was used for RT-PCR. The first strand cDNA was synthesized from the total RNA of each developmental stage by using oligo (dT) primer according to the instruction manual of High Capacity cDNA Reverse Transcription Kit (Applied Biosystems) and the resulting cDNA was used as template for RT-PCR. The S. japonicum actin gene was used for internal quality assurance. The cDNA sequences of some selected SST identified secreted candidate genes were amplified using pairs of sequence specific primers designed according to the S. japonicum transcriptome data  in the NCBI public database. All RT-PCR amplicons were analyzed using gel electrophoresis and confirmed by sequencing using the BigDye Terminator v1.1 Cycle Sequencing Kit (Applied Biosystem).
Southern hybridization was performed following standard procedures  using the DIG nonradioactive labeling and detection system (Roche, Germany). Briefly, the hybridization probe labeled with DIG-dUTP was synthesized using PCR DIG synthesis kit (Roche, Germany) according to the manufacturer’s instructions, and labeling was confirmed by size disparity with unlabeled amplicon as a result of slower migration in agarose gel due to digoxigenin labeling. Genomic DNA from different species of Schistosoma (Sh, Sm, Smk, Sj Japanese (Yamanashi), Sj Chinese (Jiangsu) and Sj Philippines (Leyte, Mindanao and Mondoro isolates) were double digested with three different pairs of restriction enzymes (Eco RI + Eco RV, Eco RI + Hind III and Bam HI + Hind III) to achieve the best possible fragmentation of the genomic DNA. The digested genomic DNA fragments were electrophoresed through 1 % (w/v) agarose gel, depurinated in 250 mM HCl, and denatured by incubating in two changes of denaturing solution for 15 min each (0.5 M NaOH, 1.5 M NaCl). The gels were then neutralized by incubation in two changes of neutralizing solution (0.5 M Tris-HCl at pH7.5, 1.5 M NaCl) for 15 min each, and DNA was transferred to a positively charged nylon membrane (Roche, Germany) by capillary action overnight using 20x SSC solution (3 M NaCl, 300 mM sodium citrate at pH 7.0). The transferred DNA was fixed to the membrane by baking in an oven at 80 °C for 2 hours after rinsing briefly in 2x SSC. After prehybridizing the membrane in 10 ml hybridization buffer (5x SSC, 0.1 % N-lauroylsarcosine (w/v), 0.02 % SDS (w/v), 1 % blocking solution (Roche, Germany)) for 30mins in a hybridization bag, 5μl of the PCR generated hybridization probe was mixed in 50μl of double deionized water, denatured by boiling for 5mins and introduced into the hybridization bag and incubated overnight with shaking at 50 °C. The membrane was washed in two changes of low stringent wash buffer (2x SSC, 0.1 % SDS) for 5mins each at RT, and twice in high stringent wash buffers (0.5x SSC, 0.1 % SDS) for 15 min each at 65 °C. The hybridized probe was then detected using anti-Digoxigenin antibody (Roche, Germany) using CSPD as the chemiluminiscent substrate according to the manufacturer’s instructions. The blot was then visualized by exposing to chemiluminiscence for 10 min in a LAS-4000 mini image reader (Fujifilm).
ECM participated in the conception and design of the study, in-silico analyses, molecular experiments, data analysis and interpretation and drafted the manuscript. YC carried out the signal sequence trap (SST) and participated in in-silico analyses. MK participated in the design of the study, SST, in-silico analyses, molecular experiments and data interpretation. MNS participated in in-silico analyses, molecular experiments, data interpretation and revised the manuscript. DB participated in molecular experiments and data analyses. MK2, NH, YC2 and YO maintained parasite life cycle and participated in molecular experiments. SH participated in data interpretation, supervision and revised manuscript for intellectual content. KH participated in the conception and design of the study, SST, in-silico analyses, data interpretation, revised the manuscript and general coordination. All authors approved final version of the manuscript.
We would like to thank Prof. Kenji Hirayama’s lab members (Department of Immunogenetics, Institute of Tropical Medicine, Nagasaki University) for insightful discussions. ECM is a recipient of the Japanese Government Ministry of Education, Culture, Sports, Science and Technology (MEXT) PhD fellowship. This study was supported in part by the Global Center of Excellence (GCOE) Program (2008-2011); Grant-in-Aid for 21st century COE program (2003-2008), Nagasaki University; and Grant-in-Aid for Scientific Research B (22406009) and C (23590489) from the Japanese Government Ministry of Education, Culture, Sports, Science and Technology (MEXT). The funding agency played no role in conducting the study, drafting the manuscript and the decision to publish.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.