Search

Author(s): Mehta, Sanjay R; Murrell, Ben; Anderson, Christy M; Kosakovsky Pond, Sergei L; Wertheim, Joel O; Young, Jason A; Freitas, Lorri; Richman, Douglas D; Mathews, W Chris; Scheffler, Konrad; Little, Susan J; Smith, Davey M | Abstract: BackgroundBecause recently infected individuals disproportionately contribute to the spread of human immunodeficiency virus (HIV), we evaluated the impact of a primary HIV screening program (the Early Test) implemented in San Diego.MethodsThe Early Test program used combined nucleic acid and serology testing to screen for primary infection targeting local high-risk individuals. Epidemiologic, HIV sequence, and geographic data were obtained from the San Diego County Department of Public Health and the Early Test program. Poisson regression analysis was performed to determine whether the Early Test program was temporally and geographically associated with changes in incident HIV diagnoses. Transmission chains were inferred by phylogenetic analysis of sequence data.ResultsOver time, a decrease in incident HIV diagnoses was observed proportional to the number primary HIV infections diagnosed in each San Diego region (P l .001). Molecular network analyses also showed that transmission chains were more likely to terminate in regions where the program was marketed (P = .002). Although, individuals in these zip codes had infection diagnosed earlier (P = .08), they were not treated earlier (P = .83).ConclusionsThese findings suggests that early HIV diagnoses by this primary infection screening program probably contributed to the observed decrease in new HIV diagnoses in San Diego, and they support the expansion and evaluation of similar programs.

Author(s): Yek, Christina; Gianella, Sara; Plana, Montserrat; Castro, Pedro; Scheffler, Konrad; Garcia, Felipe; Massanella, Marta; Smith, Davey M | Abstract: ObjectivesCurative strategies using agents to perturb the HIV reservoir have demonstrated only modest activity, whereas increases in viremia after standard vaccination have been described. We investigated whether vaccination against non-HIV pathogens can induce HIV transcription and thereby play a role in future eradication strategies.DesignA randomized controlled trial (NCT00329251) was performed to compare the effects of clinical vaccines with placebo on HIV transcription and immune activation.MethodsTwenty-six HIV-infected individuals on suppressive antiretroviral therapy were randomized to receive a vaccination schedule (n = 13) or placebo (n = 13). Cell-associated RNA and DNA were extracted from peripheral blood mononuclear cells, and HIV was quantified by droplet digital PCR using primers for gag and 2-LTR (for HIV DNA), unspliced gag RNA (gag usRNA), multispliced tat-rev RNA (tat-rev msRNA) and polyA mRNA.ResultsSignificant increases in gag usRNA after influenza/hepatitis B vaccination (P = 0.02) and in gag usRNA (P = 0.04) and polyA mRNA (P = 0.04) after pneumococcus/hepatitis B vaccination were seen in vaccinees but not controls. HIV DNA and plasma HIV RNA did not change in either group. Increases in CD4 and CD8 T-cell activation markers (P = 0.08 and P l 0.001, respectively) and HIV-specific CD8 responses (P = 0.04 for p24 gag, P = 0.01 for p17 gag and P = 0.04 for total gag) were seen in vaccinees but not controls.ConclusionIn this study, vaccination was associated with increases in HIV cell-associated RNA and HIV-specific responses during antiretroviral therapy. Using standard vaccines to stimulate HIV transcription may therefore be a useful component of future eradication strategies.

Standard genotypic antiretroviral resistance testing, performed by bulk sequencing, does not readily detect variants that comprise 500 copies/ml) of a first-line nonnucleoside reverse transcriptase inhibitor-based regimen, with at least the M184V mutation (lamivudine resistance), and second-line failure of a lopinavir/ritonavir (LPV/r)-based regimen. An amplicon-sequencing approach on the Roche 454 system was used. Six patients with viral loads of >90,000 copies/ml and one patient with a viral load of 520 copies/ml were included. Mutations not detectable by bulk sequencing during first- and second-line failure were detected by deep sequencing during second-line failure. Low-frequency variants (>0.5% of the sequence population) harboring major protease inhibitor resistance mutations were found in 5 of 7 patients despite poor adherence to the LPV/r-based regimen. In patients with intermittent adherence to a boosted PI regimen, deep sequencing may detect minority PI-resistant variants, which likely represent early events in resistance selection. In patients with poor or intermittent adherence, there may be low evolutionary impetus for such variants to reach fixation, explaining the low prevalence of PI resistance.

The RV144 clinical trial showed the partial efficacy of a vaccine regimen with an estimated vaccine efficacy (VE) of 31% for protecting low-risk Thai volunteers against acquisition of HIV-1. The impact of vaccine-induced immune responses can be investigated through sieve analysis of HIV-1 breakthrough infections (infected vaccine and placebo recipients). A V1/V2-targeted comparison of the genomes of HIV-1 breakthrough viruses identified two V2 amino acid sites that differed between the vaccine and placebo groups. Here we extended the V1/V2 analysis to the entire HIV-1 genome using an array of methods based on individual sites, k-mers and genes/proteins. We identified 56 amino acid sites or “signatures” and 119 k-mers that differed between the vaccine and placebo groups. Of those, 19 sites and 38 k-mers were located in the regions comprising the RV144 vaccine (Env-gp120, Gag, and Pro). The nine signature sites in Env-gp120 were significantly enriched for known antibody-associated sites (p = 0.0021). In particular, site 317 in the third variable loop (V3) overlapped with a hotspot of antibody recognition, and sites 369 and 424 were linked to CD4 binding site neutralization. The identified signature sites significantly covaried with other sites across the genome (mean = 32.1) more than did non-signature sites (mean = 0.9) (p < 0.0001), suggesting functional and/or structural relevance of the signature sites. Since signature sites were not preferentially restricted to the vaccine immunogens and because most of the associations were insignificant following correction for multiple testing, we predict that few of the genetic differences are strongly linked to the RV144 vaccine-induced immune pressure. In addition to presenting results of the first complete-genome analysis of the breakthrough infections in the RV144 trial, this work describes a set of statistical methods and tools applicable to analysis of breakthrough infection genomes in general vaccine efficacy trials for diverse pathogens., Author Summary We present an analysis of the genomes of the HIV viruses that infected some participants of the RV144 Thai trial, which was the first study to show efficacy of a vaccine to prevent HIV infection. We analyzed the HIV genomes of infected vaccine recipients and infected placebo recipients, and found differences between them. These differences coincide with previously-studied genetic features that are relevant to the biology of HIV infection, including features involved in immune recognition of the virus. The findings presented here generate testable hypotheses about the mechanism of the partial protection seen in the Thai trial, and may ultimately lead to improved vaccines. The article also presents a toolkit of methods for computational analyses that can be applied to other vaccine efficacy trials.

Motivation: Accurate detection of positive Darwinian selection can provide important insights to researchers investigating the evolution of pathogens. However, many pathogens (particularly viruses) undergo frequent recombination and the phylogenetic methods commonly applied to detect positive selection have been shown to give misleading results when applied to recombining sequences. We propose a method that makes maximum likelihood inference of positive selection robust to the presence of recombination. This is achieved by allowing tree topologies and branch lengths to change across detected recombination breakpoints. Further improvements are obtained by allowing synonymous substitution rates to vary across sites. Results: Using simulation we show that, even for extreme cases where recombination causes standard methods to reach false positive rates >90%, the proposed method decreases the false positive rate to acceptable levels while retaining high power. We applied the method to two HIV-1 datasets for which we have previously found that inference of positive selection is invalid owing to high rates of recombination. In one of these (env gene) we still detected positive selection using the proposed method, while in the other (gag gene) we found no significant evidence of positive selection. Availability: A HyPhy batch language implementation of the proposed methods and the HIV-1 datasets analysed are available at . The HyPhy package is available at , and it is planned that the proposed methods will be included in the next distribution. RDP2 is available at . Contact: konrad@cbio.uct.ac.za, cathal@science.uct.ac.za

Background Sexually transmitted infections spread across contact networks. Partner elicitation and notification are commonly used public health tools to identify, notify, and offer testing to persons linked in these contact networks. For HIV-1, a rapidly evolving pathogen with low per-contact transmission rates, viral genetic sequences are an additional source of data that can be used to infer or refine transmission networks. Methods and Findings The New York City Department of Health and Mental Hygiene interviews individuals newly diagnosed with HIV and elicits names of sexual and injection drug using partners. By law, the Department of Health also receives HIV sequences when these individuals enter healthcare and their physicians order resistance testing. Our study used both HIV sequence and partner naming data from 1342 HIV-infected persons in New York City between 2006 and 2012 to infer and compare sexual/drug-use named partner and genetic transmission networks. Using these networks, we determined a range of genetic distance thresholds suitable for identifying potential transmission partners. In 48% of cases, named partners were infected with genetically closely related viruses, compatible with but not necessarily representing or implying, direct transmission. Partner pairs linked through the genetic similarity of their HIV sequences were also linked by naming in 53% of cases. Persons who reported high-risk heterosexual contact were more likely to name at least one partner with a genetically similar virus than those reporting their risk as injection drug use or men who have sex with men. Conclusions We analyzed an unprecedentedly large and detailed partner tracing and HIV sequence dataset and determined an empirically justified range of genetic distance thresholds for identifying potential transmission partners. We conclude that genetic linkage provides more reliable evidence for identifying potential transmission partners than partner naming, highlighting the importance and complementarity of both epidemiological and molecular genetic surveillance for characterizing regional HIV-1 epidemics., Author Summary Understanding the path over which viruses such as HIV have been transmitted may be crucial for directing public health resources and guiding policy decisions. Contact tracing of named sexual and injection drug-use partners of people recently diagnosed with HIV is an indispensible tool for reconstructing this transmission network. Viral genetic sequence data—routinely collected by public health agencies—can also be used to infer the dynamics of HIV transmission. We analyzed partner naming and viral genetic sequence data in 1342 people living with HIV in New York City reported to the New York City Department of Health and Mental Hygiene between 2006 and 2012. Genetically linked partners were more likely to be named partners than named partners were to be genetically linked. This finding indicates that genetic sequence data are better than partner naming data for reconstructing this viral transmission network. Importantly, the success rate in naming a genetically linked partner varied by transmission risk category (e.g., men who have sex with men, heterosexuals, and injection drug users). This study validates the use viral genetic sequences in reconstructing these viral transmission networks in a public health surveillance setting.

Since its identification in 1983, HIV-1 has been the focus of a research effort unprecedented in scope and difficulty, whose ultimate goals — a cure and a vaccine – remain elusive. One of the fundamental challenges in accomplishing these goals is the tremendous genetic variability of the virus, with some genes differing at as many as 40% of nucleotide positions among circulating strains. Because of this, the genetic bases of many viral phenotypes, most notably the susceptibility to neutralization by a particular antibody, are difficult to identify computationally. Drawing upon open-source general-purpose machine learning algorithms and libraries, we have developed a software package IDEPI (IDentify EPItopes) for learning genotype-to-phenotype predictive models from sequences with known phenotypes. IDEPI can apply learned models to classify sequences of unknown phenotypes, and also identify specific sequence features which contribute to a particular phenotype. We demonstrate that IDEPI achieves performance similar to or better than that of previously published approaches on four well-studied problems: finding the epitopes of broadly neutralizing antibodies (bNab), determining coreceptor tropism of the virus, identifying compartment-specific genetic signatures of the virus, and deducing drug-resistance associated mutations. The cross-platform Python source code (released under the GPL 3.0 license), documentation, issue tracking, and a pre-configured virtual machine for IDEPI can be found at https://github.com/veg/idepi.

The evolution of substitutions conferring drug resistance to HIV-1 is both episodic, occurring when patients are on antiretroviral therapy, and strongly directional, with site-specific resistant residues increasing in frequency over time. While methods exist to detect episodic diversifying selection and continuous directional selection, no evolutionary model combining these two properties has been proposed. We present two models of episodic directional selection (MEDS and EDEPS) which allow the a priori specification of lineages expected to have undergone directional selection. The models infer the sites and target residues that were likely subject to directional selection, using either codon or protein sequences. Compared to its null model of episodic diversifying selection, MEDS provides a superior fit to most sites known to be involved in drug resistance, and neither one test for episodic diversifying selection nor another for constant directional selection are able to detect as many true positives as MEDS and EDEPS while maintaining acceptable levels of false positives. This suggests that episodic directional selection is a better description of the process driving the evolution of drug resistance., Author Summary When exposed to treatment, HIV-1 and other rapidly evolving viruses have the capacity to acquire drug resistance mutations (DRAMs), which limit the efficacy of antivirals. There are a number of experimentally well characterized HIV-1 DRAMs, but many mutations whose roles are not fully understood have also been reported. In this manuscript we construct evolutionary models that identify the locations and targets of mutations conferring resistance to antiretrovirals from viral sequences sampled from treated and untreated individuals. While the evolution of drug resistance is a classic example of natural selection, existing analyses fail to detect the majority of DRAMs. We show that, in order to identify resistance mutations from sequence data, it is necessary to recognize that in this case natural selection is both episodic (it only operates when the virus is exposed to the drugs) and directional (only mutations to a particular amino-acid confer resistance while allowing the virus to continue replicating). The new class of models that allow for the episodic and directional nature of adaptive evolution performs very well at recovering known DRAMs, can be useful at identifying unknown resistance-associated mutations, and is generally applicable to a variety of biological scenarios where similar selective forces are at play.

Codon models of evolution have facilitated the interpretation of selective forces operating on genomes. These models, however, assume a single rate of non-synonymous substitution irrespective of the nature of amino acids being exchanged. Recent developments have shown that models which allow for amino acid pairs to have independent rates of substitution offer improved fit over single rate models. However, these approaches have been limited by the necessity for large alignments in their estimation. An alternative approach is to assume that substitution rates between amino acid pairs can be subdivided into rate classes, dependent on the information content of the alignment. However, given the combinatorially large number of such models, an efficient model search strategy is needed. Here we develop a Genetic Algorithm (GA) method for the estimation of such models. A GA is used to assign amino acid substitution pairs to a series of rate classes, where is estimated from the alignment. Other parameters of the phylogenetic Markov model, including substitution rates, character frequencies and branch lengths are estimated using standard maximum likelihood optimization procedures. We apply the GA to empirical alignments and show improved model fit over existing models of codon evolution. Our results suggest that current models are poor approximations of protein evolution and thus gene and organism specific multi-rate models that incorporate amino acid substitution biases are preferred. We further anticipate that the clustering of amino acid substitution rates into classes will be biologically informative, such that genes with similar functions exhibit similar clustering, and hence this clustering will be useful for the evolutionary fingerprinting of genes., Author Summary Evolution in protein-coding DNA sequences can be modeled at three levels: nucleotides, amino acids or codons that encode the amino acids. Codon models incorporate nucleotide and amino acid information, and allow the estimation of the rate at which amino acids are replaced () versus the rate at which they are preserved (). The ratio has been used in thousands of studies to detect molecular footprints of natural selection. A serious limitation of most codon models is the unrealistic assumption that all non-synonymous substitutions occur at the same rate. Indeed, amino acid models have consistently demonstrated that different residues are exchanged more or less frequently, depending on incompletely understood factors. We derive and validate a computational approach for inferring codon models which combine the power to investigate natural selection with data-driven amino acid substitution biases from alignments. The addition of amino acid properties can lead to more powerful and accurate methods for studying natural selection and the evolutionary history of protein-coding sequences. The pattern of amino acid substitutions specific to a given alignment can be used to compare and contrast the evolutionary properties of different genes, providing an evolutionary analog to protein family comparisons.

Over time, natural selection molds every gene into a unique mosaic of sites evolving rapidly or resisting change—an “evolutionary fingerprint” of the gene. Aspects of this evolutionary fingerprint, such as the site-specific ratio of nonsynonymous to synonymous substitution rates (dN/dS), are commonly used to identify genetic features of potential biological interest; however, no framework exists for comparing evolutionary fingerprints between genes. We hypothesize that protein-coding genes with similar protein structure and/or function tend to have similar evolutionary fingerprints and that comparing evolutionary fingerprints can be useful for discovering similarities between genes in a way that is analogous to, but independent of, discovery of similarity via sequence-based comparison tools such as Blast. To test this hypothesis, we develop a novel model of coding sequence evolution that uses a general bivariate discrete parameterization of the evolutionary rates. We show that this approach provides a better fit to the data using a smaller number of parameters than existing models. Next, we use the model to represent evolutionary fingerprints as probability distributions and present a methodology for comparing these distributions in a way that is robust against variations in data set size and divergence. Finally, using sequences of three rapidly evolving RNA viruses (HIV-1, hepatitis C virus, and influenza A virus), we demonstrate that genes within the same functional group tend to have similar evolutionary fingerprints. Our framework provides a sound statistical foundation for efficient inference and comparison of evolutionary rate patterns in arbitrary collections of gene alignments, clustering homologous and nonhomologous genes, and investigation of biological and functional correlates of evolutionary rates.

Background The cytotoxic T-lymphocyte immune response is important in controlling HIV-1 replication in infected humans. In this immune pathway, viral peptides within infected cells are presented to T-lymphocytes by the polymorphic human leukocyte antigens (HLA). HLA alleles exert selective pressure on the peptide regions and immune escape mutations that occur at some of the targeted sites can enable the virus to adapt to the infected host. The pattern of ongoing immune escape and reversion associated with several human HLA alleles has been studied extensively. Such mutations revert upon transmission to a host without the HLA allele because the escape mutation incurs a fitness cost. However, to-date there has been little attempt to study permanent loss of CTL epitopes due to escape mutations without an effect on fitness. Results Here, we set out to determine the extent of adaptation of HIV-1 to three well-characterized HLA alleles during the initial exposure of the virus to the human cytotoxic immune responses following transmission from chimpanzee. We generated a chimpanzee consensus sequence to approximate the virus sequence that was initially transmitted to the human host and used a method based on peptide binding affinity to HLA crystal structures to predict peptides that were potentially targeted by the HLA alleles on this sequence. Next, we used codon-based phylogenetic models to quantify the average selective pressure that acted on these regions during the period immediately following the zoonosis event, corresponding to the branch of the phylogenetic tree leading to the common ancestor of all of the HIV-1 sequences. Evidence for adaptive evolution during this period was observed at regions recognised by HLA A*6801 and A*0201, both of which are common in African populations. No evidence of adaptive evolution was observed at sites targeted by HLA-B*2705, which is a rare allele in African populations. Conclusion Our results suggest that the ancestral HIV-1 virus experienced a period of positive selective pressure due to immune responses associated with HLA alleles that were common in the infected human population. We propose that this resulted in permanent escape from immune responses targeting unconstrained regions of the virus.

Background Positive selection pressure acting on protein-coding sequences is usually inferred when the rate of nonsynonymous substitution is greater than the synonymous rate. However, purifying selection acting directly on the nucleotide sequence can lower the synonymous substitution rate. This could result in false inference of positive selection because when synonymous changes at some sites are under purifying selection, the average synonymous rate is an underestimate of the neutral rate of evolution. Even though HIV-1 coding sequences contain a number of regions that function at the nucleotide level, and are thus likely to be affected by purifying selection, studies of positive selection assume that synonymous substitutions can be used to estimate the neutral rate of evolution. Results We modelled site-to-site variation in the synonymous substitution rate across coding regions of the HIV-1 genome. Synonymous substitution rates were found to vary significantly within and between genes. Surprisingly, regions of the genome that encode proteins in more than one frame had significantly higher synonymous substitution rates than regions coding in a single frame. We found evidence of strong purifying selection pressure affecting synonymous mutations in fourteen regions with known functions. These included an exonic splicing enhancer, the rev-responsive element, the poly-purine tract and a transcription factor binding site. A further five highly conserved regions were located within known functional domains. We also found four conserved regions located in env and vpu which have not been characterized previously. Conclusion We provide the coordinates of genomic regions with markedly lower synonymous substitution rates, which are putatively under the influence of strong purifying selection pressure at the nucleotide level as well as regions encoding proteins in more than one frame. These regions should be excluded from studies of positive selection acting on HIV-1 coding regions.

Host immune responses against infectious pathogens exert strong selective pressures favouring the emergence of escape mutations that prevent immune recognition. Escape mutations within or flanking functionally conserved epitopes can occur at a significant cost to the pathogen in terms of its ability to replicate effectively. Such mutations come under selective pressure to revert to the wild type in hosts that do not mount an immune response against the epitope. Amino acid positions exhibiting this pattern of escape and reversion are of interest because they tend to coincide with immune responses that control pathogen replication effectively. We have used a probabilistic model of protein coding sequence evolution to detect sites in HIV-1 exhibiting a pattern of rapid escape and reversion. Our model is designed to detect sites that toggle between a wild type amino acid, which is susceptible to a specific immune response, and amino acids with lower replicative fitness that evade immune recognition. Through simulation, we show that this model has significantly greater power to detect selection involving immune escape and reversion than standard models of diversifying selection, which are sensitive to an overall increased rate of non-synonymous substitution. Applied to alignments of HIV-1 protein coding sequences, the model of immune escape and reversion detects a significantly greater number of adaptively evolving sites in env and nef. In all genes tested, the model provides a significantly better description of adaptively evolving sites than standard models of diversifying selection. Several of the sites detected are corroborated by association between Human Leukocyte Antigen (HLA) and viral sequence polymorphisms. Overall, there is evidence for a large number of sites in HIV-1 evolving under strong selective pressure, but exhibiting low sequence diversity. A phylogenetic model designed to detect rapid toggling between wild type and escape amino acids identifies a larger number of adaptively evolving sites in HIV-1, and can in some cases correctly identify the amino acid that is susceptible to the immune response., Author Summary Viruses, such as HIV, are able to evade host immune responses through escape mutations, yet sometimes they do so at a cost. This cost is the reduction in the ability of the virus to replicate, and thus selective pressure exists for a virus to revert to its original state in the absence of the host immune response that caused the initial escape mutation. This pattern of escape and reversion typically occurs when viruses are transmitted between individuals with different immune responses. We develop a phylogenetic model of immune escape and reversion and provide evidence that it outperforms existing models for the detection of selective pressure associated with host immune responses. Finally, we demonstrate that amino acid toggling is a pervasive process in HIV-1 evolution, such that many of the positions in the virus that evolve rapidly, under the influence of positive Darwinian selection, nonetheless display quite low sequence diversity. This highlights the limitations of HIV-1 evolution, and sites such as these are potentially good targets for HIV-1 vaccines.

Understanding how pathogens acquire resistance to drugs is important for the design of treatment strategies, particularly for rapidly evolving viruses such as HIV-1. Drug treatment can exert strong selective pressures and sites within targeted genes that confer resistance frequently evolve far more rapidly than the neutral rate. Rapid evolution at sites that confer resistance to drugs can be used to help elucidate the mechanisms of evolution of drug resistance and to discover or corroborate novel resistance mutations. We have implemented standard maximum likelihood methods that are used to detect diversifying selection and adapted them for use with serially sampled reverse transcriptase (RT) coding sequences isolated from a group of 300 HIV-1 subtype C-infected women before and after single-dose nevirapine (sdNVP) to prevent mother-to-child transmission. We have also extended the standard models of codon evolution for application to the detection of directional selection. Through simulation, we show that the directional selection model can provide a substantial improvement in sensitivity over models of diversifying selection. Five of the sites within the RT gene that are known to harbor mutations that confer resistance to nevirapine (NVP) strongly supported the directional selection model. There was no evidence that other mutations that are known to confer NVP resistance were selected in this cohort. The directional selection model, applied to serially sampled sequences, also had more power than the diversifying selection model to detect selection resulting from factors other than drug resistance. Because inference of selection from serial samples is unlikely to be adversely affected by recombination, the methods we describe may have general applicability to the analysis of positive selection affecting recombining coding sequences when serially sampled data are available.

BackgroundMeasurement of HIV DNA-bearing cells in cerebrospinal fluid (CSF) is challenging because few cells are present. We present a novel application of the sensitive droplet digital (dd)PCR in this context.MethodsWe analyzed CSF cell pellets and paired peripheral blood mononuclear cells (PBMC) from 28 subjects, 19 of whom had undetectable HIV RNA (ResultsHIV DNA was detected in 18/28 (64%) CSF cell pellets, including 10/19 (53%) samples with undetectable HIV RNA. HIV DNA levels in CSF cell pellets were not correlated with RPP30 (p = 0.3), but correlated positively with HIV RNA in CSF (p = 0.04) and HIV DNA in PBMC (p = 0.03). Cellular HIV DNA in CSF was detected in comparable levels in HIV RNA-suppressed and unsuppressed subjects (p = 0.14). In contrast, HIV DNA levels in PBMC were significantly lower in HIV RNA-suppressed than in unsuppressed subjects (p = 0.014). Among subjects with detectable HIV DNA in both compartments, HIV DNA levels in CSF were significantly higher than in PBMC (pConclusionsDespite low mononuclear cell numbers in CSF, HIV DNA was detected in most virally suppressed individuals. In contrast to PBMC, suppressive ART was not associated with lower HIV DNA levels in CSF cells, compared to no ART, perhaps due to poorer ART penetration, slower decay of HIV DNA, or enrichment of HIV DNA-bearing mononuclear cells into the CSF, compared to blood. Future studies should determine what fraction of HIV DNA is replication-competent in CSF leukocytes, compared to PBMC.

To the Editor—A recent article by Redd et al [1] reported preferential transmission of human immunodeficiency virus type 1 (HIV-1) most closely related to strains sampled at least 2 years prior to transmission (ancestral), compared with strains sampled just after transmission (contemporary), in heterosexual couples in Rakai, Uganda. Characterization of the viral population in the recipient is essential when determining both the source of the transmitted virus in the donor and establishing the viral characteristics that may favor successful transmission. We contend that the data do not support the conclusion by Redd et al because of a common misinterpretation of relatedness in an evolutionary framework [2] and statistical issues associated with computing genetic distances between short, highly similar sequences. The closest relative of a given taxon—whether it be nucleotide sequence, strain, or species—is the taxon (or taxa) with which it shares a most recent common ancestor. Once a phylogeny has been inferred, genetic distance is inconsequential when determining relatedness. For example, humans are more closely related to both chimpanzees and bonobos than we are to gorillas, because we share a common ancestor with chimpanzees and bonobos more recently than we share a common ancestor with gorillas (Figure ​(Figure11A) [3]. Furthermore, even though the genetic distance separating gorillas and chimpanzees (0.245 substitutions/site) is shorter than the distance between gorillas and humans (0.282 substitutions/site), gorillas are equally related to both humans and chimpanzees. Figure 1. Depictions of phylogenetic relationships. (A) Phylogeny of African apes inferred from the coding region of mitochondrial genomes. Nucleotide substitutions per site are shown on each branch. (B) Idealized phylogeny in which recipient viruses are monophyletic ... When using phylogenetic analysis to determine the source of HIV-1 transmission, these points are essential. The viral source population in the donor (ancestral vs contemporary) can be identified by determining which donor viruses share the most recent common ancestor with virus in the recipient. We reanalyzed viral sequences from the 9 couples subjected to 454 next-generation sequencing (Figures 3 and 4 in the article by Redd et al [1]), before and after transmission, using a molecular clock in a Bayesian Markov chain Monte Carlo framework [4]. In 4 couples (couples 8–11), viruses from the recipient were monophyletic and shared a most recent common ancestor with contemporary viruses from the donor (posterior probability, ≥ .98; Figure ​Figure11B). In 3 couples, (couples 1, 12, and 14) viruses from the recipient were intermixed with contemporary donor viruses and were therefore more closely related to contemporary donor viruses than ancestral donor viruses (posterior probability, ≥ .99; Figure ​Figure11C). In 2 couples (couples 6 and 15), the phylogenies were poorly resolved because of insufficient differentiation between the ancestral and contemporary donor viruses. In no couple was there phylogenetic evidence to support recipient viruses being the closest relatives of ancestral donor viruses (Figure ​(Figure11D). Therefore, our analysis finds it unlikely that ancestral viruses were preferentially transmitted from a sequestered, long-lived reservoir or from a population persisting at low levels in the serum, as suggested by Redd et al [1]. In Figure 1 in the article by Redd et al [1], similarity between ancestral donor and recipient sequences was illustrated by plotting point estimates of genetic distances between donor sequences, sampled at different times, and recipient sequences. Genetic distances between 2 sequences are estimated using statistical procedures, and distances inferred from relatively short nucleotide sequences are imprecise and should be compared with care. For example, in 1 couple ({"type":"entrez-nucleotide","attrs":{"text":"F04331","term_id":"670985","term_text":"F04331"}}F04331 and {"type":"entrez-nucleotide","attrs":{"text":"H50053","term_id":"989894","term_text":"H50053"}}H50053), the recipient sequence had 1 nucleotide difference from the ancestral donor sequence and 3 nucleotide differences from the contemporary donor sequence. These differences yielded distance estimates (Tamura-Nei 93 [5]) of 0.25% (95% bootstrap confidence interval [CI], 0.0%–1.25%) and 0.76% (95% bootstrap CI, 0.23%–1.8%), respectively. Because of wide overlapping bootstrap CIs, one cannot determine which donor sequence is more genetically similar to the recipient sequence (posterior probability, .51, by the likelihood ratio test [6]). Indeed, we found sufficient signal to assert that the recipient sequence is genetically more similar to the ancestral donor sequence in only 1 of 22 total couples (posterior probability, ≤.05; no correction for multiple testing). It remains an open question whether viruses in some recipients are genetically more similar to the ancestral donor population, compared with the contemporary donor population. An increase in the evolutionary rate in donor virus later in infection or a slowdown in the rate of viral evolution in the recipient, due to changes in natural selection or demography (ie, bottleneck followed by population expansion), would account for this phenomenon. Alternatively, convergent evolution (ie, reversion to an ancestral genotype/phenotype) is a possible explanation for decreased genetic distance separating recipient sequences from the ancestral donor viruses [7–8]. Nevertheless, none of these scenarios point to the preferential transmission of ancestral viruses.