Recombination smooths the time-signal disrupted by latency in within-host HIV phylogenies

Abstract

Within-host HIV evolution involves latency and re-activation of integrated provirus that has the potential to disrupt the temporal signal induced by the evolutionary race between host immune responses and viral evolution. Yet, within-host HIV phylogenies tend to show clear, ladder-like trees structured by the time of sampling. Recombination complicates this dynamic by allowing latent HIV viruses to re-integrate as fragments in the genomes of contemporary virus populations. That is, recombination violates the fundamental assumption made by the phylogenetic methods typically used to study within-host HIV sequence data that evolutionary history can be represented by a single bifurcating tree. In this paper we develop a coalescent-based simulator of within-host HIV evolution that includes, latency, recombination, and population dynamics that allows us to study the relationship between the true, complex genealogy of within-host HIV, encoded as an Ancestral Recombination Graph (ARG), and the observed phylogenetic tree. We show how recombination recovers the disruption of the temporal signal of within-host HIV evolution caused by latency by mixing fragments of ancestral, latent genomes into the contemporary population through recombination. In effect, recombination averages over extant heterogeneity, whether it stems from mixed time-signals or population bottlenecks. Further, we establish that the signals of latency and recombination can be observed in phylogenetic trees despite being an incorrect representation of the true evolutionary history. Using an Approximate Bayesian Computation method, we develop a set of statistical probes to tune our simulation model to nine longitudinally-sampled within-host HIV phylogenies, finding evidence for recombination rates at the lower end of published estimates and relatively small latent pool sizes ranging from about 1000 to 2500 cells.

Migration coupled with recombination explains disparate HIV-1 anatomical compartmentalization signals

“…When analyzing the real HIV-1 anatomical tissue sequence data, for consistency, we used bionj [36] in the R [37] package ape [38] under a TN93 substitution model, with 1000 bootstrap replicates. While neighbor-joining trees have been shown to be inferior to maximum likelihood trees in HIV research [39], we recently showed that in terms of both topological and branch length summary statistics, they give comparable results when recombination is operating [33].…”

“…While the ARG displays the correct evolutionary history, it is difficult to view, and is still very difficult if not impossible to infer on larger real data [43][44][45][46]. On the other hand, standard bifurcating phylogenies cannot properly display recombination, but they are, nevertheless, affected by recombination [33,47]. Therefore, we used a recently developed method to resolve the ARG into an expected standard bifurcating tree to compare our modelling to the familiar phylogeny [33].…”

“…On the other hand, standard bifurcating phylogenies cannot properly display recombination, but they are, nevertheless, affected by recombination [33,47]. Therefore, we used a recently developed method to resolve the ARG into an expected standard bifurcating tree to compare our modelling to the familiar phylogeny [33].…”

“…Typically, within-host HIV sequences are used to infer a phylogenetic tree to which various tests such as the Slatkin-Maddison test of gene flow [23,24] are applied. To generate comparable data, we project the ARG a bifurcating tree using the methods described in detail in Castro et al [33]. This involves first decomposing the ARG into a population of bifurcating trees corresponding to each nucleotide site in the aligned sequences, and then computing a consensus tree from the averaged distance matrix over the set of decomposed trees by assuming a minimum evolution criterion, similar to a neighbor-joining tree [34,35].…”

Elevated HIV viral load is associated with higher recombination ratein vivo

“…In contrast, low viral loads in individuals on incompletely suppressive ART may lead to correspondingly low recombination rates, mitigating recombination of distinct drug resistance mutations on different genetic backgrounds as a path to multi-drug resistance (speculated in [13]). Lower rates of recombination in lower viral load populations may also make phylogenetic approaches that rely on treelike evolutionary divergence more tractable to apply without requiring ancestral recombination graphs [8]. While we did not have access to samples with extremely low viral loads near the limit of detection, future investigations could determine if previously reported average recombination rates ( ≈ 1.4 × 10 − 5 ) can explain low viral load patterns of linkage decay.…”