MBE Advance Access originally published online on December 17, 2007
Molecular Biology and Evolution 2008 25(3):526-535; doi:10.1093/molbev/msm278
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Articles |
Investigation of Loss and Gain of Introns in the Compact Genomes of Pufferfishes (Fugu and Tetraodon)
Institute of Molecular and Cell Biology, Agency for Science, Technology, and Research, Biopolis, Singapore
E-mail: mcbbv{at}imcb.a-star.edu.sg.
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionConclusionsSupplementary MaterialAcknowledgementsReferences |
|---|
We have investigated intron evolution in the compact genomes of 2 closely related species of pufferfishes, Fugu rubripes and Tetraodon nigroviridis, that diverged about 32 million years ago (MYA). Analysis of 148,028 aligned intron positions in 13,547 gene pairs using human as an outgroup identified 57 and 24 intron losses in Tetraodon and fugu lineages, respectively, and no gain in either lineage. For comparison, we analyzed 144,545 intron positions in 12,866 orthologous pairs of genes in human and mouse that diverged about 61 MYA using dog as an outgroup and identified 51 intron losses in mouse and 3 losses in human and no gain. The rate of intron loss in Tetraodon is higher than that in fugu, mouse, and human but lower than the previous estimates for other eukaryotes. The introns lost in pufferfishes and mammals are significantly shorter than the mean size of introns in the genome. One intron deleted in fugu and another in Tetraodon have left behind 6 and 3 nucleotides, respectively, suggesting that they were lost due to genomic deletions. Such losses of introns are likely to be the result of a higher rate of DNA deletions experienced by the genomes of pufferfishes compared with mammals. The shorter generation time of Tetraodon compared with fugu, and the rich diversity and higher activity of transposable elements in pufferfishes compared with mammals, may be responsible for the higher rate of intron loss in Tetraodon. Our findings indicate that overall very little intron turnover has occurred in pufferfishes and mammals during recent evolution and that intron gain is an extremely rare event in vertebrate evolution.
Key Words: intron loss and gain • fugu • Tetraodon • transposable elements
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsSupplementary MaterialAcknowledgementsReferences |
|---|
Spliceosomal introns are genomic sequences that interrupt coding sequences of eukaryotes. They are excised from RNA transcripts prior to translation via a mechanism that involves a spliceosomal complex. Ever since the discovery of spliceosomal introns about 30 years ago, the evolutionary origin and significance of spliceosomal introns have been the subject of a lively debate (Jeffares et al. 2006
; Roy and Gilbert 2006). The identification of components of the spliceosome in some basal eukaryotes supports the notion that introns have an ancient origin (Collins and Penny 2005). The number of introns in eukaryote genomes varies dramatically from a few introns in some protozoans (Nixon et al. 2002) to more than 150,000 in vertebrates (Jeffares et al. 2006). Such wide variations in intron numbers across taxa imply that there must have been extensive loss and gain of introns during the evolution of eukaryotes. However, the relative importance of loss and gain of introns in the evolution of eukaryote genomes is not clear.
Genome-wide comparisons of several closely related species of eukaryotes have indicated that intron losses have prevailed over gains during recent evolution (Robertson 1998; Roy et al. 2003; Cho et al. 2004; Nielsen et al. 2004; Roy and Hartl 2006; Stajich and Dietrich 2006; Roy and Penny 2006a, 2007b; Putnam et al. 2007). By contrast, comparisons of some distant species have concluded that intron gains were more prevalent than intron losses (Qiu et al. 2004; Nguyen et al. 2005). Similar contradictory results have also been reported for paralogous genes. Losses were found to be more common in paralogous genes in rice and Arabidopsis (Lin et al. 2006; Roy and Penny 2007a), whereas gains were prevalent in paralogous genes in human, fruit fly, nematode worm, and fission yeast (Babenko et al. 2004). Probabilistic models based on 391 sets of conserved genes in 19 eukaryotes have suggested that 3 distinct scenarios might be occurring in different lineages: elevated intron loss, elevated intron gain, and balanced loss and gain (Carmel et al. 2007).
The mechanisms by which introns are lost or gained are not well understood. Two main models for the loss of introns have been proposed: recombination of genomic sequence with a reverse-transcribed copy of mRNA (Fink 1987; Sverdlov et al. 2004; Roy and Gilbert 2005) and deletion of intronic sequence from the genomic DNA (Robertson 1998). In the case of intron gain, at least 5 mechanisms have been proposed. They include insertion of a transposable element (TE) (Crick 1979), insertion of a reverse-transcribed intron into a new position (Cavalier-Smith 1985; Palmer and Logsdon 1991), tandem duplication of an exon (Rogers 1989; Venkatesh et al. 1999), intron transfer between paralogs through recombination (Hankeln et al. 1997), and insertion of a self-splicing type II intron via reverse splicing (Cavalier-Smith 1991).
Vertebrate genomes are typically large and intron rich. Interestingly, intron studies in mammals have indicated that very little intron turnover has occurred in vertebrates with convincing evidence for only loss of introns. Comparisons of 1,560 human–mouse orthologs and 360 mouse–rat orthologs identified only 5 losses in the mouse and 1 loss in the rat (Roy et al. 2003). A recent genome-wide comparison of introns in human, mouse, rat, and dog could identify only losses: 4 in human, 29 in mouse, 46 in rat, and 7 in dog (Coulombe-Huntington and Majewski 2007). Pufferfishes (family Tetraodontidae) are unique among vertebrates because of their small genomes. On average, the pufferfish genomes are about one-eighth the size of the genomes of mammals. It would be interesting to examine whether the forces that have led to the compaction of the pufferfish genomes have influenced the dynamics of intron loss and gain. The whole-genome sequences of 2 pufferfishes, Fugu rubripes and Tetraodon nigroviridis, have been determined (Aparicio et al. 2002; Jaillon et al. 2004). This gave us a unique opportunity to investigate not only intron dynamics in the compact genomes of pufferfishes but also to identify intron loss and gain in 2 closely related species of vertebrates. In this study, we have analyzed intron evolution in fugu and Tetraodon, using human as an outgroup. For comparison, we also analyzed discordant introns in the whole-genome sequences of human and mouse using dog as an outgroup. The 2 pufferfishes are estimated to have diverged about 32 million years ago (MYA), whereas the human and mouse lineages diverged about 61 MYA (Benton and Donoghue 2007) (fig. 1). Our comparisons of fugu–Tetraodon orthologs uncovered evidence for 57 losses in Tetraodon and 24 losses in fugu and no gain in either lineage. Likewise, comparisons of human–mouse orthologs identified only losses: 51 in mouse and 3 in human. These results indicate that very little intron turnover has occurred in pufferfishes during recent evolution. Among the 4 vertebrates, Tetraodon has experienced the highest rate of intron loss. Two introns lost in pufferfishes have been deleted imprecisely suggesting that they were most likely lost due to spontaneous genomic deletions.
|
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsSupplementary MaterialAcknowledgementsReferences |
|---|
Sequence Data and Intron Annotation
Sequence data for fugu, Tetraodon, zebrafish, human, mouse, and dog were extracted from the February 2006 release of Ensembl (version 37) (http://www.ensembl.org). For each protein-coding gene, the sequence of the longest transcript was used to represent the gene. Only introns that are
45 bp were considered. The final data sets used for the analysis consist of 22,008 fugu genes (159,801 introns; mean length 667 bp), 28,005 Tetraodon genes (164,198 introns; mean length 573 bp), 22,209 human genes (164,851 introns; mean length 5,430 bp), and 24,754 mouse genes (160,237 introns; mean length 4,732 bp). Orthologous protein sequences were identified based on reciprocal BlastP searches with an e value cutoff of 10–10. Ortholog sets that contained any intronless gene were excluded because intronless genes are often the result of retrotransposition and would bias the analysis.
Ortholog Alignment and Gain/Loss Assignment
Fugu–Tetraodon and human–mouse orthologous protein sequences were aligned using ClustalW, following which intron positions and phases ("phase 0" when present between 2 codons; "phase 1" when present between the first and second base of a codon and "phase 2" when present between the second and third base of a codon) were mapped onto the sequence alignment. The mapped alignments were then filtered computationally as described below to identify putative discordant intron positions. The discordant intron positions were inferred as a gain or a loss based on parsimony to the outgroup sequence. To identify putative discordant introns, a Perl script was written to perform several successive filtering steps for each of the aligned intron positions in order to assign one of 3 possible classifications ("ambiguous", "conserved", or "discordant"). First, an intron-containing alignment position was classified as ambiguous if another intron-containing position was detected within 5 amino acids upstream or downstream to filter out possible cases of intron sliding events or misalignments. Next, the alignment quality of the 10 amino acids upstream and downstream (total 20 amino acids) was determined, and if there was less than 50% identity in either upstream or downstream sequences, such positions were also classified as ambiguous. Following this, if an intron was present in every species in the alignment, a conserved assignment was made. The remaining alignment positions were classified as discordant.
Manual Inspection and Curation
All alignment positions with discordant introns were manually inspected and curated in order to confer the highest degree of confidence for intron discordance. The manual inspection included confirmation of orthology assignment by BlastP search of individual sequences against the proteome of the other species, synteny comparison, verification of good overall alignment quality, presence of at least one conserved intron position in all 3 species, checking for possible intron mispredictions, and visual inspection of sequences of "gained" introns. Introns comprising mostly N's are likely to be assembly artifacts and were excluded.
Verification of Loss and Gain in Pufferfish Using Other Fishes as Outgroups
To verify the gain and loss of introns in Tetraodon and fugu inferred using human as the outgroup, we compared the discordant intron positions in pufferfishes with their homologs (identified by BlastP with an e value cutoff of 10–10) predicted in the genome assemblies of zebrafish, medaka, and stickleback (Ensembl version 46; http://www.ensembl.org).
Rates of Intron Loss
The rate of intron loss (Rl) was calculated using the formula Rl = L/((D + C) x T), where L is the observed number of losses, D is the number of discordant intron sites, C is the number of conserved intron sites, and T is the time since divergence.
Polymerase Chain Reaction Amplification and Sequencing
To verify the sequence of a putative intron gain in a Tetraodon gene (Ensembl ID, GSTENP00004545001), a fragment of the gene was amplified by polymerase chain reaction (PCR) using Tetraodon genomic DNA (kindly provided by Laszlo Orban) or DNA extracted from a bacterial artificial chromosome (BAC) clone (ID C0AA022CG10) that was used by Genoscope for generating the Tetraodon genome sequence as template and primers flanking the intron position (forward primer, CCG CGC CAA GAA TAT AAA AGA TAC C and reverse primer, GAA GGT CTT GGG GTC AAA GTC GTC). The same primer pair was used to amplify cDNA prepared using SMART RACE cDNA Amplification kit (Clontech, Mount View, CA). The cDNA was prepared from total RNA extracted from gills using Trizol reagent (Invitrogen, Carlsbad, CA). PCR products were sequenced directly on an ABI 3730xl DNA analyzer using BigDye Terminator chemistry.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsSupplementary MaterialAcknowledgementsReferences |
|---|
Discordant Introns in Fugu and Tetraodon
A total of 13,547 fugu–Tetraodon orthologous protein sequences were aligned to produce 148,028 aligned intron positions. Our objective was to identify a high-confidence set of discordant intron positions in the 2 pufferfishes using human as an outgroup. Toward this end, we used a series of stringent computational filtering and manual curation steps and classified the fugu–Tetraodon intron positions into 91,255 conserved, 56,674 ambiguous, and 99 discordant positions (see Materials and Methods). Three of the discordant positions had no convincing ortholog in human, whereas sequences of putative introns at 14 positions were found to be the result of misassembly (containing multiple N's or sequences identical to flanking exon). The remaining 82 positions were inferred to be the result of 57 losses and 4 gains in Tetraodon and 21 losses in fugu. However, when we compared the 4 putative gained intron positions in Tetraodon with their homologs predicted in the zebrafish, stickleback, and medaka genome assemblies (Ensembl version 46), we found that whereas all identifiable zebrafish homologs (3 out of 4 successfully detected) lack introns at these positions similar to fugu and human, stickleback homologs contained introns at 3 of these intron positions (in GSTENP00027886001, GSTENP00003442001, and GSTENP00023170001). The medaka assembly contained homologs for only 3 of these Tetraodon genes and 2 of them (GSTENP00027886001 and GSTENP00023170001) contained the gained introns. Thus, we conclude that 3 of the gains identified in Tetraodon genes (GSTENP00027886001, GSTENP00003442001, and GSTENP00023170001) are actually the result of independent losses in fugu and zebrafish. They were reclassified as instances of intron loss in fugu. We were finally left with a single instance of gain in Tetraodon (in GSTENP00004545001, nucleobindin 2 protein). This putative intron, located in phase 2 is 63-bp long, GC-rich, and contains the canonical splice sites (gt.ag). We also noticed that the 5'-flanking exon of this intron codes for low-complexity sequence (strings of Glu and Gln). Because this was the only gain identified, we sought to verify its sequence by PCR amplification using genomic DNA, DNA from a BAC from this locus (ID C0AA022CG10 kindly supplied by Genoscope), and cDNA as template. To our surprise, sequencing of the PCR products showed that the sequence of this putative intron reported in Ensembl assembly is wrong. It turned out to be a 24-bp long sequence that contained an open reading in-frame with the flanking exonic sequences. The cDNA sequence (amplified from gills) confirmed that this sequence is part of the exon and not an intron. Thus, this gained intron is not a real intron but a short expansion of the coding sequence. It was, therefore, not counted as a gain.
The above findings, particularly the presence of parallel intron losses in other fishes, prompted us to compare the positions of the 57 losses in Tetraodon and 21 losses in fugu with their homologs predicted in the zebrafish, stickleback, and medaka genome assemblies (Ensembl version 46). Homologs for these pufferfish genes were predicted in at least 2 of these fish genome assemblies and they contained introns at these positions. This provided strong support to our inference that the intron losses identified in pufferfishes were lost specifically in the fugu or the Tetraodon lineage. Thus, our analysis identified a final set of 57 losses in Tetraodon, 24 losses in fugu, and no gain in either lineage (table 1). A representative alignment with an inferred intron loss in the Tetraodon lineage is shown in figure 2. The details of all the intron losses are given in supplementary table 1 (Supplementary Material online), and their marked alignments are given in supplementary figure 1 (Supplementary Material online).
|
|
Discordant Introns in Mouse and Human
A total of 12,866 human–mouse orthologous sequences were aligned to produce 144,545 aligned intron positions. They were grouped into 112,739 conserved, 31,748 ambiguous, and 58 discordant positions. By comparing with orthologous positions in dog, 56 discordant positions were inferred to be the result of 51 losses and 2 gains in mouse and 3 losses in human. The remaining 2 discordant positions lacked convincing orthologs in dog. Manual inspection and database searches revealed that one of the mouse gains (in znf447 gene) is an insertion of a 1,056-bp coding sequence (also present in rat but 1,293-bp long) and not an intron. The other gain is a polymorphic insertion of an long terminal repeats (LTR) retrotransposon into the wiz gene. This insertion is present only in C57BL/6J (whose genome sequence was analyzed in this study) and C57BR/cdJ strains of mice, and there is no expressed sequence tag (EST) evidence to show that it is spliced out. Furthermore, reverse transcriptase–polymerase chain reaction analysis has shown that the insertion has adversely affected the stability of the transcripts (Baust et al. 2002
). Thus, we do not consider it as a real intron. The details of the discordant introns in mouse and human are given in supplementary table 2 (Supplementary Material online), and the marked alignments of their sequences are given in supplementary figure 2 (Supplementary Material online). Of the 51 introns lost in mouse, 31 are also lost in the rat, indicating that 20 introns were lost in the mouse lineage after it split from the rat lineage. The 3 introns lost in the human lineage are absent in the chimpanzee and macaque genomes. Thus, all 3 losses have occurred prior to the divergence of the human lineage from the Old World monkeys about 23 MYA. The number of intron losses in mouse and human identified in our study is similar to that reported by Coulombe-Huntington and Majewski (4 losses in human and 63 losses in mouse since they diverged from a common ancestor) who used a different data set (RefSeq) of human and mouse sequences and a different strategy for identifying discordant introns (Coulombe-Huntington and Majewski 2007).
Number of Losses in the 4 Lineages
The number of losses in Tetraodon is significantly higher than that in fugu and human lineages (chi-squared value, 13.45 and 61.31; degrees of freedom [df] = 1; P = 
10–4 and 10–15, respectively) but not significantly higher than that in mouse (chi-squared value, 2.82; df = 1; P = 0.0931). The numbers of losses in fugu and mouse are significantly higher than that in the human lineage (chi-squared value, 21.28 and 42.68; df = 1; P = 10–6 and 10–11, respectively), whereas the number of losses in mouse is significantly higher than in fugu (chi-squared value, 4.93; df = 1; P = 0.02641). The rates of intron loss in the 4 vertebrate lineages equate to a rate of 4.32 x 10–7 to 1.94 x 10–5 intron loss per intron per Myr, with the Tetraodon lineage experiencing the highest rate of intron loss (table 1).
Patterns of Intron Loss
The mean sizes of existing introns corresponding to the introns lost in fugu and Tetraodon (113 bp) and to the introns lost in mouse and human (764 bp) are significantly shorter (Student's t-test; pufferfishes, t = –46.98, 2-tailed P < 10–67, df = 94; mammals, t = –23.6, 2-tailed P < 10–30, df = 56) than the overall mean size of introns in pufferfishes (620 bp) and in human and mouse (5,086 bp), respectively (table 2). Indeed, the majority of introns corresponding to lost introns in pufferfishes (64 out of 81; 79%) are shorter than 120 bp and more than two-thirds of lost human and mouse introns (37 out of 54) are shorter than 300 bp. Thus, the mechanism of intron loss appears to preferentially target smaller introns. The loss of short introns is consistent with loss by recombination because recombination between cDNA and genomic DNA is likely to be more efficient if loops of short introns are involved (Roy et al. 2003; Cho et al. 2004). Of the 17 introns lost in human and mouse that were larger than 300 bp, 8 were adjacent introns in a single gene (mouse Kin17) and were likely to have been lost simultaneously by recombination with cDNA. All losses in the 4 vertebrates, with the exception of 2 in pufferfishes, were precise, which is also consistent with loss by recombination.
|
The frequencies of phases of introns lost in pufferfishes were not significantly different from the expected frequencies (table 2). However, the fraction of phase 0 introns lost in human and mouse was higher than that expected (35 against 25 expected; chi-squared value, 4.12; df = 1; P = 0.0424) (table 2). A similar higher-than-expected loss of phase 0 introns has been previously reported for several eukaryote genomes (Roy and Gilbert 2005
). Its significance remains unknown.
The mechanism of intron loss by recombination with cDNA predicts an excess loss of introns from the 3'-end of genes (Sverdlov et al. 2004; Roy and Gilbert 2005). The introns deleted in pufferfishes show neither a 5'- nor a 3'-bias (chi-squared value, 2.79; df = 1; P = 0.095). However, an excess of introns deleted in human and mouse was located in the 3'-portion of genes (chi-squared value, 5.57; df = 1; P = 0.0183) (table 3), supporting intron loss due to recombination with cDNA. Indeed, the loss of eight 3'-most adjacent introns from the mouse Kin17 gene, in which the four 5'-most introns are still intact, is strong evidence for the loss of introns due to recombination with cDNA. Previously, simultaneous loss of 10 adjacent introns has been reported in a putative RNA helicase gene of the fungus, Cryptococcus neoformans var. grubii (Stajich and Dietrich 2006). However, the eight 5'-most and three 3'-most introns of this gene are still intact (Stajich and Dietrich 2006). Such a pattern of loss indicates that the middle region of a cDNA was involved in the recombination rather than the 3'-end region.
|
Loss of introns due to recombination with cDNA is predicted to result occasionally in the simultaneous loss of adjacent introns. With the exception of mouse Kin17 gene that has experienced loss of 8 introns, the majority of intron losses in the 4 vertebrates are individual losses. Loss of individual introns has also been found to be more frequent than loss of adjacent introns in Caenorhabditis elegans and some other genomes (Robertson 1998
, 2000; Wada et al. 2002). One explanation for excess loss of individual introns is that short single introns might be targeted for recombination because genomic DNA–cDNA alignment and recombination would be more efficient (Roy et al. 2003; Cho et al. 2004). Alternatively, single introns might be lost due to deletion of intronic sequences from the genomic DNA (Robertson 1998).
An interesting observation with regard to the introns lost in pufferfishes is that 2 of them have been deleted imprecisely. One intron deleted in Tetraodon has left behind 3 nucleotides, whereas an intron lost in fugu has left behind 6 nucleotides, thereby adding 1 and 2 codons, respectively, to the original exons (fig. 3). Indeed, the 6 nucleotides left behind in the fugu gene are identical to the 6 nucleotides at the 5'-end of the intact intron in Tetraodon, indicating that the sequence left behind is most likely from the original intron in fugu. Such imprecise losses of introns rules out the possibility of intron loss through recombination involving an mRNA intermediate because such a mechanism will result in the precise deletion of the entire sequence of the intron. The additional sequences left behind following the loss of introns in the pufferfish genes are consistent with a model of loss through genomic deletion at the DNA level. Genomic deletions of intronic sequences can occur either through nonhomologous recombination mediated by the occurrence of short direct repeats at the 3'-end of an exon and the 3'-end of the succeeding intron as proposed by (Robertson 1998) or due to spontaneous genomic deletions. The latter is more likely to result in an imprecise deletion of introns, as is the case with the 2 introns lost in pufferfishes.
|
Functions and Expression Patterns of Genes with Discordant Introns
Statistical analysis of gene ontology (GO) terms using the GOstat package (Beissbarth and Speed 2004
) showed that no particular GO term was overrepresented among the genes that lost introns. The mechanism of intron loss by recombination with cDNA predicts that the gene is expressed in the germ line. Because expression patterns of most pufferfish genes are unknown, we analyzed expression patterns of only mammalian genes. Out of 46 mammalian genes with losses, associated UniGene entries were available for 28 genes. Only 16 of these genes are expressed ( > 10 transcripts per million ESTs) in the ovary or testis. The remaining 12 genes might have originally expressed in the germ line and then lost the germ line expression subsequent to the loss of introns. An alternative explanation is that introns in these genes might be lost by mechanisms other than recombination such as the deletion of genomic DNA. |
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsSupplementary MaterialAcknowledgementsReferences |
|---|
Losses versus Gains
By comparing 13,547 pairs of fugu–Tetraodon genes and 12,866 pairs of human–mouse genes, we uncovered 81 instances of intron loss and no instance of intron gain in the 2 pufferfishes and 54 instances of intron loss and no gain in the human and mouse lineages. The implied overall rates of intron loss in the pufferfishes (0.0815% and 0.194% introns per 100 Myr) are comparable with that in the human and mouse lineages (0.004% and 0.0735% introns per 100 Myr) (table 1) but considerably lower than the previous estimates of 0.3–3.2% loss per 100 Myr for apicomplexan parasites (Plasmodium and Theileria) (Roy and Hartl 2006
; Roy and Penny 2006a), basidomycete fungi (Cryptococcus) (Stajich and Dietrich 2006), euascomycetous fungi (Neurospora, Magnaporthe, and Fusarium) (Nielsen et al. 2004), and plants (Arabidopsis and rice) (Lin et al. 2006; Roy and Penny 2007a). We failed to identify intron gain either in fugu or Tetraodon lineage or in human and mouse lineages. It should be noted that no intron gains were identified in the genomes of Plasmodium and Cryptococcus fungi (Roy and Hartl 2006; Stajich and Dietrich 2006). The only convincing instances of intron gains have been identified in genome-wide comparisons of Entamoeba histolytica (2.2 x 10–4 intron per gene per Myr) (Roy et al. 2006), euascomycetous fungi (3 to 6 x 10–4 intron per gene per Myr) (Nielsen et al. 2004), nematode worm (1 x 10–5 intron per gene per Myr) (Roy and Penny 2006b), Arabidopsis (1.3 to 3.9 x 0–4 intron per gene per Myr) (Roy and Penny 2007a), and rice (1.3 x 10–4 intron per gene per Myr) (Lin et al. 2006).
Overall, our findings indicate that very little intron turnover has occurred during the recent evolution of pufferfishes and mammals. Furthermore, although intron losses were detected in the mammalian lineages, the human lineage does not seem to have experienced any intron loss since it diverged from the Old World monkeys about 23 MYA. All 3 losses identified in the human lineage occurred prior to the divergence of the Old World monkeys. This indicates that exon–intron structures of primates have been virtually static since they diverged from the Old World monkeys. It is also very striking that no convincing instances of intron gain were identified either in the fugu or Tetraodon lineage that diverged about 32 MYA or in the mouse and human lineage that diverged about 60 MYA. Previous analysis of human and mouse lineages using a different data set and/or different approaches had also failed to identify any intron gain in these lineages (Roy et al. 2003; Coulombe-Huntington and Majewski 2007). These studies together suggest that the intron-rich genomes of vertebrates are not permissive to the mechanism of intron gain. The previously reported instances of intron gain in vertebrates such as the 2 introns gained by RAG1 gene in teleost fishes (Venkatesh et al. 1999) are likely to be extremely rare events. As such, intron gains can serve as valuable molecular markers for unambiguously defining branch points and clades in vertebrate evolution.
Intron losses have been found to dominate over intron gains in most studies of closely related species of eukaryotes such as nematode worms (Robertson 1998, 2000; Cho et al. 2004; Kiontke et al. 2004; Roy and Penny 2006b), apicomplexan parasites (Plasmodium and Theileria) (Roy and Hartl 2006; Roy and Penny 2006a), Cryptococcus fungi (Stajich and Dietrich 2006), and euascomycete fungi (Nielsen et al. 2004). The same trend is also evident in the 2 pufferfishes and the human and mouse lineages, indicating that intron losses have continued to dominate over intron gain during the recent evolution of vertebrates. The 3 instances of parallel losses of introns in fugu and zebrafish lineages identified in our analysis provide further support that there is a continued loss of introns in different lineages of vertebrates. Nevertheless, it should be noted that the overall rates of loss are extremely low in these vertebrates. Furthermore, a vast majority of intron positions are conserved between pufferfishes and mammals (Roy et al. 2003). Thus, it can be concluded that most of the introns in the present day vertebrates already existed in the last common ancestor of mammals and fishes. This implies that the common ancestor of all bony vertebrates was also intron rich similar to the present day vertebrates. Most of the vertebrate introns are likely to be much more ancient than the bony vertebrate ancestor because about two-thirds of approximately 230 introns analyzed in a marine annelid, Platynereis dumerilii, were found to be conserved in human and fugu (Raible et al. 2005) and 82% of introns in the starlet sea anemone (Nematostella vectensis) are found in homologous positions in the human genome (Putnam et al. 2007). Thus, most of the introns in extant vertebrates are likely to be very ancient introns that have been retained because the ancestral eumetazoan genome.
Differences in Losses in the Pufferfish and Mammalian Lineages
Between the 2 pufferfishes, Tetraodon has experienced a significantly higher number of losses (57 losses) than fugu (24 losses). The 2 pufferfishes share most of the biological traits and their genomes are highly identical. One biological trait in which they show considerable difference is the generation time. The generation time of Tetraodon is much shorter (about 1 year) compared with fugu (3–4 years). The mechanism of intron loss by recombination predicts that the rate of loss depends on the overall rate of recombination (Roy and Gilbert 2006). Because most recombination in vertebrates appears to occur during meiosis, vertebrates with shorter generation time will experience more rounds of meiosis per unit time; hence, there are higher chances of intron losses (Jeffares et al. 2006). Comparisons of intron densities and generation times across a broad range of eukaryotes have shown that eukaryotes with a shorter generation time tend to have fewer introns than eukaryotes with a longer generation time. Such a trend has been hypothesized to be the result of selection against introns to facilitate rapid cell division, rapid gene expression, or a combination of both (Jeffares et al. 2006). It is therefore possible that the shorter generation time of Tetraodon compared with fugu could be a key factor responsible for the higher rate of intron loss in the Tetraodon lineage. Likewise, the shorter generation time of mouse (2 months) compared with human may be responsible for the significantly higher rate of intron loss in mouse (51 losses) compared with human (3 losses).
TEs are widely believed to be mediators of intron loss as well as intron gain in eukaryotes. TEs are the major source of reverse transcriptase that can reverse-transcribe transcripts and spliced introns and facilitate loss and gain of introns, respectively. TEs can also insert directly into coding sequences giving rise to novel introns, as in the case with short interspersed elements-derived novel introns in the maize Sh2 gene and the catalaseA gene of rice (Giroux et al. 1994; Iwamoto et al. 1998). Although TEs are less abundant in pufferfishes (<10%) compared with mammals (47%), pufferfishes contain a markedly higher diversity of TEs. Indeed, fugu and Tetraodon contain 23 and 16 clades of retrotransposons, respectively, compared with only 6 clades in human and mouse. Some of the retroelements found in pufferfishes are also present in invertebrates (e.g., Ty3/Gypsy) but absent or present only as "fossils" in human and mouse genomes (Volff et al. 2003). Furthermore, there is evidence that many pufferfish retroelements have been active recently (Bouneau et al. 2003; Fischer et al. 2005). This is in contrast to most of the retroelements in the human genome that have been inactivated through mutations. Thus, although pufferfishes contain a low copy number of TEs, the overall diversity and activity of TEs are considerably higher in pufferfishes than in mammals. The higher activity of retrotransposons in Tetraodon combined with its relatively shorter generation time may explain the higher rate of intron loss in the Tetraodon lineage. The higher activity of TE in pufferfishes also predicts a higher rate of intron gain compared with mammals. However, we failed to identify any intron gain either in fugu or in Tetraodon. Thus, pufferfishes seem to be resistant to the TE-mediated mechanism of intron gain.
Evidence for an Unusual Mechanism of Intron Loss
The most common hypothesis for the loss of introns is by recombination with a reverse-transcribed copy of a transcript. Such a mechanism of intron loss predicts that introns are lost precisely, lost introns tend to be located at the 3'-end of genes, and adjacent introns will have a higher probability of being lost together. The intron losses identified in most of the previous studies, as well as in our present study, are consistent with some of these predictions and thus are hypothesized to be lost by recombination. However, we have identified 2 rare instances of intron loss, one in fugu and another in Tetraodon, that are unlikely to have been due to recombination. The 2 introns have been deleted incompletely; thus, the losses are not the result of recombination with a reverse-transcribed copy of a transcript. Only one such instance of imprecise loss of intron has been reported previously. An intron lost in only one allele of the jingwei gene of Drosophila teissieri resulting in an intron presence–absence polymorphism was found to have left behind 12 nucleotides from the 3'-end of the original intron (Llopart et al. 2002). Such imprecise losses of introns strongly suggest that they were lost due to spontaneous genomic deletions. Spontaneous deletions of noncoding sequences involving up to a few hundred nucleotides are known to occur in genomes of eukaryotes such as Drosophila (Petrov et al. 1996; Petrov and Hartl 1998) and C. elegans (Robertson 2000). Such deletions have the potential to cause complete or near complete deletion of introns, particularly the smaller introns, in the genome. More importantly, such a mechanism of intron loss does not require the genes to be expressed at high levels in the germ line and can target introns of any gene in the genome. Thus, the loss of introns by this mechanism may be more common than previously thought.
Compaction of Pufferfish Genome and the Dynamics of Intron Loss and Gain
Because the genomes of pufferfishes are the smallest among vertebrates, a common ancestor of fugu and Tetraodon must have experienced an extreme streamlining process resulting in a significant reduction in the amount of noncoding sequence. Our genome-wide comparisons of intron positions in fugu, Tetraodon, and human have shown that a majority of intron positions are conserved in pufferfishes and humans and that even though pufferfishes have experienced intron losses during recent evolution, the rate of intron loss is extremely low. These data suggest the compaction of the ancestral pufferfish genome was not accomplished by a higher rate of loss of entire introns. The compaction seems to have been accomplished mainly by a reduction in the length of introns (and intergenic regions). Previous analyses of indels in the genomes of insects have shown that insects with smaller genomes such as Drosophila experience a higher rate of DNA loss than larger genomes (Petrov et al. 1996, 2000). The smaller genomes of pufferfishes also seem to experience a higher rate of DNA loss. Comparisons of patterns of substitutions, insertions, and deletions of pseudogenes in Tetraodon and mammals have indicated that nonessential DNA is subjected to a higher rate of deletions in Tetraodon than in human and mouse (Dasilva et al. 2002). These findings have led to the hypothesis that a higher rate of DNA deletion may be responsible for maintaining the genome size of pufferfishes small (Dasilva et al. 2002). The 2 instances of imprecise deletion of introns identified by us in pufferfishes could be the result of a higher rate of DNA deletion experienced by the pufferfish genomes.
|
Conclusions |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsSupplementary MaterialAcknowledgementsReferences |
|---|
We have shown that intron losses have occurred in 2 closely related species of pufferfishes (fugu and Tetraodon) since they diverged about 32 MYA. However, no intron gain was identified in either lineage. Comparison of human and mouse genomes also identified only losses and no gain, indicating that intron gain is an extremely rare event in vertebrate evolution. The overall rates of intron loss in the 2 pufferfishes, mouse, and human are extremely low. Thus, most of the introns in the extant vertebrates are likely to be of ancient origin. The significantly higher rate of intron loss in Tetraodon compared with fugu could be related to its shorter generation time. The imprecise loss of 2 introns in pufferfishes suggests that they were lost due to genomic deletion. This mechanism of intron loss may be a common mechanism of intron loss particularly in small genomes that experience a higher rate of loss of nonessential DNA.
|
Supplementary Material |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsSupplementary MaterialAcknowledgementsReferences |
|---|
Supplementary figures 1 and 2 and tables 1 and 2 are available at Molecular Biology and Evolution online (http://www.mbe.oxfordjournals.org/).
|
Acknowledgements |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsSupplementary MaterialAcknowledgementsReferences |
|---|
We thank Genoscope, France for supplying a Tetraodon BAC clone and Laszlo Orban for providing Tetraodon genomic DNA. Thanks to zebrafish, medaka, and stickleback genome project groups for making available the genome assemblies for comparative analysis. We also thank Patrick Gilligan for helpful comments on the manuscript. We would also like to thank 3 anonymous reviewers for their critical comments on earlier versions of this manuscript. This project was supported by the Biomedical Research Council of A*STAR (Agency for Science, Technology and Research), Singapore. B.V. is an adjunct staff member of the Department of Pediatrics, Yong Loo Lin School of Medicine, National University of Singapore.
|
Footnotes |
|---|
Barbara Holland, Associate Editor
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsSupplementary MaterialAcknowledgementsReferences |
|---|
Aparicio S, Chapman J, Stupka E, et al, (41 co-authors). Whole-genome shotgun assembly and analysis of the genome of Fugu rubripes. Science (2002) 297:1301–1310.
Babenko VN, Rogozin IB, Mekhedov SL, Koonin EV. Prevalence of intron gain over intron loss in the evolution of paralogous gene families. Nucleic Acids Res (2004) 32:3724–3733.
Baust C, Baillie GJ, Mager DL. Insertional polymorphisms of ETn retrotransposons include a disruption of the wiz gene in C57BL/6 mice. Mamm Genome (2002) 13:423–428.[CrossRef][Web of Science][Medline]
Beissbarth T, Speed TP. GOstat: find statistically overrepresented Gene Ontologies within a group of genes. Bioinformatics (2004) 20:1464–1465.
Benton MJ, Donoghue PC. Paleontological evidence to date the tree of life. Mol Biol Evol (2007) 24:26–53.
Bouneau L, Fischer C, Ozouf-Costaz C, Froschauer A, Jaillon O, Coutanceau JP, Korting C, Weissenbach J, Bernot A, Volff JN. An active non-LTR retrotransposon with tandem structure in the compact genome of the pufferfish Tetraodon nigroviridis. Genome Res (2003) 13:1686–1695.
Carmel L, Wolf YI, Rogozin IB, Koonin EV. Three distinct modes of intron dynamics in the evolution of eukaryotes. Genome Res (2007) 17:1034–1044.
Cavalier-Smith T. Selfish DNA and the origin of introns. Nature (1985) 315:283–284.[Medline]
Cavalier-Smith T. Intron phylogeny: a new hypothesis. Trends Genet (1991) 7:145–148.[Web of Science][Medline]
Cho S, Jin SW, Cohen A, Ellis RE. A phylogeny of caenorhabditis reveals frequent loss of introns during nematode evolution. Genome Res (2004) 14:1207–1220.
Collins L, Penny D. Complex spliceosomal organization ancestral to extant eukaryotes. Mol Biol Evol (2005) 22:1053–1066.
Coulombe-Huntington J, Majewski J. Characterization of intron loss events in mammals. Genome Res (2007) 17:23–32.
Crick F. Split genes and RNA splicing. Science (1979) 204:264–271.
Dasilva C, Hadji H, Ozouf-Costaz C, Nicaud S, Jaillon O, Weissenbach J, Roest Crollius H. Remarkable compartmentalization of transposable elements and pseudogenes in the heterochromatin of the Tetraodon nigroviridis genome. Proc Natl Acad Sci USA (2002) 99:13636–13641.
Fink GR. Pseudogenes in yeast? Cell (1987) 49:5–6.[CrossRef][Web of Science][Medline]
Fischer C, Bouneau L, Coutanceau JP, Weissenbach J, Ozouf-Costaz C, Volff JN. Diversity and clustered distribution of retrotransposable elements in the compact genome of the pufferfish Tetraodon nigroviridis. Cytogenet Genome Res (2005) 110:522–536.[CrossRef][Medline]
Giroux MJ, Clancy M, Baier J, Ingham L, McCarty D, Hannah LC. De novo synthesis of an intron by the maize transposable element Dissociation. Proc Natl Acad Sci USA (1994) 91:12150–12154.
Hankeln T, Friedl H, Ebersberger I, Martin J, Schmidt ER. A variable intron distribution in globin genes of Chironomus: evidence for recent intron gain. Gene (1997) 205:151–160.[CrossRef][Web of Science][Medline]
Iwamoto M, Maekawa M, Saito A, Higo H, Higo K. Evolutionary relationship of plant catalase genes inferred from intron-exon structures: isozyme divergence after the separation of monocots and dicots. Theor Appl Genet (1998) 97:9–19.[CrossRef][Web of Science]
Jaillon O, Aury JM, Brunet F, et al, (61 co-authors). Genome duplication in the teleost fish Tetraodon nigroviridis reveals the early vertebrate proto-karyotype. Nature (2004) 431:946–957.[CrossRef][Medline]
Jeffares DC, Mourier T, Penny D. The biology of intron gain and loss. Trends Genet (2006) 22:16–22.[CrossRef][Web of Science][Medline]
Kiontke K, Gavin NP, Raynes Y, Roehrig C, Piano F, Fitch DH. Caenorhabditis phylogeny predicts convergence of hermaphroditism and extensive intron loss. Proc Natl Acad Sci USA (2004) 101:9003–9008.
Lin H, Zhu W, Silva JC, Gu X, Buell CR. Intron gain and loss in segmentally duplicated genes in rice. Genome Biol (2006) 7:R41.[CrossRef][Medline]
Llopart A, Comeron JM, Brunet FG, Lachaise D, Long M. Intron presence-absence polymorphism in Drosophila driven by positive Darwinian selection. Proc Natl Acad Sci USA (2002) 99:8121–8126.
Nguyen HD, Yoshihama M, Kenmochi N. New maximum likelihood estimators for eukaryotic intron evolution. PLoS Comput Biol (2005) 1:e79.[CrossRef][Medline]
Nielsen CB, Friedman B, Birren B, Burge CB, Galagan JE. Patterns of intron gain and loss in fungi. PLoS Biol (2004) 2:e422.[CrossRef][Medline]
Nixon JE, Wang A, Morrison HG, McArthur AG, Sogin ML, Loftus BJ, Samuelson J. A spliceosomal intron in Giardia lamblia. Proc Natl Acad Sci USA (2002) 99:3701–3705.
Palmer JD, Logsdon JM Jr. The recent origins of introns. Curr Opin Genet Dev (1991) 1:470–477.[CrossRef][Medline]
Petrov DA, Hartl DL. High rate of DNA loss in the Drosophila melanogaster and Drosophila virilis species groups. Mol Biol Evol (1998) 15:293–302.[Abstract]
Petrov DA, Lozovskaya ER, Hartl DL. High intrinsic rate of DNA loss in Drosophila. Nature (1996) 384:346–349.[CrossRef][Medline]
Petrov DA, Sangster TA, Johnston JS, Hartl DL, Shaw KL. Evidence for DNA loss as a determinant of genome size. Science (2000) 287:1060–1062.
Putnam NH, Srivastava M, Hellsten U, et al, (20 co-authors). Sea anemone genome reveals ancestral eumetazoan gene repertoire and genomic organization. Science (2007) 317:86–94.
Qiu WG, Schisler N, Stoltzfus A. The evolutionary gain of spliceosomal introns: sequence and phase preferences. Mol Biol Evol (2004) 21:1252–1263.
Raible F, Tessmar-Raible K, Osoegawa K, et al, (12 co-authors). Vertebrate-type intron-rich genes in the marine annelid Platynereis dumerilii. Science (2005) 310:1325–1326.
Robertson HM. Two large families of chemoreceptor genes in the nematodes Caenorhabditis elegans and Caenorhabditis briggsae reveal extensive gene duplication, diversification, movement, and intron loss. Genome Res (1998) 8:449–463.
Robertson HM. The large srh family of chemoreceptor genes in Caenorhabditis nematodes reveals processes of genome evolution involving large duplications and deletions and intron gains and losses. Genome Res (2000) 10:192–203.
Rogers JH. How were introns inserted into nuclear genes? Trends Genet (1989) 5:213–216.[CrossRef][Web of Science][Medline]
Roy SW, Fedorov A, Gilbert W. Large-scale comparison of intron positions in mammalian genes shows intron loss but no gain. Proc Natl Acad Sci USA (2003) 100:7158–7162.
Roy SW, Gilbert W. The pattern of intron loss. Proc Natl Acad Sci USA (2005) 102:713–718.
Roy SW, Gilbert W. The evolution of spliceosomal introns: patterns, puzzles and progress. Nat Rev Genet (2006) 7:211–221.[CrossRef][Web of Science][Medline]
Roy SW, Hartl DL. Very little intron loss/gain in Plasmodium: intron loss/gain mutation rates and intron number. Genome Res (2006) 16:750–756.
Roy SW, Irimia M, Penny D. Very little intron gain in Entamoeba histolytica genes laterally transferred from prokaryotes. Mol Biol Evol (2006) 23:1824–1827.
Roy SW, Penny D. Large-scale intron conservation and order-of-magnitude variation in intron loss/gain rates in apicomplexan evolution. Genome Res (2006a) 16:1270–1275.
Roy SW, Penny D. Smoke without fire: most reported cases of intron gain in nematodes instead reflect intron losses. Mol Biol Evol (2006b) 23:2259–2262.
Roy SW, Penny D. Patterns of intron loss and gain in plants: intron loss-dominated evolution and genome-wide comparison of O. sativa and A. thaliana. Mol Biol Evol (2007a) 24:171–181.
Roy SW, Penny D. A very high fraction of unique intron positions in the intron-rich diatom Thalassiosira pseudonana indicates widespread intron gain. Mol Biol Evol (2007b) 24:1447–1457.
Stajich JE, Dietrich FS. Evidence of mRNA-mediated intron loss in the human-pathogenic fungus Cryptococcus neoformans. Eukaryot Cell (2006) 5:789–793.
Sverdlov AV, Babenko VN, Rogozin IB, Koonin EV. Preferential loss and gain of introns in 3' portions of genes suggests a reverse-transcription mechanism of intron insertion. Gene (2004) 338:85–91.[CrossRef][Web of Science][Medline]
Venkatesh B, Ning Y, Brenner S. Late changes in spliceosomal introns define clades in vertebrate evolution. Proc Natl Acad Sci USA (1999) 96:10267–10271.
Volff JN, Bouneau L, Ozouf-Costaz C, Fischer C. Diversity of retrotransposable elements in compact pufferfish genomes. Trends Genet (2003) 19:674–678.[CrossRef][Web of Science][Medline]
Wada H, Kobayashi M, Sato R, Satoh N, Miyasaka H, Shirayama Y. Dynamic insertion-deletion of introns in deuterostome EF-1alpha genes. J Mol Evol (2002) 54:118–128.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. Hussain, L. R. Saraiva, and S. I. Korsching Positive Darwinian selection and the birth of an olfactory receptor clade in teleosts PNAS, March 17, 2009; 106(11): 4313 - 4318. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||