Secondary Logo

Journal Logo

Disentangling the impact of within-host evolution and transmission dynamics on the tempo of HIV-1 evolution

Vrancken, Brama; Baele, Guya; Vandamme, Anne-Miekea,b; van Laethem, Kristela; Suchard, Marc A.c,d,e; Lemey, Philippea

doi: 10.1097/QAD.0000000000000731
Epidemiology and Social
Free
SDC

Objective: To determine how HIV-1 risk groups impact transmitted diversity and the tempo of viral evolution at a population scale.

Methods: We investigated a set of previously described transmission chains (n = 70) using a population genetic approach, and tested whether the expected differences in proportions of multivariant transmissions are reflected by varying proportions of transmitted diversity between men having sex with men (MSM) and heterosexual (HET) subpopulations – the largest contributors to HIV spread. To assess evolutionary rate differences among the different risk groups, we compiled risk group datasets for subtypes A1, B and CRF01_AE, and directly compared the absolute substitution rate and its synonymous and non-synonymous components.

Results: There was sufficient demographic signal to inform the transmission model in Bayesian evolutionary analysis by sampling trees using env data to compare the transmission bottleneck size between the MSM and HET risk groups. We found no indications for a different proportion of transmitted genetic diversity at the population level between these groups. In the direct rate comparisons between the risk groups, however, we consistently recovered a higher evolutionary rate in the male-dominated risk group compared to the HET datasets.

Conclusion: We find that the risk group composition affects the viral evolutionary rate and therefore potentially also the adaptation rate. In particular, risk group-specific sex ratios, and the variation in within-host evolutionary rates between men and women, impose evolutionary rate differences at the epidemic level, but we cannot exclude a role of varying transmission rates.

Supplemental Digital Content is available in the text

aDepartment of Microbiology and Immunology, Rega Institute, KU Leuven, Leuven, Belgium

bCentro de Malária e Outras Doenças Tropicais Instituto de Higiene e Medicina Tropical and Unidade de Microbiologia, Universidade Nova de Lisboa, Lisboa, Portugal

cDepartment of Biomathematics

dDepartment of Human Genetics, David Geffen School of Medicine at UCLA

eDepartment of Biostatistics, UCLA Fielding School of Public Health, University of California, Los Angeles, California, USA.

Correspondence to Bram Vrancken, Rega Instituut, Minderbroedersstraat 10, 3000 Leuven, Belgium. Tel: +32 16 332892; fax: +32 16 332131; e-mail: bram.vrancken@rega.kuleuven.be

Received 5 December, 2014

Revised 20 April, 2015

Accepted 20 April, 2015

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Website (http://www.AIDSonline.com).

Back to Top | Article Outline

Introduction

The determinants of the HIV type 1 (HIV-1) evolutionary rate and its variability has been subject to extensive investigation. Because the process that controls which genetic variants survive the genetic bottleneck at transmission determines how the enormous within-host evolutionary potential of HIV-1 is translated into long-term evolution at a population-scale level, the impact of transmission dynamics is central to our understanding of HIV evolution. Various observations suggest that transmission has both a drift and a selective component [1–18]; the latter was recently highlighted by a rigorous examination of many transmission pairs by Carlson et al.[19].

Since transmission is not fully deterministic, differences in the magnitude of the virus population bottleneck at transmission – with multivariant transmission estimated to be twice as common in men having sex with men (MSM) (∼40%) than in heterosexual (HET) contacts (∼20%) [2] – may affect the long-term evolution of HIV at the epidemic level. As a case in point, when investigating the within-versus-between host rate difference for HIV-1, we found a markedly lower rate difference in our subtype C transmission chain (an approximately two-fold rate difference) [15] as compared to the earlier results on the basis of subtype B data (a ∼4–5-fold rate difference) [14]. Following the argumentation of Lythgoe and Fraser [13], we hypothesized that the dissimilarity in the magnitude of the rate difference follows from the differences in the underlying biological characteristics associated with transmission between the largely MSM-driven subtype B epidemic and the predominantly HET-driven subtype C epidemic. That is, the smaller the number of viruses being transmitted (in HET), the higher the chance that new infections will be established by variants that avoided the accumulation of mutations in the donors. Consequently, the association between the number of transmitted variants and risk group can bring about risk group composition-related differences in the tempo of HIV evolution. This is not limited to the host human leukocyte antigen background, but can also involve resistance to combination antiretroviral therapy (cART) and vaccines.

In addition to transmission dynamics, the factors that influence the overall amount of divergence accumulating between the founder strain(s) and transmitted virus are also likely to impact the among-host evolutionary rate. Within-host evolutionary rates have, for example, been shown to vary with disease progression [20]. In this respect, it is interesting to note that men tend to have higher set-point viral loads (spVL) – a predictor of disease progression [21] – than women [22–24], a difference that can persist for several years [25]. This suggests a potentially complex interplay between transmission and within-host evolutionary dynamics in determining the tempo of HIV-1 evolution.

Here, we investigated whether the risk group composition can affect the viral evolutionary rate at a population scale. We did this by examining what factor – the transmission dynamics, within-host evolution, or a combination of both – imposes HIV-1 evolutionary rate differences among different risk groups. To explore the impact of the transmission bottleneck, we used a population genetic approach to contrast the loss of genetic diversity at transmission between risk groups. To this end, we compiled a dataset of the previously described HIV-1 transmission chains (n = 70; see Table 1). We described the viral evolutionary histories with a recently introduced transmission model in Bayesian evolutionary analysis by sampling trees (BEAST) [15], and tested for the transmission bottleneck size differences with a Bayesian hierarchical phylogenetic model (HPM) approach [26] that incorporates fixed effects [27]. We found no support for a difference in the loss of genetic diversity between the HET and the MSM groups. To assess the impact of within-host evolution, we compiled risk group-specific datasets of subtypes A1, B and CRF01_AE, and tested for differences in substitution rate (Table 2). For subtype B and CRF01_AE, we found that HIV evolves slower in the HET than in the MSM epidemics, and that for subtype A1, the evolutionary rate is also lower in the HET than in the injecting drug user (IDU) subepidemics, and this may be associated with the varying proportions of men in the examined datasets. These estimates indicate that the within-host evolutionary processes can impact differences in between-host evolutionary rates.

Table 1

Table 1

Table 2

Table 2

Back to Top | Article Outline

Methods

Dataset compilation

For the bottleneck size comparison between the risk groups, we collected time-stamped clonal sequence data for HIV-1 transmission chains available from the HIV database (http://www.hiv.lanl.gov/), including HET, MSM and blood contact risk groups (see Supplementary Text 1, http://links.lww.com/QAD/A706). A summary of the transmission chain data we retrieved is presented in Table 1, and dataset details are listed in Supplementary Table 1, http://links.lww.com/QAD/A738. To compare the evolutionary rates between the risk groups for particular subtypes, we downloaded near-complete genome datasets for subtype A1, B and CRF01_AE from the Los Alamos HIV database (see Supplementary Text 1, http://links.lww.com/QAD/A706). The characteristics of the subtype-specific risk group datasets are provided in Table 2.

Back to Top | Article Outline

Phylogenetic inference

We used the BEAST software package [28] for all phylogenetic and population genetic inferences. The substitution process was described using a Hasegawa–Kishino–Yano (HKY [29]) substitution model with discrete Γ-distribution to model among-site rate heterogeneity [30,31]. Sampling and transmission time uncertainty was integrated out over a known time interval (see Supplementary Text 1, http://links.lww.com/QAD/A706). We used a recently described transmission model [15] to accommodate the transmission history and to quantify the transmission bottleneck in terms of the loss of genetic diversity at transmission. We adopted a Bayesian HPM procedure to share information across transmission chains, and specified hierarchical prior distributions on all parameters of the coalescent model. To formally test for differences in bottleneck size among the risk groups, we specified a fixed effect on the bottleneck size parameter [27]. This approach only models the evolutionary dynamics throughout the individual transmission chains and does not consider the evolutionary process between the different datasets. Although we collected data from gag, pol and env from the HET, MSM and blood contact risk groups, we eventually only retained env data from the HET and MSM groups because many transmission chain datasets do not capture the early population dynamics (see Supplementary Text 1 for more details, http://links.lww.com/QAD/A706).

The sparse within-host sampling scheme of most transmission chains (Supplementary Table 1, http://links.lww.com/QAD/A738) resulted in poor temporal information to calibrate the molecular clock in many datasets. We therefore followed a slight modification of the approach used by Keele et al.[1] to arrive at an informative prior distribution on the evolutionary rate on very short time scales. Specifically, we transformed the standard generation time curve [32] to a rate distribution using the experimentally derived reverse transcriptase error rate [33], and set the upper limit to the rate at a generation time of 18 h [32].

Tutorials describing how to set up the various components of the above analysis can be found at http://beast.bio.ed.ac.uk/ and example XMLs for the various analyses are provided as Supplementary Material. All XMLs used for the analyses are available from the authors upon request.

Back to Top | Article Outline

Among risk group evolutionary rate comparison

We visually inspected a regression of root-to-tip divergences as a function of sampling time using Path-O-Gen (http://tree.bio.ed.ac.uk/software/pathogen/). This revealed that sufficient temporal signal was present in all but the full genome subtype B datasets. To remove the noise from this dataset, we followed the same procedure as in [15] to arrive at a balanced dataset with clear temporal signal (R2 for both > 0.48). The evolutionary rate for all datasets was estimated under a relaxed clock model using a lognormal distribution [34] by fitting the HKY model [29] while allowing for gamma-distributed among-site rate heterogeneity [30,31]. The genome was split into gene-specific partitions to also allow among-gene rate variation. A Skygrid model was specified as a flexible tree prior [35]. To estimate absolute synonymous and nonsynonymous rates for each dataset, we applied a renaissance counting procedure [20] following [15] to the non-overlapping parts of the open reading frames (except for nef for which data was missing for many taxa).

Back to Top | Article Outline

Results

Dataset compilation

We collected data for 70 transmission chains representing the most important risk groups in HIV spread: the MSM, HET and blood contact risk groups. The distribution of the transmission chains by risk group, subtype and the genomic fragments sequenced are listed in Table 1. Because data from the gag and pol genomic regions proved uninformative for estimating the transmission-associated loss in genetic diversity (Supplementary Text 1, http://links.lww.com/QAD/A706), we only report results based on the env datasets.

Back to Top | Article Outline

Support for a transmission bottleneck

We first set out to test whether the env transmission chain datasets we collected (n = 69) support a bottleneck at transmission. To this purpose, we apply a genealogical transmission chain model that allows the effective viral population size to change upon transmission according to different coalescent models [15]: constant population size (CON, no bottleneck), exponential growth (EXP, the population size upon transmission is an estimable proportion of the donor population size and grows exponentially in the new recipient) and logistic growth (LOG, as in EXP, but with logistic growth in the recipient). We independently fitted the transmission model with the three different demographic functions to each dataset and compared their model fit using marginal likelihood testing (Supplementary Text 1, http://links.lww.com/QAD/A706). This indicates that a model accommodating a bottleneck (EXP or LOG) is supported by 91.3% (63/69) of the env transmission chain datasets (Supplementary Fig. 1, http://links.lww.com/QAD/A737). There is, however, a marked difference between the risk groups: whereas we find strong Bayes factor support in 82.35% (28/34) of the HET and 64.3% (18/28) of the MSM transmission events for a model with a bottleneck, this drops to 14.3% (1/7) for the blood contact risk group. This likely reflects less informative datasets due to longer times between time of infection and time of sampling for the recipients in the blood contact transmission pairs (see Supplementary Text 1 and Supplementary Table 1, http://links.lww.com/QAD/A706).

Back to Top | Article Outline

Similar amounts of drift and selection in all risk groups

Unambiguously estimating bottleneck sizes using the transmission model proved difficult for many datasets because of reasons described in Supplementary Text 1 (http://links.lww.com/QAD/A706). We therefore restricted our approach to test differences between the risk groups to the most informative subset of transmission chains for the env region, which were only available for the HET (n = 17) and MSM (n = 11) risk groups (Supplementary Text 1, http://links.lww.com/QAD/A706). The demographic function was parameterized according to the best fitting demographic model in the ‘best fit’ analysis. To test the robustness of our bottleneck size estimates to demographic model specification, we also performed an analysis consistently applying either an EXP or a LOG function to all datasets in the ‘exponential’ and ‘logistic’ analysis, respectively. Using our population genetic approach, the magnitude of the bottleneck is estimated as the proportion of the donor effective population size that is transmitted to the recipient, but we report the complement of this proportion as the percentage of loss in diversity at transmission.

In our test approach, we allowed sharing of information on the demographic parameters across individuals, but modelled potential differences in bottleneck sizes among the risk groups using a fixed effect. We did not find Bayes factor support for the risk group fixed effect, indicating no difference in the loss of genetic diversity in env for both the risk groups, which is unlikely to be caused by differences in donor viral diversity between both the risk groups (Supplementary Text 1, http://links.lww.com/QAD/A706). The severity of the bottleneck was estimated to be 96.8, 96.8 and 98% for the ‘best fit’, ‘exponential’ and ‘logistic’ analyses, respectively, with individual patient estimates ranging from 54 to 99.9%. The average difference in ancestral proportion size estimates between the ‘exponential’ and ‘logistic’ analyses was less than 1%, and there was no trend for either model to consistently estimate higher or lower values for this parameter, indicating that our estimates are robust to the demographic parameterization.

Back to Top | Article Outline

Sex ratio drives risk group evolutionary rate differences

To test for HIV-1 evolutionary rate differences among the different risk groups, we collected near-complete genome data from the HET risk groups for subtype A1, B and CRF01_AE, from the MSM risk groups for subtype B and CRF01_AE and from IDUs for subtype A1 (Table 2). We consistently found slower HIV evolutionary rates in the HET datasets as compared to the MSM datasets. For subtype A1, the HIV evolutionary rate was also slower in the HET risk group compared to the IDU risk group (Fig. 1). In order to asses whether the rate differences reflect variation in selective pressure and/or replication rate (generation time), we followed the approach from [20] to obtain posterior estimates of the absolute synonymous (μS) and nonsynonymous (μN) rates for all risk group datasets (see ‘Methods’ section). This reveals both elevated μS and μN rates in the MSM and IDU datasets when compared to the HET groups for all subtypes, suggesting that the underlying replication rate is lower in the HET groups. By comparing the rates with the proportion of men in the datasets, we saw a consistently higher rate for a higher proportion of men within each subtype (Fig. 2), and found strong Bayes factor support for this (Bayes factors were 8.2, 333.5 and 203.2 for subtypes A1, B and CRF01_AE, respectively). There was, however, no clear linear relationship between the rate and proportion of men independent of the subtype (Fig. 2), suggesting that other factors confound evolutionary rate differences associated with sex composition at the epidemiological level.

Fig. 1

Fig. 1

Fig. 2

Fig. 2

Back to Top | Article Outline

Discussion

We set out to investigate whether the population genetic dynamics associated with transmission, within-host viral replication or a combination of both, are responsible for evolutionary rate differences between the risk groups. For the transmission dynamics to impose such differences, we would expect a difference in bottleneck size among the risk groups and we evaluated this based on a collection of HIV-1 transmission chains for the three most important risk groups (HET, MSM and blood contact) involved in HIV spread. Although 91.3% env transmission chains provide evidence for a transmission-associated population bottleneck, we had to restrict the estimation of its size to the most informative subset for the HET and the MSM risk groups. Formal testing between the two risk groups did not provide any evidence for differences in loss of diversity at transmission, which is very high in both the HET and the MSM groups.

The absence of a bottleneck size difference is in agreement with many founder effects in HET (∼80%) and MSM (∼60%) transmission [3], a scenario under which many transmissions will generally be represented by a single phylogenetic branch connecting the donor and the recipient viral populations. In such circumstances, it is expected the measured loss in genetic diversity at transmission relates to the sampled diversity in the donor and recipient. In this respect, it is important to point out that our comparative analysis included both plasma and peripheral blood mononuclear cell (PBMC) samples. Plasma samples reflect the freely circulating virus, which is usually interpreted as recently generated diversity. PBMCs, however, may represent an archive of both past and current HIV-1 diversity. These cells can also be co-cultured before DNA extraction and the resulting in-vitro picture may not accurately mirror the in-vivo diversity. Reassuringly, however, it has been shown that the different experimental sampling approaches lead to a similar viral genetic composition [36], and there were no specific biases of either plasma or PBMC between our risk group samples. PCR errors or recombination associated with traditional cloning approaches, which can now be avoided by the use of single genome amplification [1], also represent potential sources of error. Other biases may result from differences in the length of infection [37], as well as the therapy history differences in the donor. The lack of support for a difference in the loss of diversity at transmission, however, implies that these potentially confounding factors did not bias the estimates for the risk groups in a particular direction. Also, we measured the bottleneck as the proportion of diversity that is lost at transmission, and in combination with the many founder effects, this explains why, despite the larger share of recently infected donors in the MSM sample, we did not find a difference in the size of the bottleneck between the HET and the MSM risk groups (Supplementary Text 1, http://links.lww.com/QAD/A706). Finally, although the Los Alamos HIV sequence database contains every HIV sequence the world scientific community has ever published and deposited in GenBank, the representativeness of our work depends on what the scientific field has investigated so far.

We estimated the size of the bottleneck in env to approximately 97%. This is in close agreement with the findings by Edwards et al.[39], who inferred a decrease of more than 99% in env and gag genetic diversity at transmission in one MSM couple. By comparing the estimated diversity at transmission in nine MSM and 27 mother-to-child (MTC) transmissions, they also found that the mode of transmission does not seem to impact the severeness of the transmission-associated bottleneck [39]. Taken together with our findings, this indicates that the difference in frequency of multivariant transmission (∼20% between MSM and HET, ∼10% between MSM and MTCT [3]) is too limited to set apart the overall transmitted diversity between the risk groups. The comparable amounts of transmitted diversity imply a similar interplay of drift and selection at transmission in the different risk groups and are therefore not expected to lead to evolutionary rate differences at the epidemic level.

The evolutionary rate comparison among the risk groups indicates differences between HET and MSM, and between HET and IDU, but the direction of these differences may seem counterintuitive in the light of previous findings. There is increasing evidence that the transmission/establishment advantage of ancestral variants leaves its footprint by slowing down the divergence rate among hosts [13–15,40]. Given that this effect seems larger for subtype B than for subtype C [15], we hypothesized that differences in proportion of multivariant transmission may play an important role. Following this reasoning, we would expect the following overall ranking in the evolutionary rate between the risk groups: HET > MSM > IDU. We however, found that the evolutionary rate is lower in HET compared to MSM, and also lower in HET compared to IDU, and this is the case for both absolute synonymous and nonsynonymous rates (Fig. 1). Because synonymous substitution rates, as a marker of viral replication rates, are associated with HIV-1 disease progression and men are predicted to have higher disease progression rates by their viral set-point, we hypothesized that sex ratio may be key to explaining the risk group-associated evolutionary rate differences. This implies that within-host divergence rates, together with a general transmission-associated rate slowdown, impact the tempo of evolution at the population level.

Despite higher rates for the risk groups with a larger proportion of men within each subtype, this relationship is less clear, independent of the subtypes. On one hand, the individuals sampled in our datasets may not accurately reflect the composition of the risk group population from which the samples were taken. On the other hand, the impact of differences in sex-specific within-host evolution may also be confounded by varying transmission rates by stage of infection. This was put forward by Maljkovic Berry et al.[41] to explain the lower subtype A1 rate estimates among IDUs in the former Soviet Union (FSU) as compared to heterosexual transmission in Africa. Although we found opposite differences here between our HET and IDU dataset, this does not necessarily contradict their findings because our IDU sampling is fundamentally different and, as a consequence, does not specifically reflect IDU transmission.

More replication cycles per unit of time in male-dominated epidemics imply more opportunities for nonsynonymous mutations to arise and subsequently become fixed under conditions under which they are beneficial. Because the propagation of mutations in a population depends on several factors, however, this may not easily translate into faster adaptation in, for example, MSM-driven epidemics. One of these factors is the viral genetic background in which mutations are emerging. Also, population genetic theory predicts that the fixation of mutations will mainly be determined by chance events (drift) in small populations, whereas in large populations, selection will determine the fate of mutations. In this respect, the spread of beneficial mutations will be confounded by population structure, which introduces drift. In the UK, for example, the size of heterosexual transmission networks was found to be smaller than among MSM [42,43]. Moreover, in a US-based survey, it was found that MSM not only have more partners but also have more concurrent relationships [44]. Similar to the variation in rates of partner exchange [45], this may impose different levels of drift. The fact that behavioural differences between risk groups are not universal further complicates predictions. In Quebec, for example, transmission clustering did not associate with the mode of transmission [46]. It has also been proposed that a high concurrency of partnerships in heterosexuals may explain why HIV has spread so much more extensively in sub-Saharan Africa, than in other regions [47,48].

In summary, we show that the effect of within-host evolution on the between-host rate dominates over transmission-related events. Because of this, we propose varying sex ratios as driving risk group-related evolutionary rate differences. Knowledge of the factors underlying the differences in the evolutionary dynamics of HIV-1 subtypes may help to predict how the virus will evolve in response to large-scale public health interventions such as vaccination or treatment up-scaling.

Back to Top | Article Outline

Acknowledgements

B.V. conceived and performed the experiments, drafted and wrote the manuscript. M.A.S. implemented the computational developments. P.L. participated in the design of the study and helped with the analysis. G.B. helped with the analyses. P.L., G.B., K.V.L. and A.V.D. contributed to writing the manuscript. All authors read and approved the final manuscript. We thank Dr Thomas Leitner for background information on the contents of the Los Alamos HIV sequence database.

This study was made possible by funding of the Onderzoeksfonds KU Leuven Research Fund KU Leuven. The research leading to these results has received funding from the European Union Seventh Framework Programme [FP7/2007–2013] under Grant Agreement nr. 278433-PREDEMICS and ERC Grant Agreement nr. 260864. This study was supported in part by grants from the Fonds voor Wetenschappelijk Onderzoek Vlaanderen (FWO G.0692.14), by a grant from the Interuniversity Attraction Poles Programme, Belgian State, Belgian Science Policy (IUAP-VI P6/41), by the European Community's Seventh Framework Programme (FP7/2007–2013) under the project ‘Collaborative HIV and Anti-HIV Drug Resistance Network’ (CHAIN, grant 223131), by KU Leuven (Program Financing no. PF/10/018). Some of the computational resources (Stevin Supercomputer Infrastructure) and services used in this work were provided by Ghent University, the Hercules Foundation and the Flemish Government – department EWI.

B.V. was supported by a PhD grant from the Institute for the Promotion of Innovation through Science and Technology in Flanders (IWT Vlaanderen). M.A.S. was partially supported by National Science Foundation grant DMS-1264153.

Back to Top | Article Outline

Conflicts of interest

The authors declare that they have no competing interests.

Back to Top | Article Outline

References

1. Keele BF, Giorgi EE, Salazar-Gonzalez JF, Decker JM, Pham KT, Salazar MG, et al. Identification and characterization of transmitted and early founder virus envelopes in primary HIV-1 infection. Proc Natl Acad Sci U S A 2008; 105:7552–7557.
2. Li H, Bar KJ, Wang S, Decker JM, Chen Y, Sun C, et al. High multiplicity infection by HIV-1 in men who have sex with men. PLoS Pathog 2010; 6:e1000890.
3. Shaw GM, Hunter E. HIV transmission. Cold Spring Harb Perspect Med 2012; 2:
4. Soulie C, Calvez V, Marcelin A-G. Coreceptor usage in different reservoirs. Curr Opin HIV AIDS 2012; 7:450–455.
5. English S, Katzourakis A, Bonsall D, Flanagan P, Duda A, Fidler S, et al. Phylogenetic analysis consistent with a clinical history of sexual transmission of HIV-1 from a single donor reveals transmission of highly distinct variants. Retrovirology 2011; 8:54.
6. Fouda GG, Mahlokozera T, Salazar-Gonzalez JF, Salazar MG, Learn G, Kumar SB, et al. Postnatally-transmitted HIV-1 Envelope variants have similar neutralization-sensitivity and function to that of nontransmitted breast milk variants. Retrovirology 2013; 10:3.
7. Parrish NF, Gao F, Li H, Giorgi EE, Barbian HJ, Parrish EH, et al. Phenotypic properties of transmitted founder HIV-1. Proc Natl Acad Sci U S A 2013; 110:6626–6633.
8. Frange P, Meyer L, Jung M, Goujard C, Zucman D, Abel S, et al. Sexually-transmitted/founder HIV-1 cannot be directly predicted from plasma or PBMC-derived viral quasispecies in the transmitting partner. PLoS One 2013; 8:e69144.
9. Wolinsky SM, Wike CM, Korber BT, Hutto C, Parks WP, Rosenblum LL, et al. Selective transmission of human immunodeficiency virus type-1 variants from mothers to infants. Science 1992; 255:1134–1137.
10. Wolfs TF, Zwart G, Bakker M, Goudsmit J. HIV-1 genomic RNA diversification following sexual and parenteral virus transmission. Virology 1992; 189:103–110.
11. Zhu T, Mo H, Wang N, Nam DS, Cao Y, Koup RA, et al. Genotypic and phenotypic characterization of HIV-1 patients with primary infection. Science 1993; 261:1179–1181.
12. Sagar M. HIV-1 transmission biology: selection and characteristics of infecting viruses. J Infect Dis 2010; 202 (Suppl 2):S289–S296.
13. Lythgoe KA, Fraser C. New insights into the evolutionary rate of HIV-1 at the within-host and epidemiological levels. Proc Biol Sci 2012; 279:3367–3375.
14. Alizon S, Fraser C. Within-host and between-host evolutionary rates across the HIV-1 genome. Retrovirology 2013; 10:49.
15. Vrancken B, Rambaut A, Suchard MA, Drummond A, Baele G, Van Wijngaerden E, et al. The genealogical population dynamics of HIV-1 in a large transmission chain: bridging within and among host evolutionary rates. PLoS Computational Biology 2014; 10:e1003505.
16. Lythgoe KA, Pellis L, Fraser C. Is HIV short-sighted? Insights from a multistrain nested model. Evolution 2013; 67:2769–2782.
17. Fraser C, Lythgoe K, Leventhal GE, Shirreff G, Hollingsworth TD, Alizon S, et al. Virulence and pathogenesis of HIV-1 infection: an evolutionary perspective. Science 2014; 343:1243727.
18. Gonzalez MW, DeVico AL, Lewis GK, Spouge JL. Conserved molecular signatures in gp120 are associated with the genetic bottleneck during SIV, SHIV, and HIV-1 transmission. J Virol 2015; 89:3619–3629.
19. Carlson JM, Schaefer M, Monaco DC, Batorsky R, Claiborne DT, Prince J, et al. HIV transmission. Selection bias at the heterosexual HIV-1 transmission bottleneck. Science 2014; 345:1254031.
20. Lemey P, Kosakovsky Pond SL, Drummond AJ, Pybus OG, Shapiro B, Barroso H, et al. Synonymous substitution rates predict HIV disease progression as a result of underlying replication dynamics. PLoS Comput Biol 2007; 3:e29.
21. Mellors JW, Rinaldo CR Jr, Gupta P, White RM, Todd JA, Kingsley LA. Prognosis in HIV-1 infection predicted by the quantity of virus in plasma. Science 1996; 272:1167–1170.
22. Farzadegan H, Hoover DR, Astemborski J, Lyles CM, Margolick JB, Markham RB, et al. Sex differences in HIV-1 viral load and progression to AIDS. Lancet 1998; 352:1510–1514.
23. Langford SE, Ananworanich J, Cooper DA. Predictors of disease progression in HIV infection: a review. AIDS Res Ther 2007; 4:11.
24. Alizon S, von Wyl V, Stadler T, Kouyos RD, Yerly S, Hirschel B, et al. Phylogenetic approach reveals that virus genotype largely determines HIV set-point viral load. PLoS Pathog 2010; 6:e1001123.
25. Sterling TR, Vlahov D, Astemborski J, Hoover DR, Margolick JB, Quinn TC. Initial plasma HIV-1 RNA levels and progression to AIDS in women and men. N Engl J Med 2001; 344:720–725.
26. Suchard MA, Kitchen CMR, Sinsheimer JS, Weiss RE. Hierarchical phylogenetic models for analyzing multipartite sequence data. Syst Biol 2003; 52:649–664.
27. Edo-Matas D, Lemey P, Tom JA, Serna-Bolea C, van den Blink AE, van ’t Wout AB, et al. Impact of CCR5delta 32 host genetic background and disease progression on HIV-1 intrahost evolutionary processes: efficient hypothesis testing through hierarchical phylogenetic models. Mol Biol Evol 2011; 28:1605–1616.
28. Drummond AJ, Suchard MA, Xie D, Rambaut A. Bayesian phylogenetics with BEAUti and the BEAST 1.7. Mol Biol Evol 2012; 29:1969–1973.
29. Hasegawa M, Kishino H, Yano T. Dating of the human-ape splitting by a molecular clock of mitochondrial DNA. J Mol Evol 1985; 22:160–174.
30. Yang Z. Among-site rate variation and its impact on phylogenetic analyses. Trends Ecol Evol 1996; 11:367–372.
31. Yang Z. Maximum-likelihood models for combined analyses of multiple sequence data. J Mol Evol 1996; 42:587–596.
32. Althaus CL, De Vos AS, De Boer RJ. Reassessing the human immunodeficiency virus type 1 life cycle through age-structured modeling: life span of infected cells, viral generation time, and basic reproductive number, R0. J Virol 2009; 83:7659–7667.
33. Mansky LM, Temin HM. Lower in vivo mutation rate of human immunodeficiency virus type 1 than that predicted from the fidelity of purified reverse transcriptase. J Virol 1995; 69:5087–5094.
34. Drummond AJ, Ho SYW, Phillips MJ, Rambaut A. Relaxed phylogenetics and dating with confidence. PLoS Biol 2006; 4:e88.
35. Gill MS, Lemey P, Faria NR, Rambaut A, Shapiro B, Suchard MA. Improving Bayesian population dynamics inference: a coalescent-based model for multiple loci. Mol Biol Evol 2013; 30:713–724.
36. Edo-Matas D, van Gils MJ, Bowles EJ, Navis M, Rachinger A, Boeser-Nunnink B, et al. Genetic composition of replication competent clonal HIV-1 variants isolated from peripheral blood mononuclear cells (PBMC), HIV-1 proviral DNA from PBMC and HIV-1 RNA in serum in the course of HIV-1 infection. Virology 2010; 405:492–504.
37. Shankarappa R, Margolick JB, Gange SJ, Rodrigo AG, Upchurch D, Farzadegan H, et al. Consistent viral evolutionary changes associated with the progression of human immunodeficiency virus type 1 infection. J Virol 1999; 73:10489–10502.
38. Struck D, Lawyer G, Ternes AM, Schmit JC, Bercoff DP. COMET: adaptive context-based modeling for ultrafast HIV-1 subtype identification. Nucleic Acids Res 2014; 42:e144.
39. Edwards CTT, Holmes EC, Wilson DJ, Viscidi RP, Abrams EJ, Phillips RE, et al. Population genetic estimation of the loss of genetic diversity during horizontal transmission of HIV-1. BMC Evol Biol 2006; 6:28.
40. Redd AD, Collinson-Streng AN, Chatziandreou N, Mullis CE, Laeyendecker O, Martens C, et al. Previously transmitted HIV-1 strains are preferentially selected during subsequent sexual transmissions. J Infect Dis 2012; 206:1433–1442.
41. Maljkovic Berry I, Ribeiro R, Kothari M, Athreya G, Daniels M, Lee HY, et al. Unequal evolutionary rates in the human immunodeficiency virus type 1 (HIV-1) pandemic: the evolutionary rate of HIV-1 slows down when the epidemic rate increases. J Virol 2007; 81:10625–10635.
42. Hughes GJ, Fearnhill E, Dunn D, Lycett SJ, Rambaut A, Leigh Brown AJ, et al. Molecular phylodynamics of the heterosexual HIV epidemic in the United Kingdom. PLoS Pathog 2009; 5:e1000590.
43. Lewis F, Hughes GJ, Rambaut A, Pozniak A, Leigh Brown AJ. Episodic sexual transmission of HIV revealed by molecular phylodynamics. PLoS Med 2008; 5:e50.
44. Glick SN, Morris M, Foxman B, Aral SO, Manhart LE, Holmes KK, et al. A comparison of sexual behavior patterns among men who have sex with men and heterosexual men and women. J Acquir Immune Defic Syndr 2012; 60:83–90.
45. Rambaut A, Posada D, Crandall KA, Holmes EC. The causes and consequences of HIV evolution. Nat Rev Genet 2004; 5:52–61.
46. Brenner BG, Roger M, Routy JP, Moisi D, Ntemgwa M, Matte C, et al. High rates of forward transmission events after acute/early HIV-1 infection. J Infect Dis 2007; 195:951–959.
47. Halperin DT, Epstein H. Concurrent sexual partnerships help to explain Africa's high HIV prevalence: implications for prevention. Lancet 2004; 364:4–6.
48. Samuelsen H, Norgaard O, Ostergaard LR. Social and cultural aspects of HIV and AIDS in West Africa: a narrative review of qualitative research. SAHARA J 2012; 9:64–73.
Keywords:

evolutionary rate; genetic diversity; HIV; risk group; transmission bottleneck

Supplemental Digital Content

Back to Top | Article Outline
Copyright © 2015 Wolters Kluwer Health, Inc.