Anellovirus evolution during long-term chronic infection

Abstract

Human anelloviruses (AVs) are extremely genetically diverse, are widespread in the human population, and cause chronic infections. However, the evolutionary dynamics of AVs within single hosts is currently unknown, and it is unclear whether these changes have an implication on the long-term persistence of AVs in the host. Here, we assessed the evolutionary dynamics of six AV lineages during 30 years of chronic infection at single host resolution. The total number of substitutions and the number of variable sites increased over time. However, not all substitutions reached population fixation, showing that AV lineages form heterogeneous swarms within the host. Most substitutions occurred within a hypervariable region (HVR) located between nucleotide positions 800 and 1,300 of ORF1, which is known to be located within the spike domain. Different regions of the ORF1 gene undergo either positive or negative selection pressure. Sites under strong diversifying selection pressure were detected in the HVR, while the majority of the sites under purifying selection were detected outside this region. The HVR may play the role of an immunological decoy that prevents antibodies from binding to more vulnerable parts of ORF1. Moreover, the frequent substitutions in this region may increase the chances of AV particles escaping immune recognition.

Keywords: anellome; anellovirus; genetic variability; selection pressure; viral swarm; virus evolution.

Figure 1.

The number of time points included in the analysis for each lineage and the age of the subjects.

Figure 2.

Estimation of the HVR. (A) Histogram of the total number of substitutions at each ORF1 position. The logarithm of (B) the total number of substitutions per site and per year, (C) non-synonymous substitutions, (D) and synonymous substitutions, for each selected lineage, within and outside the HVR. The number of substitutions was calculated by comparing the reference genome (the first time point) with the other time points. The P-values (estimated using Wilcoxon sum-rank test) are shown on the top of the boxplots.

Figure 3.

The accumulation of substitutions over time. The relationship between the number of substitutions compared to the reference (first time point of the follow-up) and the age of the subjects. The rho (ρ) and P-values were estimated using Spearman’s rank correlation test. No statistical test was performed for TTV-AMS-S2-04 due to a low number of samples (three time points). The best-fit lines were generated using the linear model.

Figure 4.

Heterogeneity of lineage populations in time. Nucleotide changes in time for (A) TTV-AMS-S1-01, (B) TTV-AMS-S1-23, (C) TTMV-AMS-S1-41, (D) TTV-AMS-S2-01, (E) TTV-AMS-S2-03, and (F) TTV-AMS-S2-04. In each panel, the upper part shows the whole ORF1 length, while the lower part (if present) shows only the hypervariable region. Different colors represent different nucleotide changes, and the strength of the color represents the frequency of the variant.

Figure 5.

Selection pressure in the ORF1 protein. The location of the hypervariable region is indicated with a dashed-line box. The approximate locations of the protein domains were defined based on the results of Liou, Cohen, and Zhang et al. (2022).