Limits...
Recurrent loss of specific introns during angiosperm evolution.

Wang H, Devos KM, Bennetzen JL - PLoS Genet. (2014)

Bottom Line: The two larger genomes, maize and sorghum, were found to have a higher rate of both recurrent loss and overall loss and/or gain than foxtail millet, rice or Brachypodium.Adjacent introns and small introns were found to be preferentially lost.This last result suggests that epigenetic status, as evidenced by a loss of methylated CG dinucleotides, may play a role in the process of intron loss.

View Article: PubMed Central - PubMed

Affiliation: Department of Genetics, University of Georgia, Athens, Georgia, United States of America.

ABSTRACT
Numerous instances of presence/absence variations for introns have been documented in eukaryotes, and some cases of recurrent loss of the same intron have been suggested. However, there has been no comprehensive or phylogenetically deep analysis of recurrent intron loss. Of 883 cases of intron presence/absence variation that we detected in five sequenced grass genomes, 93 were confirmed as recurrent losses and the rest could be explained by single losses (652) or single gains (118). No case of recurrent intron gain was observed. Deep phylogenetic analysis often indicated that apparent intron gains were actually numerous independent losses of the same intron. Recurrent loss exhibited extreme non-randomness, in that some introns were removed independently in many lineages. The two larger genomes, maize and sorghum, were found to have a higher rate of both recurrent loss and overall loss and/or gain than foxtail millet, rice or Brachypodium. Adjacent introns and small introns were found to be preferentially lost. Intron loss genes exhibited a high frequency of germ line or early embryogenesis expression. In addition, flanking exon A+T-richness and intron TG/CG ratios were higher in retained introns. This last result suggests that epigenetic status, as evidenced by a loss of methylated CG dinucleotides, may play a role in the process of intron loss. This study provides the first comprehensive analysis of recurrent intron loss, makes a series of novel findings on the patterns of recurrent intron loss during the evolution of the grass family, and provides insight into the molecular mechanism(s) underlying intron loss.

No MeSH data available.


Related in: MedlinePlus

Chromosomal locations of (a) all genes in five grass genomes and those that have undergone (b) recurrent intron loss or (c) any PA intron variation (gain or loss).Short arms and long arms of chromosomes are normalized separately. The centromere is located at 0 and the short and long arm termini at -1 and 1, respectively. The locations of genes are calculated as distance from centromere divided by total length of chromosome arm, where distance is a negative number for the short arm. The normalized chromosome is partitioned into 20 intervals (X-axis) and Y-axis values are the percentage of genes in these intervals. Error bars in (b) and (c) represent one sd from interval mean values (circle), where mean and sd are calculated by resampling with replacement (1000 times) from the entire intron set.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4256211&req=5

pgen-1004843-g004: Chromosomal locations of (a) all genes in five grass genomes and those that have undergone (b) recurrent intron loss or (c) any PA intron variation (gain or loss).Short arms and long arms of chromosomes are normalized separately. The centromere is located at 0 and the short and long arm termini at -1 and 1, respectively. The locations of genes are calculated as distance from centromere divided by total length of chromosome arm, where distance is a negative number for the short arm. The normalized chromosome is partitioned into 20 intervals (X-axis) and Y-axis values are the percentage of genes in these intervals. Error bars in (b) and (c) represent one sd from interval mean values (circle), where mean and sd are calculated by resampling with replacement (1000 times) from the entire intron set.

Mentions: We investigated the distribution of intron turnover across chromosomes (Fig. 4 and Figure S9) by normalization of chromosome size. The centromere was located at 0; and then the short arms and long arms of chromosomes were normalized separately, with the short and long arm termini located at -1 and 1, respectively. Locations of genes belonging to the same intron group were counted independently. The distribution of the whole gene set of the five genomes (Fig. 4a) showed a smooth ā€œVā€ shape with the lowest gene density located at the centromeric/pericentromeric region. The distributions of genes with detected recurrent loss (Fig. 4b) and total intron turnover (Fig. 4c) exhibited a similar overall trend. However, it should be noted that plant genes are highly mobile over evolutionary time, and even centromeres can be found in different positions in closely related lineages [31], so the current genomic location is not a perfect predictor for any single gene of its location when an intron was gained or lost.


Recurrent loss of specific introns during angiosperm evolution.

Wang H, Devos KM, Bennetzen JL - PLoS Genet. (2014)

Chromosomal locations of (a) all genes in five grass genomes and those that have undergone (b) recurrent intron loss or (c) any PA intron variation (gain or loss).Short arms and long arms of chromosomes are normalized separately. The centromere is located at 0 and the short and long arm termini at -1 and 1, respectively. The locations of genes are calculated as distance from centromere divided by total length of chromosome arm, where distance is a negative number for the short arm. The normalized chromosome is partitioned into 20 intervals (X-axis) and Y-axis values are the percentage of genes in these intervals. Error bars in (b) and (c) represent one sd from interval mean values (circle), where mean and sd are calculated by resampling with replacement (1000 times) from the entire intron set.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4256211&req=5

pgen-1004843-g004: Chromosomal locations of (a) all genes in five grass genomes and those that have undergone (b) recurrent intron loss or (c) any PA intron variation (gain or loss).Short arms and long arms of chromosomes are normalized separately. The centromere is located at 0 and the short and long arm termini at -1 and 1, respectively. The locations of genes are calculated as distance from centromere divided by total length of chromosome arm, where distance is a negative number for the short arm. The normalized chromosome is partitioned into 20 intervals (X-axis) and Y-axis values are the percentage of genes in these intervals. Error bars in (b) and (c) represent one sd from interval mean values (circle), where mean and sd are calculated by resampling with replacement (1000 times) from the entire intron set.
Mentions: We investigated the distribution of intron turnover across chromosomes (Fig. 4 and Figure S9) by normalization of chromosome size. The centromere was located at 0; and then the short arms and long arms of chromosomes were normalized separately, with the short and long arm termini located at -1 and 1, respectively. Locations of genes belonging to the same intron group were counted independently. The distribution of the whole gene set of the five genomes (Fig. 4a) showed a smooth ā€œVā€ shape with the lowest gene density located at the centromeric/pericentromeric region. The distributions of genes with detected recurrent loss (Fig. 4b) and total intron turnover (Fig. 4c) exhibited a similar overall trend. However, it should be noted that plant genes are highly mobile over evolutionary time, and even centromeres can be found in different positions in closely related lineages [31], so the current genomic location is not a perfect predictor for any single gene of its location when an intron was gained or lost.

Bottom Line: The two larger genomes, maize and sorghum, were found to have a higher rate of both recurrent loss and overall loss and/or gain than foxtail millet, rice or Brachypodium.Adjacent introns and small introns were found to be preferentially lost.This last result suggests that epigenetic status, as evidenced by a loss of methylated CG dinucleotides, may play a role in the process of intron loss.

View Article: PubMed Central - PubMed

Affiliation: Department of Genetics, University of Georgia, Athens, Georgia, United States of America.

ABSTRACT
Numerous instances of presence/absence variations for introns have been documented in eukaryotes, and some cases of recurrent loss of the same intron have been suggested. However, there has been no comprehensive or phylogenetically deep analysis of recurrent intron loss. Of 883 cases of intron presence/absence variation that we detected in five sequenced grass genomes, 93 were confirmed as recurrent losses and the rest could be explained by single losses (652) or single gains (118). No case of recurrent intron gain was observed. Deep phylogenetic analysis often indicated that apparent intron gains were actually numerous independent losses of the same intron. Recurrent loss exhibited extreme non-randomness, in that some introns were removed independently in many lineages. The two larger genomes, maize and sorghum, were found to have a higher rate of both recurrent loss and overall loss and/or gain than foxtail millet, rice or Brachypodium. Adjacent introns and small introns were found to be preferentially lost. Intron loss genes exhibited a high frequency of germ line or early embryogenesis expression. In addition, flanking exon A+T-richness and intron TG/CG ratios were higher in retained introns. This last result suggests that epigenetic status, as evidenced by a loss of methylated CG dinucleotides, may play a role in the process of intron loss. This study provides the first comprehensive analysis of recurrent intron loss, makes a series of novel findings on the patterns of recurrent intron loss during the evolution of the grass family, and provides insight into the molecular mechanism(s) underlying intron loss.

No MeSH data available.


Related in: MedlinePlus