Limits...
Patterns of recombination in HIV-1M are influenced by selection disfavouring the survival of recombinants with disrupted genomic RNA and protein structures.

Golden M, Muhire BM, Semegni Y, Martin DP - PLoS ONE (2014)

Bottom Line: We confirm that chimaeric Gag p24, reverse transcriptase, integrase, gp120 and Nef proteins that are expressed by natural HIV-1 recombinants have significantly lower degrees of predicted folding disruption than randomly generated recombinants.Similarly, we use a novel single-stranded RNA folding disruption test to show that there is significant, albeit weak, evidence that natural HIV recombinants tend to have genomic secondary structures that more closely resemble parental structures than do randomly generated recombinants.These results are consistent with the hypothesis that natural selection has acted both in the short term to purge recombinants with disrupted RNA and protein folds, and in the longer term to modify the genome architecture of HIV to ensure that recombination prone sites correspond with those where recombination will be minimally deleterious.

View Article: PubMed Central - PubMed

Affiliation: Department of Statistics, University of Oxford, Oxford, United Kingdom.

ABSTRACT
Genetic recombination is a major contributor to the ongoing diversification of HIV. It is clearly apparent that across the HIV-genome there are defined recombination hot and cold spots which tend to co-localise both with genomic secondary structures and with either inter-gene boundaries or intra-gene domain boundaries. There is also good evidence that most recombination breakpoints that are detectable within the genes of natural HIV recombinants are likely to be minimally disruptive of intra-protein amino acid contacts and that these breakpoints should therefore have little impact on protein folding. Here we further investigate the impact on patterns of genetic recombination in HIV of selection favouring the maintenance of functional RNA and protein structures. We confirm that chimaeric Gag p24, reverse transcriptase, integrase, gp120 and Nef proteins that are expressed by natural HIV-1 recombinants have significantly lower degrees of predicted folding disruption than randomly generated recombinants. Similarly, we use a novel single-stranded RNA folding disruption test to show that there is significant, albeit weak, evidence that natural HIV recombinants tend to have genomic secondary structures that more closely resemble parental structures than do randomly generated recombinants. These results are consistent with the hypothesis that natural selection has acted both in the short term to purge recombinants with disrupted RNA and protein folds, and in the longer term to modify the genome architecture of HIV to ensure that recombination prone sites correspond with those where recombination will be minimally deleterious.

Show MeSH

Related in: MedlinePlus

The predicted sensitivity of HIV-1M proteins to recombinational disruption.(A) Depicted are the means (black lines) and ranges (gray backgrounds) of predicted degrees of recombination-induced folding disruption in various HIV-1 proteins (those for which suitable atomic resolution three dimensional structures are available). The white areas interspersed between the gray areas are positions where there was no protein structure data available or where there were extra amino acids inserted into the alignment that were not present in the protein structure used. For all genome regions that had associated protein structure data, all conceivable single breakpoint recombinants were simulated using parental sequences that resembled as closely as possible the parental sequences of actual recombinant viruses with single detectable recombination breakpoints in these genome regions. Amino acid substitution rates and breakpoint positions occurring in these actual HIV-1 recombinants are displayed at the top of the figure. (B) Recombination breakpoint density plot illustrating breakpoint positions detected across 434 detectable HIV-1M recombination events (After [14]). Light and dark grey areas respectively indicate the 95% and 99% confidence intervals of breakpoint numbers that would have been detectable in different genome locations under random recombination. The grey areas undulate with degrees of sequence conservation because recombination events are more easily detectable in genome regions that are genetically diverse. Note firstly that the peaks of the plots in A indicate recombination breakpoint positions that are predicted to have the greatest disruptive effects on protein folding, and secondly that in actual recombinant HIV-1M genomes sampled from nature these “disruptive breakpoint positions” tend to correspond in plot B with regions of low recombination breakpoint densities.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4061080&req=5

pone-0100400-g002: The predicted sensitivity of HIV-1M proteins to recombinational disruption.(A) Depicted are the means (black lines) and ranges (gray backgrounds) of predicted degrees of recombination-induced folding disruption in various HIV-1 proteins (those for which suitable atomic resolution three dimensional structures are available). The white areas interspersed between the gray areas are positions where there was no protein structure data available or where there were extra amino acids inserted into the alignment that were not present in the protein structure used. For all genome regions that had associated protein structure data, all conceivable single breakpoint recombinants were simulated using parental sequences that resembled as closely as possible the parental sequences of actual recombinant viruses with single detectable recombination breakpoints in these genome regions. Amino acid substitution rates and breakpoint positions occurring in these actual HIV-1 recombinants are displayed at the top of the figure. (B) Recombination breakpoint density plot illustrating breakpoint positions detected across 434 detectable HIV-1M recombination events (After [14]). Light and dark grey areas respectively indicate the 95% and 99% confidence intervals of breakpoint numbers that would have been detectable in different genome locations under random recombination. The grey areas undulate with degrees of sequence conservation because recombination events are more easily detectable in genome regions that are genetically diverse. Note firstly that the peaks of the plots in A indicate recombination breakpoint positions that are predicted to have the greatest disruptive effects on protein folding, and secondly that in actual recombinant HIV-1M genomes sampled from nature these “disruptive breakpoint positions” tend to correspond in plot B with regions of low recombination breakpoint densities.

Mentions: Figure 2A illustrates degrees of intra-protein amino-acid – amino-acid interaction disruption (where higher E-values equate with greater disruption) that are predicted to occur within various HIV-1 proteins between HIV variants that have previously been observed to recombine (Figure 2B). The average degrees of predicted recombination-induced folding disruption in the gp120 and Nef proteins are appreciably higher than those predicted in the other HIV-1 proteins for which extensive high resolution structural data is available. Consistent with the notion that these proteins may be particularly sensitive to recombination-induced folding disruption is the fact that, in actual HIV-1M recombinants, breakpoints only very rarely occur at sites where they are anticipated to have a maximally disruptive effect on the folding of these proteins. A plausible explanation for gp120 and Nef being particularly sensitive to recombination-induced folding disruption relative to the other HIV-1M proteins examined here, is that they are less conserved than these other proteins (see amino acid substitution rate plot in Figure 2A) and recombinant versions of gp120 and Nef will therefore tend to have many more potentially disruptive amino acid combinations.


Patterns of recombination in HIV-1M are influenced by selection disfavouring the survival of recombinants with disrupted genomic RNA and protein structures.

Golden M, Muhire BM, Semegni Y, Martin DP - PLoS ONE (2014)

The predicted sensitivity of HIV-1M proteins to recombinational disruption.(A) Depicted are the means (black lines) and ranges (gray backgrounds) of predicted degrees of recombination-induced folding disruption in various HIV-1 proteins (those for which suitable atomic resolution three dimensional structures are available). The white areas interspersed between the gray areas are positions where there was no protein structure data available or where there were extra amino acids inserted into the alignment that were not present in the protein structure used. For all genome regions that had associated protein structure data, all conceivable single breakpoint recombinants were simulated using parental sequences that resembled as closely as possible the parental sequences of actual recombinant viruses with single detectable recombination breakpoints in these genome regions. Amino acid substitution rates and breakpoint positions occurring in these actual HIV-1 recombinants are displayed at the top of the figure. (B) Recombination breakpoint density plot illustrating breakpoint positions detected across 434 detectable HIV-1M recombination events (After [14]). Light and dark grey areas respectively indicate the 95% and 99% confidence intervals of breakpoint numbers that would have been detectable in different genome locations under random recombination. The grey areas undulate with degrees of sequence conservation because recombination events are more easily detectable in genome regions that are genetically diverse. Note firstly that the peaks of the plots in A indicate recombination breakpoint positions that are predicted to have the greatest disruptive effects on protein folding, and secondly that in actual recombinant HIV-1M genomes sampled from nature these “disruptive breakpoint positions” tend to correspond in plot B with regions of low recombination breakpoint densities.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4061080&req=5

pone-0100400-g002: The predicted sensitivity of HIV-1M proteins to recombinational disruption.(A) Depicted are the means (black lines) and ranges (gray backgrounds) of predicted degrees of recombination-induced folding disruption in various HIV-1 proteins (those for which suitable atomic resolution three dimensional structures are available). The white areas interspersed between the gray areas are positions where there was no protein structure data available or where there were extra amino acids inserted into the alignment that were not present in the protein structure used. For all genome regions that had associated protein structure data, all conceivable single breakpoint recombinants were simulated using parental sequences that resembled as closely as possible the parental sequences of actual recombinant viruses with single detectable recombination breakpoints in these genome regions. Amino acid substitution rates and breakpoint positions occurring in these actual HIV-1 recombinants are displayed at the top of the figure. (B) Recombination breakpoint density plot illustrating breakpoint positions detected across 434 detectable HIV-1M recombination events (After [14]). Light and dark grey areas respectively indicate the 95% and 99% confidence intervals of breakpoint numbers that would have been detectable in different genome locations under random recombination. The grey areas undulate with degrees of sequence conservation because recombination events are more easily detectable in genome regions that are genetically diverse. Note firstly that the peaks of the plots in A indicate recombination breakpoint positions that are predicted to have the greatest disruptive effects on protein folding, and secondly that in actual recombinant HIV-1M genomes sampled from nature these “disruptive breakpoint positions” tend to correspond in plot B with regions of low recombination breakpoint densities.
Mentions: Figure 2A illustrates degrees of intra-protein amino-acid – amino-acid interaction disruption (where higher E-values equate with greater disruption) that are predicted to occur within various HIV-1 proteins between HIV variants that have previously been observed to recombine (Figure 2B). The average degrees of predicted recombination-induced folding disruption in the gp120 and Nef proteins are appreciably higher than those predicted in the other HIV-1 proteins for which extensive high resolution structural data is available. Consistent with the notion that these proteins may be particularly sensitive to recombination-induced folding disruption is the fact that, in actual HIV-1M recombinants, breakpoints only very rarely occur at sites where they are anticipated to have a maximally disruptive effect on the folding of these proteins. A plausible explanation for gp120 and Nef being particularly sensitive to recombination-induced folding disruption relative to the other HIV-1M proteins examined here, is that they are less conserved than these other proteins (see amino acid substitution rate plot in Figure 2A) and recombinant versions of gp120 and Nef will therefore tend to have many more potentially disruptive amino acid combinations.

Bottom Line: We confirm that chimaeric Gag p24, reverse transcriptase, integrase, gp120 and Nef proteins that are expressed by natural HIV-1 recombinants have significantly lower degrees of predicted folding disruption than randomly generated recombinants.Similarly, we use a novel single-stranded RNA folding disruption test to show that there is significant, albeit weak, evidence that natural HIV recombinants tend to have genomic secondary structures that more closely resemble parental structures than do randomly generated recombinants.These results are consistent with the hypothesis that natural selection has acted both in the short term to purge recombinants with disrupted RNA and protein folds, and in the longer term to modify the genome architecture of HIV to ensure that recombination prone sites correspond with those where recombination will be minimally deleterious.

View Article: PubMed Central - PubMed

Affiliation: Department of Statistics, University of Oxford, Oxford, United Kingdom.

ABSTRACT
Genetic recombination is a major contributor to the ongoing diversification of HIV. It is clearly apparent that across the HIV-genome there are defined recombination hot and cold spots which tend to co-localise both with genomic secondary structures and with either inter-gene boundaries or intra-gene domain boundaries. There is also good evidence that most recombination breakpoints that are detectable within the genes of natural HIV recombinants are likely to be minimally disruptive of intra-protein amino acid contacts and that these breakpoints should therefore have little impact on protein folding. Here we further investigate the impact on patterns of genetic recombination in HIV of selection favouring the maintenance of functional RNA and protein structures. We confirm that chimaeric Gag p24, reverse transcriptase, integrase, gp120 and Nef proteins that are expressed by natural HIV-1 recombinants have significantly lower degrees of predicted folding disruption than randomly generated recombinants. Similarly, we use a novel single-stranded RNA folding disruption test to show that there is significant, albeit weak, evidence that natural HIV recombinants tend to have genomic secondary structures that more closely resemble parental structures than do randomly generated recombinants. These results are consistent with the hypothesis that natural selection has acted both in the short term to purge recombinants with disrupted RNA and protein folds, and in the longer term to modify the genome architecture of HIV to ensure that recombination prone sites correspond with those where recombination will be minimally deleterious.

Show MeSH
Related in: MedlinePlus