Mechanisms of intron gain and loss in Cryptococcus
© Sharpton et al.; licensee BioMed Central Ltd. 2008
Received: 6 September 2007
Accepted: 30 January 2008
Published: 30 January 2008
Genome comparisons across deep phylogenetic divergences have revealed that spliceosomal intron gain and loss are common evolutionary events. However, because of the deep divergences involved in these comparisons, little is understood about how these changes occur, particularly in the case of intron gain. To ascertain mechanisms of intron gain and loss, we compared five relatively closely related genomes from the yeast Cryptococcus.
We observe a predominance of intron loss over gain and identify a relatively slow intron loss rate in Cryptococcus. Some genes preferentially lose introns and a large proportion of intron losses occur in the middle of genes (so called internal intron loss). Finally, we identify a gene that displays a differential number of introns in a repetitive DNA region.
Based the observed patterns of intron loss and gain, population resequencing and population genetic analysis, it appears that recombination causes the widely observed but poorly understood phenomenon of internal intron loss and that DNA repeat expansion can create new introns in a population.
The evolutionary dynamics of spliceosomal introns has remained a subject of considerable debate since their discovery over 30 years ago . Although the dynamic nature of gene structure is well documented - intron position is not always conserved between orthologs - the frequency of intron gain and loss and the mechanisms behind these processes remain controversial subjects. Indeed, single gene analyses have identified examples of both gain and loss, but such single case studies are limited in scope and, consequently, trends and patterns are difficult to identify. With comparative genomics, many orthologs can be evaluated simultaneously. Thus, enough events can be characterized to identify trends and to infer mechanisms pertinent to intron evolution. Several studies have employed such techniques across deep phylogenetic distances to evaluate the relative contributions of intron gain and loss [2–4]. While successful at demonstrating the plasticity of eukaryotic gene structure evolution, the long times for evolutionary events associated with comparisons of deep divergence can confound estimation of the true number of events and elucidation of mechanisms responsible for these events . Comparative analysis of recently diverged taxa may improve estimation of the number of intron gains and losses and identify the mechanisms responsible for the evolution of this aspect of gene structure. In this regard, the Cryptococcus species complex offers the ideal system to study the mechanisms of intron gain and loss.
Estimation of divergence time
In a previous study, Xu et al.  used a four-gene phylogeny to estimate that C. neoformans and C. gatii diverged roughly 37 million years ago. We evaluated this prediction by estimating the silent substitution rate (dS) from all aligned orthologs from the five available Cryptococcus genomes. Pairwise comparisons between serotypes A/D and B identified dS estimates of approximately 0.32. Previous work found that the Eurotiomycete neutral mutation rate ranges from 1e-8 to 1e-9 mutations per year depending on the method of rate calculation . Thus, assuming midpoint rooting and a molecular clock rate of evolution similar to the Eurotiomycetes, we estimate that C. neoformans and C. gatii diverged between 16 and 160 million years ago. Employing a commonly applied neutral mutation rate of 2 × 10-9 suggests that the two species diverged 80 million years ago, doubling previous estimates that applied these same assumptions [7, 11]. However, because of the difficulty of equating sequence divergence with time, we report all rate calculations and comparisons in terms of this date approximation as well as dS. For the purposes of identifying fixed structural changes by parsimony, we consider D1 and D2 to be the same species, and, therefore, consider the branch subtending these two strains to be a terminal branch. We calculate the total terminal branch length to be 0.29 substitutions per silent site, and use this figure to calculate the rate of observed gene structure changes on terminal branches.
Intron loss predominates
The reference strain, D1, has an intron-dense genome with 36,275 introns across 6,954 genes (average of 5.5 introns per gene). Cryptococcus thus exhibits a density more similar to humans (7.8) than to other fungi, such as Saccharomyces cerevisiae (0.05), Neurospora crassa (1.77) or Ustilago maydis (0.75) (from ). This relative intron abundance may be driven by intron gain, a rarely observed phenomenon across such short phylogenetic distances. Thus, we sought to determine the relative contribution of intron gain and loss in the evolution of Cryptococcus gene structure.
Some genes preferentially lose introns
Introns have been popular phylogenetic markers under the assumption that gains and losses are relatively rare and unlikely to be the result of parallel evolution . However, analysis of the Dipteran white gene warned that some genes might be prone to parallel gene structure changes . Additionally, studies have shown that some highly expressed genes are prone to losing introns and may be evolving more rapidly than previously assumed . We examined Cryptococcus gene structure evolution to identify genes that rapidly undergo intron gain or loss. Because of the phylogenetic structure of the taxa involved in this study and the restriction of the analysis to serotype D introns, it is impossible to resolve true parallel loss of an intron in serotype A and in serotype B taxa (as it would appear to be a gain in the serotype D clade). Thus, we searched for loci that lost different introns along multiple branches. Assuming that the 49 gene structure changes are all losses, the probability that a gene will lose an intron is 1.5e-3. Assuming, then, that there are five introns in a gene, the probability of a single gene losing two independent introns is 1.13e-5. Because such events are unlikely to occur by chance alone, the identification of such events suggests that some genes preferentially lose introns. We identified two loci that appear to rapidly evolve their gene structure, losing single, different introns along independent lineages. CNE00990, a ubiquinone biosynthesis-related protein that ancestrally exhibited 5 introns, lost its first intron in B1 and its fifth intron in A. Similarly, CNF02410, a cytoplasmic protein containing five ancestral introns, lost its second intron in A and its fourth intron in B1. Both loci are single copy in the genome and are supported with ESTs. If we treat the adjacent intron losses observed as single loss events, these relatively rapid gene structure changes constitute 11.1-22.2% (depending on whether the ambiguous events are true losses) of the losses observed. Though the sample size is small, this suggests that a sizable number of gene structure changes in Cryptococcus may not be stochastic, and may help to explain why parsimony-based inferences of intron loss over larger evolutionary distances tend to underestimate parallel losses and, consequently, overestimate intron gains .
Evaluation of a common intron loss mechanism
First, we investigated adjacent intron loss by searching for multiple, adjacent intron losses and then determined the likelihood that the losses were due to a single evolutionary event. Some genes have lost more than one intron, as evidenced by the observation that 31 parsimoniously explained intron losses occurred in only 16 loci. Specifically, genes CNI01550, CNN02320, and CNA01350 unambiguously demonstrate multiple, consecutive intron losses (Additional data file 1). Genes CNK02730 and CNB02430 also have consecutive introns missing in orthologs, but it is uncertain if these changes are the result of intron loss. In their study of fungal introns, Stajich and Dietrich  employed a comparative method that resolved the events in CNK02730 as losses. Here, given the relative rarity of intron gain and the lack of any clear mechanism that would enable multiple adjacent gains, we treat the changes in these two genes as losses. However, without the necessary phylogenetic support, characterization of these particular events is not conclusive.
Multiple adjacent intron loss may be caused by a single evolutionary event (that is, the spliced mRNA based model) or by multiple, independent losses. To distinguish between these hypotheses, we generated a loss distribution for each gene demonstrating adjacent loss via a resampling analysis under the assumption that losses were random and independent across a gene (see Materials and methods). These distributions were used to test the null hypothesis that the observed losses were the result of independent loss events within the gene (Additional data file 1). This analysis thus evaluates how intron loss is distributed across the length of a gene regardless of that gene's intron loss rate. In every gene except CNA03150, the null hypothesis could be rejected, supporting the hypothesis that one event accounted for all intron losses within a single gene. This observation of simultaneous adjacent loss suggests that an mRNA mediated loss mechanism exists in Cryptococcus.
Identification of a polymorphic, gained intron
We found no clear examples of completely new D1 introns; however, we did find a D1 gene with additional identical introns compared to A and D2. By submitting every intron to a BLASTn homology search against the rest of the genome, we identified the gene CNN00420, a single copy, six-intron gene in D1 in which introns 3, 4 and 5 are identical and differ from intron 2 by a single point mutation; intron 6 has additional 3' sequence. The introns in this gene are supported by unique ESTs, indicating that the gene is functional and that splicing occurs at the canonical splice sites (GT|AG). The orthology detection method employed in the previously discussed comparative analysis failed to detect this locus due to the lack of identifiable orthologs in the serotype B genomes. By relaxing this criterion, orthologs were identified in A (CNAG_06312 ) and D2 (GenBank: CNBN0400). The subsequent alignment identified two D1 intron sequences (introns 3 and 5) that correspond to gaps in the orthologs, implicating gains in the D1 lineage. Examination of CNN00420 reveals two types of repeats in the gene: short, internal exonic repeats (approximately 39 nucleotides in length) and longer repeats that include the ends of exons as well as a full length intron (approximately 107 nucleotides in length) (Additional data file 4). The alignment between the D1 and D2 orthologs indicates that the creation of extra copies of this longer repeat was responsible for the intron gains in the D1 sequence.
To rule out genome sequencing and contig assembly artifacts, we resequenced CNN00420 from the Cryptococcus serotype D population (strains TP0603, WM629, JEC20 (sib to D1), B3502 (sib to D1 and parent to D2) and genomic strains D2 and D1). Our new sequence verified that the genome sequences are correct and showed that the intron in D1 is shared with sib JEC20 and parent B3502 (Additional data file 5). Sequence from the additional individuals demonstrates a differential number of D1 intron containing repeats among the strains, suggesting that an expansion or contraction in the number of repeats accounts for the unique sequence in D1 (Additional data file 6). Because A lacks the introns and because the introns lay within a repeat, the most parsimonious explanation for the unique intron sequence is that it was gained by duplication in D1.
Cryptococcus genes are more intron-dense than those of other fungi , suggesting that Cryptococcus either gained introns at a faster rate or lost introns at a slower rate than other fungi. A screen of 33,473 introns across five taxa identified 31 intron losses and no intron gains. Because of a lack of phylogenetic resolution, an additional 18 gene structure changes could not be unequivocally attributed to intron gain or loss. While some of these unresolved events could be gains, a previous analysis of Cryptococcus gene structure evolution that screened for intron gain in the serotype A and B lineages also found no clear cases of gain . If introns have been gained since Cryptococcus diverged, then they have done so at a much lower rate than losses or in such a way that makes detection difficult. Intron loss appears to be a rare event (0.09% of introns evaluated were lost) in Cryptococcus, occurring, under conditions assuming the most possible loss, at a rate of 1.04 × 10-11 introns per year (5.2 × 10-3 time dS), or 4.33 × 10-12 introns per year (2.17 × 10-3) if adjacent losses are single events. Conversely, the intron loss rate in S. pombe is 2 × 10-9 introns per year . It should be noted that the S. pombe rate, the only other intron loss rate calculated in fungi, was estimated by comparing S. pombe to the Opisthikont ancestor (estimated divergence approximately 1,500 million years ago). As the rate of intron loss may not be constant, there could have been moments during the much longer period assessed for S. pombe that matched the rate that we calculated over 80 million years of Cryptococcus evolution. These results suggest that the relatively large intron density of Cryptococcus compared to other fungi is not the result of a faster rate of intron gain, but rather a slower rate of intron loss.
Not all genes in the genome, however, appear to lose introns at the same rate. We identified two loci that have lost different introns along multiple, independent lineages. These gene structure changes are unlikely the result of chance and may be driven by some gene specific bias to lose introns. Whatever the cause, this finding suggests that introns may be lost in parallel over relatively recent time scales. This should raise caution about using an intron as a phylogenetic marker, a relatively common technique in evolutionary analysis. Further studies of such genes may help identify any selective pressures associated with relatively unstable gene structures.
We identified five genes that lost several adjacent introns, a pattern indicative of a spliced mRNA-intermediate intron loss mechanism (Figure 3a) . If reverse transcriptase creates cDNA by initiating at the poly-A tail of the transcript and if homologous recombination between the ends of the spliced cDNA and the intron-containing gene causes intron loss, then we would expect to identify a bias towards 3' intron loss and a single loss event should frequently remove multiple introns and always include the 3' terminal intron in the loss. We do observe a bias towards 3' intron loss (Figure 3a), but when we evaluate where intron losses occur, they are most frequently internal; we rarely observe the loss of the 3' terminal intron. While it may be argued that natural selection may favor 3' intron retention, studies on this subject indicate otherwise, observing only a 5' intron retention bias . It is noteworthy that Cryptococcus is not the only organism in which internal intron loss predominates. A previous analysis found that extraordinarily long exons, which result from intron loss, accumulate internally in the genes of many Eukaryotes  and a survey of Ascomycete gene structure evolution identified an internal intron loss bias . Given the diversity of these observations, a general mechanism of internal intron loss should be considered.
Several mechanisms could explain internal intron loss via a spliced mRNA intermediate. First, the points of homologous recombination between the cDNA and the gene may occur upstream of the 3' most intron, resulting in its persistence by being excluded from the double cross-over event. Given that the mechanism proposed for intron loss involving mRNA is the same as that proposed for the origin of processed pseudogenes, if internal recombination was common, one would expect to find processed pseudogenes that are internal segments of genes. However, to our knowledge, there are few examples of processed pseudogenes that are internal gene segments . Another point arguing against internal recombination is the model that the ends of cDNA molecules, and not internal regions, which necessarily exclude the 3' intron, promote cross-over through double strand break machinery . In the same publication, Hu proposes a different model for internal intron loss. Here, intron loss could occur when an intron experiences a double strand break and if cDNA is used to repair the damage. However, this model would result only in single intron losses and could not explain our observations of adjacent intron loss.
Perhaps the mechanism with the best support is so-called self-primed reverse transcription. Here, internal initiation of reverse transcriptase occurs because the 3' end of the mRNA folds back onto the transcript and complementary pairs with some upstream region of mRNA (resulting in a hairpin in the secondary structure). cDNA is then synthesized from the region upstream of this base-pairing. This mechanism is suspected to cause the 4f-rnp intron loss polymorphism in Drosophila  and is implicated in the loss of internal introns in many eukaryotes, including several fungi (Aspergillus nidulans, Fusarium graminearum, Magnaporthe grisea, N. crassa, Plasmodium falciparum, S. pombe, and U. maydis) . While subsequent mutation impedes our ability to determine if this mechanism contributes to intron loss (for example, erosion of the complementary sequence), the pattern of loss observed in Cryptococcus suggests a different mechanism contributes to internal intron loss. Because there is no known bias regarding where, along the length of a transcript, the 3' end can complementary pair, this hairpin-based loss mechanism should produce intron losses along the length of a gene with equal probability. The majority of losses, however, cluster toward the 3' end of the gene (Figure 4a). Additionally, while there are losses that occur at extreme distances from the 3' end of the gene in Cryptococcus (Figure 4b), all losses are at distances smaller than the known, active fungal retrotransposons. Reverse transcriptase is, then, capable of synthesizing cDNA from the transcript's poly-A tail to these distant positions. Indeed, the only observation that does not fit the requirements of the poly-A primed mechanism is that intron loss events frequently exclude 3' terminal introns. Given this loss pattern, we propose a simple explanation to account for the observation of internal intron loss (shown in Figure 3b). If, within a population, an allele that had lost a series of introns via recombination with a poly-A initiated cDNA were to undergo meiotic recombination with an allele that retained 3' introns, the pattern of internal intron loss could be created.
To determine if this mechanism is likely, we employed an allelic sojourn time density function derived from a diffusion approximation to estimate the frequency with which an initially single-copy, 3' poly-A primed intron loss allele recombines with a wild-type allele before being absorbed or fixed in the population (see Materials and methods). While we cannot be certain about historical recombination rates, we can make a reasonable estimation by employing the measured meiotic recombination rate determined for C. neoformans serotype D . Our analysis estimates that 'internal intron loss' alleles are generated at a rate 2.08 times that of '3'-generated intron loss' alleles. Assuming equal probability of fixation, we therefore might expect approximately twice as many genes to exhibit internal intron loss as genes exhibiting a 3' loss of two or more introns, even though all intron losses are originally caused by reverse transcription from the 3' poly-A tail. This model for internal intron loss is built upon an experimentally verified method of cDNA creation (initiation at the poly-A tail), explains the relative clustering of 3' intron loss and, assuming meiotic recombination occurs at random positions between the cDNA and the gene, accounts for the diverse distribution of losses observed (Additional data file 3). To our knowledge, this is the first time such a mechanism has been described to explain the observation of internal intron loss.
While we found no intron gains between the two Cryptococcus species, we did identify a gene that has gained introns within C. neoformans. The D1 gene CNN00420 contains unique intron sequences relative to orthologs in A and D2 (no orthologs could be identified in serotype B). Resequencing strains from the serotype D population confirmed that these introns in the D1 allele are unique, implicating a sequence gain. Because each gained intron has exact identity to three other introns in the gene (and near exact identity to two others), it is likely that unequal crossing over, gene conversion or repeat expansion through replication strand slippage caused the gains. While we cannot discern which of these DNA duplicating mechanisms caused the intron gains in question, several studies have shown that replication strand slippage is a common and frequent means of varying intergenic repeat copy number. In Escherichia coli, RecA independent duplications and deletions between repeats of several hundred bases have been shown to occur at high frequency (10-5 to 10-4 per cell generation) . In an elegant analysis of FLO gene repeat variation in yeast, Verstrepen et al.  concluded that replication slippage is the cause of the observed length polymorphisms and that the repeats could be at least 100 nucleotides in length. Here, we propose its role in generating new introns in CNN00420 (Additional data file 7).
To our knowledge, there is only one other example of an intron gain occurring through repeat duplication. Knowles and McLysaght  suggested that this mechanism is responsible for a gained intron in the TOUCH3 gene relative to its paralog in Arabidopsis. However, though there is debate on this point , paralogs may undergo wholly different gene structure evolution than orthologs . Additionally, the role of the Arabidopsis whole genome duplication on gene structure evolution is uncertain. Our analysis verifies that this mechanism can create introns in orthologs and contributes to gene structure divergence between species. Furthermore, this is the first time a DNA repeat expansion has been observed to create an intron polymorphism.
In his study of intron gain rates, Roy noted that certain ancient lineages underwent big bursts of intron gain while subsequent lineages experienced a precipitous decline in intron gain rates . This suggests that these early lineage introns may have been created via a wholly different mechanism than those created in contemporary times. While we do not suspect that intron containing repeat expansion is the mechanism responsible for the origin of most introns, it appears to be a more widespread mechanism than previously expected and may be an important contemporary means of creating new intronic sequence in the genome. As additional closely related genome sequences become available, future comparative genomic studies should verify whether this mechanism of gain is ubiquitous and identify other contemporary mechanisms that create introns.
Materials and methods
Genome sequences utilized
The C. neoformans serotype D strain JEC21 (D1) genome sequence and annotation was obtained from GenBank as deposited by TIGR (January 2005, 20 Mb, 14 chromosomes). A second serotype D strain sequence, B3501 (D2), was obtained from the Stanford Genome Technology Center (June 2004 assembly, 18.5 Mb, 70 contigs). The C. neoformans serotype A strain H99 (A) was sequenced by the Broad Institute (October 2004 assembly, 19.5 Mb, 210 contigs). The Broad Institute also sequenced the C. gatii serotype B strain R265 (B1) (January 2005 assembly, 17.2 Mb, 28 contigs). The second serotype B strain WM276 (B2) was obtained from the BC Genome sequencing center (assembly dated March 2004, 18.0 Mb, 33 contigs).
Whole genome alignments were created through a multi-step process using D1 as a reference. First, pairwise alignments between D1 and the other sequenced strains were created using PatternHunter . Blocks of four-way homologous contigs were then identified using a hierarchical synteny-clustering algorithm as described in . Multiple alignments of homologous regions were generated using Multi-LAGAN .
As described in , the annotation for D1 is well supported because an extensive EST library was used to train the gene structure predictive algorithms. Thus, we aligned orthologs to each D1 reference gene and evaluated the conservation of each annotated D1 intron across the alignment (Figure 2) . We identified orthologs of D1 genes by searching for best all-way reciprocal BLASTn hits against the corresponding genomes that also demonstrated syntenic conservation. Because previous studies have demonstrated differential tempos and modes of intron evolution between paralogs , we elected to use this type of conservative ortholog detection method . To enable comparative analysis, each orthology cluster had to include a sequence from serotype B, which represented the outgroup of the comparison.
Clusters of orthologs meeting these criteria were aligned separately from the whole genome alignment using the mLAGAN multiple sequence alignment algorithm and the phylogeny shown in Figure 1 as the guide tree. To identify introns in these alignments, Perl scripts were constructed to overlay the D1 gene annotation onto the alignment. We subsequently identified sequences with gaps that completely spanned an intron position. When we detected missing introns, the quality of the gene alignment was evaluated by eye and the principles of parsimony were used to infer whether the intron in question had been gained or lost. For example, if A is missing an intron sequence that is found in D1, D2, B1 and B2, then we score the sequence as an intron loss in A. It should be noted that parsimony has been shown to overestimate the relative contribution of intron gain at large phylogenetic divergences by effectively undercounting the number of parallel loss events . The close evolutionary distance involved in this study should mitigate this problem.
The unrooted phylogenetic topology among the five sequenced strains was determined by a neighbor-joining analysis  on a concatenation of all orthologous open reading frames. Branch lengths based on silent site divergence observed across a concatenation of 5,700 aligned ortholog clusters were estimated using the codeml application (model = 0, NSsites = 0, ncatG = 1) from the PAML 3.15 package .
Genome sequenced strains JEC21(D1) and B3051(D2), as well as additional serotype D individuals TP0603, WM629, JEC20 and B3502, were generously provided by W Meyer. The method of Bolano was used to extract genomic DNA from strains grown in liquid culture . Conserved regions of the alignment between the D1 and D2 genome sequences were used to construct PCR primers for touchdown PCR. Amplified target DNA was cloned with Invitrogen's TopoTA cloning kit (Invitrogen, Carlsbad, CA, USA) and subject to standard BigDye (Applied Biosystems, Foster City, CA, USA) sequencing procedures as in . Sequencher (v4.2.2, Gene Codes Corp., Ann Arbor, MI, USA) was used to identify high quality sequence reads and BioEdit's CAP contig assembly tool was used to construct full length sequences . Initial alignments were generated by ClustalW and edited by hand .
Here, r is the recombination rate and L is the average exon length. In Cryptococcus, these values are 7.58e-7 recombination events per base-pair per generation  and 253 bp, respectively. Incorporating these values returns an expectation of 2.08 recombination events that would create 'internal intron loss' alleles for every '3 intron loss' allele generated by mutation, suggesting that, on average, 'internal intron loss' alleles are generated at a rate approximately 2.08 times that of '3' intron loss' alleles.
Resampling analysis was conducted to determine intron loss rate 95% confidence intervals. A total population of 33,473 introns (equal to the total number of screened orthologous introns) was established, where 49 introns in the population corresponded to loss events and the remainder were intron conservation events. We sampled, with replacement, 10,000 introns from this population and then calculated the frequency and subsequent rate. Calculations were conducted 10,000 times and a normal curve approximation of the subsequent rate distribution was used to calculate the 95% confidence interval.
Resampling analysis was also conducted for each gene demonstrating adjacent intron loss to determine the probability that the adjacent losses were caused by random, independent events. For each gene, a population equal to the number of introns in the gene was established, where each individual in the population has a value corresponding to an intron position within the gene. Samples equal to the number of observed losses in the gene were taken from this population, without replacement, and the longest consecutive run of adjacent introns was counted. After 10,000 runs, a histogram of the longest runs was created and used to evaluate the null hypothesis that the observed data were due to random, independent loss events.
Additional data files
The following additional data are available. Additional data file 1 is a table documenting the consecutive intron loss events. Additional data file 2 is a histogram of the relative distribution of introns within genes in the D1 genome. Additional data file 3 is a histogram of the order number of a lost intron, from 5' to 3' along the length of the gene, divided by the total number of introns in the gene. Additional data file 4 is a dot plot of CNN00420 against itself. Additional data file 5 provides the multiple sequence alignment (FASTA format) of CNN00420 orthologs from the serotype D population. Additional data file 6 shows CNN00420 gained introns along the D1 lineage. Additional data file 7 shows the mechanism of intron gain via strand slippage. Additional data file 8 provides our Mathematica source code for the sojourn time density approximation.
C. neoformans serotype A strain H99
C. gattii serotype B strain R265
C. gattii serotype B strain WM276
C. neoformans serotype D strain JEC21
C. neoformans serotype D strain B3501
silent substitution rate
expressed sequence tag.
The authors would like to thank WM Meyer for contributing Cryptococcus strains and JE Stajich for his invaluable comments.
- Gilbert W: Why genes in pieces. Nature. 1978, 271: 501-10.1038/271501a0.PubMedView ArticleGoogle Scholar
- 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. 10.1093/nar/gkh686.PubMedPubMed CentralView ArticleGoogle Scholar
- Roy SW, Gilbert W: Rates of intron loss and gain: implications for early eukaryotic evolution. Proc Natl Acad Sci USA. 2005, 102: 5773-5778. 10.1073/pnas.0500383102.PubMedPubMed CentralView ArticleGoogle Scholar
- Nielsen CB, Friedman B, Birren B, Burge CB, Galagan JE: Patterns of intron gain and loss in fungi. PLoS Biol. 2004, 2: 2234-2242. 10.1371/journal.pbio.0020422.View ArticleGoogle Scholar
- Belshaw R, Bensasson D: The rise and falls of introns. Heredity. 2006, 96: 208-213. 10.1038/sj.hdy.6800791.PubMedView ArticleGoogle Scholar
- Katsu M, Kidd S, Ando A, Moretti-Branchini ML, Mikami Y, Nishimura K, Meyer W: The internal transcribed spacers and 5.8S rRNA gene show extensive diversity among isolates of the Cryptococcus neoformans species complex. FEMS Yeast Res. 2004, 4: 377-388. 10.1016/S1567-1356(03)00176-4.PubMedView ArticleGoogle Scholar
- Xu JP, Vilgalys R, Mitchell TG: Multiple gene genealogies reveal recent dispersion and hybridization in the human pathogenic fungus Cryptococcus neoformans. Mol Ecol. 2000, 9: 1471-1481. 10.1046/j.1365-294x.2000.01021.x.PubMedView ArticleGoogle Scholar
- Casadevall A, Perfect JR: Cryptococcus neoformans. 1998, Washington, DC: American Society for MicrobiologyView ArticleGoogle Scholar
- Loftus BJ, Fung E, Roncaglia P, Rowley D, Amedeo P, Bruno D, Vamathevan J, Miranda M, Anderson IJ, Fraser JA, Allen JE, Bosdet IE, Brent MR, Chiu R, Doering TL, Donlin MJ, D'Souza CA, Fox DS, Grinberg V, Fu J, Fukushima M, Haas BJ, Huang JC, Janbon G, Jones SJ, Koo HL, Krzywinski MI, Kwon-Chung JK, Lengeler KB, Maiti R, et al: The genome of the basidiomycetous yeast and human pathogen Cryptococcus neoformans. Science. 2005, 307: 1321-1324. 10.1126/science.1103773.PubMedPubMed CentralView ArticleGoogle Scholar
- Kasuga T, White TJ, Taylor JW: Estimation of nucleotide substitution rates in eurotiomycete fungi. Mol Biol Evol. 2002, 19: 2318-2324.PubMedView ArticleGoogle Scholar
- Nei M, Stephens JC, Saitou N: Methods for computing the standard errors of branching points in an evolutionary tree and their application to molecular-data from humans and apes. Mol Biol Evol. 1985, 2: 66-85.PubMedGoogle Scholar
- Broad Institute Fungal Genome Initiative Database. [http://www.broad.mit.edu/annotation/fgi]
- 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. 10.1073/pnas.96.18.10267.PubMedPubMed CentralView ArticleGoogle Scholar
- Krzywinski J, Besansky NJ: Frequent intron loss in the white gene: a cautionary tale for phylogeneticists. Mol Biol Evol. 2002, 19: 362-366.PubMedView ArticleGoogle Scholar
- Urrutia AO, Hurst LD: The signature of selection mediated by expression on human genes. Genome Res. 2003, 13: 2260-2264. 10.1101/gr.641103.PubMedPubMed CentralView ArticleGoogle Scholar
- Roy SW, Gilbert W: The pattern of intron loss. Proc Natl Acad Sci USA. 2005, 102: 713-718. 10.1073/pnas.0408274102.PubMedPubMed CentralView ArticleGoogle Scholar
- Roy SW, Gilbert W: The evolution of spliceosomal introns: patterns, puzzles and progress. Nat Rev Genet. 2006, 7: 211-221.PubMedGoogle Scholar
- Lewin R: How mammalian RNAa returns to its genome. Science. 1983, 219: 1052-1054. 10.1126/science.6186029.PubMedView ArticleGoogle Scholar
- Mourier T: Reverse transcription in genome evolution. Cytogenet Genome Res. 2005, 110: 56-62. 10.1159/000084938.PubMedView ArticleGoogle Scholar
- Fink GR: Pseudogenes in yeast?. Cell. 1987, 49: 5-6. 10.1016/0092-8674(87)90746-X.PubMedView ArticleGoogle Scholar
- Stajich JE, Dietrich FS: Evidence of mRNA-mediated intron loss in the human-pathogenic fungus Cryptococcus neoformans. Eukaryot Cell. 2006, 5: 789-793. 10.1128/EC.5.5.789-793.2006.PubMedPubMed CentralView ArticleGoogle Scholar
- Fedorova L, Fedorov A: Introns in gene evolution. Genetica. 2003, 118: 123-131. 10.1023/A:1024145407467.PubMedView ArticleGoogle Scholar
- Niu DK, Hou WR, Li SW: mRNA-mediated intron losses: evidence from extraordinarily large exons. Mol Biol Evol. 2005, 22: 1475-1481. 10.1093/molbev/msi138.PubMedView ArticleGoogle Scholar
- Ophir R, Graur D: Patterns and rates of indel evolution in processed pseudogenes from humans and murids. Gene. 1997, 205: 191-202. 10.1016/S0378-1119(97)00398-3.PubMedView ArticleGoogle Scholar
- Hu K: Intron exclusion and the mystery of intron loss. FEBS Lett. 2006, 580: 6361-6365. 10.1016/j.febslet.2006.10.048.PubMedView ArticleGoogle Scholar
- Feiber AL, Rangarajan J, Vaughn JC: The evolution of single-copy Drosophila nuclear 4f-rnp genes: spliceosomal intron losses create polymorphic alleles. J Mol Evol. 2002, 55: 401-413. 10.1007/s00239-002-2336-y.PubMedView ArticleGoogle Scholar
- Marra RE, Huang JC, Fung E, Nielsen K, Heitman J, Vilgalys R, Mitchell TG: A genetic linkage map of Cryptococcus neoformans variety neoformans serotype D (Filobasidiella neoformans). Genetics. 2004, 167: 619-631. 10.1534/genetics.103.023408.PubMedPubMed CentralView ArticleGoogle Scholar
- Lovett ST: Encoded errors: mutations and rearrangements mediated by misalignment at repetitive DNA sequences. Mol Microbiol. 2004, 52: 1243-1253. 10.1111/j.1365-2958.2004.04076.x.PubMedView ArticleGoogle Scholar
- Verstrepen KJ, Jansen A, Lewitter F, Fink GR: Intragenic tandem repeats generate functional variability. Nat Genet. 2005, 37: 986-990. 10.1038/ng1618.PubMedPubMed CentralView ArticleGoogle Scholar
- Knowles DG, McLysaght A: High rate of recent intron gain and loss in simultaneously duplicated Arabidopsis genes. Mol Biol Evol. 2006, 23: 1548-1557. 10.1093/molbev/msl017.PubMedView ArticleGoogle Scholar
- 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. 2007, 24: 171-181. 10.1093/molbev/msl159.PubMedView ArticleGoogle Scholar
- Ma B, Tromp J, Li M: PatternHunter: faster and more sensitive homology search. Bioinformatics. 2002, 18: 440-445. 10.1093/bioinformatics/18.3.440.PubMedView ArticleGoogle Scholar
- Galagan JE, Calvo SE, Cuomo C, Ma LJ, Wortman JR, Batzoglou S, Lee SI, Baştürkmen M, Spevak CC, Clutterbuck J, Kapitonov V, Jurka J, Scazzocchio C, Farman M, Butler J, Purcell S, Harris S, Braus GH, Draht O, Busch S, D'Enfert C, Bouchier C, Goldman GH, Bell-Pedersen D, Griffiths-Jones S, Doonan JH, Yu J, Vienken K, Pain A, Freitag M, et al: Sequencing of Aspergillus nidulans and comparative analysis with A. fumigatus and A. oryzae. Nature. 2005, 438: 1105-1115. 10.1038/nature04341.PubMedView ArticleGoogle Scholar
- Brudno M, Do CB, Cooper GM, Kim MF, Davydov E, NISC Comparative Sequencing Program, Green ED, Sidow A, Batzoglou S: LAGAN and Multi-LAGAN: Efficient tools for large-scale multiple alignment of genomic DNA. Genome Res. 2003, 13: 721-731. 10.1101/gr.926603.PubMedPubMed CentralView ArticleGoogle Scholar
- Gilbert W, Glynias M: On the ancient nature of introns. Gene. 1993, 135: 137-144. 10.1016/0378-1119(93)90058-B.PubMedView ArticleGoogle Scholar
- Saitou N, Nei M: The neighbor-joining method - a new method for reconstructing phylogenetic trees. Mol Biol Evol. 1987, 4: 406-425.PubMedGoogle Scholar
- Yang ZH, Nielsen R: Estimating synonymous and nonsynonymous substitution rates under realistic evolutionary models. Mol Biol Evol. 2000, 17: 32-43.PubMedView ArticleGoogle Scholar
- Bolano A, Stinchi S, Preziosi R, Bistoni F, Allegrucci M, Baldelli F, Martini A, Cardinali G: Rapid methods to extract DNA and RNA from Cryptococcus neoformans. FEMS Yeast Res. 2001, 1: 221-224.PubMedGoogle Scholar
- Johannesson H, Townsend JP, Hung CY, Cole GT, Taylor JW: Concerted evolution in the repeats of an immunomodulating cell surface protein, SOWgp, of the human pathogenic fungi Coccidioides immitis and C. posadasii. Genetics. 2005, 171: 109-117. 10.1534/genetics.105.040923.PubMedPubMed CentralView ArticleGoogle Scholar
- Huang XQ: A contig assembly program based on sensitive detection of fragment overlaps. Genomics. 1992, 14: 18-25. 10.1016/S0888-7543(05)80277-0.PubMedView ArticleGoogle Scholar
- Thompson JD, Higgins DG, Gibson TJ: CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 1994, 22: 4673-4680. 10.1093/nar/22.22.4673.PubMedPubMed CentralView ArticleGoogle Scholar
- Ewens WH: Mathematical Population Genetics. 1979, New York: Springer-VerlagGoogle Scholar
- Nagylaki T: The moments of stochastic integrals and the distribution of sojourn times. Proc Natl Acad Sci USA. 1974, 71: 746-749. 10.1073/pnas.71.3.746.PubMedPubMed CentralView ArticleGoogle Scholar
- Petrov DA, Aminetzach YT, Davis JC, Bensasson D, Hirsh AE: Size matters: non-LTR retrotransposable elements and ectopic recombination in Drosophila. Mol Biol Evol. 2003, 20: 880-892. 10.1093/molbev/msg102.PubMedView ArticleGoogle Scholar
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.