Consistent risk group-associated differences in human immunodeficiency virus type 1 vpr, vpu and V3 sequences despite independent evolution

Abstract

Human immunodeficiency virus type 1 vpr, vpu and V3 sequences from 15 homosexual men and 19 intravenous drug users in the Amsterdam Cohort studies were analysed. Previously, we reported that V3 domains of viruses from drug users are distinguishable from those of homosexual men on the basis of two silent mutations. Phylogenetic analysis of vpr, vpu and V3 shows that differences in all three regions correlate with risk group. Two positions in both vpr and vpu were found to differ significantly between the risk groups. The distinguishing positions were confirmed for sequences from 11 Scottish and four German samples. The three regions show relatively independent evolution patterns; they resulted in different phylogenies, the only stable clustering being that based on the risk group distinction. Pairwise differences between sequences of the genes were moderately correlated (around 0.30). Surprisingly, when only silent changes were counted, the correlations dropped almost to zero, indicating that the evolution towards independence was more advanced in the silent than in the non-silent positions. This suggests that selection at the amino acid level is not the primary driving force for the independent evolutionary behaviour of the genes. Recombination, combined with restrictions on certain amino acids because of epistatic interactions between the genes, could be an alternative explanation of this phenomenon.

Viral evolution and epidemiology

“…Other common problems with interpretations have been caused by sampling bias [3,4], statistically questionable correlations [5] or too small a sample size; a widely publicized paper [2 • ] used only one sample (and sequences ranging in size from 38-152 bp) to demonstrate the origin of the 1918 influenza pandemic. Sometimes assumptions about the gene nucleotide substitution rates are included in the phylogeny program which might impact on the outcome [6,7 • ] such as tree branching pattern or bootstrap values, but the effect of these data are hard to evaluate.…”

Genetic differences between human immunodeficiency virus type 1 subpopulations in faeces and serum.

“…We used recently described methods for analysing groups of sequences in order to identify amino acid positions which distinguish groups : calculation of the S-index for separation, and determination of the Pearson correlation coefficients (Kuiken et al, , 1996. The S-index for separation can identify positions that contain different amino acids in the faecal sequences compared to sequences from serum, as well as positions that are conserved in one group and variable in the other.…”

“…The S-index for separation can identify positions that contain different amino acids in the faecal sequences compared to sequences from serum, as well as positions that are conserved in one group and variable in the other. To evaluate the strength of the association between the amino acid positions and the two groups of sequences, a Pearson correlation coefficient was calculated from a faecal sequence matrix and a serum sequence matrix (Kuiken et al, 1996). The only amino acid positions showing a significant S-index and a Pearson correlation above the cut-off were positions 331 and 333.…”

“…Probability (P) values were determined by using a two-tailed Fisher's exact test. Separation indexes (S-indexes) and the Pearson correlation coefficients were calculated as described by Kuiken et al (1995Kuiken et al ( , 1996. To determine the cut-off value for the Pearson correlation coefficient, sequences were divided in two random groups 100 times, and for all positions the (random) correlation was calculated.…”

“…The V3 region was sequenced directly from 33 paired faecal and serum samples, and the difference between faeces and serum was determined. Individual clones obtained from 13 persons were sequenced, and recently described methods for analysing groups of sequences were employed (Kuiken et al, , 1996 to identify amino acid motifs possibly involved in tissue specificity.…”