Pseudogenes as an alternative source of natural antisense transcripts

Abstract

Background: Naturally occurring antisense transcripts (NATs) are non-coding RNAs that may regulate the activity of sense transcripts to which they bind because of complementarity. NATs that are not located in the gene they regulate (trans-NATs) have better chances to evolve than cis-NATs, which is evident when the sense strand of the cis-NAT is part of a protein coding gene. However, the generation of a trans-NAT requires the formation of a relatively large region of complementarity to the gene it regulates.

Results: Pseudogene formation may be one evolutionary mechanism that generates trans-NATs to the parental gene. For example, this could occur if the parental gene is regulated by a cis-NAT that is copied as a trans-NAT in the pseudogene. To support this we identified human pseudogenes with a trans-NAT to the parental gene in their antisense strand by analysis of the database of expressed sequence tags (ESTs). We found that the mutations that appeared in these trans-NATs after the pseudogene formation do not show the flat distribution that would be expected in a non functional transcript. Instead, we found higher similarity to the parental gene in a region nearby the 3' end of the trans-NATs.

Conclusions: Our results do not imply a functional relation of the trans-NAT arising from pseudogenes over their respective parental genes but add evidence for it and stress the importance of duplication mechanisms of genetic material in the generation of non-coding RNAs. We also provide a plausible explanation for the large transcripts that can be found in the antisense strand of some pseudogenes.

Figure 1

Procedure of analysis followed in this study. (A) Selection of transcripts with PAS and other evidence for transcription (trans-NAT, dark green box; see Methods) in antisense of pseudogenes (hatched box). (B) Next, the corresponding region in the antisense of the parental gene (light green box) is obtained. (C) The alignments of 77 selected trans-NATs to their corresponding regions in the antisense parental genes were used to study their identity levels as represented in Figure 2 below.

Figure 2

Fraction of 77 pseudogenes (putatively expressing an EST in anti-sense) that have a mutation respect to the homologous position in the parental gene at a given position in their sequence. Blue curve: distance is taken upstream the putative PAS of the EST +10 nt (the PAS is at position × = 10); the curve indicates that there is a region of high conservation in a 50 nt region upstream the putative PAS (gray box). Red curve: distance is taken downstream the 5'-end of the EST -100 nt; no region of high conservation is present. Values are averaged in a window of five nucleotides. The difference between the two distributions is statistically significant (p-value = 4.482 × 10^-7 from a Kolmogorov-Smirnov test).

Figure 3

A trans-NAT expressed antisense of a pseudogene. (A) Genomic features in a region of human chromosome 3 (negative strand, coordinates decrease from left to right). EST AI803540 is expressed in this region from positions 75,547,155 to 75,547,600 and represents a trans-NAT expressed antisense of a pseudogene. Another five ESTs have an identical or very close 3'-end position and support the same antisense transcript. "Pooled" indicates that the EST was obtained from pooled human melanocyte, fetal heart, and pregnant uterus. (B) This genomic region is highly identical to the parental gene region in Chromosome 16, from positions 5,068,153 to 5,068,597, as indicated by sequence alignment (partially represented in the figure). The PAS of the trans-NAT ("AATAAA", in yellow) is not conserved in the parental gene. (C) The region of similarity is located in an intron of gene ALG1 and it is antisense of the direction of transcription. No evidence of antisense transcription was found in this region of the parental gene and some mutations happened in the positions aligning to the PAS of the trans-NAT.