Limits...
Structural conservation versus functional divergence of maternally expressed microRNAs in the Dlk1/Gtl2 imprinting region.

Kircher M, Bock C, Paulsen M - BMC Genomics (2008)

Bottom Line: Further, our analyses did not uncover a functional relation between imprinted gene regulation of this microRNA-encoding region, expression patterns or functions of predicted target genes.Specifically, our results indicate that these microRNAs do not regulate a particular set of genes.We conclude that these imprinted microRNAs do not regulate a particular set of genes.

View Article: PubMed Central - HTML - PubMed

Affiliation: Max-Planck-Institut für Informatik, Saarbrücken, Germany. Martin.Kircher@eva.mpg.de

ABSTRACT

Background: MicroRNAs play an important functional role in post-transcriptional gene regulation. One of the largest known microRNA clusters is located within the imprinted Dlk1/Gtl2 region on human chromosome 14 and mouse chromosome 12. This cluster contains more than 40 microRNA genes that are expressed only from the maternal chromosome in mouse.

Results: To shed light on the function of these microRNAs and possible crosstalk between microRNA-based gene regulation and genomic imprinting, we performed extensive in silico analyses of the microRNAs in this imprinted region and their predicted target genes.Bioinformatic analysis reveals that these microRNAs are highly conserved in both human and mouse. Whereas the microRNA precursors at this locus mostly belong to large sequence families, the mature microRNAs sequences are highly divergent. We developed a target gene prediction approach that combines three widely used prediction methods and achieved a sufficiently high prediction accuracy. Target gene sets predicted for individual microRNAs derived from the imprinted region show little overlap and do not differ significantly in their properties from target genes predicted for a group of randomly selected microRNAs. The target genes are enriched with long and GC-rich 3' UTR sequences and are preferentially annotated to development, regulation processes and cell communication. Furthermore, among all analyzed human and mouse genes, the predicted target genes are characterized by consistently higher expression levels in all tissues considered.

Conclusion: Our results suggest a complex evolutionary history for microRNA genes in this imprinted region, including an amplification of microRNA precursors in a mammalian ancestor, and a rapid subsequent divergence of the mature sequences. This produced a broad spectrum of target genes. Further, our analyses did not uncover a functional relation between imprinted gene regulation of this microRNA-encoding region, expression patterns or functions of predicted target genes. Specifically, our results indicate that these microRNAs do not regulate a particular set of genes. We conclude that these imprinted microRNAs do not regulate a particular set of genes. Rather, they seem to stabilize expression of a variety of genes, thereby being an integral part of the genome-wide microRNA gene regulatory network.

Show MeSH

Related in: MedlinePlus

GC content and length distribution of 3' UTR sequences. Shown is the density distribution of GC content and length in the set of all 3' UTR sequences, in the set of predicted human "Vari" targets of at least three Dlk1/Gtl2 microRNAs and in the set of human antitargets of at least 16 Dlk1/Gtl2 microRNAs. The frequencies of the bins are normalized to equal one for each set of sequences (density distribution), which corrects for the different number of sequences in each set. For a better visualization of the 3' UTR length, the natural logarithm of the length was used.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2500034&req=5

Figure 5: GC content and length distribution of 3' UTR sequences. Shown is the density distribution of GC content and length in the set of all 3' UTR sequences, in the set of predicted human "Vari" targets of at least three Dlk1/Gtl2 microRNAs and in the set of human antitargets of at least 16 Dlk1/Gtl2 microRNAs. The frequencies of the bins are normalized to equal one for each set of sequences (density distribution), which corrects for the different number of sequences in each set. For a better visualization of the 3' UTR length, the natural logarithm of the length was used.

Mentions: In addition to exploring the target and antitarget lists of individual microRNAs, we also analyzed the merged target and antitarget lists for both the Dlk1/Gtl2 microRNAs and the randomly selected reference set. This analysis facilitates a further investigation for a common function of the microRNAs in the imprinted cluster. To reduce the impact of transcripts only predicted for a small subset of microRNAs, we removed all sequences that were predicted for fewer than three microRNAs from the merged target lists and fewer than 16 microRNAs from the merged antitarget lists. In analyzing these targets and antitargets, we notice that the average target 3' UTR length is about 2829 nt in human (2870 nt in mouse) with the "Vari" filter – three times greater than the average 3' UTR length of all transcripts analyzed – and that the GC content of the targets is about 6.5% higher in human (4.0% in mouse) than the average. For the antitargets, we see the same shift in the other direction for the 3' UTR length and GC content (Figure 5). These observations are not unexpected given the known preference of target predictors for guanine and cytosine bases and that long sequences have a higher probability of containing target sites. Similar results were obtained for the targets and antitargets of the reference set.


Structural conservation versus functional divergence of maternally expressed microRNAs in the Dlk1/Gtl2 imprinting region.

Kircher M, Bock C, Paulsen M - BMC Genomics (2008)

GC content and length distribution of 3' UTR sequences. Shown is the density distribution of GC content and length in the set of all 3' UTR sequences, in the set of predicted human "Vari" targets of at least three Dlk1/Gtl2 microRNAs and in the set of human antitargets of at least 16 Dlk1/Gtl2 microRNAs. The frequencies of the bins are normalized to equal one for each set of sequences (density distribution), which corrects for the different number of sequences in each set. For a better visualization of the 3' UTR length, the natural logarithm of the length was used.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2500034&req=5

Figure 5: GC content and length distribution of 3' UTR sequences. Shown is the density distribution of GC content and length in the set of all 3' UTR sequences, in the set of predicted human "Vari" targets of at least three Dlk1/Gtl2 microRNAs and in the set of human antitargets of at least 16 Dlk1/Gtl2 microRNAs. The frequencies of the bins are normalized to equal one for each set of sequences (density distribution), which corrects for the different number of sequences in each set. For a better visualization of the 3' UTR length, the natural logarithm of the length was used.
Mentions: In addition to exploring the target and antitarget lists of individual microRNAs, we also analyzed the merged target and antitarget lists for both the Dlk1/Gtl2 microRNAs and the randomly selected reference set. This analysis facilitates a further investigation for a common function of the microRNAs in the imprinted cluster. To reduce the impact of transcripts only predicted for a small subset of microRNAs, we removed all sequences that were predicted for fewer than three microRNAs from the merged target lists and fewer than 16 microRNAs from the merged antitarget lists. In analyzing these targets and antitargets, we notice that the average target 3' UTR length is about 2829 nt in human (2870 nt in mouse) with the "Vari" filter – three times greater than the average 3' UTR length of all transcripts analyzed – and that the GC content of the targets is about 6.5% higher in human (4.0% in mouse) than the average. For the antitargets, we see the same shift in the other direction for the 3' UTR length and GC content (Figure 5). These observations are not unexpected given the known preference of target predictors for guanine and cytosine bases and that long sequences have a higher probability of containing target sites. Similar results were obtained for the targets and antitargets of the reference set.

Bottom Line: Further, our analyses did not uncover a functional relation between imprinted gene regulation of this microRNA-encoding region, expression patterns or functions of predicted target genes.Specifically, our results indicate that these microRNAs do not regulate a particular set of genes.We conclude that these imprinted microRNAs do not regulate a particular set of genes.

View Article: PubMed Central - HTML - PubMed

Affiliation: Max-Planck-Institut für Informatik, Saarbrücken, Germany. Martin.Kircher@eva.mpg.de

ABSTRACT

Background: MicroRNAs play an important functional role in post-transcriptional gene regulation. One of the largest known microRNA clusters is located within the imprinted Dlk1/Gtl2 region on human chromosome 14 and mouse chromosome 12. This cluster contains more than 40 microRNA genes that are expressed only from the maternal chromosome in mouse.

Results: To shed light on the function of these microRNAs and possible crosstalk between microRNA-based gene regulation and genomic imprinting, we performed extensive in silico analyses of the microRNAs in this imprinted region and their predicted target genes.Bioinformatic analysis reveals that these microRNAs are highly conserved in both human and mouse. Whereas the microRNA precursors at this locus mostly belong to large sequence families, the mature microRNAs sequences are highly divergent. We developed a target gene prediction approach that combines three widely used prediction methods and achieved a sufficiently high prediction accuracy. Target gene sets predicted for individual microRNAs derived from the imprinted region show little overlap and do not differ significantly in their properties from target genes predicted for a group of randomly selected microRNAs. The target genes are enriched with long and GC-rich 3' UTR sequences and are preferentially annotated to development, regulation processes and cell communication. Furthermore, among all analyzed human and mouse genes, the predicted target genes are characterized by consistently higher expression levels in all tissues considered.

Conclusion: Our results suggest a complex evolutionary history for microRNA genes in this imprinted region, including an amplification of microRNA precursors in a mammalian ancestor, and a rapid subsequent divergence of the mature sequences. This produced a broad spectrum of target genes. Further, our analyses did not uncover a functional relation between imprinted gene regulation of this microRNA-encoding region, expression patterns or functions of predicted target genes. Specifically, our results indicate that these microRNAs do not regulate a particular set of genes. We conclude that these imprinted microRNAs do not regulate a particular set of genes. Rather, they seem to stabilize expression of a variety of genes, thereby being an integral part of the genome-wide microRNA gene regulatory network.

Show MeSH
Related in: MedlinePlus