Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

Genomic characterization and molecular evolution analysis of subtype B and BF recombinant HIV-1 strains among Argentinean men who have sex with men reveal a complex scenario

Genomic characterization and molecular evolution analysis of subtype B and BF recombinant HIV-1... characterization and molecular evolution analysis of subtype B and BF recombinant HIV-1 strains Currently, data on HIV-1 circulating strains among men who have sex with men (MSM) in among Argentinean men who have sex with men Argentina is scarce. In South America, the distribution and the prevalence of BF recombi- reveal a complex scenario. PLoS ONE 12(12): nants are dissimilar and exhibit an underappreciated heterogeneity of recombinant struc- e0189705. https://doi.org/10.1371/journal. pone.0189705 tures. Here, we studied for the first time the genetic diversity of HIV-1 BF recombinants and their evolution over time through in-depth phylogenetic analysis and multiple recombination Editor: Douglas F. Nixon, George Washington University, UNITED STATES detection methods involving 337 HIV-1 nucleotide sequences (25 near full-length (NFL) and 312 partial pol gene) obtained from Argentinean MSM. The recombination profiles were Received: May 5, 2017 studied using multiple in silico tools to characterize the genetic mosaicism, and phylogenetic Accepted: November 30, 2017 approaches to infer their relationships. The evolutionary history of BF recombinants and Published: December 15, 2017 subtype B sequences was reconstructed by a Bayesian coalescent-based method. By Copyright:© 2017 Cevallos et al. This is an open phylogenetic inference, 81/312 pol sequences clustered within BF clade. Of them, 46 access article distributed under the terms of the sequences showed a genetic mosaic with CRF12_BF-like patterns, including plausible sec- Creative Commons Attribution License, which ond-generation recombinants. Other CRFs_BF like (CRF17, 28, 29, 39, 42, 44, 47) and permits unrestricted use, distribution, and reproduction in any medium, provided the original probable URFs_BF were less frequently found. Phylogenetic and recombination analyses author and source are credited. on NFL sequences allowed a meticulous definition of new BF mosaics of genomic patterns. Data Availability Statement: All relevant data are The Bayesian analyses pointed out quite consistent onset dates for the CRFs_BF clade within the paper. All nucleotide sequences are based on B and F gene datasets (~1986 and ~1991 respectively). These results indicate deposited in GenBank database (accession that the CRFs_BF variants have been circulating among Argentinean MSM for about 30 numbers DQ383746- DQ38375, KY968394- years. This study reveals, through growing evidence showing the importance of MSM in the KY968408, EF120266, EF120267, EF120268, EF120269, EF120270, EF120271, EF120272, dynamics of the HIV-1 epidemic in Argentina, the coexistence of CRF12_BF-like and high EF120274, EF120275, EF120276, EF120277, diversity of strains exhibiting several BF mosaic patterns, including non-reported URFs that EF120278, EF120279, EF120280, EF120281, may reflect active clusters as potential intervention targets to hinder HIV-1 transmission. EF120282, EF120283, EF120284, EF120285, EF120286, EF120287, EF120288, EF120289, PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 1 / 12 HIV-1 BF recombinants among Argentinean MSM EF120290, EF120291, EF120292, EF120293, Background EF120294, EF120295, EF120296, EF120297, The HIV-1 epidemic in Argentina is mostly represented by subtype B, plus a wide variety of EF120298, EF120300, EF120301, EF120302, recombinant strains between subtypes B and F. Other less frequent HIV-1 subtypes (A, C, F) EF120303, EF120304, EF120305, EF120306, EF120307, EF120308, EF120309, EF120310, and recombinant forms (BC, BCF, A2D, and AG) have also been characterized [1±5]. Since EF120311, EF120312, EF120313, EF120314, the onset of the HIV/AIDS epidemic, men who have sex with men (MSM) have remained a EF120315, EF120316, EF120317, EF120318, recognized high-risk group for HIV-1 infection. It is well known worldwide that many MSM EF120319, EF120320, EF120321, EF120322, are at high±risk for infection, hence, HIV-1 prevalence and incidence in MSM are consistently EF120323, EF120324, EF120325, EF120326, higher than in the general population [6]. According to official data published in 2016 by the EF120327, EF120328, EF120329, EF120330, EF120331, EF120332, EF120333, EF120334, Ministry of Health, Argentina, the current prevalence in MSM is between 12 and 15% [7]. The EF120335, EF120351, EF120353, EF120354, genomic characterization of HIV-1 among Argentinean MSM was performed by phylogenetic EF120356, KY968342- KY968393, KY968409- inference [3, 8, 9]. Since then, no other evidence has been provided regarding HIV-1 genetic KY968571). variability associated with BF recombinants among MSM. Funding: This study was supported by grants In South America, the prevalence rates of BF recombinant sequences, either Circulating obtained by MMA from the Agencia Nacional de Recombinant Forms (CRF_BF) or Unique Recombinant Forms (URF_BF), have been found Promocio Ân Cientifica y Tecnologica-Fondo to be significant. Other CRF_BF have also been described in South America: CRF28_BF and Nacional para la Ciencia y Tecnica (grant number CRF29_BF in Sao Paulo (Brazil) [10]; CRF39_BF and CRF40_BF in Rio de Janeiro (Brazil) PICT 2013/2229) and the Consejo Nacional de [11]; CRF38_BF in Uruguay [12]; and CRF44_BF in Chile [13]. In Argentina after the identifi- Investigaciones Cientificas y Tecnicas (grant number PIP 2012/1048) to MMA. The funders had cation of the CRF12_BF and CRF17_BF in 2001 [14, 15], several studies showed that most BF no role in study design, data collection and recombinants circulating locally were closely associated with CRF12_BF or diverse BF recom- analysis, decision to publish, or preparation of the binants related to it [1, 16±18]. manuscript. The aim of the present study was to evaluate the heterogeneity of the HIV-1 associated Competing interests: The authors have declared with subtype B and BF recombinants and their evolution over time through in-depth phylo- that no competing interests exist. genetic and Bayesian coalescent analysis and multiple recombination detection methods among Argentinean MSM. Using near-full length and pol-gene sequences, this study reveals a complex scenario for BF recombinants among MSM where CRF12_BF-like and high diversity of strains exhibiting several BF mosaic patterns -including non-reported URFs- coexist. Materials and methods HIV-1 nucleotide sequences datasets Two datasets were constructed involving a total of 337 HIV-1 nucleotide sequences. These include 312 partial pol gene sequences (defined from nucleotide 2253 up to 3720 according to HXB-2 numbering) and 25 near full-length (NFL) sequences. These were obtained from 429 HIV-1 seropositive Argentinean men who have sex with men (MSM) recruited in previous studies performed during the period 2000±2009. They are mainly residents of Buenos Aires city and surroundings areas with a median age of 30 years, and most of them were not on anti- viral therapy at the time of the studies. As documented previously, the consumption of illicit drugs presented by this population was very infrequent. Blood collection was performed once a written informed consent was obtained as stated in each of the original studies [2, 4, 8, 9, 19]. The methods used for extraction of the DNA and sequencing are described in the respective references. All the sequences used in this work were obtained from projects previously approved by the Ethic Committee from School of Medicine of Buenos Aires University or Ethic Committee of NEXO AC (IRB 00005349). In all cases the consent of each participant, confidentiality of the information as well as laboratory results, were obtained according to international recommendation and current legislation (National AIDS Law No. 23,798 and its regulatory decree) and the recommended international ethical guidelines for epidemiological studies were strictly followed, (CIOMS, Ginebra 2009). PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 2 / 12 HIV-1 BF recombinants among Argentinean MSM HIV-1 subtype references strains available in Los Alamos Database (http://www.hiv.lanl. gov/content/sequence/HIV/mainpage.html) were used to make alignments with the Argentin- ean sequences under analysis. The alignments were performed with MAFFT software included in http://www.hiv.lanl.gov/content/sequence/VIRALIGN/viralign.html [20]. Phylogenetic analysis Phylogenetic analyses were carried out in order to classify the viruses by HIV-1 subtype or CRFs. Phylogenetic trees were constructed with a GTR nucleotide substitution model among site rate heterogeneity and a proportion of invariable site selected by jModelTest v.2.1.10 [21] using the Maximum Likelihood (ML) method implemented in PhyML v3.0 employing an online web server (http://www.atgc-montpellier.fr/phyml/) [22]. The best SPR and NNI heu- ristic options were selected. The reliability of tree topologies was assessed by two methods according to dataset, such as bootstrapping using 500 replications (values 70% were consid- ered significant) for NFL dataset, and approximate likelihood ratios (aLRT) 0.90 based on Shimodaira-Hasegawa-like procedure were used to assign the subtype/CRF designation of the strains for pol dataset [23]. Detection of recombination events and resistance mutations The HIV-1 intersubtype recombination patterns were analyzed separately for each sequence by determining the genetic mosaic structure, location of breakpoints, and parental subtypes involved. For this purpose, several in silico methods were selected considering both the type of analy- sis to be carried out and the nature of the database [24]. Initially, pol sequences were analyzed by three different methods or algorithms including: (i) Jumping profile Hidden Markov Model (jpHMM) method. It is based on a pre-calculated multiple alignment of the major HIV-1 subtypes where each one is represented by a profile hidden Markov model further con- nected by empirical probabilities, allowing the detection of possible recombinants and plausi- ble breakpoints. The HIV-1 subtype B and subtype F have adequate sampling to build well informed profiles offering enough data to construct a good model of sequence variation [25], (ii) Recombination Detection Program (RDP4), a software package for statistical identifi- cation and characterization of recombination events in DNA sequences that uses several non- parametric recombination detection methods. Among multiple approaches to identify recom- bination signals, RDP, BOOTSCAN, GENECONV, MAXCHI, and SISCAN methods were carried out. All methods are described in detail in reference [26]. This tool considers every sequence included in the analysis as a potential recombinant and monitors systematically each sequence triplets or quartets in order to identify those that contain a recombinant as well as those two sequences that could serve as parents while performing a statistical evaluation of recombination signals. The highest acceptable P-value was set to 0.05 and the other parameters were default RDP4 settings. Recombination events detected by at least three different methods (Pvalues of0.05 after Bonferroni correction for multiple comparisons implemented in RDP4) were taken as credible evidence of possible HIV-1 recombination. Lastly, (iii) the recombination analysis using cost optimization (RECCO) was performed. This software com- putes the number of ªsavingsº, i.e., number of mismatches saved when explaining the read by a recombination event between two viral strains and mutations, rather than by a single strain and mutations only. In this analysis, sequences were defined as recombinant when RECCO assigns a savings value higher than 5 [27]. When conflicting results emerged in recombination events, the criteria were to consider as actual event in which at least two out of the three meth- ods used would match. Only for near full length sequences, bootscanning analysis was also PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 3 / 12 HIV-1 BF recombinants among Argentinean MSM performed separately to detect recombination as implemented in the SimPlot software version 3.5.1 (http://www.med.jhu.edu/deptmed/sray/download/) [28]. The presence of resistance mutations was detected by the Stanford algorithm (https:// hivdb.stanford.edu/hivseq/by-mutations/) Concordance analysis This analysis was performed to determine whether the partial pol gene sequence contained enough information to define the presence of a BF recombination phenomenon after compar- ing against the result based on NFL sequences analyses. For this goal, we assessed the degree of agreement by calculating the Kappa of Fleiss index. It was carried out among the 25 near full length sequences and with the pol region belonging to them using SPSS version 22. The Kappa degree of agreement was defined according to [29]. Estimation of the time to the most recent common ancestor (tMRCA) for HIV-1 BF recombinant and subtype B strains Estimation of tMRCAs and substitution rates were performed by the Bayesian Markov chain Monte Carlo (MCMC) phylogenetic analysis as implemented in the program BEAUTI/BEAST v2 [30] that allows the substitution rate to vary among branches in the tree. The nucleotide substitution models were obtained by the jModelTest program v.2.1.10 [21] and then, they were compared using the Bayes Factor test to find the best fit with Tracer V1.5 [31]. Two clock models were used in this analysis: strict clock model, that assumes a single evolutionary rate for all the lineages, and the uncorrelated lognormal relaxed clock that allows variable evolution rates among lineages within lognormal relaxed. These two clock models were statistically com- −3 pared using Bayes Factor test to find the best fit. The empirical prior of 2.5 X 10 substitution per nucleotide site per year was used to analyze the pol gene [32]. As recombination could affect molecular clock estimation, we performed an evolutionary reconstruction of subtype B and F fragments from BF recombinant MSM sequences, independently of each other [33]. For subtype B, the region analyzed contained 669 bases in length, located in the partial pol gene between positions 3051 and 3719 according to HXB-2 numbering (n = 41). Besides, for sub- type F, the region analyzed contained 650 bases in length, located in the partial pol gene between position 2253 and 2904 referred to HXB-2 (n = 30). Comparatively, we analyzed 40 partial pol B sequences containing 1041 bases in length, located between positions 2253 and 3294 referred to as HXB-2. In order to infer the epidemiological history of our study popula- tion, three demographic models were included (constant population size, exponential growth, and Bayesian Skyline) and compared with the Bayes Factor. Finally, three independent Bayes- ian MCMC runs for 100 million generations were performed. The convergence of parameters was assessed through the effective sampling size, with all parameters for each run with values >100 indicating a sufficient level of sampling. The mean time of the tMRCA and the maxi- mum clade credibility of the phylogenetic trees were then calculated after removing 10% of the samples. Results Phylogenetic analysis and recombination detection of partial pol and near full length sequences The phylogenetic relatedness after ML analysis showed 81 out of the 312 Argentine HIV-1 pol sequences clustering in CRF_BF clade, including 46 closely related to CRF12, 14 related with other CRFs_BF (CRF17, 28, 29, 39, 42, 44, 47), and 21 sequences -defined as URFs_BF- PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 4 / 12 HIV-1 BF recombinants among Argentinean MSM Fig 1. Phylogenetic tree of pol sequences and genetic mosaic for the recombinant sequences misclassified by phylogeny. (A) Maximum likelihood phylogenetic tree of the HIV-1 pol (1467 nt) sequences obtained from 312 Argentinean MSM. Reference sequences include BF recombinant and pure subtype sequences and the tree was rooted using SIVcpzGAB2 as outgroup. The branch support values are indicated with an asterisk (aLRT> 0.9) at key nodes. The seven pol sequences indicated with a red dot at terminal node were subsequently characterized as BF recombinant by means of NFL analysis. Red boxes are clusters of nucleotide sequences exhibiting discrepancies between pol gene- and NFL-based subtype assignment are zoomed in (Fig.1B), (Fig.1C), and (Fig.1D). Genetic mosaic maps for #83109 (box I) and #143170 (box II) pol sequences are consistent with phylogeny shown in Fig.1A but the NFL-sequence genetic mosaics exhibit as BF recombinants (Fig.1B and 1C, respectively). The #237902, #237903 pol sequences (box III) are phylogenetically ascribed to pure subtype F (Fig.1A) but their genetic mosaic exhibited them as BF recombinants. The #160677 pol sequence (box III) also phylogenetically classified as pure subtype F, shows a NFL-based genetic mosaic characterized as B-C-F recombinant (Fig.1D). https://doi.org/10.1371/journal.pone.0189705.g001 segregated in an independent clade. Other 2 sequences were closely related with subtype C, and 3 sequences grouped into subtype F clade. Additionally, 3 sequences did not group any of the major clades, and the remaining 223 sequences grouped into subtype B clade (Fig 1). When investigating the level of intermixing of HIV-1 strains from MSM with those charac- terized previously from the Argentinean general population, pol gene sequences appeared intermingled. Furthermore, the phylogenetic relatedness of the HIV-1 pol sequences from MSM appeared intermingled along the BF clade independently of their recombination pattern. All these associations were sustained with high support values (Fig 1). Interestingly, two out of three pol sequences ascribed previously to subtype F by phyloge- netic analysis (#237902, #237903), appeared as a BF recombinant after recombination detec- tion methods were applied. The remaining pol sequence (#160677) classified as subtype F was PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 5 / 12 HIV-1 BF recombinants among Argentinean MSM Fig 2. Predominant BF recombination patterns detected by different methods. Maps of BF recombinant genetic mosaic patterns (A and B) and breakpoint locations at pol gene. Nucleotide sequences named with their ID are grouped according to their concordance ( ) or discrepancy (X) in breakpoint locations defined according to the three in silico methods used (jpHMM, RECCO, and RDP4). https://doi.org/10.1371/journal.pone.0189705.g002 finally characterized as triple B/C/F recombinant when the NFL sequence was analyzed (Fig 1D). The genetic mosaicism at pol gene was defined by using different in silico tools able to estab- lish the breakpoint locations. Different recombination patterns were characterized. Two of them, named A and B, exhibited one breakpoint at different locations. These two breakpoints are present in the CRF12_BF pattern plus another one at position 2641. In patterns A and B, breakpoints were located at nucleotide position 3000±25 and 2461±22 respectively. The break- point in the pattern B is also present in CRF38_BF. Thirty-six sequences showed pattern A (Fig 2A), 12 exhibited pattern B (Fig 2B), 6 exhibited non-A non-B recombination patterns and, the remaining 32 sequences showed undefined mosaic patterns that were multiple but not associ- ated with any of the previous patterns. Primary associated-resistance mutations (M184V and/or K103L) were detected in a minority of pol sequences ascribed to pattern A (n = 3) and B (n = 2). PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 6 / 12 HIV-1 BF recombinants among Argentinean MSM Fig 3. Phylogenetic tree of near full length sequences and recombination patterns corresponding to BF recombinant sequences. (A) Maximum likelihood phylogenetic tree of 25 HIV-1 NFL sequences obtained from Argentinean MSM. Those sequences characterized as BF recombinants once genetic mosaic was defined, are indicated with a red dot in the terminal node. Bootstrap values 70 are indicated in the nodes. (B) Maps of the genetic mosaic patterns including breakpoint locations of the 7 BF recombinant NFL sequences. https://doi.org/10.1371/journal.pone.0189705.g003 Phylogenetic relatedness between the 25 HIV-1 near full length (NFL) sequences depicted 4 sequences clustered within BFs recombinant clade. Of them, one was closely related to CRF12_BF and one to CRF42_BF reference sequences (#115455 and #88560 respectively), and the remaining two segregated alone (#158637 and #160677). The other 21 NFL sequences were clustered into the subtype B clade (Fig 3A). When recombination breakpoints location was defined by in silico analysis, 7 out of 25 NFL sequences showed evidence of inter-subtype recombination. Coincidentally with phylogenetic inference, the sequence #115455 shared breakpoints compatible with CRF12_BF but, as previ- ously mentioned, with pattern B at pol gene instead of pattern A. Similarly, the sequence #158637 depicts also similar breakpoints with CRF12 but exhibited the pattern A at pol gene. As mentioned above, sequence #160677 exhibited a genetic mosaic evidencing B-C-F triple recombinant. Sequence #88560 shared breakpoints with CRF29_BF (Fig 3B). Interestingly, the other three sequences that characterized phylogenetically as subtype B appeared as BF recombinants. Sequences #83109 and #143170, classified previously as B by phylogeny and recombination detection analysis on pol gene fragments (Fig 1B and 1C, PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 7 / 12 HIV-1 BF recombinants among Argentinean MSM respectively) as well as #163052 sequence included subtype F fragments with dissimilar size (from 200-to-700bp) and genomic location (gag, pol, vpr-env) (Fig 3B). These recombination patterns did not match any of the CRF_BF described previously, and may represent plausible new URFs_BF. Concordance analysis The concordance recombination detection and breakpoint positions between partial HIV-1 pol sequences and near-full length sequences data were analyzed in more detail by computing the level of agreement between the two sources of genomic evaluation. The relatively low kappa score indicates that the agreement was good (0.689±0.4 to 0.98; p<0.001) [34]. Estimation of the time to the most recent common ancestor (tMRCA) Our analysis performed over pure subtype B fragment sequences from BF recombinants (nucleotide positions 3051 and 3719 according to HXB-2 numbering) found the maximum probability for the ancestor to that clade in the late 1980s (tMRCA: 1986; 95% HPD: 11.21± 55.38) (Fig 4A). Almost coincidently, the estimation based on pure subtype F fragment sequences from BF recombinants (nucleotide positions 2253 and 2904 according to HXB-2 Fig 4. Bayesian analysis for BF recombinant sequences and pure B sequences from MSM population. Time-calibrated maximum clade credibility trees (MCCT) of HIV-1 BF and B clades identified in the pol ML analyses. (A) For BF recombinant sequences we obtained two MCCT corresponding to the analysis of pure subtype B and F fragments independently of each other. (B) MCCT obtained by analyzing the subtype B pure sequences. The posterior probabilities are indicated at key nodes ( 0.90). In each tree, the node indicating the most recent common ancestor of the MSM Argentinian B and F sequences fragments is indicated with a black circle. https://doi.org/10.1371/journal.pone.0189705.g004 PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 8 / 12 HIV-1 BF recombinants among Argentinean MSM numbering) was performed in the early 1990s (tMRCA: 1991; 95% HPD: 8.66±36.31) (Fig 4A). Regarding pure subtype B sequences (nucleotide positions 2253 and 3293 according to HXB-2 numbering), the analysis found the maximum probability for the ancestor to that clade in the early 1980s (tMRCA: 1979; 95%HPD: 20.1851±46.9062) (Fig 4B). The substitution rates found -3 -4 -3 -3 -4 -3 were 3x10 (5.27x10 ±5.77x10 ), 2.9x10 (3.18x10 ±6.17x10 ) substitutions/site/year for -3 the subtype B and F fragments from BF recombinant sequences and 1.839x10 (95%HPD: -3 9.6x10-4±2.6931x10 ) substitutions/site/year for the subtype B pure sequences, thus remain- −3 ing within the order of magnitude of 10 expected for the HIV-1 pol gene [32, 33, 35, 36]. Discussion Here we performed for the first time an extensive in-depth analysis of the genetic diversity of HIV-1 strains circulating in the MSM population from Argentina. Such characterization implied HIV-1 pol gene and near full length sequences used for phylodynamic analysis and definition of the genetic mosaic structure. In Argentina, HIV-1 BF recombinants and subtype B viruses were found with dissimilar prevalence according to the at-risk study population [8, 9, 19]. Firstly, BF recombinants have revealed the presence of three different recombination patterns after the analysis of 284 BF pol sequences from adult patients [16]. Recently, other recombinant structures have been described in two studies involving a small group of pregnant women, and a group of HIV-1 vertically-infected children [1, 18]. The present study addresses a larger population of Argentinean MSM allowing characteriz- ing 337 HIV-1 sequences. Of them, CRF12_BF-like sequences were the most common form identified (>56%) by phylogenetic inference based on pol and near full-length sequences. In a lesser extent, other BF recombinant sequences characterized at pol gene appeared to be closely related with CRFs_BF reported previously in Argentina (CRF17_BF-like) [1, 14], as well as others unrecognized previously (CRF28, 29, 39, 42, 44, and 47) but reported in neighboring countries such as Uruguay [12], Brazil [10], and Chile [13]. This finding is plausible consider- ing the fluent population exchange between these countries. Additionally, numerous recombi- nant sequences defined as URFs_BF were also found. Furthermore, BF recombinants were deeply characterized after defining the genetic mosaicism patterns using several recombina- tion detection programs simultaneously. These in silico strategies showed an acceptable coinci- dence among each other and allowed a rigorous analysis of the HIV-1 recombinants but also denote plausible inconsistencies between them: location of breakpoints, minimum length of inserted genomic fragments able to be detected and phylogenetic uncertainty. The analysis of genetic mosaicism by multiple strategies suggests the presence of unrecognized BF recombina- tion patterns allowing a deeper and unbiased molecular characterization. Nevertheless, the occurrence of recombination based on pol gene sequences exhibited a good degree of agree- ment when comparing with results based on NFL sequences, but a larger dataset is required to reach more conclusive results. Taken together, these analyses depict a complex heterogeneity for HIV-1 BF recombinant forms, including the presence of gene mosaics sharing some -but not all- recombination break- points with well-defined CRFs_BF. Thus, new genetic mosaics showing larger subtype B seg- ments appeared among MSM in Argentina. These variants may be considered as ªsecond generation recombinantsº originated by successive rounds of recombination occurred between the CRF12_BF and the subtype B circulating locally, as proposed previously [37]. Regarding the Bayesian analysis, our estimation of the tMRCA performed specifically over the BF recombinant sequences circulating in the MSM population showed a relatively recent coalescence time, consistent with previous assumptions [14]. The introduction of both B and F PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 9 / 12 HIV-1 BF recombinants among Argentinean MSM subtypes is probably a recent event dated in the end of 1980s-early 1990s. Hence, the CRFs_BF variants have been circulating among Argentinean MSM for about 30 years, although its pres- ence was not detected until recently due to its lower prevalence compared with subtype B, which appeared to be introduced in the early 1980s. Previously it was reported that the tMRCA based on subtype B fragments from CRF12_BF recombinants was introduced in the early 1970s [38]. Such a difference may be explained by the diverse experimental designs used between this study and that by Dilernia et al. [38] included individuals from the whole popula- tion of infected individuals, whereas we focused only on MSM. In addition, the time of period encompassed by Dilernia et al. was broader than the period analyzed by us. Conclusion The characterization of HIV-1 BF recombinant forms from MSM in Argentina revealed the presence of CRF12-like variants accompanied by a heterogeneous mix of strains exhibiting several mosaic patterns, including non-reported URFs. Their phylodynamic analysis showed transmission expansion and provides evidence of BF recombination as a dynamic and persis- tent phenomenon spreading among MSM in Argentina allowing the detection of active clus- ters as potential intervention targets to hinder HIV-1 transmission. Acknowledgments The authors would like to thank Andrea Rubio and Lucas Zapata for their technical contribu- tion in the performance of this work and Sergio Mazzini for his contribution to manuscript preparation. Author Contributions Conceptualization: Cintia G. Cevallos, Jean K. Carr, Maria M. Avila, Jorge Quarleri. Data curation: Leandro R. Jones, Maria A. Pando, Maria M. Avila, Jorge Quarleri. Formal analysis: Cintia G. Cevallos, Leandro R. Jones, Jorge Quarleri. Funding acquisition: Maria M. Avila, Jorge Quarleri. Investigation: Cintia G. Cevallos, Leandro R. Jones, Jorge Quarleri. Methodology: Cintia G. Cevallos, Leandro R. Jones. Project administration: Jorge Quarleri. Resources: Jorge Quarleri. Software: Leandro R. Jones. Supervision: Maria M. Avila, Jorge Quarleri. Validation: Leandro R. Jones, Jean K. Carr. Writing ± original draft: Cintia G. Cevallos, Leandro R. Jones, Maria A. Pando, Jean K. Carr, Maria M. Avila, Jorge Quarleri. Writing ± review & editing: Jean K. Carr, Maria M. Avila, Jorge Quarleri. References 1. Aulicino PC, Bello G, Guimaraes ML, Ruchansky D, Rocco C, Mangano A, et al. Longitudinal analysis of HIV-1 BF1 recombinant strains in vertically infected children from Argentina reveals a decrease in PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 10 / 12 HIV-1 BF recombinants among Argentinean MSM CRF12_BF pol gene mosaic patterns and high diversity of BF unique recombinant forms. Infect Genet Evol. 2011; 11(2):349±57. https://doi.org/10.1016/j.meegid.2010.11.008 PMID: 21130901. 2. Carrion G, Eyzaguirre L, Montano SM, Laguna-Torres V, Serra M, Aguayo N, et al. Documentation of subtype C HIV Type 1 strains in Argentina, Paraguay, and Uruguay. AIDS Res Hum Retroviruses. 2004; 20(9):1022±5. https://doi.org/10.1089/aid.2004.20.1022 PMID: 15585091. 3. Dilernia DA, Gomez AM, Lourtau L, Marone R, Losso MH, Salomon H, et al. HIV type 1 genetic diversity surveillance among newly diagnosed individuals from 2003 to 2005 in Buenos Aires, Argentina. AIDS Res Hum Retroviruses. 2007; 23(10):1201±7. https://doi.org/10.1089/aid.2007.0068 PMID: 17961105. 4. Pando MA, Eyzaguirre LM, Segura M, Bautista CT, Marone R, Ceballos A, et al. First report of an HIV-1 triple recombinant of subtypes B, C and F in Buenos Aires, Argentina. Retrovirology. 2006; 3:59. https:// doi.org/10.1186/1742-4690-3-59 PMID: 16959032. 5. Gomez-Carrillo M, Quarleri JF, Rubio AE, Carobene MG, Dilernia D, Carr JK, et al. Drug Resistance Testing Provides Evidence of the Globalization of HIV Type 1: A New Circulating Recombinant Form. AIDS Res Hum Retroviruses. 2004; 20(8): 885±8. https://doi.org/10.1089/0889222041725172 PMID: 6. Pando MA, Balan IC, Marone R, Dolezal C, Leu CS, Squiquera L, et al. HIV and other sexually transmit- ted infections among men who have sex with men recruited by RDS in Buenos Aires, Argentina: high HIV and HPV infection. PloS one. 2012; 7(6):e39834. https://doi.org/10.1371/journal.pone.0039834 PMID: 22768137. 7. Argentinean National Health Ministry; AIDS and STD Department. 33.th Bulletin of HIV-AIDS in Argen- tina. Buenos Aires, 2016: http://www.msal.gob.ar/images/stories/bes/graficos/0000000918cnt-2016- 11-30_boletin-vih-sida-2016.pdf. 8. Segura M, Sosa Estani S, Marone R, Bautista CT, Pando MA, Eyzaguirre L, et al. Buenos Aires cohort of men who have sex with men: prevalence, incidence, risk factors, and molecular genotyping of hiv type 1. AIDS Res Hum Retroviruses 2007; 23(11):1322±9. https://doi.org/10.1089/aid.2007.0063 PMID: 18184073 9. Pando MA, Gomez-Carrillo M, Vignoles M, Rubio AE, dos Ramos Farias MS, Vila M, et al. Incidence of HIV type 1 infection, antiretroviral drug resistance, and molecular characterization in newly diagnosed individuals in Argentina: A Global Fund Project. AIDS Res Hum Retroviruses. 2011; 27(1):17±23. https://doi.org/10.1089/aid.2010.0013 PMID: 20860532. 10. De Sa Filho DJ, Sucupira MC, Caseiro MM, Sabino EC, Diaz RS, Janini LM. Identification of two HIV type 1 circulating recombinant forms in Brazil. AIDS Res Hum Retroviruses. 2006; 22(1):1±13. https:// doi.org/10.1089/aid.2006.22.1 PMID: 16438639. 11. Guimaraes ML, Eyer-Silva WA, Couto-Fernandez JC, Morgado MG. Identification of two new CRF_BF in Rio de Janeiro State, Brazil. AIDS (London, England). 2008; 22(3):433±5. https://doi.org/10.1097/ QAD.0b013e3282f47ad0 PMID: 18195572. 12. Ruchansky D, Casado C, Russi JC, Arbiza JR, Lopez-Galindez C. Identification of a new HIV Type 1 cir- culating recombinant form (CRF38_BF1) in Uruguay. AIDS Res Hum Retroviruses. 2009; 25(3):351±6. https://doi.org/10.1089/aid.2008.0248 PMID: 19327055. 13. Delgado E, Rios M, Fernandez J, Perez-Alvarez L, Najera R, Thomson MM. Identification of a new HIV type 1 BF intersubtype circulating recombinant form (CRF44_BF) in Chile. AIDS Res Hum Retrovi- ruses. 2010; 26(7):821±6. https://doi.org/10.1089/aid.2010.0006 PMID: 20618101. 14. Carr JK, Avila M, Gomez Carrillo M, Salomon H, Hierholzer J, Watanaveeradej V, et al. Diverse BF recombinants have spread widely since the introduction of HIV-1 into South America. AIDS (London, England). 2001; 15(15):F41±7. PMID: 11600844. 15. Aulicino PC, Gomez-Carrillo M, Bello G, Rocco C, Mangano A, Carr J, et al. Characterization of full- length HIV-1 CRF17_BF genomes and comparison to the prototype CRF12_BF strains. Infect Genet Evol. 2012; 12(2):443±7. https://doi.org/10.1016/j.meegid.2012.01.003 PMID: 22266022. 16. Quarleri JF, Rubio A, Carobene M, Turk G, Vignoles M, Harrigan RP, et al. HIV type 1 BF recombinant strains exhibit different pol gene mosaic patterns: descriptive analysis from 284 patients under treat- ment failure. AIDS Res Hum Retroviruses. 2004; 20(10):1100±7. https://doi.org/10.1089/aid.2004.20. 1100 PMID: 15585101. 17. Thomson MM, Delgado E, Herrero I, Villahermosa ML, Vazquez-de Parga E, Cuevas MT, et al. Diver- sity of mosaic structures and common ancestry of human immunodeficiency virus type 1 BF intersub- type recombinant viruses from Argentina revealed by analysis of near full-length genome sequences. The Journal of general virology. 2002; 83(Pt 1):107±19. https://doi.org/10.1099/0022-1317-83-1-107 PMID: 11752707. 18. Gomez-Carrillo M, Pampuro S, Duran A, Losso M, Harris DR, Read JS, et al. Analysis of HIV type 1 diversity in pregnant women from four Latin American and Caribbean countries. AIDS Res Hum Retrovi- ruses. 2006; 22(11):1186±91. https://doi.org/10.1089/aid.2006.22.1186 PMID: 17147509. PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 11 / 12 HIV-1 BF recombinants among Argentinean MSM 19. Avila MM, Pando MA, Carrion G, Peralta LM, Salomon H, Carrillo MG, et al. Two HIV-1 epidemics in Argentina: different genetic subtypes associated with different risk groups. Journal of acquired immune deficiency syndromes (1999). 2002; 29(4):422±6. PMID: 11917249. 20. Katoh K, Standley DM. MAFFT: iterative refinement and additional methods. Methods in molecular biol- ogy (Clifton, NJ). 2014; 1079:131±46. https://doi.org/10.1007/978-1-62703-646-7_8 PMID: 24170399. 21. Darriba D, Taboada GL, Doallo R, Posada D. jModelTest 2: more models, new heuristics and parallel computing. Nature methods. 2015; 9(8):772. https://doi.org/10.1038/nmeth.2109 PMID: 22847109. 22. Guindon S, Dufayard JF, Lefort V, Anisimova M, Hordijk W, Gascuel O. New Algorithms and Methods to Estimate Maximum-Likelihood Phylogenies: Assessing the Performance of PhyML 3.0. Society of Systematic Biologists. 2010. 23. Anisimova M, Gascuel O. Approximate likelihood-ratio test for branches: A fast, accurate, and powerful alternative. Systematic biology. 2006; 55(4):539±52. https://doi.org/10.1080/10635150600755453 PMID: 16785212. 24. Posada D. Evaluation of methods for detecting recombination from DNA sequences: empirical data. Molecular biology and evolution. 2002; 19(5):708±17. PMID: 11961104. 25. Schultz AK, Zhang M, Bulla I, Leitner T, Korber B, Morgenstern B, et al. jpHMM: improving the reliability of recombination prediction in HIV-1. Nucleic acids research. 2009; 37(Web Server issue):W647±51. https://doi.org/10.1093/nar/gkp371 PMID: 19443440. 26. Martin DP. RDP4: Detection and analysis of recombination patterns in virus genomes. Virus Evolution. 2015; 1(1):vev003. https://doi.org/10.1093/ve/vev003 PMID: 27774277 27. Maydt J, Lengauer T. Recco: recombination analysis using cost optimization. Bioinformatics. 2006; 22 (9):1064±71. https://doi.org/10.1093/bioinformatics/btl057 PMID: 16488909. 28. Salminen MO, Carr JK, Burke DS, McCutchan FE. Identification of Breakpoints in Intergenotypic Recombinants of HIV Type 1 by Bootscanning. AIDS Res Hum Retroviruses. 1995; 11(1):1423±5. 29. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977; 33(1):159±74. PMID: 843571. 30. Bouckaert R, Heled J, Kuhnert D, Vaughan T, Wu CH, Xie D, et al. BEAST 2: a software platform for Bayesian evolutionary analysis. PLoS computational biology. 2014; 10(4):e1003537. https://doi.org/10. 1371/journal.pcbi.1003537 PMID: 24722319. 31. Suchard MA, Weiss RE, Sinsheimer JS. Bayesian selection of continuous time Markov chain evolutio- naty models. Molecular Biology Evolution. 2001; 18:1001±13. PMID: 11371589 32. Bello G, Aulicino PC, Ruchansky D, Guimaraes ML, Lopez-Galindez C, Casado C, et al. Phylodynamics of HIV-1 circulating recombinant forms 12_BF and 38_BF in Argentina and Uruguay. Retrovirology. 2010; 7:22. https://doi.org/10.1186/1742-4690-7-22 PMID: 20307282. 33. Faria NR, Rambaut A, Suchard MA, Baele G, Bedford T, Ward MJ, et al. HIV epidemiology. The early spread and epidemic ignition of HIV-1 in human populations. Science (New York, NY). 2014; 346 (6205):56±61. PMID: 25278604. 34. Lantz CA, Nebenzahl E. Behavior and interpretation of the kappa statistic: resolution of the two para- doxes. Journal of clinical epidemiology. 1996; 49(4):431±4. PMID: 8621993. 35. Ristic N, Zukurov J, Alkmim W, Diaz RS, Janini LM, Chin MP. Analysis of the origin and evolutionary his- tory of HIV-1 CRF28_BF and CRF29_BF reveals a decreasing prevalence in the AIDS epidemic of Bra- zil. PloS one. 2011; 6(3):e17485. https://doi.org/10.1371/journal.pone.0017485 PMID: 21390250. 36. Hue S, Pillay D, Clewley JP, Pybus OG. Genetic analysis reveals the complex structure of HIV-1 trans- mission within defined risk groups. Proceedings of the National Academy of Sciences of the United States of America. 2005; 102(12):4425±9. https://doi.org/10.1073/pnas.0407534102 PMID: 15767575. 37. Sierra M, Thomson MM, Rios M, Casado G, Castro RO, Delgado E, et al. The analysis of near full- length genome sequences of human immunodeficiency virus type 1 BF intersubtype recombinant viruses from Chile, Venezuela and Spain reveals their relationship to diverse lineages of recombinant viruses related to CRF12_BF. Infect Genet Evol. 2005; 5(3):209±17. https://doi.org/10.1016/j.meegid. 2004.07.010 PMID: 15737911. 38. Dilernia DA, Jones LR, Pando MA, Rabinovich RD, Damilano GD, Turk G, et al. Analysis of HIV type 1 BF recombinant sequences from South America dates the origin of CRF12_BF to a recombination event in the 1970s. AIDS Res Hum Retroviruses. 2011; 27(5):569±78. https://doi.org/10.1089/aid.2010. 0118 PMID: 20919926. PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 12 / 12 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png PLoS ONE Pubmed Central

Genomic characterization and molecular evolution analysis of subtype B and BF recombinant HIV-1 strains among Argentinean men who have sex with men reveal a complex scenario

Loading next page...
 
/lp/pubmed-central/genomic-characterization-and-molecular-evolution-analysis-of-subtype-b-ki6EaQJJ8R

References (78)

Publisher
Pubmed Central
Copyright
© 2017 Cevallos et al
ISSN
1932-6203
eISSN
1932-6203
DOI
10.1371/journal.pone.0189705
Publisher site
See Article on Publisher Site

Abstract

characterization and molecular evolution analysis of subtype B and BF recombinant HIV-1 strains Currently, data on HIV-1 circulating strains among men who have sex with men (MSM) in among Argentinean men who have sex with men Argentina is scarce. In South America, the distribution and the prevalence of BF recombi- reveal a complex scenario. PLoS ONE 12(12): nants are dissimilar and exhibit an underappreciated heterogeneity of recombinant struc- e0189705. https://doi.org/10.1371/journal. pone.0189705 tures. Here, we studied for the first time the genetic diversity of HIV-1 BF recombinants and their evolution over time through in-depth phylogenetic analysis and multiple recombination Editor: Douglas F. Nixon, George Washington University, UNITED STATES detection methods involving 337 HIV-1 nucleotide sequences (25 near full-length (NFL) and 312 partial pol gene) obtained from Argentinean MSM. The recombination profiles were Received: May 5, 2017 studied using multiple in silico tools to characterize the genetic mosaicism, and phylogenetic Accepted: November 30, 2017 approaches to infer their relationships. The evolutionary history of BF recombinants and Published: December 15, 2017 subtype B sequences was reconstructed by a Bayesian coalescent-based method. By Copyright:© 2017 Cevallos et al. This is an open phylogenetic inference, 81/312 pol sequences clustered within BF clade. Of them, 46 access article distributed under the terms of the sequences showed a genetic mosaic with CRF12_BF-like patterns, including plausible sec- Creative Commons Attribution License, which ond-generation recombinants. Other CRFs_BF like (CRF17, 28, 29, 39, 42, 44, 47) and permits unrestricted use, distribution, and reproduction in any medium, provided the original probable URFs_BF were less frequently found. Phylogenetic and recombination analyses author and source are credited. on NFL sequences allowed a meticulous definition of new BF mosaics of genomic patterns. Data Availability Statement: All relevant data are The Bayesian analyses pointed out quite consistent onset dates for the CRFs_BF clade within the paper. All nucleotide sequences are based on B and F gene datasets (~1986 and ~1991 respectively). These results indicate deposited in GenBank database (accession that the CRFs_BF variants have been circulating among Argentinean MSM for about 30 numbers DQ383746- DQ38375, KY968394- years. This study reveals, through growing evidence showing the importance of MSM in the KY968408, EF120266, EF120267, EF120268, EF120269, EF120270, EF120271, EF120272, dynamics of the HIV-1 epidemic in Argentina, the coexistence of CRF12_BF-like and high EF120274, EF120275, EF120276, EF120277, diversity of strains exhibiting several BF mosaic patterns, including non-reported URFs that EF120278, EF120279, EF120280, EF120281, may reflect active clusters as potential intervention targets to hinder HIV-1 transmission. EF120282, EF120283, EF120284, EF120285, EF120286, EF120287, EF120288, EF120289, PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 1 / 12 HIV-1 BF recombinants among Argentinean MSM EF120290, EF120291, EF120292, EF120293, Background EF120294, EF120295, EF120296, EF120297, The HIV-1 epidemic in Argentina is mostly represented by subtype B, plus a wide variety of EF120298, EF120300, EF120301, EF120302, recombinant strains between subtypes B and F. Other less frequent HIV-1 subtypes (A, C, F) EF120303, EF120304, EF120305, EF120306, EF120307, EF120308, EF120309, EF120310, and recombinant forms (BC, BCF, A2D, and AG) have also been characterized [1±5]. Since EF120311, EF120312, EF120313, EF120314, the onset of the HIV/AIDS epidemic, men who have sex with men (MSM) have remained a EF120315, EF120316, EF120317, EF120318, recognized high-risk group for HIV-1 infection. It is well known worldwide that many MSM EF120319, EF120320, EF120321, EF120322, are at high±risk for infection, hence, HIV-1 prevalence and incidence in MSM are consistently EF120323, EF120324, EF120325, EF120326, higher than in the general population [6]. According to official data published in 2016 by the EF120327, EF120328, EF120329, EF120330, EF120331, EF120332, EF120333, EF120334, Ministry of Health, Argentina, the current prevalence in MSM is between 12 and 15% [7]. The EF120335, EF120351, EF120353, EF120354, genomic characterization of HIV-1 among Argentinean MSM was performed by phylogenetic EF120356, KY968342- KY968393, KY968409- inference [3, 8, 9]. Since then, no other evidence has been provided regarding HIV-1 genetic KY968571). variability associated with BF recombinants among MSM. Funding: This study was supported by grants In South America, the prevalence rates of BF recombinant sequences, either Circulating obtained by MMA from the Agencia Nacional de Recombinant Forms (CRF_BF) or Unique Recombinant Forms (URF_BF), have been found Promocio Ân Cientifica y Tecnologica-Fondo to be significant. Other CRF_BF have also been described in South America: CRF28_BF and Nacional para la Ciencia y Tecnica (grant number CRF29_BF in Sao Paulo (Brazil) [10]; CRF39_BF and CRF40_BF in Rio de Janeiro (Brazil) PICT 2013/2229) and the Consejo Nacional de [11]; CRF38_BF in Uruguay [12]; and CRF44_BF in Chile [13]. In Argentina after the identifi- Investigaciones Cientificas y Tecnicas (grant number PIP 2012/1048) to MMA. The funders had cation of the CRF12_BF and CRF17_BF in 2001 [14, 15], several studies showed that most BF no role in study design, data collection and recombinants circulating locally were closely associated with CRF12_BF or diverse BF recom- analysis, decision to publish, or preparation of the binants related to it [1, 16±18]. manuscript. The aim of the present study was to evaluate the heterogeneity of the HIV-1 associated Competing interests: The authors have declared with subtype B and BF recombinants and their evolution over time through in-depth phylo- that no competing interests exist. genetic and Bayesian coalescent analysis and multiple recombination detection methods among Argentinean MSM. Using near-full length and pol-gene sequences, this study reveals a complex scenario for BF recombinants among MSM where CRF12_BF-like and high diversity of strains exhibiting several BF mosaic patterns -including non-reported URFs- coexist. Materials and methods HIV-1 nucleotide sequences datasets Two datasets were constructed involving a total of 337 HIV-1 nucleotide sequences. These include 312 partial pol gene sequences (defined from nucleotide 2253 up to 3720 according to HXB-2 numbering) and 25 near full-length (NFL) sequences. These were obtained from 429 HIV-1 seropositive Argentinean men who have sex with men (MSM) recruited in previous studies performed during the period 2000±2009. They are mainly residents of Buenos Aires city and surroundings areas with a median age of 30 years, and most of them were not on anti- viral therapy at the time of the studies. As documented previously, the consumption of illicit drugs presented by this population was very infrequent. Blood collection was performed once a written informed consent was obtained as stated in each of the original studies [2, 4, 8, 9, 19]. The methods used for extraction of the DNA and sequencing are described in the respective references. All the sequences used in this work were obtained from projects previously approved by the Ethic Committee from School of Medicine of Buenos Aires University or Ethic Committee of NEXO AC (IRB 00005349). In all cases the consent of each participant, confidentiality of the information as well as laboratory results, were obtained according to international recommendation and current legislation (National AIDS Law No. 23,798 and its regulatory decree) and the recommended international ethical guidelines for epidemiological studies were strictly followed, (CIOMS, Ginebra 2009). PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 2 / 12 HIV-1 BF recombinants among Argentinean MSM HIV-1 subtype references strains available in Los Alamos Database (http://www.hiv.lanl. gov/content/sequence/HIV/mainpage.html) were used to make alignments with the Argentin- ean sequences under analysis. The alignments were performed with MAFFT software included in http://www.hiv.lanl.gov/content/sequence/VIRALIGN/viralign.html [20]. Phylogenetic analysis Phylogenetic analyses were carried out in order to classify the viruses by HIV-1 subtype or CRFs. Phylogenetic trees were constructed with a GTR nucleotide substitution model among site rate heterogeneity and a proportion of invariable site selected by jModelTest v.2.1.10 [21] using the Maximum Likelihood (ML) method implemented in PhyML v3.0 employing an online web server (http://www.atgc-montpellier.fr/phyml/) [22]. The best SPR and NNI heu- ristic options were selected. The reliability of tree topologies was assessed by two methods according to dataset, such as bootstrapping using 500 replications (values 70% were consid- ered significant) for NFL dataset, and approximate likelihood ratios (aLRT) 0.90 based on Shimodaira-Hasegawa-like procedure were used to assign the subtype/CRF designation of the strains for pol dataset [23]. Detection of recombination events and resistance mutations The HIV-1 intersubtype recombination patterns were analyzed separately for each sequence by determining the genetic mosaic structure, location of breakpoints, and parental subtypes involved. For this purpose, several in silico methods were selected considering both the type of analy- sis to be carried out and the nature of the database [24]. Initially, pol sequences were analyzed by three different methods or algorithms including: (i) Jumping profile Hidden Markov Model (jpHMM) method. It is based on a pre-calculated multiple alignment of the major HIV-1 subtypes where each one is represented by a profile hidden Markov model further con- nected by empirical probabilities, allowing the detection of possible recombinants and plausi- ble breakpoints. The HIV-1 subtype B and subtype F have adequate sampling to build well informed profiles offering enough data to construct a good model of sequence variation [25], (ii) Recombination Detection Program (RDP4), a software package for statistical identifi- cation and characterization of recombination events in DNA sequences that uses several non- parametric recombination detection methods. Among multiple approaches to identify recom- bination signals, RDP, BOOTSCAN, GENECONV, MAXCHI, and SISCAN methods were carried out. All methods are described in detail in reference [26]. This tool considers every sequence included in the analysis as a potential recombinant and monitors systematically each sequence triplets or quartets in order to identify those that contain a recombinant as well as those two sequences that could serve as parents while performing a statistical evaluation of recombination signals. The highest acceptable P-value was set to 0.05 and the other parameters were default RDP4 settings. Recombination events detected by at least three different methods (Pvalues of0.05 after Bonferroni correction for multiple comparisons implemented in RDP4) were taken as credible evidence of possible HIV-1 recombination. Lastly, (iii) the recombination analysis using cost optimization (RECCO) was performed. This software com- putes the number of ªsavingsº, i.e., number of mismatches saved when explaining the read by a recombination event between two viral strains and mutations, rather than by a single strain and mutations only. In this analysis, sequences were defined as recombinant when RECCO assigns a savings value higher than 5 [27]. When conflicting results emerged in recombination events, the criteria were to consider as actual event in which at least two out of the three meth- ods used would match. Only for near full length sequences, bootscanning analysis was also PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 3 / 12 HIV-1 BF recombinants among Argentinean MSM performed separately to detect recombination as implemented in the SimPlot software version 3.5.1 (http://www.med.jhu.edu/deptmed/sray/download/) [28]. The presence of resistance mutations was detected by the Stanford algorithm (https:// hivdb.stanford.edu/hivseq/by-mutations/) Concordance analysis This analysis was performed to determine whether the partial pol gene sequence contained enough information to define the presence of a BF recombination phenomenon after compar- ing against the result based on NFL sequences analyses. For this goal, we assessed the degree of agreement by calculating the Kappa of Fleiss index. It was carried out among the 25 near full length sequences and with the pol region belonging to them using SPSS version 22. The Kappa degree of agreement was defined according to [29]. Estimation of the time to the most recent common ancestor (tMRCA) for HIV-1 BF recombinant and subtype B strains Estimation of tMRCAs and substitution rates were performed by the Bayesian Markov chain Monte Carlo (MCMC) phylogenetic analysis as implemented in the program BEAUTI/BEAST v2 [30] that allows the substitution rate to vary among branches in the tree. The nucleotide substitution models were obtained by the jModelTest program v.2.1.10 [21] and then, they were compared using the Bayes Factor test to find the best fit with Tracer V1.5 [31]. Two clock models were used in this analysis: strict clock model, that assumes a single evolutionary rate for all the lineages, and the uncorrelated lognormal relaxed clock that allows variable evolution rates among lineages within lognormal relaxed. These two clock models were statistically com- −3 pared using Bayes Factor test to find the best fit. The empirical prior of 2.5 X 10 substitution per nucleotide site per year was used to analyze the pol gene [32]. As recombination could affect molecular clock estimation, we performed an evolutionary reconstruction of subtype B and F fragments from BF recombinant MSM sequences, independently of each other [33]. For subtype B, the region analyzed contained 669 bases in length, located in the partial pol gene between positions 3051 and 3719 according to HXB-2 numbering (n = 41). Besides, for sub- type F, the region analyzed contained 650 bases in length, located in the partial pol gene between position 2253 and 2904 referred to HXB-2 (n = 30). Comparatively, we analyzed 40 partial pol B sequences containing 1041 bases in length, located between positions 2253 and 3294 referred to as HXB-2. In order to infer the epidemiological history of our study popula- tion, three demographic models were included (constant population size, exponential growth, and Bayesian Skyline) and compared with the Bayes Factor. Finally, three independent Bayes- ian MCMC runs for 100 million generations were performed. The convergence of parameters was assessed through the effective sampling size, with all parameters for each run with values >100 indicating a sufficient level of sampling. The mean time of the tMRCA and the maxi- mum clade credibility of the phylogenetic trees were then calculated after removing 10% of the samples. Results Phylogenetic analysis and recombination detection of partial pol and near full length sequences The phylogenetic relatedness after ML analysis showed 81 out of the 312 Argentine HIV-1 pol sequences clustering in CRF_BF clade, including 46 closely related to CRF12, 14 related with other CRFs_BF (CRF17, 28, 29, 39, 42, 44, 47), and 21 sequences -defined as URFs_BF- PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 4 / 12 HIV-1 BF recombinants among Argentinean MSM Fig 1. Phylogenetic tree of pol sequences and genetic mosaic for the recombinant sequences misclassified by phylogeny. (A) Maximum likelihood phylogenetic tree of the HIV-1 pol (1467 nt) sequences obtained from 312 Argentinean MSM. Reference sequences include BF recombinant and pure subtype sequences and the tree was rooted using SIVcpzGAB2 as outgroup. The branch support values are indicated with an asterisk (aLRT> 0.9) at key nodes. The seven pol sequences indicated with a red dot at terminal node were subsequently characterized as BF recombinant by means of NFL analysis. Red boxes are clusters of nucleotide sequences exhibiting discrepancies between pol gene- and NFL-based subtype assignment are zoomed in (Fig.1B), (Fig.1C), and (Fig.1D). Genetic mosaic maps for #83109 (box I) and #143170 (box II) pol sequences are consistent with phylogeny shown in Fig.1A but the NFL-sequence genetic mosaics exhibit as BF recombinants (Fig.1B and 1C, respectively). The #237902, #237903 pol sequences (box III) are phylogenetically ascribed to pure subtype F (Fig.1A) but their genetic mosaic exhibited them as BF recombinants. The #160677 pol sequence (box III) also phylogenetically classified as pure subtype F, shows a NFL-based genetic mosaic characterized as B-C-F recombinant (Fig.1D). https://doi.org/10.1371/journal.pone.0189705.g001 segregated in an independent clade. Other 2 sequences were closely related with subtype C, and 3 sequences grouped into subtype F clade. Additionally, 3 sequences did not group any of the major clades, and the remaining 223 sequences grouped into subtype B clade (Fig 1). When investigating the level of intermixing of HIV-1 strains from MSM with those charac- terized previously from the Argentinean general population, pol gene sequences appeared intermingled. Furthermore, the phylogenetic relatedness of the HIV-1 pol sequences from MSM appeared intermingled along the BF clade independently of their recombination pattern. All these associations were sustained with high support values (Fig 1). Interestingly, two out of three pol sequences ascribed previously to subtype F by phyloge- netic analysis (#237902, #237903), appeared as a BF recombinant after recombination detec- tion methods were applied. The remaining pol sequence (#160677) classified as subtype F was PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 5 / 12 HIV-1 BF recombinants among Argentinean MSM Fig 2. Predominant BF recombination patterns detected by different methods. Maps of BF recombinant genetic mosaic patterns (A and B) and breakpoint locations at pol gene. Nucleotide sequences named with their ID are grouped according to their concordance ( ) or discrepancy (X) in breakpoint locations defined according to the three in silico methods used (jpHMM, RECCO, and RDP4). https://doi.org/10.1371/journal.pone.0189705.g002 finally characterized as triple B/C/F recombinant when the NFL sequence was analyzed (Fig 1D). The genetic mosaicism at pol gene was defined by using different in silico tools able to estab- lish the breakpoint locations. Different recombination patterns were characterized. Two of them, named A and B, exhibited one breakpoint at different locations. These two breakpoints are present in the CRF12_BF pattern plus another one at position 2641. In patterns A and B, breakpoints were located at nucleotide position 3000±25 and 2461±22 respectively. The break- point in the pattern B is also present in CRF38_BF. Thirty-six sequences showed pattern A (Fig 2A), 12 exhibited pattern B (Fig 2B), 6 exhibited non-A non-B recombination patterns and, the remaining 32 sequences showed undefined mosaic patterns that were multiple but not associ- ated with any of the previous patterns. Primary associated-resistance mutations (M184V and/or K103L) were detected in a minority of pol sequences ascribed to pattern A (n = 3) and B (n = 2). PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 6 / 12 HIV-1 BF recombinants among Argentinean MSM Fig 3. Phylogenetic tree of near full length sequences and recombination patterns corresponding to BF recombinant sequences. (A) Maximum likelihood phylogenetic tree of 25 HIV-1 NFL sequences obtained from Argentinean MSM. Those sequences characterized as BF recombinants once genetic mosaic was defined, are indicated with a red dot in the terminal node. Bootstrap values 70 are indicated in the nodes. (B) Maps of the genetic mosaic patterns including breakpoint locations of the 7 BF recombinant NFL sequences. https://doi.org/10.1371/journal.pone.0189705.g003 Phylogenetic relatedness between the 25 HIV-1 near full length (NFL) sequences depicted 4 sequences clustered within BFs recombinant clade. Of them, one was closely related to CRF12_BF and one to CRF42_BF reference sequences (#115455 and #88560 respectively), and the remaining two segregated alone (#158637 and #160677). The other 21 NFL sequences were clustered into the subtype B clade (Fig 3A). When recombination breakpoints location was defined by in silico analysis, 7 out of 25 NFL sequences showed evidence of inter-subtype recombination. Coincidentally with phylogenetic inference, the sequence #115455 shared breakpoints compatible with CRF12_BF but, as previ- ously mentioned, with pattern B at pol gene instead of pattern A. Similarly, the sequence #158637 depicts also similar breakpoints with CRF12 but exhibited the pattern A at pol gene. As mentioned above, sequence #160677 exhibited a genetic mosaic evidencing B-C-F triple recombinant. Sequence #88560 shared breakpoints with CRF29_BF (Fig 3B). Interestingly, the other three sequences that characterized phylogenetically as subtype B appeared as BF recombinants. Sequences #83109 and #143170, classified previously as B by phylogeny and recombination detection analysis on pol gene fragments (Fig 1B and 1C, PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 7 / 12 HIV-1 BF recombinants among Argentinean MSM respectively) as well as #163052 sequence included subtype F fragments with dissimilar size (from 200-to-700bp) and genomic location (gag, pol, vpr-env) (Fig 3B). These recombination patterns did not match any of the CRF_BF described previously, and may represent plausible new URFs_BF. Concordance analysis The concordance recombination detection and breakpoint positions between partial HIV-1 pol sequences and near-full length sequences data were analyzed in more detail by computing the level of agreement between the two sources of genomic evaluation. The relatively low kappa score indicates that the agreement was good (0.689±0.4 to 0.98; p<0.001) [34]. Estimation of the time to the most recent common ancestor (tMRCA) Our analysis performed over pure subtype B fragment sequences from BF recombinants (nucleotide positions 3051 and 3719 according to HXB-2 numbering) found the maximum probability for the ancestor to that clade in the late 1980s (tMRCA: 1986; 95% HPD: 11.21± 55.38) (Fig 4A). Almost coincidently, the estimation based on pure subtype F fragment sequences from BF recombinants (nucleotide positions 2253 and 2904 according to HXB-2 Fig 4. Bayesian analysis for BF recombinant sequences and pure B sequences from MSM population. Time-calibrated maximum clade credibility trees (MCCT) of HIV-1 BF and B clades identified in the pol ML analyses. (A) For BF recombinant sequences we obtained two MCCT corresponding to the analysis of pure subtype B and F fragments independently of each other. (B) MCCT obtained by analyzing the subtype B pure sequences. The posterior probabilities are indicated at key nodes ( 0.90). In each tree, the node indicating the most recent common ancestor of the MSM Argentinian B and F sequences fragments is indicated with a black circle. https://doi.org/10.1371/journal.pone.0189705.g004 PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 8 / 12 HIV-1 BF recombinants among Argentinean MSM numbering) was performed in the early 1990s (tMRCA: 1991; 95% HPD: 8.66±36.31) (Fig 4A). Regarding pure subtype B sequences (nucleotide positions 2253 and 3293 according to HXB-2 numbering), the analysis found the maximum probability for the ancestor to that clade in the early 1980s (tMRCA: 1979; 95%HPD: 20.1851±46.9062) (Fig 4B). The substitution rates found -3 -4 -3 -3 -4 -3 were 3x10 (5.27x10 ±5.77x10 ), 2.9x10 (3.18x10 ±6.17x10 ) substitutions/site/year for -3 the subtype B and F fragments from BF recombinant sequences and 1.839x10 (95%HPD: -3 9.6x10-4±2.6931x10 ) substitutions/site/year for the subtype B pure sequences, thus remain- −3 ing within the order of magnitude of 10 expected for the HIV-1 pol gene [32, 33, 35, 36]. Discussion Here we performed for the first time an extensive in-depth analysis of the genetic diversity of HIV-1 strains circulating in the MSM population from Argentina. Such characterization implied HIV-1 pol gene and near full length sequences used for phylodynamic analysis and definition of the genetic mosaic structure. In Argentina, HIV-1 BF recombinants and subtype B viruses were found with dissimilar prevalence according to the at-risk study population [8, 9, 19]. Firstly, BF recombinants have revealed the presence of three different recombination patterns after the analysis of 284 BF pol sequences from adult patients [16]. Recently, other recombinant structures have been described in two studies involving a small group of pregnant women, and a group of HIV-1 vertically-infected children [1, 18]. The present study addresses a larger population of Argentinean MSM allowing characteriz- ing 337 HIV-1 sequences. Of them, CRF12_BF-like sequences were the most common form identified (>56%) by phylogenetic inference based on pol and near full-length sequences. In a lesser extent, other BF recombinant sequences characterized at pol gene appeared to be closely related with CRFs_BF reported previously in Argentina (CRF17_BF-like) [1, 14], as well as others unrecognized previously (CRF28, 29, 39, 42, 44, and 47) but reported in neighboring countries such as Uruguay [12], Brazil [10], and Chile [13]. This finding is plausible consider- ing the fluent population exchange between these countries. Additionally, numerous recombi- nant sequences defined as URFs_BF were also found. Furthermore, BF recombinants were deeply characterized after defining the genetic mosaicism patterns using several recombina- tion detection programs simultaneously. These in silico strategies showed an acceptable coinci- dence among each other and allowed a rigorous analysis of the HIV-1 recombinants but also denote plausible inconsistencies between them: location of breakpoints, minimum length of inserted genomic fragments able to be detected and phylogenetic uncertainty. The analysis of genetic mosaicism by multiple strategies suggests the presence of unrecognized BF recombina- tion patterns allowing a deeper and unbiased molecular characterization. Nevertheless, the occurrence of recombination based on pol gene sequences exhibited a good degree of agree- ment when comparing with results based on NFL sequences, but a larger dataset is required to reach more conclusive results. Taken together, these analyses depict a complex heterogeneity for HIV-1 BF recombinant forms, including the presence of gene mosaics sharing some -but not all- recombination break- points with well-defined CRFs_BF. Thus, new genetic mosaics showing larger subtype B seg- ments appeared among MSM in Argentina. These variants may be considered as ªsecond generation recombinantsº originated by successive rounds of recombination occurred between the CRF12_BF and the subtype B circulating locally, as proposed previously [37]. Regarding the Bayesian analysis, our estimation of the tMRCA performed specifically over the BF recombinant sequences circulating in the MSM population showed a relatively recent coalescence time, consistent with previous assumptions [14]. The introduction of both B and F PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 9 / 12 HIV-1 BF recombinants among Argentinean MSM subtypes is probably a recent event dated in the end of 1980s-early 1990s. Hence, the CRFs_BF variants have been circulating among Argentinean MSM for about 30 years, although its pres- ence was not detected until recently due to its lower prevalence compared with subtype B, which appeared to be introduced in the early 1980s. Previously it was reported that the tMRCA based on subtype B fragments from CRF12_BF recombinants was introduced in the early 1970s [38]. Such a difference may be explained by the diverse experimental designs used between this study and that by Dilernia et al. [38] included individuals from the whole popula- tion of infected individuals, whereas we focused only on MSM. In addition, the time of period encompassed by Dilernia et al. was broader than the period analyzed by us. Conclusion The characterization of HIV-1 BF recombinant forms from MSM in Argentina revealed the presence of CRF12-like variants accompanied by a heterogeneous mix of strains exhibiting several mosaic patterns, including non-reported URFs. Their phylodynamic analysis showed transmission expansion and provides evidence of BF recombination as a dynamic and persis- tent phenomenon spreading among MSM in Argentina allowing the detection of active clus- ters as potential intervention targets to hinder HIV-1 transmission. Acknowledgments The authors would like to thank Andrea Rubio and Lucas Zapata for their technical contribu- tion in the performance of this work and Sergio Mazzini for his contribution to manuscript preparation. Author Contributions Conceptualization: Cintia G. Cevallos, Jean K. Carr, Maria M. Avila, Jorge Quarleri. Data curation: Leandro R. Jones, Maria A. Pando, Maria M. Avila, Jorge Quarleri. Formal analysis: Cintia G. Cevallos, Leandro R. Jones, Jorge Quarleri. Funding acquisition: Maria M. Avila, Jorge Quarleri. Investigation: Cintia G. Cevallos, Leandro R. Jones, Jorge Quarleri. Methodology: Cintia G. Cevallos, Leandro R. Jones. Project administration: Jorge Quarleri. Resources: Jorge Quarleri. Software: Leandro R. Jones. Supervision: Maria M. Avila, Jorge Quarleri. Validation: Leandro R. Jones, Jean K. Carr. Writing ± original draft: Cintia G. Cevallos, Leandro R. Jones, Maria A. Pando, Jean K. Carr, Maria M. Avila, Jorge Quarleri. Writing ± review & editing: Jean K. Carr, Maria M. Avila, Jorge Quarleri. References 1. Aulicino PC, Bello G, Guimaraes ML, Ruchansky D, Rocco C, Mangano A, et al. Longitudinal analysis of HIV-1 BF1 recombinant strains in vertically infected children from Argentina reveals a decrease in PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 10 / 12 HIV-1 BF recombinants among Argentinean MSM CRF12_BF pol gene mosaic patterns and high diversity of BF unique recombinant forms. Infect Genet Evol. 2011; 11(2):349±57. https://doi.org/10.1016/j.meegid.2010.11.008 PMID: 21130901. 2. Carrion G, Eyzaguirre L, Montano SM, Laguna-Torres V, Serra M, Aguayo N, et al. Documentation of subtype C HIV Type 1 strains in Argentina, Paraguay, and Uruguay. AIDS Res Hum Retroviruses. 2004; 20(9):1022±5. https://doi.org/10.1089/aid.2004.20.1022 PMID: 15585091. 3. Dilernia DA, Gomez AM, Lourtau L, Marone R, Losso MH, Salomon H, et al. HIV type 1 genetic diversity surveillance among newly diagnosed individuals from 2003 to 2005 in Buenos Aires, Argentina. AIDS Res Hum Retroviruses. 2007; 23(10):1201±7. https://doi.org/10.1089/aid.2007.0068 PMID: 17961105. 4. Pando MA, Eyzaguirre LM, Segura M, Bautista CT, Marone R, Ceballos A, et al. First report of an HIV-1 triple recombinant of subtypes B, C and F in Buenos Aires, Argentina. Retrovirology. 2006; 3:59. https:// doi.org/10.1186/1742-4690-3-59 PMID: 16959032. 5. Gomez-Carrillo M, Quarleri JF, Rubio AE, Carobene MG, Dilernia D, Carr JK, et al. Drug Resistance Testing Provides Evidence of the Globalization of HIV Type 1: A New Circulating Recombinant Form. AIDS Res Hum Retroviruses. 2004; 20(8): 885±8. https://doi.org/10.1089/0889222041725172 PMID: 6. Pando MA, Balan IC, Marone R, Dolezal C, Leu CS, Squiquera L, et al. HIV and other sexually transmit- ted infections among men who have sex with men recruited by RDS in Buenos Aires, Argentina: high HIV and HPV infection. PloS one. 2012; 7(6):e39834. https://doi.org/10.1371/journal.pone.0039834 PMID: 22768137. 7. Argentinean National Health Ministry; AIDS and STD Department. 33.th Bulletin of HIV-AIDS in Argen- tina. Buenos Aires, 2016: http://www.msal.gob.ar/images/stories/bes/graficos/0000000918cnt-2016- 11-30_boletin-vih-sida-2016.pdf. 8. Segura M, Sosa Estani S, Marone R, Bautista CT, Pando MA, Eyzaguirre L, et al. Buenos Aires cohort of men who have sex with men: prevalence, incidence, risk factors, and molecular genotyping of hiv type 1. AIDS Res Hum Retroviruses 2007; 23(11):1322±9. https://doi.org/10.1089/aid.2007.0063 PMID: 18184073 9. Pando MA, Gomez-Carrillo M, Vignoles M, Rubio AE, dos Ramos Farias MS, Vila M, et al. Incidence of HIV type 1 infection, antiretroviral drug resistance, and molecular characterization in newly diagnosed individuals in Argentina: A Global Fund Project. AIDS Res Hum Retroviruses. 2011; 27(1):17±23. https://doi.org/10.1089/aid.2010.0013 PMID: 20860532. 10. De Sa Filho DJ, Sucupira MC, Caseiro MM, Sabino EC, Diaz RS, Janini LM. Identification of two HIV type 1 circulating recombinant forms in Brazil. AIDS Res Hum Retroviruses. 2006; 22(1):1±13. https:// doi.org/10.1089/aid.2006.22.1 PMID: 16438639. 11. Guimaraes ML, Eyer-Silva WA, Couto-Fernandez JC, Morgado MG. Identification of two new CRF_BF in Rio de Janeiro State, Brazil. AIDS (London, England). 2008; 22(3):433±5. https://doi.org/10.1097/ QAD.0b013e3282f47ad0 PMID: 18195572. 12. Ruchansky D, Casado C, Russi JC, Arbiza JR, Lopez-Galindez C. Identification of a new HIV Type 1 cir- culating recombinant form (CRF38_BF1) in Uruguay. AIDS Res Hum Retroviruses. 2009; 25(3):351±6. https://doi.org/10.1089/aid.2008.0248 PMID: 19327055. 13. Delgado E, Rios M, Fernandez J, Perez-Alvarez L, Najera R, Thomson MM. Identification of a new HIV type 1 BF intersubtype circulating recombinant form (CRF44_BF) in Chile. AIDS Res Hum Retrovi- ruses. 2010; 26(7):821±6. https://doi.org/10.1089/aid.2010.0006 PMID: 20618101. 14. Carr JK, Avila M, Gomez Carrillo M, Salomon H, Hierholzer J, Watanaveeradej V, et al. Diverse BF recombinants have spread widely since the introduction of HIV-1 into South America. AIDS (London, England). 2001; 15(15):F41±7. PMID: 11600844. 15. Aulicino PC, Gomez-Carrillo M, Bello G, Rocco C, Mangano A, Carr J, et al. Characterization of full- length HIV-1 CRF17_BF genomes and comparison to the prototype CRF12_BF strains. Infect Genet Evol. 2012; 12(2):443±7. https://doi.org/10.1016/j.meegid.2012.01.003 PMID: 22266022. 16. Quarleri JF, Rubio A, Carobene M, Turk G, Vignoles M, Harrigan RP, et al. HIV type 1 BF recombinant strains exhibit different pol gene mosaic patterns: descriptive analysis from 284 patients under treat- ment failure. AIDS Res Hum Retroviruses. 2004; 20(10):1100±7. https://doi.org/10.1089/aid.2004.20. 1100 PMID: 15585101. 17. Thomson MM, Delgado E, Herrero I, Villahermosa ML, Vazquez-de Parga E, Cuevas MT, et al. Diver- sity of mosaic structures and common ancestry of human immunodeficiency virus type 1 BF intersub- type recombinant viruses from Argentina revealed by analysis of near full-length genome sequences. The Journal of general virology. 2002; 83(Pt 1):107±19. https://doi.org/10.1099/0022-1317-83-1-107 PMID: 11752707. 18. Gomez-Carrillo M, Pampuro S, Duran A, Losso M, Harris DR, Read JS, et al. Analysis of HIV type 1 diversity in pregnant women from four Latin American and Caribbean countries. AIDS Res Hum Retrovi- ruses. 2006; 22(11):1186±91. https://doi.org/10.1089/aid.2006.22.1186 PMID: 17147509. PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 11 / 12 HIV-1 BF recombinants among Argentinean MSM 19. Avila MM, Pando MA, Carrion G, Peralta LM, Salomon H, Carrillo MG, et al. Two HIV-1 epidemics in Argentina: different genetic subtypes associated with different risk groups. Journal of acquired immune deficiency syndromes (1999). 2002; 29(4):422±6. PMID: 11917249. 20. Katoh K, Standley DM. MAFFT: iterative refinement and additional methods. Methods in molecular biol- ogy (Clifton, NJ). 2014; 1079:131±46. https://doi.org/10.1007/978-1-62703-646-7_8 PMID: 24170399. 21. Darriba D, Taboada GL, Doallo R, Posada D. jModelTest 2: more models, new heuristics and parallel computing. Nature methods. 2015; 9(8):772. https://doi.org/10.1038/nmeth.2109 PMID: 22847109. 22. Guindon S, Dufayard JF, Lefort V, Anisimova M, Hordijk W, Gascuel O. New Algorithms and Methods to Estimate Maximum-Likelihood Phylogenies: Assessing the Performance of PhyML 3.0. Society of Systematic Biologists. 2010. 23. Anisimova M, Gascuel O. Approximate likelihood-ratio test for branches: A fast, accurate, and powerful alternative. Systematic biology. 2006; 55(4):539±52. https://doi.org/10.1080/10635150600755453 PMID: 16785212. 24. Posada D. Evaluation of methods for detecting recombination from DNA sequences: empirical data. Molecular biology and evolution. 2002; 19(5):708±17. PMID: 11961104. 25. Schultz AK, Zhang M, Bulla I, Leitner T, Korber B, Morgenstern B, et al. jpHMM: improving the reliability of recombination prediction in HIV-1. Nucleic acids research. 2009; 37(Web Server issue):W647±51. https://doi.org/10.1093/nar/gkp371 PMID: 19443440. 26. Martin DP. RDP4: Detection and analysis of recombination patterns in virus genomes. Virus Evolution. 2015; 1(1):vev003. https://doi.org/10.1093/ve/vev003 PMID: 27774277 27. Maydt J, Lengauer T. Recco: recombination analysis using cost optimization. Bioinformatics. 2006; 22 (9):1064±71. https://doi.org/10.1093/bioinformatics/btl057 PMID: 16488909. 28. Salminen MO, Carr JK, Burke DS, McCutchan FE. Identification of Breakpoints in Intergenotypic Recombinants of HIV Type 1 by Bootscanning. AIDS Res Hum Retroviruses. 1995; 11(1):1423±5. 29. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977; 33(1):159±74. PMID: 843571. 30. Bouckaert R, Heled J, Kuhnert D, Vaughan T, Wu CH, Xie D, et al. BEAST 2: a software platform for Bayesian evolutionary analysis. PLoS computational biology. 2014; 10(4):e1003537. https://doi.org/10. 1371/journal.pcbi.1003537 PMID: 24722319. 31. Suchard MA, Weiss RE, Sinsheimer JS. Bayesian selection of continuous time Markov chain evolutio- naty models. Molecular Biology Evolution. 2001; 18:1001±13. PMID: 11371589 32. Bello G, Aulicino PC, Ruchansky D, Guimaraes ML, Lopez-Galindez C, Casado C, et al. Phylodynamics of HIV-1 circulating recombinant forms 12_BF and 38_BF in Argentina and Uruguay. Retrovirology. 2010; 7:22. https://doi.org/10.1186/1742-4690-7-22 PMID: 20307282. 33. Faria NR, Rambaut A, Suchard MA, Baele G, Bedford T, Ward MJ, et al. HIV epidemiology. The early spread and epidemic ignition of HIV-1 in human populations. Science (New York, NY). 2014; 346 (6205):56±61. PMID: 25278604. 34. Lantz CA, Nebenzahl E. Behavior and interpretation of the kappa statistic: resolution of the two para- doxes. Journal of clinical epidemiology. 1996; 49(4):431±4. PMID: 8621993. 35. Ristic N, Zukurov J, Alkmim W, Diaz RS, Janini LM, Chin MP. Analysis of the origin and evolutionary his- tory of HIV-1 CRF28_BF and CRF29_BF reveals a decreasing prevalence in the AIDS epidemic of Bra- zil. PloS one. 2011; 6(3):e17485. https://doi.org/10.1371/journal.pone.0017485 PMID: 21390250. 36. Hue S, Pillay D, Clewley JP, Pybus OG. Genetic analysis reveals the complex structure of HIV-1 trans- mission within defined risk groups. Proceedings of the National Academy of Sciences of the United States of America. 2005; 102(12):4425±9. https://doi.org/10.1073/pnas.0407534102 PMID: 15767575. 37. Sierra M, Thomson MM, Rios M, Casado G, Castro RO, Delgado E, et al. The analysis of near full- length genome sequences of human immunodeficiency virus type 1 BF intersubtype recombinant viruses from Chile, Venezuela and Spain reveals their relationship to diverse lineages of recombinant viruses related to CRF12_BF. Infect Genet Evol. 2005; 5(3):209±17. https://doi.org/10.1016/j.meegid. 2004.07.010 PMID: 15737911. 38. Dilernia DA, Jones LR, Pando MA, Rabinovich RD, Damilano GD, Turk G, et al. Analysis of HIV type 1 BF recombinant sequences from South America dates the origin of CRF12_BF to a recombination event in the 1970s. AIDS Res Hum Retroviruses. 2011; 27(5):569±78. https://doi.org/10.1089/aid.2010. 0118 PMID: 20919926. PLOS ONE | https://doi.org/10.1371/journal.pone.0189705 December 15, 2017 12 / 12

Journal

PLoS ONEPubmed Central

Published: Dec 15, 2017

There are no references for this article.