Moyano et al. Retrovirology (2022) 19:6 https://doi.org/10.1186/s12977-022-00591-7 Retrovirology RESEARCH Open Access Immunoescape of HIV‑1 in Env‑EL9 CD8 + T cell response restricted by HLA‑B*14:02 in a Non progressor who lost twenty‑seven years of HIV‑1 control Ana Moyano1,2, Oscar Blanch‑Lombarte3,4†, Laura Tarancon‑Diez5,6†, Nuria Pedreño‑Lopez1,3, Miguel Arenas7,8,9, Tamara Alvaro1, Concepción Casado1, Isabel Olivares1, Mar Vera10, Carmen Rodriguez10, Jorge del Romero10, Cecilio López‑Galíndez1, Ezequiel Ruiz‑Mateos5, Julia G. Prado3,11*† and María Pernas1*† Abstract Background: Long-Term Non-Progressors (LTNPs) are untreated Human Immunodeficiency virus type 1 (HIV-1) infected individuals able to control disease progression for prolonged periods. However, the LTNPs status is tempo‑ rary, as viral load increases followed by decreases in CD4 + T-cell counts. Control of HIV-1 infection in LTNPs viremic controllers, have been associated with effective immunodominant HIV-1 Gag-CD8 + T-cell responses restricted by protective HLA-B alleles. Individuals carrying HLA-B*14:02 control HIV-1 infection is related to an immunodominant Env-CD8 + T-cell response. Limited data are available on the contribution of HLA-B*14:02 CD8 + T -cells in LTNPs. Results: In this study, we performed a virological and immunological detailed analysis of an HLA-B*14:02 LNTP individual that lost viral control (LVC) 27 years after HIV-1 diagnosis. We analysed viral evolution and immune escape in HLA-B*14:02 restricted CD8 + T -cell epitopes and identified viral evolution at the Env-EL9 epitope selecting the L592R mutation. By IFN-γ ELISpot and immune phenotype, we characterized HLA- B*14:02 HIV-1 CD8 + T cell responses targeting, Gag-DA9 and Env-EL9 epitopes before and after LVC. We observed an immunodominant response against the Env-EL9 epitope and a decreased of the CD8 T + cell response over time with LVC. Loss of Env-EL9 responses was concomitant with selecting K588R + L592R mutations at Env-EL9. Finally, we evaluated the impact of Env-EL9 escape mutations on HIV-1 infectivity and Env protein structure. The K588R + L592R escape variant was directly related to HIV-1 increase replicative capacity and stability of Env at the LVC. Conclusions: These findings support the contribution of immunodominant Env-EL9 CD8 + T-cell responses and the imposition of immune escape variants with higher replicative capacity associated with LVC in this LNTP. These data *Correspondence: jgarciaprado@irsicaixa.es; mpernas@isciii.es † Oscar Blanch-Lombarte, Laura Tarancon-Diez, Julia G. Prado and María Pernas contributed equally to this work 1 Virología Molecular, Laboratorio de Referencia e Investigación en Retrovirus, Centro Nacional de Microbiología, Instituto de Salud Carlos III, Carretera de Pozuelo a Majadahonda Km 2, 28220 Madrid, Spain 3 IrsiCaixa AIDS Research Institute, Crta Canyet SN, Badalona, 08916 Barcelona, Spain Full list of author information is available at the end of the article © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. The Creative Commons Public Domain Dedication waiver (http://​creat​iveco​ mmons.​org/​publi​cdoma​in/​zero/1.​0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Moyano et al. Retrovirology (2022) 19:6 Page 2 of 14 highlight the importance of Env-EL9 specific-CD8 + T-cell responses restricted by the HLA-B*14:02 and brings new insights into understanding long-term HIV-1 control mediated by Env mediated CD8 + T-cell responses. Keywords: Long-term non-progressor (LTNP), Loss of viral control (LVC), Env-EL9 escape HLA-B*14:02, CD8 + T-cells Background HLA-B*14:02 restricted CD8 + T-cell responses target Long-term non-progressors (LTNPs) constitute a small an epitope located in Gp41, the Env-EL9 [16–18]. Due to proportion (5–10%) of untreated HIV-1 infected indi- the high plasticity of the Env protein, it has been hypoth- viduals that remain with CD4 + T-cell counts higher esized that Env-EL9 escape mutations may not confer a than 500 cells/mm3. A subset of LTNPs termed viraemic significant impact on viral fitness, although some previ- controllers, maintain viral loads between 50 and 2000 ous studies describe mutations in this epitope related to HIV-RNA copies/ml [1, 2]. However, a third of these loss of CD8 + T-cell recognition and lower viral fitness individuals eventually experience an immunological and [17, 19]. clinical decline over time [3] and lose their status [4]. Here, we identified an HLA-B*14:02 viraemic LTNP Only 1–4% of HIV-infected LNTPs are progression-free that lost spontaneous HIV-1 control after more than survivors after 15 years of follow-up [5]. 27 years from diagnosis. To study what are the key mech- The immunological and viral factors involved in LTNP anisms to maintain the non progressor status and what status, and how they may contribute to the understand- leads to LVC, we performed an in-depth characterization ing of spontaneous HIV-1 control, constitute a topic of of clinical, virological and immunological factors using major interest in the HIV field [2]. While some studies biological samples collected over a period of 13 years. associate long-term control with attenuated viruses [6, 7], The analysis of the viral quasispecies identified the pre- others identify non progressor individuals infected with dominance of viral escape variants that target Env-EL9. replication-competent viruses [8]. These findings support Additionally, we monitored the magnitude, breadth the idea that immunological factors are also responsible and function of the Env-EL9 and Gag-DA9 specific- for the LTNPs status. The enrichment of certain HLA CD8 + T-cell responses in this individual, and deter- class I molecules, including HLA-B*57, -B*27, -B*13, mined the impact of the mutational escape in Env-EL9 -B*58 or -Cw*3 [9, 10], as well as highly functional HIV- on the virus infectivity and protein stability. 1-specific-CD8 + T-cell responses have been strongly associated with LTNPs. In fact, immunodominant HLA- Results B*27 and HLA-B*57-restricted CD8 + T-cell responses Phylogenetic analysis of HIV‑1 env in a viraemic LTNP mediate their effect by targeting highly conserved regions over 13 years. of HIV-1 Gag, which result in a high mutational viral We analysed HIV-1 evolution in a LTNP individual over fitness cost [9, 11–13]. On the contrary, in controllers a period of 13 years (from May 2003 to December 2015). without protective HLA class I alleles, Gag CD8 + T-cell As shown in Fig. 1a, we observed two different phases responses do not appear to play a significant role in viral in the clinical follow-up. From May 2003 to October control [13, 14]. 2012 (17 to 27 years after HIV-1 diagnosis), the subject Although the presence of the HLA-B*14 allele is remained asymptomatic in the absence of Antiretroviral common in viraemic individuals [10], the occurrence Treatment (ART) with a mean viral load of 2,821 HIV- of HLA-B*14:02 is more frequently associated with RNA copies/ml in plasma, slow decrease on the percent- viraemic control than HLA-B*14:01 [15]. Particularly, age of CD4 + T-cell counts (slope -0.78% ± 0.14) and an HLA-B*14:02-restricted CD8 + T-cell responses dem- increase in the percentage of CD8 + T-cells (1.14% ± 0.23) onstrate immunodominance towards HIV-1 Env [15]. per year. We collected 12 biological samples during the (See figure on next page.) Fig. 1 Phylogenetic analysis of HIV-1 env evolution during clinical follow-up. a Clinical follow-up. Viral load (circles), percentage of CD4 + (squares) and CD8 + T cells (triangles) during the patient follow-up are represented. The x axis refers to years after HIV-1 diagnosis. Colored circles indicate samples taken during the patient follow-up and that are included in the phylogenetic analysis. The dotted grey line separates samples before and after LVC. b Phylogenetic analysis revealed the presence of three HIV-1 variants. Two different viruses (A and B) and two different subpopulations (B1 and B2) are represented. A Maximum likelihood (ML) phylogenetic tree based on the C2-V5 env region was constructed with 201 unique sequences obtained from proviral DNA (circles), plasma viral RNA (triangles) and virions co-cultured RNA (squares). The sequences obtained at different times points during the patient follow-up are color coded. The identical sequences are included in boxes. Branch lengths are presented as the number of substitutions per site. Bootstrap values above 85% are shown with asterisks. ES61 and HXB2 sequences were included as outgroup viruses to root the phylogenetic tree. c. Replacement of viral populations in proviral sequences during follow-up. A, B1 and B frequency (%) obtained at different times points are represented color coded. The grey line indicates samples collected after LVC Moyano et al. Retrovirology (2022) 19:6 Page 3 of 14 Fig. 1 (See legend on previous page.) Moyano et al. Retrovirology (2022) 19:6 Page 4 of 14 2003–2012 period. In April 2013, the subject lost viral the HLA-B*14:02 restricted epitope Env-EL9, 584ERYL- control after 27 years of HIV-1 diagnosis. RDQQR592. We identified two mutations in Env-EL9 LVC was defined by two consecutive viral loads > 10,000 from Lys to Arg at position five K588R and Leu to Arg HIV-1 RNA copies/ml and was identified in four sam- at position nine (L592R). Furthermore, we examined the ples during the 2013–2015 period. After April 2013, we changes over time, observing the replacement of the Env- observed a 2.48% decrease of CD4 + T-cells and a 2.82% EL9 mutant L592R by the double mutant K588R + L592R. increase of CD8 + T-cells per year. In March 2016, the In addition, we identified changes associated with puta- individual started ART (FTC + RPV + TDF). This LNTP tive escape mutations in Gag GY9 and Pol QI9 in the individual was infected by two HIV-1 variants (here- HIV-1 proteome of the B2 variant (Table 1). We detected after referred to as A and B) as previously described by the Tyr (Y) to Phe (F) mutation at position 9 in GY9 our group [20]. Variant B was segregated into two sub- epitope before and after LVC and a secondary mutation groups, named B1 and B2 (supported by bootstrap val- at position 5, which never became dominant in the popu- ues of 97 and 95%, respectively) (Fig. 1b). We extended lation. Mutation at position 5 from Val (V) to Asp (D) in our previous findings [20] by including new sequences the QI9 epitope was observed after LVC. Previous work from peripheral blood mononuclear cells (PBMCs), correlated entropy with time to escape [21, 22]. We next plasma RNA and viral isolates obtained from PBMCs calculated Shannon entropy for the different epitopes co-cultures during a longer follow-up. The phylogenetic restricted by the patient HLAs. Of note, the Env-EL9 analysis revealed that A and B1 proviral variants were epitope had a medium entropy value, 0.2093 (medium homogenous, with high number of clonal sequences, and tercile in our data set) (Table 1) according with previous were absent from the circulating virus obtained from reports [21]. plasma samples over time (Fig. 1b). By contrast, B2 vari- Next, we extended the study to proviral DNA and ant obtained from plasma and viral isolates showed a plasma viral RNA sequences before and after LVC for the differential clustering before and after LVC. B2 plasma mapping of changes at Gag-DA9 and Env-EL9. In provi- sequences preceding LVC clustered mostly together, ral DNA, the WT Env-EL9 sequence was detected in A separated from viral isolates and proviral sequences variant while the L592R mutant appeared in 1 out the 2 obtained at the same time points. Inversely, B2 sequences sequences obtained for the B1 variant (March 04) and in after LVC (October 2014, and April, October and all sequences for the B2 from March 04 to December 15. December 2015) clustered together independently from The second mutation in Env-EL9 (K588R) was detected their origin (PBMCs, plasma, or isolates). Interestingly, in earlier in plasma samples (May 12) than proviral DNA proviral sequences, while the A variant frequency barely (October 14). In plasma viral RNA sequences, K558R was changed over time (mean 12 ± 1.5%), the percentage of a minority (2 out 5 sequences) in May 2012, preceding B1 changed from 75% in May 2003 to 0% in May 2012 LVC (Fig. 3a), but became dominant by October 2014 (in (Fig. 1c). Concomitantly, the B2 variant outgrew B1 by 4 out of the 5 sequences). This shift in plasma sequences increasing its frequency from 8% (May 2003) to 100% in was also reflected in proviral DNA sequences at the same 2015. We examined the C2-V5 env HIV-1 evolution of time. After October 2014, all sequences in B2 displayed these three variants. The overall env diversity of the B2 the combination of K588R + L592R mutations (Fig. 3a). sequences was significantly higher than the diversity of By contrast, we did not identify amino acid changes in A or B1 sequences (Fig. 2a, b). In the C2-V5 env region, the Gag-DA9 epitope over time (Fig. 3a). we found a linear correlation between divergence and Next, to further investigate the impact of Env-EL9 sampling date in B2 but not in A or B1 variants (Fig. 2b), mutations in CD8 + T-cell immune escape, we synthe- indicating a different evolution for each variant. Further- tized Gag-DA9 and Env-EL9 wild-type (WT) peptides more, C2-V5 env data of B2 showed the number the non- and Env-EL9 peptides containing K558R and L592R synonymous (dN) / synonymous substitution (dS) rate variants alone or combined (Fig. 3b). We monitored Env- ratio > 1, which indicates positive (diversifying) selection EL9 restricted CD8 + T-cell responses by IFN-γ ELISpot in C2-V5 env (Fig. 2c–d). These findings demonstrated on PBMCs, and used Gag-DA9-specific CD8 + T-cell the B2 variant imposition with extensive evolution in the responses as a reference. We detected a reduction of Env- C2-V5 region. EL9 WT responses over time and a lack of responsiveness towards the L592R variant (Fig. 3c). Meanwhile, we did Immune characterization of Env‑EL9 L592R variants not found any changes and we detected a low frequency To investigate the relationship between the acquisition of of Gag-DA9 responses at all time points. In addition, to escape mutations and LVC, we explored individual HLA- further characterize HLA-B*14:02-restricted HIV-1-spe- restricted epitopes in DNA samples from PBMCs before cific CD8 + T-cell responses, we performed a multi-para- and after LVC (Table 1). Viral variation mapped mainly in metric flow cytometry analysis based on lineage markers Moyano et al. Retrovirology (2022) 19:6 Page 5 of 14 a 15 b 0.15 <0.0001 Variants diversity Variants divergence <0.0001 10 0.10 A B1 5 0.05 B2 0 0.00 A B1 B2 20 24 28 p- 5 05 07 Ju -13 D r-14 ar 3 M -10 Ja -04 M -11 ec 5 2 5 Apn-1 M y-0 Sen-0 -1 -1 n- ay ay ay ct a Ju O M M Sampling date c 0.15 d 0.15 d dS d dS B2 variant divergence dN B2 variant diversity dN 0.10 0.10 0.05 0.05 0.00 0.00 20 24 28 20 24 28 Ju -05 p 5 Ju -06 ar 3 07 Ju 13 D r-14 M -10 M -11 2 5 Ja -04 ec 5 p 5 Ju -05 Ju -06 07 Ju 13 D r-14 ar 3 M -10 2 Ja -04 M -11 5 ec 5 Apn-1 M y-0 Sen-0 -1 -1 Apn-1 -1 M y-0 Sen-0 -1 n- - n- - n ay ay ay ct n ay ay ay ct a a O M M O M M Sampling date Sampling date Fig. 2 Estimation of viral evolution in the C2-V5 region during the patient follow-up. a The C2-V5 env diversity (y axis) was calculated as the number of base substitutions per site and averaging from all sequence pairs within each group and for each variant. For statistical evaluations, we used the Tukey’s multiple comparisons test considering significance at the 95% (p-values < 0.05 were assumed as statistically significant). b The env divergence (y axis) was estimated as the pairwise temporal distance from A, B1 and B2 time points (x axis) to the corresponding time to the MRCA. c and d panels show B2 C2-V5 genetic diversity c and genetic divergence d (y axis) over time (x axis). The plots include the global diversity (squares), number of substitutions per site d (circles), number of sites with synonymous substitution per synonymous site (dS) (triangles) and number of sites with nonsynonymous substitutions per nonsynonymous site (dN) (square) from averaging over all sequence pairs between groups Table 1 Proviral DNA sequences in patient HLA restricted epitopes from variant B2 a Predicted epitope representing the consensus sequence obtained from HIV-1 subtype B. ? denote mix of more than two aa in that position in the HIV data base (https://​www.​hiv.​lanl.​gov/​conte​nt/​seque​nce/​QUICK_​ALIGN​v2/​Quick​Align.​html) b Shannon entropy values at population level calculated as described in Methods c Date refers to the different time-points sequenced. Samples before and after LVC are separated by a black line Aa sequence conservation relative to the HXB2 subtype B consensus is indicated with dashes. Lower case letters indicate a mix of two aa, the consensus and the mutated one Moyano et al. Retrovirology (2022) 19:6 Page 6 of 14 Fig. 3 HIV-1 variants in Gag-DA9 and Env-EL9 HLA-B*14:02 restricted epitopes and HIV-specific CD8 + T cell responses. a Amino acid (aa) sequence variation in Gag-DA9 and Env-EL9 epitopes from proviral DNA and plasma RNA samples at different time points. Aa sequence conservation relative to the HXB2 sequence is indicated with dashes. The number of observed epitope variants vs the number of sequences analyzed at each time point is shown in parenthesis. A line separates samples collected before and after the LVC. b DA9 and Env-EL9 aa sequences representative of the different variants designed to perform the ELISpot and intracellular cytokine staining (ICS). c Quantification of HIV-1 specific-CD8 + T cell responses (y axis) against DA9 (blue), Env-EL9 WT (green) and Env-EL9 with the L592R mutation (red), determined by the ELISpot analysis, at different time points (x axis). d Gag-DA9, Env-EL9 WT, Env-EL9 with L592R and Env-EL9 592R + 588R specific-CD8 + T-cell responses before and after LVC as the proportion of peptide specific IFN-γ + production by CD8 + T-cells by multiparametric flow cytometry assay. e The pie charts show the poly-functionality of Env-EL9 (upper pies) and Env-EL9 with L592R + L588R mutations (bottom pies) specific-CD8 + T-cell responses before (Jun-06) and after (Oct-15) LVC. The pie charts are based on the proportions of cells producing combinations of IFN-γ, TNF-α, IL-2, CD107a and perforin. Single and double production of CD107a and perforin were excluded from the representations and cytokine production. The frequency of functional June and November 2006 (Fig. 3d). By contrast, minimal Gag-DA9 and Env-EL9 specific-CD8 + T-cell responses IFN-γ + Gag-DA9 specific CD8 + T-cell responses were was monitored by the percentage of IFN-γ + producing found in agreement with the ELISpot data. Additionally, cells. Our data showed the decline of IFN-γ + Env-EL9 we characterize functional changes of Env-EL9 HLA- specific-CD8 + T-cell responses against WT and viral B*14:02-restricted HIV-specific CD8 + T-cell response variants over time, presenting a notable decrease between before and after LVC (Oct-15) based on the single or Moyano et al. Retrovirology (2022) 19:6 Page 7 of 14 combinatorial expression of TNF-α, perforin, CD107a, obtained before LVC showed lower RC than those IFN-γ and IL-2. We identified cytokine-expressing cells obtained after LVC (Fig. 4a). Moreover, we next investi- before and after LVC, probably due to an increase of gated the infectivity of A, B1 and B2 pseudoviruses. We antigen in circulation and viral load upon LVC (Fig. 3e). found low infectivity of A and B pseudoviruses (Fig. 4b) These results support a stronger immune pressure of while B2 pseudoviruses showed similar levels of infectiv- HLA-B*14:02 restricted CD8 + T-cells towards the Env- ity as the pNL4-3 reference virus at the LVC time point EL9 epitope than to the Gag-DA9 epitope and the selec- (Fig. 4b). These results indicate a potential association tion of CD8 + T-cell HIV-1 escape variants at positions between increase of viral isolates and pseudoviruses K558R and L592R of the Env-EL9 epitope driven by infectivity and LVC. immune pressure. Interestingly, L592R variant showed lower infectiv- ity than WT (Fig. 4c). Moreover, the addition of K588R mutation to the L592R mutant resulted in a sixfold In vitro viral replication of Env‑EL9 variants (± 1.4) increase compared to the WT in infectivity assays To understand the influence of the HIV-1 replicative (Fig. 4c). Taken together, these results indicate the com- capacity (RC) of the different variants over time on the pensatory role in infectivity of the K588R mutation. LNTP phenotype, we compared the RC of co-cultured In order to identify structural changes associated with viral isolates obtained before and after LVC. Viruses the L592R and K588 + L592R mutations in Gp41, we Fig. 4 Changes in RC before and after LVC. Infectivity and protein structural stability of EL9 mutants. a RC estimated as RLU (relative units of luciferase) production (y axis) of the co-cultured viruses obtained before and after LVC (Oct-14 and Apr-15, blue bars) tested in TZM-bl cells. b Infectivity of the pseudoviruses generated with Gp160 Env expression plasmids derived from A (April-06), B1 (c4- May-12 and c7 March- 04) and B2 variants. Four B2 clones from samples previous to LVC (c4, c3, c6 and c8) and 6 B2 clones from Oct-14 (c16, c11, c15, c8, c20, c14) are included. Time points used for pseudoviruses generation was included in brackets. A pNL4-3 pseudovirus is also included as a reference. All these experiments were performed in duplicate c We compared infectivity expressed as RLU (relative units of luciferase) for WT, ESC (L592R mutation) and ESC + COM (L592R and K588R mutations) Env-EL9 sequences. Pseudoviruses were constructed from B2 c-11 with L592R mutation by site-directed mutagenesis. The infectivity values were normalized according to the RC of the pNL4-3 control virus (value = 1). Bars represent standard deviation from two experiments with three replicates per experiment d DOPE of the protein structure of the WT (black), WT with the escape mutation L592R (ESC, white) and WT with both escape and compensatory mutations (ESC + COM, dark gray). Note that in terms of protein stability, the order is the following: WT (more stable) > L592R with K588R > L592R (less stable). e Illustration of the best-fitting 3D structural models (with zoom in the mutation positions) for the WT (the orange and blue positions correspond to those that will suffer the escape and compensatory mutations, respectively), ESC (the red position corresponds to that with the escape mutation) and ESC + COM (the red and pink positions correspond to those with the escape mutation and compensatory mutations, respectively) states Moyano et al. Retrovirology (2022) 19:6 Page 8 of 14 evaluated Gp160 protein folding stability by discrete opti- inferences describe viral evolution [5] and the potential mized protein energy (DOPE) of the Env-EL9 WT and contribution of variant emergence and replacement to mutants. Compared to the WT virus, the L592R muta- LVC. We detected the presence of A and B1 variants in tion caused an unstable folding to the overall protein sta- proviral DNA 24 years after HIV-1 diagnosis. Conversely, bility. However, the incorporation of the K588R mutation these variants were absent from plasma and co-culture provided higher protein stability (Fig. 4d). A mapping sources in all time points analysed. According to previ- of potential Env-EL9 structural changes showed a close ous studies, PBMCs contain a pool of recent and archived proximity between L592R and K588R mutations, both variants throughout the infection [33], while plasma sam- located in the same α-helix (Fig. 4e), that may explain the ples harbour recently produced viral variants and the compensatory role of the K588R mutation. These results virions obtained by co-culture generally represent the suggest a relationship between the diminished infectivity best fitting variants over time [34]. Based on these previ- of the L592R mutant and sustained viral control and the ous findings, we postulate that A and B1 variants did not compensatory role of K588R in infectivity and protein actively replicate over time. This hypothesis is supported stability. by the lack of viral evolution, the absence in plasma or in co-culture samples and the low viral infectivity. It is pos- Discussion sible that these variants may be ancient archived variants Studies examining the loss of spontaneous control in maintained in proviral reservoirs by homeostatic prolif- HIV-1 infected individuals provide essential informa- eration events [35] or are defective, which are present in tion to understand the determinants of viral control and a high proportion in proviral compartments [36]. Conse- pathogenesis. However, longitudinal studies on the loss quently, these variants do not seem to have contributed of LTNP status are limited due to the lack of continuous to the LVC in our LTNP case. Therefore, we focused our clinical follow-up [19, 23, 24]. In this study, we overcome study on the B2 variant in which the evolution, measured some of these limitations by providing long term clini- by the genetic diversity and divergence, increased over cal follow-up, accompanied by the appropriate biologi- time. cal sampling, of an HIV-1-infected individual with LTNP In this study, we found a continuous shift in viral pop- status who lost viral control after 27 years from diagnosis. ulations linked to changes in CD8 + T cell responses, In the last years, the number of studies involved in as observed between June and November 2006. The the identification of predictors associated with the reduction in Env-EL9 CD8 + T cell responses paralleled loss of LTNP status has increased. Several cross-sec- changes in the proportion of A, B1 and B2 variants. The tional studies propose HIV-1 DNA load in PBMCs [25], “A” variant sharply decreased around that date (in Apr- CD4 + T-cell counts after 10 years from seroconversion 06) while escape variant became predominant after this [4], high viral loads and co-infection with hepatitis C [4], time point. The L592R mutant appeared in one out the as predictors of loss of LTNP status. Only a few longitudi- two sequences obtained for B1 variant (Mar-04) and in all nal studies have been able to directly address the LVC in sequences for the B2 from Mar-04 to Dec-15. Then, the LTNPs. In previous work, we found that the maintenance imposition of the L592R mutant in EL9 was concomitant of elite controller status was associated with the presence with the changes in CD8 + T-cell responses in 2006. of Gag-specific polyfunctional CD8 + T cell, lack of viral We also found that spite of the low proportion of evolution and low levels of pro-inflammatory cytokines WT Env-EL9 variants in proviral DNA sequences, [26]. Another study associated the LVC to a combination the CD8 + T-cell response targeting WT epitope was of viral factors, including X4 co-receptor switching, and detectable over time. The persistence of this response loss of antiviral CD8 + T-cell activity [27]. Recently, the could be explained by a continuous antigenic stimulus, impairment of HIV-1 proliferative and cytolytic T-cell as other proposed [37]. Frater et al. also suggested that effector functions has been observed prior LVC [28]. this response can maintain selective pressure for HIV-1 In this study we have analyzed a dually HIV-infected escape variants by TCR cross-reactivity [37]. Similarly, HLA-B*14:02 LTNP, who lost viral control after 27 years we observed the acquisition of the compensatory muta- of HIV-1 diagnosis, with the aim of understanding viral tion K588R concomitant with the reduction of WT and host factors involved in HIV-1 long term control. Our immune response in May 2011. Additionally, at this results demonstrate the association between the presence point, the CD8 + T-cell poly-functionality increases, of Env-EL9 immunodominant HLA-B*14:02-restricted probably due to the increase of viral replication and anti- CD8 + T-cell responses, viral evolution and immune gen circulation. escape with LVC. In contrast to other studies [29–31], HIV-1 evolves by genetic drift as well as by the selection the infection with two heterologous HIV-1 strains was pressure exerted by antibodies and antiviral CD8 + T-cell not associated with LVC [20, 24, 32]. Our phylogenetic responses [38, 39]. In LTNPs with protective HLAs class Moyano et al. Retrovirology (2022) 19:6 Page 9 of 14 I alleles, CD8 + T-cell responses mostly target highly Our results bring novel information, on the impact of conserved epitopes in Gag [40–43], which drive the Env-EL9 mutations in CD8 + T-cell recognition [15, 17]. emergence of escape mutants. In the case of the protec- In particular, our results on ELISpot and immunopheno- tive HLA-B*14 alleles, present in LTNP particularly for typing showed that a non-conservative change of L592R the HLA-B*14:02 allele, CD8 + T-cell responses operate in Env-EL9 conferred the loss of Env-EL9 HIV-1 specific against the Env-EL9 epitope [15]. These data support our CD8 + T-cell responses. In addition, L592R reduces virus findings concerning the immunodominance of Env-EL9 infectivity which it is further compensated by the acquisi- CD8 + T-cell responses over Gag-DA9. Furthermore, tion of the K588R mutation. Furthermore, the analysis of this immunodominance of Env responses is reinforced protein folding stability indicated loss of Env stability by by the identification of escape mutants at this epitope. L592R compensated by K588R and according to in vitro We focused our study in the immunodominant Env-EL9 observations of infectivity [47]. Troyer et al. found that epitope restricted by the protective allele HLA-B*14:02 amino acid changes in Gag usually involve a fitness cost and previously associated with HIV-1 control. In addi- [48]. As a result, those escape variants imposing fitness tion, CD8 + T-cell responses restricted by HLA-B are the constrains could contribute to a prolonged control of main driver of HIV-1 evolution and play a crucial role in HIV-1 replication [48]. Likewise, in this study, we con- viral control in contrast to HLA-A haplotypes [9]. Nev- clude that CD8 + T-cell escape mutations in Env-EL9 ertheless, the functional analysis of additional T cells affect viral infectivity. It is possible that its predominance responses would had given us a broader view of other T in the viral quasispecies is related to long-term viral epitopes involved in LVC. These studies were limited by control. sample availability. To overcome this limitation, we car- Here, we describe a new HIV-1 escape muta- ried out the analysis of escape mutations in the epitopes tion in Env-EL9 driven by HLA-B*14:02 restricted restricted by the individual’s HLA in the proviral DNA CD8 + T-cells. We hypothesize that the predominance sequences. This compartment reflects all replicating vari- of L592R immune escape with low infectivity allowed the ants throughout the infection [33], so if escape mutations balance between host immune response and viral replica- had occurred, such mutations should have accumulated tion, contributing to HIV-1 viraemic control for 27 years. and should have detected in the proviral sequences of When the immunodominant response directed by Env- PBMC. Beside Env-EL9 mutations, we detected addi- EL9 declined, the mutation K588R arose in association tional changes in Gag-GY9 epitope after LVC. Responses with a compensatory effect in viral infectivity leading to against this epitope restricted by HLA-A*01 has previ- disease progression. ously described as a subdominant CD8 + T cell responses during the acute phase of HIV-1 infection [44]. Although, we identified changes in the epitope over time, includ- Conclusions Our results reveal the role of Env-specific CD8 + T-cell ing selecting a double mutant variant before and after responses in HLA-B*14:02 LNTP. These results bring the LVC, the new mutant never became dominant. We new insights into the contribution of non-Gag CD8 + T acknowledge limitations of the analyses restricted to cell responses in HIV-1 control [15, 45] and support fur- epitopes matched to the patient HLAs and limited depth ther exploration of Env epitopes and its potential role in of sequencing at some time points that may underesti- HIV-1 spontaneous control. mate the presence of additional minority variants. Patterns of immunodominant T cell responses vary between patients and time. The detailed changes in Methods CD8 + T cell immunodominance in HIV-1 infection can Study participant only be monitored from acute infection and with long- The study participant was an intravenous drug user, term follow-up [21, 22]. During acute infection, immu- HIV-1 subtype B diagnosed in 1986 and without cART nodominant CD8 + T cells responses target high entropy treatment during the complete follow-up. The subject regions, often Nef and Env where the selection of HIV-1 carry HLA- A*01:01, A*29:02, and B*14:02, B*44:03 and immune escape is rapid comparing with more conserved was clinically followed between May-03 and Decem- epitopes [45]. Conversely, although Env-EL9 entropy is ber-15 (17.4–30 years after the first HIV-1 detection) in medium, escape mutations in this epitope appear very an outpatient Clinic Centre (Centro Sanitario Sandoval, early in the infection, at least in the context of HLA Hospital Clínico San Carlos, Madrid, Spain). The indi- B*14:02 patients [21]. Our data support the immuno- vidual provided a written informed consent for research dominance of the Env-EL9 CD8 + T cell responses and purposes of biological samples following the guidelines rapid escape in Env-EL9 by selection of the 592R muta- tion early in infection, as previously reported [21, 46]. Moyano et al. Retrovirology (2022) 19:6 Page 10 of 14 of the institutional ethical committees of the Institute of tree was inferred with the MEGA program (vX.10) Salud Carlos III (CEI PI 05-2010-v3). [50] under the General Time Reversible exchangeabil- ity matrix, and accounting for variation of the substitu- Proviral HIV‑1 DNA, plasma and competent‑virus RNA tion rate among sites and proportion of invariable sites, isolation (GTR + G + I) substitution model which is the substitu- We collected 19 blood samples during the patient’s fol- tion model of DNA evolution that best fitted the data. low-up and, isolated peripheral blood mononuclear The phylogenetic confidence was assessed by bootstrap cells (PBMCs) and plasma by the Ficoll-Hypaque gra- analysis using 1,000 replicates. Initial tree(s) for the heu- dient. HIV-1 DNA was extracted from PBMCs, using ristic search were obtained by applying the Neighbour- “Speedtools tissue DNA extraction kit” (Biotools B&M Joining (NJ) method to a matrix of pairwise distances Labs, Spain) following the manufacturer’s instructions. estimated with the Maximum Composite Likelihood Moreover, viral isolates were obtained from PBMCs by (MCL) approach. Viral populations were identified by co-culturing the patient’s PBMCs with PHA-stimulated bootstrap values above 85% in the phylogenetic tree and PMBCs from three healthy donors in a 1:1 ratio. Co-cul- three viral variants (A, B1 and B2) were defined. We cal- tures were maintained in RPMI 1640 (L-glutamine, gen- culated the relative percentage of A, B1 and B2 as the tamicin and SFB at 10%) supplemented with IL-2 (10 µg/ number of sequences of each variant at each time point ml). Every week, 500 µl of the culture supernatants were divided by the number of total sequences at that time discarded and fed with 5 × ­106 fresh PHA-stimulated point. donor’s PBMCs. Viruses were harvested when HIV-1 p24 The mean viral diversity (number of base substitutions antigen reached > 5000 arbitrary units (25 ng/ml) by elec- per site, by averaging from all sequence pairs within each trochemiluminescence immunoassay (Roche Diagnos- group) and overall divergence (group mean distance to its respective most recent common ancestor, MRCA) were viral isolates (n = 6) were extracted using “Nuclisens® tic, Spain). Viral RNA from plasma samples (n = 11) and calculated for the three variants in C2-V5 env using the aesyMAG ™” (bioMérieux, France) following the manu- MCL approach. We estimated the number of nonsyn- facturer’s instructions. onymous substitutions per nonsynonymous site (dN), the number of synonymous substitutions per synonymous Amplification of HIV‑1 env C2‑V5 region, sequencing site (dS) pairwise distances and dN/dS rate using longitu- and phylogenetic analysis dinal C2-V5 env sequences data of the B2 variant. Diver- Proviral DNA and viral RNA were amplified by nested sity and divergence estimation for each time point were PCR for the env gene (C2-V5) by single genome ampli- conducted using the MCL approach. These analyses were fication using 169–96 outer and 27–167 inner primers conducted using the Kimura 2-parameter (K2P) substitu- as previously described [26]. DNA was amplified using tion model of DNA evolution in MEGA vX.10. the Phusion High-Fidelity enzyme (Thermo Fisher, MA, USA). For viral RNA amplification, reverse-transcrip- tion-PCR was performed using the One-Step RT-PCR kit HIV‑1 amino acid variation in patient HLA (Qiagen, Germany). Next, a total of 1 µl of the first-PCR restricted‑CD8 + T cell epitopes product was reamplified using the Phusion High-Fidel- Twelve optimal epitopes restricted by patient HLA were ity enzyme (Thermo Fisher, MA, USA) with the inner identified from Los Alamos HIV Database https://​www.​ primers. hiv.​lanl.​gov/​conte​nt/​immun​ology/ (Table 1). Amino acid (aa) sequences for the optimal epitopes were obtained Dye™ Terminator Cycle Sequencing kit (Applied Biosys- Nucleotide sequences were determined by the Big from bulk proviral DNA by nested PCR using primers tems, Thermo Fisher Corporation, MA, USA) in an ABI described in Additional file 1 following the conditions 3730 sequencer (Applied Biosystems, Thermo Fisher, described in [51]. Aa changes were studied in longitu- MA, USA). We used SeqMan Pro v 12.3.1 (DNASTAR) dinal samples taken before (March-04, Sepember-05, and Bioedit Sequence Alignment Editor (v 7.0.5.3) for June-06, June-07, May-10, May-11) and after LVC (Octo- assembling and hand-editing [49]. Nucleotide sequences ber-13, October-14, April-15, December-15). Next, we were aligned using MAFFT method (https://​www.​hiv.​ studied viral quasispecies from proviral DNA and viral lanl.​gov/​conte​nt/​seque​nce/​V IRAL​IGN/​v iral​ign.​html). RNA samples with focus on the immunodominant HLA- Hypermutated sequences identified using Hipermut B*14:02 restricted epitopes: Gag-DA9 (DRFYKTLRA, (https:// ​ w ww. ​ h iv. ​ l anl. ​ g ov/ ​ conte ​ nt/ ​ s eque ​ n ce/​ H YPER​ aa 298–306) and Env- EL9 (ERYLKDQQL, aa 584–592). MUT/​hyper​mut.​html) were removed prior to phyloge- To perform this analysis, we carried out a single genome netic inference. Maximum likelihood (ML) phylogenetic amplification in gag and partial gp41 region by nested Moyano et al. Retrovirology (2022) 19:6 Page 11 of 14 PCR using primers described in Additional file 1 follow- using the cytofix/cytoperm kit (BD Biosciences) and ing the same PCR conditions described for C2-V5. stained intracellularly for 30 min with anti-CD3-APC- Shannon entropy of the 12 optimal epitopes restricted H7, anti-IFNγ-PE-Cy7, anti-TNF-A700, anti-IL-2-APC by patient HLA was calculated using Los Alamos data- and anti-perforin-PE (BD Biosciences) and then washed base https://​www.​hiv.​lanl.​gov/​conte​nt/​seque​nce/​ENTRO​ twice and fixed in PBS containing 4% paraformaldehyde. PY/​entro​py.​html. Subtype B sequence alignment was Unstimulated condition was included as negative control obtained from https://​www.​hiv.​lanl.​gov/​conte​nt/​seque​ in each experiment. PBMCs were acquired using a LSR nce/​QUICK_​ALIGN​v2/​Quick​Align.​html. Fortessa Cell Analyzer (BD Biosciences, Spain) and a minimum of 1.5 × ­106 total events were recorded for each Characterization of Gag‑DA9 and Env‑EL9 HIV‑1 condition. Lymphocytes were defined as low Forward/ specific‑CD8 + T cell responses Side scatter and expressed CD3 + , and/or no CD8 + , and Based on autologous viral sequences, we designed not CD19, CD14, and CD56. peptides for Gag-DA9 (298DRFYKTLRA306), Env- EL9 (584ERYLKDQQL592) and for the observed Env- Pseudoviruses generation and infectivity assays of Env‑EL9 EL9 variants at positions L592R (ERYLKDQQR) and variants K588R + L592R (ERYLRDQQR). Peptides were syn- We generated Env-expression plasmid for the different thetized by EZBiolab (USA) and resuspended follow- variants. To do this, gp160 region (HXB2 position aa ing the manufacturer’s instructions. In brief, ELISpot 2075–2931) was amplified with the primers described plates (Merck Millipore, MA, USA) were coated with in Additional file 1 using Phusion High-Fidelity enzyme monoclonal antibodies to anti-human IFN-γ (1-D1K, following amplification conditions as described [51]. Mabtech, Sweden) overnight at 4 °C. Plates were washed To ensure the presence of variants A, and B1 and B2 six times with PBS containing 1% FCS. Cryopreserved throughout the infection, we used several time points PBMCs were thawed and 1 × ­105 cells/well were cul- for cloning including March 04, April 06, May 11, May tured in duplicate overnight at 37 °C with 5% of ­CO2 in 12 and October 14. For variant selection, C2-V5 region the presence of Gag-DA9, Env-EL9 or mutated Env-EL9 from the gp160 env sequences were analyzed with a ML peptides at 0.01 μg/ml, 0.1 μg/ml, 1 μg/ml and 10 μg/ml. phylogenetic tree including representative sequences In addition, PHA (250 μg/ml, Sigma-Aldrich, USA) was from the different variants. Amplicons corresponding to (Invitrogen, MA, USA) and cloned into the pcDNA ™ 3.1 included as positive control. Finally, plates were revealed A, B1 and B2 variants were purified using the SNAP kit with biotinylated anti-human IFN-γ, streptavidin–alka- line phosphatase and its coloured substrate (Mabtech, according to the manufacturer’s instructions (Invitrogen, Sweden). IFN-γ secreting cells were quantified under MA, USA). To study the role of the mutations in Env- Immuno Capture and Immuno Spot software to calcu- EL9 epitope, R592L and K588R were introduced into an late the number of Spot Forming Cells (SFC) as previ- Env-expression plasmid containing the L592R mutation ously described [52, 53]. Wells were considered positive to generate WT pseudovirus (584ERYLKDQQL592), and if they contained at least 50 SFCs per ­106 PBMCs above escape + compensatory (ESC + COM) (K588R + L592R). the background level (2X mean + 3X standard deviation). Mutations were introduced by site directed mutagenesis For immunophenotyping, cryopreserved PBMCs using Quick Change Lightening Site-Directed Mutagene- were thawed, washed and incubated with 10U of DNa- sis Kit (Agilent Technologies, CA, USA) with the primers seI (Roche Diagnostics, Spain) containing RPMI-1640 described in Additional file 1. All mutations were verified (L-glutamine, ampicillin and 10% FCS) medium at 37 °C by plasmid sequencing. Pseudoviruses were obtained by during 2 h. PBMCs were stimulated at 37 °C with 2 μg/ml 293T co-transfection of the Env-expression plasmid in of peptides according to sample availability and quasispe- combination with an Env defective HIV-1 backbone by cies evolution in the presence of 1 μg/ml of CD28/49, calcium chloride protocol [54]. After 72 h, the pseudovi- 0.7 μg/ml of monensin (BD Biosciences, Spain), 10 μg/ ruses generation was quantified by measuring HIV-1 p24 ml of brefeldin A (Sigma Chemical Co, USA), and anti- antigen production in the supernatants by electrochemi- CD107a-BV786 (BD Biosciences, Spain) at the beginning luminiscence Immunoassay (Roche Diagnostic). of the incubation. After 6 h of stimulation, cells were sur- In addition, we studied viruses RC derived from the face stained for 30 min with LIVE/DEAD fixable Violet co-culture along the following time points: May-04, Dead Cell Stain (Life Technologies, CA, USA) for viabil- April-06, November-06, May-10, May-11, October-14 ity, anti-CD14-PB, anti-CD19-PB (Life Technologies), and April-15. To this end, we evaluated RC in ­105 TZM anti-CD56-PB (Biolegend), anti-CD8-PerCP-Cy5.5, anti- cells with 30 units of p24 antigen. TZM-bl cells (NIH- CD45RA-FITC and anti-CD27-BV605 (BD Biosciences). ARP Cat# 8129-442, RRID:CVCL_B478) were cultured Then, cells were washed and permeabilized for 30 min in DMEM supplemented with 1% fetal bovine serum Moyano et al. Retrovirology (2022) 19:6 Page 12 of 14 cultured and 20 μg/ml DEAE-dextran was added to each Authors’ contributions CC, IO, ERM, CLG, JGP and MP contributed to the study design, data interpreta‑ well. Three replicates of each virus were included in each tion and performed data analyses. AM, OB-L, LT, NP, TA, MA and MP designed assay. After 48 h, we lysed the cells and the production and performed the experiments. MV, CR and JDR carried out the, clinical of luciferase was measured “Luciferasa assay system” follow-up of the patients and the analysis of Clinical Data. JGP and MP per‑ formed the manuscript drafting and discussion. All authors read and approved (Promega). All the assays were performed in duplicated. the final manuscript. Pseudoviruses and Env-EL9 mutant’s infectivity was cal- culated following the same procedure. Funding Molecular Virology Laboratory was supported by grants SAF (2016-77894- R) from Ministerio de Economía y Competitividad (MINECO), ISCIII through Evaluation of the stability of Gp160 variants the projects PI 13/02269, PI17/00164, PI16/0684, PI19/01127 (Co-funded by European Regional Development Fund/European Social Fund "Investing in We evaluated the Gp160 stability for the WT amino your future"). The RIS-RETIC grants RD12/0017/0028, RD16/0025/0020 and acidic sequence, the sequence with the escape muta- RD16CIII/0002/0005. LTD was supported by the Instituto de Salud Carlos III tion L592R (ESC variant) and, the sequence with both (ISCIII) under grant agreement “CD20/00025” through the Sara Borrell Program. O.B.L was funded by an AGAUR-FI_B 00582 Ph.D. fellowship from the Catalan the escape L592R and compensatory K588R mutations Government and the European Social Fund. M.A. was funded by grants (ESC + COM variant). To do so, first we performed RYC-2015-18241 and PID2019-107931GA-I00 from the Spanish Government homology modelling with Swiss-Model [55] to obtain and, ED431F 2018/08 from the “Xunta de Galicia”. ERM was supported by the the best structural representation (structural model) for Spanish National Research Council (CSIC). JGP laboratory was supported by National Health Institute Carlos III grant PI17/00164 and Redes Temáticas de the WT, ESC and ESC + COM protein sequences. Next, Investigación en SIDA (ISCIII RETIC RD16/0025/0041). The funders had no role we computed the protein stability for each previously in study design, data collection and analysis, the decision to publish or draft‑ selected structural model with the discrete optimized ing of the manuscript. protein energy (DOPE); a traditional metric of protein Availability of data and materials folding stability [56] that is implemented in the well- Nucleotide sequences for C2-V5 env, gp160 and gag sequences were depos‑ established framework Modeller [57] The DOPE esti- ited in GenBank with the following accession numbers: MT 530457-530692, MT 530693–530743, MT 530744–530774. mates from every variant (WT, ESC and ESC + COM) can be compared (i.e., [58]) to evaluate the influence of Declarations every mutation on the Gp160 stability. Ethics approval and consent to participate The individual provided a written informed consent for research purposes of Statistical analysis biological samples following the guidelines of the institutional ethical com‑ A linear regression analysis for clinical and quasispecies mittees of the Institute of Salud Carlos III (CEI PI 05-2010-v3). evolution over time was performed with GraphPad Prism Consent for publication (v8.01). For env nucleotide divergence and diversity anal- All the authors revised and approved the final version of the manuscript. yses, we applied the Tukey’s multiple comparisons test. Differences in viral RC were evaluated with the unpaired Competing interests The authors declare that there is no conflict of interest. t-test. For all the analyses we considered significance at the 95% (p-values < 0.05). We constructed poly-function- Author details 1 ality pie charts using Pestle (v1.6.2) and SPICE (v5.2). Virología Molecular, Laboratorio de Referencia e Investigación en Retrovirus, Centro Nacional de Microbiología, Instituto de Salud Carlos III, Carretera de Pozuelo a Majadahonda Km 2, 28220 Madrid, Spain. 2 Max Von Pettenkofer Institute and Gene Center, Virology, National Reference Center for Retro‑ Abbreviations viruses, Faculty of Medicine, LMU München, Munich, Germany. 3 IrsiCaixa LTNPs: Long-Term Non-Progressors; HIV-1: Human immunodeficiency virus AIDS Research Institute, Crta Canyet SN, Badalona, 08916 Barcelona, Spain. type 1; LVC: Loss viral control; ART​: Antiretroviral Treatment; PBMCs: Periph‑ 4 Autonomous University of Barcelona, Cerdanyola del Vallès, Barcelona, eral blood mononuclear cells; MCL: Maximum Composite Likelihood; dN: Spain. 5 Institute of Biomedicine of Seville (IBiS)/Virgen del Rocío University Nonsynonymous; dS: Synonymous substitution; WT: Wild type; ESC: Escape; Hospital, CSIC, University of Seville, Seville, Spain. 6 Molecular Immunobiology ESC+COM: Escape+ compensatory; DOPE: Discrete optimized protein energy. Laboratory, Immunology Section, Hospital Gregorio Marañón, Madrid, Spain. 7 Department of Biochemistry, Genetics and Immunology, University of Vigo, Supplementary Information 36310 Vigo, Spain. 8 CINBIO, University of Vigo, 36310 Vigo, Spain. 9 Galicia Sur Health Research Institute (IIS Galicia Sur), 36310 Vigo, Spain. 10 Centro Sanitario The online version contains supplementary material available at https://​doi.​ Sandoval. Hospital Clínico San Carlos, IdISSC, Madrid, Spain. 11 Germans Trias I org/​10.​1186/​s12977-​022-​00591-7. Pujol Research Institute (IGTP), Badalona, Spain. Received: 14 December 2021 Accepted: 1 March 2022 Additional file 1. Oligonucleotides used for amplification. Table contain‑ ing all the oligonucleotides used for the different PCR assays described in the Methods. References Acknowledgements 1. Casado C, Colombo S, Rauch A, Martinez R, Gunthard HF, Garcia S, et al. We greatly appreciate the technical work made by Rosa Fuentes. Host and viral genetic correlates of clinical definitions of HIV-1 disease progression. PLoS ONE. 2010;5(6):e11079. Moyano et al. Retrovirology (2022) 19:6 Page 13 of 14 2. Lopez-Galindez C, Pernas M, Casado C, Olivares I, Lorenzo-Redondo 22. Liu MK, Hawkins N, Ritchie AJ, Ganusov VV, Whale V, Brackenridge S, et al. R. Elite controllers and lessons learned for HIV-1 cure. Curr Opin Virol. Vertical T cell immunodominance and epitope entropy determine HIV-1 2019;38:31–6. escape. J Clin Invest. 2013;123(1):380–93. 3. Leon A, Perez I, Ruiz-Mateos E, Benito JM, Leal M, Lopez-Galindez C, et al. 23. Fang G, Burger H, Chappey C, Rowland-Jones S, Visosky A, Chen CH, et al. Rate and predictors of progression in elite and viremic HIV-1 controllers. Analysis of transition from long-term nonprogressive to progressive AIDS. 2016;30(8):1209–20. infection identifies sequences that may attenuate HIV type 1. AIDS Res 4. van der Helm JJ, Geskus R, Lodi S, Meyer L, Schuitemaker H, Gunsenhe‑ Hum Retroviruses. 2001;17(15):1395–404. imer-Bartmeyer B, et al. Characterisation of long-term non-progression 24. Pernas M, Casado C, Arcones C, Llano A, Sanchez-Merino V, Mothe B, et al. of HIV-1 infection after seroconversion: a cohort study. Lancet HIV. Low-replicating viruses and strong anti-viral immune response associ‑ 2014;1(1):e41–8. ated with prolonged disease control in a superinfected HIV-1 LTNP elite 5. Chereau F, Madec Y, Sabin C, Obel N, Ruiz-Mateos E, Chrysos G, et al. controller. PLoS ONE. 2012;7(2):e31928. Impact of CD4 and CD8 dynamics and viral rebounds on loss of virologi‑ 25. Madec Y, Boufassa F, Porter K, Prins M, Sabin C, D’Arminio MA, et al. Natural cal control in HIV controllers. PLoS ONE. 2017;12(4):e0173893. history of HIV-control since seroconversion. AIDS. 2013;27(15):2451–60. 6. Casado C, Pernas M, Sandonis V, Alvaro-Cifuentes T, Olivares I, Fuentes 26. Pernas M, Tarancon-Diez L, Rodriguez-Gallego E, Gomez J, Prado JG, R, et al. Identification of a cluster of HIV-1 controllers infected with low Casado C, et al. Factors Leading to the Loss of Natural Elite Control of replicating viruses. PLoS ONE. 2013;8(10):e77663. HIV-1 Infection. J Virol. 2018. https://​doi.​org/​10.​1128/​JVI.​01805-​17. 7. Miura T, Brumme ZL, Brockman MA, Rosato P, Sela J, Brumme CJ, et al. 27. Rosas-Umbert M, Mothe B, Noguera-Julian M, Bellido R, Puertas MC, Impaired replication capacity of acute/early viruses in persons who Carrillo J, et al. Virological and immunological outcome of treatment become HIV controllers. J Virol. 2010;84(15):7581–91. interruption in HIV-1-infected subjects vaccinated with MVA-B. PLoS ONE. 8. Blankson JN, Bailey JR, Thayil S, Yang HC, Lassen K, Lai J, et al. Isolation and 2017;12(9):e0184929. characterization of replication-competent human immunodeficiency 28. Collins DR, Urbach JM, Racenet ZJ, Arshad U, Power KA, Newman RM, virus type 1 from a subset of elite suppressors. J Virol. 2007;81(5):2508–18. et al. Functional impairment of HIV-specific CD8(+) T cells precedes 9. Kiepiela P, Leslie AJ, Honeyborne I, Ramduth D, Thobakgale C, Chetty aborted spontaneous control of viremia. Immunity. 2021;54(10):2372–84. S, et al. Dominant influence of HLA-B in mediating the potential co- 29. Caetano DG, Cortes FH, Bello G, de Azevedo SSD, Hoagland B, Villela LM, evolution of HIV and HLA. Nature. 2004;432(7018):769–75. et al. A case report of HIV-1 superinfection in an HIV controller leading 10. Pereyra F, Jia X, McLaren PJ, Telenti A, de Bakker PI, Walker BD, et al. The to loss of viremia control: a retrospective of 10 years of follow-up. BMC major genetic determinants of HIV-1 control affect HLA class I peptide Infect Dis. 2019;19(1):588. presentation. Science. 2010;330(6010):1551–7. 30. Clerc O, Colombo S, Yerly S, Telenti A, Cavassini M. HIV-1 elite controllers: 11. Brockman MA, Brumme ZL, Brumme CJ, Miura T, Sela J, Rosato PC, et al. beware of super-infections. J Clin Virol. 2010;47(4):376–8. Early selection in Gag by protective HLA alleles contributes to reduced 31. Rachinger A, Navis M, van Assen S, Groeneveld PH, Schuitemaker H. HIV-1 replication capacity that may be largely compensated for in chronic Recovery of viremic control after superinfection with pathogenic HIV infection. J Virol. 2010;84(22):11937–49. type 1 in a long-term elite controller of HIV type 1 infection. Clin Infect 12. Martinez-Picado J, Prado JG, Fry EE, Pfafferott K, Leslie A, Chetty S, et al. Dis. 2008;47(11):e86–9. Fitness cost of escape mutations in p24 Gag in association with control of 32. Pernas M, Casado C, Sandonis V, Arcones C, Rodriguez C, Ruiz-Mateos E, human immunodeficiency virus type 1. J Virol. 2006;80(7):3617–23. et al. Prevalence of HIV-1 dual infection in long-term nonprogressor-elite 13. Prado JG, Honeyborne I, Brierley I, Puertas MC, Martinez-Picado J, Goulder controllers. J Acquir Immune Defic Syndr. 2013;64(3):225–31. PJ. Functional consequences of human immunodeficiency virus escape 33. Chun TW, Finzi D, Margolick J, Chadwick K, Schwartz D, Siliciano RF. In vivo from an HLA-B*13-restricted CD8+ T-cell epitope in p1 Gag protein. J fate of HIV-1-infected T cells: quantitative analysis of the transition to Virol. 2009;83(2):1018–25. stable latency. Nat Med. 1995;1(12):1284–90. 14. Koofhethile CK, Ndhlovu ZM, Thobakgale-Tshabalala C, Prado JG, Ismail N, 34. Mittler JE, Markowitz M, Ho DD, Perelson AS. Improved estimates for Mncube Z, et al. CD8+ T cell breadth and ex vivo virus inhibition capacity HIV-1 clearance rate and intracellular delay. AIDS. 1999;13(11):1415–7. distinguish between viremic controllers with and without protective HLA 35. Bui JK, Sobolewski MD, Keele BF, Spindler J, Musick A, Wiegand A, et al. class I alleles. J Virol. 2016;90(15):6818–31. Proviruses with identical sequences comprise a large fraction of the 15. Leitman EM, Willberg CB, Tsai MH, Chen H, Buus S, Chen F, et al. HLA- replication-competent HIV reservoir. PLoS Pathog. 2017;13(3):e1006283. B*14:02-restricted env-specific CD8(+) T-cell activity has highly potent 36. Bruner KM, Murray AJ, Pollack RA, Soliman MG, Laskey SB, Capoferri AA, antiviral efficacy associated with immune control of HIV infection. J Virol. et al. Defective proviruses rapidly accumulate during acute HIV-1 infec‑ 2017. https://​doi.​org/​10.​1128/​JVI.​00544-​17. tion. Nat Med. 2016;22(9):1043–9. 16. Goulder PJ, Altfeld MA, Rosenberg ES, Nguyen T, Tang Y, Eldridge RL, et al. 37. Frater AJ, Brown H, Oxenius A, Gunthard HF, Hirschel B, Robinson N, et al. Substantial differences in specificity of HIV-specific cytotoxic T cells in Effective T-cell responses select human immunodeficiency virus mutants acute and chronic HIV infection. J Exp Med. 2001;193(2):181–94. and slow disease progression. J Virol. 2007;81(12):6742–51. 17. Johnson RP, Trocha A, Buchanan TM, Walker BD. Identification of overlap‑ 38. Shankarappa R, Gupta P, Learn GH Jr, Rodrigo AG, Rinaldo CR Jr, Gorry ping HLA class I-restricted cytotoxic T cell epitopes in a conserved region MC, et al. Evolution of human immunodeficiency virus type 1 envelope of the human immunodeficiency virus type 1 envelope glycoprotein: sequences in infected individuals with differing disease progression definition of minimum epitopes and analysis of the effects of sequence profiles. Virology. 1998;241(2):251–9. variation. J Exp Med. 1992;175(4):961–71. 39. Rambaut A, Posada D, Crandall KA, Holmes EC. The causes and conse‑ 18. Koibuchi T, Allen TM, Lichterfeld M, Mui SK, O’Sullivan KM, Trocha A, et al. quences of HIV evolution. Nat Rev Genet. 2004;5(1):52–61. Limited sequence evolution within persistently targeted CD8 epitopes 40. Brumme ZL, Tao I, Szeto S, Brumme CJ, Carlson JM, Chan D, et al. in chronic human immunodeficiency virus type 1 infection. J Virol. Human leukocyte antigen-specific polymorphisms in HIV-1 Gag and 2005;79(13):8171–81. their association with viral load in chronic untreated infection. AIDS. 19. Kemal KS, Beattie T, Dong T, Weiser B, Kaul R, Kuiken C, et al. Transi‑ 2008;22(11):1277–86. tion from long-term nonprogression to HIV-1 disease associated with 41. Edwards BH, Bansal A, Sabbaj S, Bakari J, Mulligan MJ, Goepfert PA. escape from cellular immune control. J Acquir Immune Defic Syndr. Magnitude of functional CD8+ T-cell responses to the gag protein of 2008;48(2):119–26. human immunodeficiency virus type 1 correlates inversely with viral load 20. Casado C, Pernas M, Alvaro T, Sandonis V, Garcia S, Rodriguez C, et al. in plasma. J Virol. 2002;76(5):2298–305. Coinfection and superinfection in patients with long-term, nonprogres‑ 42. Kiepiela P, Ngumbela K, Thobakgale C, Ramduth D, Honeyborne I, sive HIV-1 disease. J Infect Dis. 2007;196(6):895–9. Moodley E, et al. CD8+ T-cell responses to different HIV proteins have 21. Ferrari G, Korber B, Goonetilleke N, Liu MK, Turnbull EL, Salazar-Gonzalez discordant associations with viral load. Nat Med. 2007;13(1):46–53. JF, et al. Relationship between functional profile of HIV-1 specific CD8 T 43. Piantadosi A, Chohan B, Panteleeff D, Baeten JM, Mandaliya K, Ndinya- cells and epitope variability with the selection of escape mutants in acute Achola JO, et al. HIV-1 evolution in gag and env is highly correlated but HIV-1 infection. PLoS Pathog. 2011;7(2):e1001273. exhibits different relationships with viral load and the immune response. AIDS. 2009;23(5):579–87. Moyano et al. Retrovirology (2022) 19:6 Page 14 of 14 44. Henn MR, Boutwell CL, Charlebois P, Lennon NJ, Power KA, Macalalad AR, et al. Whole genome deep sequencing of HIV-1 reveals the impact of early minor variants upon immune recognition during acute infection. PLoS Pathog. 2012;8(3):e1002529. 45. Liu Y, McNevin J, Cao J, Zhao H, Genowati I, Wong K, et al. Selection on the human immunodeficiency virus type 1 proteome following primary infection. J Virol. 2006;80(19):9519–29. 46. Turnbull EL, Wong M, Wang S, Wei X, Jones NA, Conrod KE, et al. Kinetics of expansion of epitope-specific T cell responses during primary HIV-1 infection. J Immunol. 2009;182(11):7131–45. 47. Burgess CA, Grant GH. Compensatory mutations and protein stability: a theoretical study. Biochem Soc Trans. 1997;25(1):120S. 48. Troyer RM, McNevin J, Liu Y, Zhang SC, Krizan RW, Abraha A, et al. Variable fitness impact of HIV-1 escape mutations to cytotoxic T lymphocyte (CTL) response. PLoS Pathog. 2009;5(4):e1000365. 49. Hall TA. Bioedit: a user-friendly biological sequence aligment editor and analysis program for Windows 95/98/NT. 1999. London, Information Retrieval Ltd. Ref Type: Generic 50. Tamura K, Stecher G, Peterson D, Filipski A, Kumar S. MEGA6: molecular evolutionary genetics analysis version 60. Mol Biol Evol. 2013;30(12):2725–9. 51. Sandonis V, Casado C, Alvaro T, Pernas M, Olivares I, Garcia S, et al. A combination of defective DNA and protective host factors are found in a set of HIV-1 ancestral LTNPs. Virology. 2009;391(1):73–82. 52. Addo MM, Yu XG, Rathod A, Cohen D, Eldridge RL, Strick D, et al. Compre‑ hensive epitope analysis of human immunodeficiency virus type 1 (HIV- 1)-specific T-cell responses directed against the entire expressed HIV-1 genome demonstrate broadly directed responses, but no correlation to viral load. J Virol. 2003;77(3):2081–92. 53. Dalmau J, Rotger M, Erkizia I, Rauch A, Reche P, Pino M, et al. Highly pathogenic adapted HIV-1 strains limit host immunity and dictate rapid disease progression. AIDS. 2014;28(9):1261–72. 54. Perucho M, Hanahan D, Wigler M. Genetic and physical linkage of exog‑ enous sequences in transformed cells. Cell. 1980;22(1 Pt 1):309–17. 55. Waterhouse A, Bertoni M, Bienert S, Studer G, Tauriello G, Gumienny R, et al. SWISS-MODEL: homology modelling of protein structures and complexes. Nucleic Acids Res. 2018;46(W1):W296–303. 56. Shen MY, Sali A. Statistical potential for assessment and prediction of protein structures. Protein Sci. 2006;15(11):2507–24. 57. Webb B, Sali A. Comparative protein structure modeling using MODEL‑ LER. Curr Protoc Bioinformatics. 2014;8(47):5–32. 58. Arenas M, Dos Santos HG, Posada D, Bastolla U. Protein evolution along phylogenetic histories under structurally constrained substitution mod‑ els. Bioinformatics. 2013;29(23):3020–8. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub‑ lished maps and institutional affiliations. Ready to submit your research ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions