Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Is the tryptophan codon of gene vif the Achilles’ heel of HIV-1?

  • Fabiola Villanova,

    Roles Data curation, Formal analysis

    Affiliation Institute of Biological Sciences, Federal University of Pará, Belém, PA, Brazil

    ⨯
  • Marta Barreiros,

    Roles Data curation, Formal analysis

    Current address: Department of Electrical Engineering, Federal University of Maranhão, São Luís, MA, Brazil

    Affiliation Institute of Biological Sciences, Federal University of Pará, Belém, PA, Brazil

    ⨯
  • Élcio Leal

    Roles Conceptualization, Formal analysis, Writing – review & editing

    elcioleal@gmail.com

    Affiliation Institute of Biological Sciences, Federal University of Pará, Belém, PA, Brazil

    ⨯

Abstract

To evaluate the impact of hypermutation on the HIV-1 dissemination at the population level we studied 7072 sequences HIV-1 gene vif retrieved from the public databank. From this dataset 854 sequences were selected because they had associated values of CD4+ T lymphocytes counts and viral loads and they were used to assess the correlation between clinical parameters and hypermutation. We found that the frequency of stop codons at sites 5, 11 and 79 ranged from 2.8x10-4 to 4.2x10-4. On the other hand, at codons 21, 38, 70, 89 and 174 the frequency of stop codons ranged from 1.4x10-3 to 2.5x10-3. We also found a correlation between clinical parameters and hypermutation where patients harboring proviruses with one or more stop codons at the tryptophan sites of the gene vif had higher CD4+ T lymphocytes counts and lower viral loads compared to the population. Our findings indicate that A3 activity potentially restrains HIV-1 replication because individuals with hypermutated proviruses tend to have lower numbers of RNA copies. However, owing to the low frequency of hypermutated sequences observed in the databank (44 out of 7072), it is unlikely that A3 has a significant impact to curb HIV-1 dissemination at the population level.

Citation: Villanova F, Barreiros M, Leal É (2020) Is the tryptophan codon of gene vif the Achilles’ heel of HIV-1? PLoS ONE 15(6): e0225563. https://doi.org/10.1371/journal.pone.0225563

Editor: Michael Schindler, University Hospital Tuebingen, GERMANY

Received: November 6, 2019; Accepted: May 5, 2020; Published: June 22, 2020

Copyright: © 2020 Villanova et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: To evaluate the impact of hypermutation to the HIV-1 dissemination at the population level we studied 7072 sequences of the vif gene obtained from the public HIV databank at the Los Alamos National Laboratory. All HIV vif gene sequences and their clinical data (CD4 levels and viral loads) that were used in this study are freely available to anyone in the Los Alamos HIV databases (https://www.hiv.lanl.gov). Other information such as tables, CVS files, etc are also included in the supporting information.

Funding: M.B was supported by Coordenação de Aperfeiçoamento de Pessoal de Nível; EL is supported by the Conselho Nacional de Desenvolvimento Científico e Tecnológico/CNPq (#302677/2019-4). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: No authors have competing interests.

1. Introduction

The apolipoprotein mRNA editing enzyme catalytic polypeptide-like 3 (APOBEC3; A3) proteins are a family of seven cytidine deaminases (A3A, A3B, A3C, A3D, A3F, A3G, and A3H) that restrict certain lentiviruses, retrotransposons, hepatitis B virus and human papillomavirus [1–4]. Many studies focused on the genes A3F, A3G, and A3H because of their innate defense ability to restrict the replication of the HIV-1 [5–21]. Particularly, A3F and A3G are incorporated into viral particles and during reverse transcription, within newly infected cells; by deamination, these proteins alter C in the viral minus-strand DNA to U. This activity of A3 is termed hypermutation because it induces high rates of G-to-A mutation in the newly synthesized plus-strand of viral DNA [2]. Besides, A3F and A3G inhibit the HIV-1 life cycle curbing the reverse transcription and the integration [2, 5, 22]. The hypermutation activity of A3 proteins is highly dependable of the nucleotide context. Particularly, A3G mutates primarily TGGs when this codon is followed by a G (TGGG). On the other hand, A3F, A3D, and A3H mutate TGG into the TGA stop codon when the TGG codon is followed by an A (TGGA) [1, 2, 8].

HIV-1 counteracts the antiviral property of A3 proteins by the activity of the viral [16, 21, 23–25] infectivity factor (Vif). The Vif proteins assemble with viral-specific E3 ubiquitin ligase by interacting with the cellular Cullin5 (Cul5)-ElonginB-ElonginC proteins, inducing ubiquitination of A3G and consequent degradation by the proteasomal complex [17,21,23–25].

It has been proposed that hypermutation is not enough to repress HIV-1 infection within host because proviruses with varying amounts of G→A mutations are commonly observed in the cells of infected individuals [10, 26]. So, it is reasonable that when G→A mutations are ineffective in neutralizing viral genomes, A3G activity can actually increase HIV-1 diversification [9, 27].

In this work, we used 7042 sequences to estimate the rate of stop codons at tryptophan (TGG) residues of the vif gene. By assuming that A3 hypermutations are context-dependent and RT mutations are context-independent this approach enables us to distinguish the impact of A3 hypermutation from the RT activity in the vif gene.

We found that patients harboring proviruses with one or more stop codons at the tryptophan sites of the gene vif had higher CD4+ T lymphocytes counts and lower viral loads compared to the population.

2. Materials and methods

2.1. Sequence processing

To evaluate the impact of hypermutation to the HIV-1 dissemination at the population level we studied 7072 sequences of the vif gene obtained from the public HIV databank at the Los Alamos National Laboratory (https://www.hiv.lanl.gov). From this dataset 854 sequences that had associated values of CD4+ T lymphocytes counts were selected to assess the correlation between clinical parameters and hypermutation. Sequence alignment and editing. The sequences were aligned using the ClustalX program [28]. In addition, the SE-AL program, version 2.0 (Department of Zoology, Oxford University; http://tree.bio.ed.ac.uk/software/seal/) was used to edit the alignment in order to keep all reading frames opened.

2.2. Mutation rates and statistical analysis

To compute the stop codons induced by A3 proteins we used the same approach described by Cuevas et al., 2015. Briefly, it is well characterized by the preferential targets of distinct A3 proteins while editing the minus strand of viral DNA [27]. The A3G protein converts GG to AG primarily in the GGG triplet, thus in this context tryptophan codons (TGG) can be converted to the stop codons TAG or TGA. On the other hand, A3F/D/H convert GG to GA in the target GGA, consequently, tryptophan codons (TGG) will be converted into TGA stop codons. Following this conception, it is possible to estimate the amount of A3 mutations compared to the baseline mutations induced by the reverse transcriptase RT. Since vif gene of HIV-1 is essential to the life cycle of the virus and the protein has eight canonical tryptophan codons at the sites 5, 11, 21, 38, 70, 79, 89 and 174 we counted the number of stop codons in each of these sites [7, 19, 24, 25]. The Bayesian independent Welch test was used to correlate clinical data and mutations rates. All statistical tests were performed using JASP software v.0.11.1 (https://jasp-stats.org/). Boxplots were constructed using R software v 3.5.1 (www.r-project.org).

3. Results

3.1. Rates of stop codons

Estimates of mutations rates indicated there are variable (i.e., 83 and 70) and more conserved (i.e., 5) sites in vif (Fig 1). The frequency of stop codons at sites 5, 11 and 79 ranged from 2.8x10-4 to 4.2x10-4 while at codons 21, 38, 70, 89 and 174 the frequency of stop codons ranged from 1.4x10-3 to 2.5x10-3. The differences in mutation rates induced by the A3 (dark gray bars in Fig 1) and by the RT (light gray bars in Fig 1) revealed that sites not targeted by A3 proteins (i.e., 5, 11 and 79) were those with lower mutations rates. Conversely, the sites 38 and 70, which are targets of the A3G (at this site TGG is followed by G; TGGG), had higher rates (Fig 1). It is worth mentioning that in site 70 of the Vif all patients presented the context TGGG meanwhile in site 38 the context TGGG was presented in 205 of patients while TGGT was present in 70%. The context TGGT can be mutated either by the A3 activity, converting TGGT into TAGT, or by the RT activity, converting TGGT into TGAT. Besides, at the sites 89 and 174 the tryptophan codon TGG is followed by A, this context (TGGA) is targeted by the A3D/F/H proteins. The above results showed that tryptophan sites 5, 11 and 79, not targeted by A3 proteins, are those with lower rates of stop codons. Conversely, sites targeted by A3 proteins have higher rates of stop codons.

thumbnail
Download:
Fig 1. Mutation rates in tryptophan codons of the vif gene of HIV-1.

At each tryptophan site (TGG) of the gene vif the number of amino acid changes was estimated. Dark gray horizontal bars indicate the number of changes from TGG to stop codons (i.e., TAA, TAG, TGA) and light gray bars represents changes from TGG to any other codons such as AGG, TTG, TCA, etc. The next nucleotide after the TGG is shown in each codon. These nucleotides determine the context target by the human proteins A3 protein (see next figure). Codons 21 and 38 have distinct percentages of a certain nucleotide that is indicated in the figure.

https://doi.org/10.1371/journal.pone.0225563.g001

3.2. Rates of other mutations

We also estimated the overall mutation rates considering the TGG context regardless of the position it was located in the vif protein. These measurements were summarized in Fig 2 that shows that mutations from TGG (tryptophan codon) to the stop codons TGA or TAG are higher when the tryptophan codon is followed by Gs or As (TGGG). Tryptophan codons that are targeted by A3 proteins (i.e., TGGG or TGGA) have rates of 1x10-3 to 3.1x10-3 while codons target by the RT has rates of 2x10-4 to 6x10-4.

thumbnail
Download:
Fig 2. The context associated mutation rates of apobec3.

The Context refers to cDNA sequences edited by members of apobec3 (A3) protein family. The human A3G protein targets the CC sequences on minus-strand cDNA, thus causing GG to AG mutations in the positive strand of HIV. A3G mutates TGG codons mainly to TAG stop codon when the TGG codon is followed by a G (TGGG). A3F/H/D mutate TGG to stop codons when TGG is followed by an A (TGGA). Reverse transcriptase has no preferences and can change TGG into any codon, including stop codons. Arrows indicate a mutation from TGG to a certain stop codon and numbers are the estimated mutation rates of stop codons. Boxes close to the arrows indicate the enzyme likely associate with a mutation indicated by the arrow. RT = reverse transcriptase. TGG = Tryptophan, TGA, TAA, TGA = Stop codons.

https://doi.org/10.1371/journal.pone.0225563.g002

3.3. Stop codons and clinical status

To assess the correlation between clinical parameters and mutations we used 854 sequences that had associated values of CD4+ T lymphocytes counts and viral loads. We found eleven sequences having one or more stop codons at the tryptophan sites of the gene vif. Notably, these sequences presented lower viral loads (posterior probability = 0.097) and higher levels of CD4+ lymphocytes (posterior probability = 0.071) compared with the overall values of 854 dataset (Table 1). The median viral load in patients with hypermutation was equal to 7,864.00 copies/mL (variance = 1.624E11) with a mean of 216,392.36 copies/mL, but in patients without hypermutation the median was 50,709.00 copies/mL (variance = 2.865E11) and mean of 225,917.78 copies/mL. The median CD4 + T lymphocytes for the group with hypermutation was 434.00 (variance = 67968.82) cells/mm3 and mean of 485.72 cells/mm3, while the samples without hypermutation median CD4 + T cells was 403.00 (variance = 65505.90) cells/mm3 and mean of 434.61 cells/mm3 (Fig 3).

thumbnail
Download:
Fig 3. RNA levels and CD4+ cell counts.

Comparisons of RNA levels per ml (upper boxes) and counts of CD4+ cells per ml (lower boxes) between the sequences with stop codons at the tryptophan (grey boxes) and sequences without stop codons (white boxes). In the boxes the center lines show the medians; box limits indicate the 25th and 75th percentiles. Whiskers extend 1.5 times the interquartile range from the 25th and 75th percentiles, outliers are represented by open dots. RNA levels (upper boxes) are shown in log scale.

https://doi.org/10.1371/journal.pone.0225563.g003

thumbnail
Download:
Table 1. Hypermutation in vif gene and clinical parameters of HIV-1 infected patients.

https://doi.org/10.1371/journal.pone.0225563.t001

4. Discussion

The amount of A3G hypermutation varies considerably along HIV-1 genome and this gradient of G-to-A substitutions correlates with the time the minus strand remains as a single-stranded molecule during replication [26]. One consequence of the hypermutation gradient is that some genes are more affected than others. The vif gene has the lowest amount of A3G-associated mutation compared to other HIV-1 genes [27]. Vif also has tryptophan residues (W) at the specific positions 5, 11, 21, 38, 70, 89 and 174 that are involved in A3G and A3F binding. These codons will be target by the A3 activity and the TGG codon will be changed into a stop codon (e.g., TAG, TGA, TAA). Equally, the TGG codon will be targeted by RT activity converting it into stop codons and also into others codons such as TTG, TGT, AGG, etc.

Vif has some conserved residues, notably in the motifs 14DRMR17, 21WK/NSLVK26, 40YRHHY44 and 161PPLP164. For example, we found the overall 16% of conserved residues in 317 HIV-1 subtype B vif sequences in Brazil [19]. This lower diversity in some residues has been related to the very strong purifying selection detected on this viral protein, thus indicating that vif is essential to the HIV life cycle [4,7,19,21,23]. While hypermutation induced by A3G activity is a natural barrier against retroviruses it is not enough to restrain HIV-1 infection. Since HIV-1 infection is characterized by multiple strains forming a quasispecies, then it is likely that hypermutated strains can benefit from circulation of or even reservoir viruses in distinct tissues. It is likely that A3G activity can actually increase HIV-1 diversification when G-to-A hypermutation is ineffective in neutralizing all viral genomes within a host [9].

Colson et al., [29] showed that in long-term non-progressors patients A3 activity is able to restrain HIV-1 replication by changing of tryptophan (TGG) codons into stop codons (TAG/TGA) mainly on the gene vif. However, the effect of hypermutation to the spread of HIV-1 is not known yet. We studied this subject by using the tryptophan codons of Vif as a proxy to evaluate the A3 activity to potentially reduce the chances of HIV spread between individuals. Our analysis indicated a correlation between clinical parameters and hypermutation where patients harboring proviruses with one or more stop codons at the tryptophan sites of the gene vif had higher CD4+ T lymphocytes counts and lower viral loads compared to the population. We found a correlation between clinical parameters and hypermutation in patients harboring proviruses with one or more stop codons at the tryptophan sites of the gene vif had higher CD4+ T lymphocytes counts and lower viral loads compared to the population. Thus our findings indicate that A3 activity potentially restrains HIV-1 replication because individuals with hypermutated proviruses tend to have a lower number of RNA copies. However owing to the low frequency of hypermutated sequences observed in the databank (44 out of 7072), it is unlikely that A3 has a significant impact to curb HIV-1 dissemination at the population level.

Our findings are in consent with the observation that A3G hypermutation is more frequent among elite controllers [30]. It is also worth to mention that CD4+ T lymphocytes counts are related to selective diversity and hypermutation [11, 13, 31].

References

  1. 1. An P, Penugonda S, Thorball CW, Bartha I, Goedert JJ, Donfield S, et al. Role of APOBEC3F Gene Variation in HIV-1 Disease Progression and Pneumocystis Pneumonia. PLoS Genet. 2016;12(3):e1005921. pmid:26942578; PubMed Central PMCID: PMC4778847.
  2. 2. Ebrahimi D, Alinejad-Rokny H, Davenport MP. Insights into the motif preference of APOBEC3 enzymes. PLoS One. 2014;9(1):e87679. pmid:24498164; PubMed Central PMCID: PMC3909203.
  3. 3. Moris A, Murray S, Cardinaud S. AID and APOBECs span the gap between innate and adaptive immunity. Front Microbiol. 2014;5:534. pmid:25352838; PubMed Central PMCID: PMC4195361.
  4. 4. Sato K, Izumi T, Misawa N, Kobayashi T, Yamashita Y, Ohmichi M, et al. Remarkable lethal G-to-A mutations in vif-proficient HIV-1 provirus by individual APOBEC3 proteins in humanized mice. J Virol. 2010;84(18):9546–56. pmid:20610708; PubMed Central PMCID: PMC2937654.
  5. 5. Arias JF, Koyama T, Kinomoto M, Tokunaga K. Retroelements versus APOBEC3 family members: No great escape from the magnificent seven. Front Microbiol. 2012;3:275. pmid:22912627; PubMed Central PMCID: PMC3418512.
  6. 6. Binning JM, Smith AM, Hultquist JF, Craik CS, Caretta Cartozo N, Campbell MG, et al. Fab-based inhibitors reveal ubiquitin independent functions for HIV Vif neutralization of APOBEC3 restriction factors. PLoS Pathog. 2018;14(1):e1006830. pmid:29304101; PubMed Central PMCID: PMC5773222.
  7. 7. Bizinoto MC, Yabe S, Leal E, Kishino H, Martins Lde O, de Lima ML, et al. Codon pairs of the HIV-1 vif gene correlate with CD4+ T cell count. BMC Infect Dis. 2013;13:173. pmid:23578255; PubMed Central PMCID: PMC3637627.
  8. 8. Chen J, MacCarthy T. The preferred nucleotide contexts of the AID/APOBEC cytidine deaminases have differential effects when mutating retrotransposon and virus sequences compared to host genes. PLoS Comput Biol. 2017;13(3):e1005471. pmid:28362825; PubMed Central PMCID: PMC5391955.
  9. 9. Jern P, Russell RA, Pathak VK, Coffin JM. Likely role of APOBEC3G-mediated G-to-A mutations in HIV-1 evolution and drug resistance. PLoS Pathog. 2009;5(4):e1000367. pmid:19343218; PubMed Central PMCID: PMC2659435.
  10. 10. Kim EY, Lorenzo-Redondo R, Little SJ, Chung YS, Phalora PK, Maljkovic Berry I, et al. Human APOBEC3 induced mutation of human immunodeficiency virus type-1 contributes to adaptation and evolution in natural infection. PLoS Pathog. 2014;10(7):e1004281. pmid:25080100; PubMed Central PMCID: PMC4117599.
  11. 11. Kobayashi T, Koizumi Y, Takeuchi JS, Misawa N, Kimura Y, Morita S, et al. Quantification of deaminase activity-dependent and -independent restriction of HIV-1 replication mediated by APOBEC3F and APOBEC3G through experimental-mathematical investigation. J Virol. 2014;88(10):5881–7. pmid:24623435; PubMed Central PMCID: PMC4019142.
  12. 12. Land AM, Ball TB, Luo M, Pilon R, Sandstrom P, Embree JE, et al. Human immunodeficiency virus (HIV) type 1 proviral hypermutation correlates with CD4 count in HIV-infected women from Kenya. J Virol. 2008;82(16):8172–82. pmid:18550667; PubMed Central PMCID: PMC2519552.
  13. 13. Mussil B, Sauermann U, Motzkus D, Stahl-Hennig C, Sopper S. Increased APOBEC3G and APOBEC3F expression is associated with low viral load and prolonged survival in simian immunodeficiency virus infected rhesus monkeys. Retrovirology. 2011;8:77. pmid:21955401; PubMed Central PMCID: PMC3192745.
  14. 14. Ronsard L, Raja R, Panwar V, Saini S, Mohankumar K, Sridharan S, et al. Genetic and functional characterization of HIV-1 Vif on APOBEC3G degradation: First report of emergence of B/C recombinants from North India. Sci Rep. 2015;5:15438. pmid:26494109; PubMed Central PMCID: PMC4616021.
  15. 15. Russell RA, Moore MD, Hu WS, Pathak VK. APOBEC3G induces a hypermutation gradient: purifying selection at multiple steps during HIV-1 replication results in levels of G-to-A mutations that are high in DNA, intermediate in cellular viral RNA, and low in virion RNA. Retrovirology. 2009;6:16. pmid:19216784; PubMed Central PMCID: PMC2657108.
  16. 16. Sadler HA, Stenglein MD, Harris RS, Mansky LM. APOBEC3G contributes to HIV-1 variation through sublethal mutagenesis. J Virol. 2010;84(14):7396–404. pmid:20463080; PubMed Central PMCID: PMC2898230.
  17. 17. Sheehy AM, Gaddis NC, Malim MH. The antiretroviral enzyme APOBEC3G is degraded by the proteasome in response to HIV-1 Vif. Nat Med. 2003;9(11):1404–7. pmid:14528300.
  18. 18. Venkatesan S, Rosenthal R, Kanu N, McGranahan N, Bartek J, Quezada SA, et al. Perspective: APOBEC mutagenesis in drug resistance and immune escape in HIV and cancer evolution. Ann Oncol. 2018;29(3):563–72. pmid:29324969; PubMed Central PMCID: PMC5888943.
  19. 19. Villanova F, Barreiros M, Janini LM, Diaz RS, Leal E. Genetic Diversity of HIV-1 Gene vif Among Treatment-Naive Brazilians. AIDS Res Hum Retroviruses. 2017;33(9):952–9. pmid:28443724.
  20. 20. Yamashita T, Kamada K, Hatcho K, Adachi A, Nomaguchi M. Identification of amino acid residues in HIV-1 Vif critical for binding and exclusion of APOBEC3G/F. Microbes Infect. 2008;10(10–11):1142–9. pmid:18603011.
  21. 21. Zhang W, Du J, Evans SL, Yu Y, Yu XF. T-cell differentiation factor CBF-beta regulates HIV-1 Vif-mediated evasion of host restriction. Nature. 2011;481(7381):376–9. pmid:22190036.
  22. 22. Ara A, Love RP, Chelico L. Different mutagenic potential of HIV-1 restriction factors APOBEC3G and APOBEC3F is determined by distinct single-stranded DNA scanning mechanisms. PLoS Pathog. 2014;10(3):e1004024. pmid:24651717; PubMed Central PMCID: PMC3961392.
  23. 23. Stanley BJ, Ehrlich ES, Short L, Yu Y, Xiao Z, Yu XF, et al. Structural insight into the human immunodeficiency virus Vif SOCS box and its role in human E3 ubiquitin ligase assembly. J Virol. 2008;82(17):8656–63. pmid:18562529; PubMed Central PMCID: PMC2519636.
  24. 24. Tian C, Yu X, Zhang W, Wang T, Xu R, Yu XF. Differential requirement for conserved tryptophans in human immunodeficiency virus type 1 Vif for the selective suppression of APOBEC3G and APOBEC3F. J Virol. 2006;80(6):3112–5. pmid:16501124; PubMed Central PMCID: PMC1395459.
  25. 25. Wolfe LS, Stanley BJ, Liu C, Eliason WK, Xiong Y. Dissection of the HIV Vif interaction with human E3 ubiquitin ligase. J Virol. 2010;84(14):7135–9. pmid:20463065; PubMed Central PMCID: PMC2898223.
  26. 26. Armitage AE, Katzourakis A, de Oliveira T, Welch JJ, Belshaw R, Bishop KN, et al. Conserved footprints of APOBEC3G on Hypermutated human immunodeficiency virus type 1 and human endogenous retrovirus HERV-K(HML2) sequences. J Virol. 2008;82(17):8743–61. pmid:18562517; PubMed Central PMCID: PMC2519685.
  27. 27. Cuevas JM, Geller R, Garijo R, Lopez-Aldeguer J, Sanjuan R. Extremely High Mutation Rate of HIV-1 In Vivo. PLoS Biol. 2015;13(9):e1002251. pmid:26375597; PubMed Central PMCID: PMC4574155.
  28. 28. Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, McWilliam H, et al. Clustal W and Clustal X version 2.0. Bioinformatics. 2007;23(21):2947–8. pmid:17846036.
  29. 29. Colson P, Ravaux I, Tamalet C, Glazunova O, Baptiste E, Chabriere E, et al. HIV infection en route to endogenization: two cases. Clin Microbiol Infect. 2014;20(12):1280–8. pmid:25366539; PubMed Central PMCID: PMC4360783.
  30. 30. Eyzaguirre LM, Charurat M, Redfield RR, Blattner WA, Carr JK, Sajadi MM. Elevated hypermutation levels in HIV-1 natural viral suppressors. Virology. 2013;443(2):306–12. pmid:23791226; PubMed Central PMCID: PMC3762252.
  31. 31. Leal E, Casseb J, Hendry M, Busch MP, Diaz RS. Relaxation of adaptive evolution during the HIV-1 infection owing to reduction of CD4+ T cell counts. PLoS One. 2012;7(6):e39776. pmid:22768122; PubMed Central PMCID: PMC3387245.