Near-full-length genomic characterization and phylogenetic analysis of an HIV-1 CRF02_AG/G second-generation recombinant in Ningxia, China

Abstract

With the rapid evolution of global HIV-1 recombinant strains, existing diagnostic and therapeutic strategies are under threat. This study aims to conduct molecular characterization and origin tracing of a CRF02_AG-derived second-generation recombinant detected in Ningxia, China. In 2024, during genotyping screening of the pol region in HIV-1-infected individuals from Ningxia, preliminary evidence of CRF02_AG-related signals was identified in one sample. Subsequently, the near-full-length genome sequence of this strain (designated A01) was successfully obtained. Phylogenetic analysis revealed that its gag, pol, and env gene regions clustered within the A1 and G subtypes, respectively. Recombination analysis using jpHMM and Simplot software confirmed that this strain utilizes the CRF02_AG (A1/G primary recombinant) as its genomic backbone, with two G subtype fragments inserted into the env region (HXB2 sites: 6194–7137 and 7788–8532). Consequently, this strain was defined as a novel CRF02_AG/G second-generation recombinant. Bayesian phylogeographic analysis further revealed that this Ningxia strain shares close genetic relationships with CRF02_AG strains prevalent in the United Kingdom and South Korea. This study marks the first identification of an HIV-1 CRF02_AG/G second-generation recombinant in Ningxia (singular case), underscoring the importance of sustained high-resolution molecular surveillance in populations with mobility for early warning of imported and locally transmitted novel recombinant strains.

Introduction

Acquired immunodeficiency syndrome (AIDS) is an infectious disease caused by infection with the human immunodeficiency virus (HIV). Since its first reported cases in the 1980s, AIDS has claimed over 40 million lives worldwide. The rapid and continuous genetic evolution of HIV-1 has led to the emergence of circulating recombinant forms (CRFs) and unique recombinant forms (URFs), which have become dominant strains in certain regions. According to the Los Alamos HIV Sequence Database (https://www.hiv.lanl.gov/content/sequence/HIV/mainpage.html), over 167 CRFs and numerous URFs have been identified worldwide. In China, CRFs constitute the vast majority of HIV-1 infections. For example, data from Wang D et al. [1]. Indicate that CRFs account for 88.6% of all HIV-1 infections, with URF detection rates also showing an upward trend. However, the globally prevalent CRF02_AG recombinant strain is extremely rare in China and has never been documented in Ningxia before this study.

The fourth molecular epidemiological survey revealed that the predominant HIV-1 strains currently circulating in China are CRF07_BC (41.3%), CRF01_AE (32.7%), CRF08_BC (11.3%), and B' (4.0%) subtypes, with CRF01_AE and CRF07_BC being the most prevalent [2]. In the Ningxia region, CRF07_BC and CRF01_AE are the predominant circulating subtypes [3]. The widespread circulation of recombinant strains complicates HIV prevention and control efforts and poses significant challenges to antiretroviral therapy (ART) and vaccine development.

According to a study by Wang D et al. [1], HIV-1 strains other than CRF01_AE, CRF07_BC, CRF08_BC, and CRF55_01B together account for fewer than 6% of cases, suggesting that CRF02_AG is a rare strain in China. Previously, this strain had only been sporadically reported in coastal provinces (e.g., Guangdong [4], Shanghai [5]) within China, with no confirmed cases identified in Ningxia. The detection of the CRF02_AG/G second-generation recombinant in Ningxia holds significant epidemiological importance: it not only enhances China's surveillance data system for URFs and reveals the geographical distribution of this strain within China, but also underscores the necessity of strengthening surveillance efforts in inland regions.

The identification of recombinant forms is crucial for tracking viral evolution, optimizing treatment strategies, and informing public health interventions [1]. In this study, we performed subtype genotyping of the pol, gag, and env region sequences of an identified HIV-1 recombinant strain from Ningxia in 2024. We determined the recombination patterns using near-full-length genome (NFLG) sequences. This study aims to conduct molecular surveillance of HIV-1 subtypes in Ningxia, perform near-full-length genome sequencing of the newly identified URF, and explore its epidemiological significance through phylogenetic and phylogeographic analyses.

Materials and methods

Study participants

This study conducted a cross-sectional investigation in July 2025, utilizing blood samples (n = 324) collected from all newly diagnosed HIV-1/AIDS participants in Ningxia during January to December 2024 from the biorepository. Inclusion criteria for this study were as follows: (1) participants had to be HIV-1/AIDS participants residing in NHAR; (2) included both HIV-1/AIDS participants on ART and those not on ART; (3)underwent viral load testing between January 2024 and December 2024 with a viral load ≥ 400 copies/mL; (4) demographic information could be correctly matched with laboratory data; and (5) all participants had given informed consent to participate in the survey. On the other hand, the exclusion criteria included: (1) samples with failed HIV nucleic acid extraction, amplification, or sequencing; (2) duplicate sequences from the same HIV-1/AIDS participant; (3) viral load < 400 copies/mL; and (4) samples with missing or incorrect demographic data.

The participant ultimately identified for this study was a 42-year-old unmarried male infected through homosexual transmission (MSM). The sample was collected on April 8, 2024, from the Yinchuan Center for Disease Control and Prevention and designated as A01(GenBank accession C_AA134007.1). The participant has signed a written informed consent form authorizing subsequent analysis.

Laboratory tests

Plasma samples were stored at − 80 °C within 2 h of collection and transported on dry ice to the laboratory for subsequent analysis. Viral RNA was extracted from plasma using the automated nucleic acid extraction and purification system, along with the HIV-1 viral load assay according to the manufacturer’s protocol (Zhuhai Livzon Diagnostics Inc.) [6]. Partial pol sequences (yielding a product of approximately 1100 bp) corresponding to codons 1 to 99 aa of the protease and 1–299 aa of the reverse transcriptase were amplified by in-house nested reverse transcription PCR using the One Step RNA PCR Kit (Takara, China) [7]. The amplification products were verified by 1% agarose gel electrophoresis and subsequently sent to Beijing Novogene Bioinformatics Technology Co., Ltd., Beijing, China, for purification and sequencing [8]. A sample identified as a novel recombinant strain underwent near-full-length genome sequencing using NGS technology by Beijing Dehong Changyuan Biotechnology Co., Ltd. (sequencing details are provided in Appendix I).

Sequence analysis

The obtained HIV-1 sequences were analyzed for preliminary subtyping using the BLAST tool on the NCBI HIV database website (https://blast.ncbi.nlm.nih.gov/Blast.cgi). Subsequently, the sequences were aligned using Clustal W in MEGA11 software. The best-fit nucleotide substitution model was determined using model selection tools in MEGA11. Based on this, a phylogenetic tree was constructed via the Neighbor-Joining method using the Tamura-Nei model. Branch support was assessed with 1000 bootstrap replicates. Recombination analysis was performed using the jpHMM online tool (http://jphmm.gobics.de). This tool calculates the posterior probability of each sequence position belonging to different reference subtypes based on Hidden Markov Models (HMMs). It outputs the most probable recombination pathway (i.e., the maximum a posteriori decoding path). We adopted the optimal pathway result generated by this tool as the preliminary recombination map. The reference sequence panel employed was the built-in HIV-1 Group M standard reference subtype set within jpHMM. This panel encompasses major subtypes (A–D, F–H, J–K) and common recombinant forms, maintaining version synchronization with the Los Alamos HIV database to ensure analytical standardization and result comparability. The results from jpHMM were enhanced by confirming recombination sites through similarity analysis and step-scan analysis in Simplot 3.5.1 software, with a window size of 400 and a step size of 40. If discrepancies existed, the Simplot results were used as the basis for the conclusion, and the website’s recombinant drawing tool was used to visualize the mosaic structure and gene recombination breakpoints of NFLG. This combined approach ensured precise detection of complex events. The target sequence (A01) was ultimately uploaded to the HIV database website via the HIV BLAST online tool (https://blast.ncbi.nlm.nih.gov/Blast.cgi). CRF02_AG sequences were selected that shared ≥ 95% genetic similarity with the target sequence and contained ≥ 900 bp of the target gene fragment (while excluding duplicate sequences from the same participant). The final dataset comprised 27 sequences from 7 countries, sampled between 1998 and 2016. By integrating HIV sequence sampling dates with geographic information, Bayesian phylogenetic analysis was conducted using BEAST v10.5.0 software. Model settings were as follows: the nucleotide substitution model employed the Generalized Temporally Reversible (GTR) model, with evolutionary rates estimated using an Uncorrelated Relaxed Clock. A Geographic Discrete Trait Model was integrated to infer transmission pathways. The Markov Chain Monte Carlo (MCMC) ran for 20,000,000 generations, sampling every 1,000 generations, with the first 10% of samples discarded as burn-in. Chain convergence was assessed using Tracer v1.7.2 software. The effective sample size (ESS) for all key parameters exceeded 200, indicating sufficient posterior sample size and reliable analysis results.

Results

Fragment-based phylogenetic analysis

Downloaded reference sequences for HIV-1 Group M from the HIV BLAST database. Using MEGA11 software, the gag, pol, and env gene region sequences of A01 were aligned against the reference sequences, and a Neighbor-Joining phylogenetic tree was constructed based on these alignments. The stability of each node was assessed through bootstrap analysis with 1000 repetitions, where bootstrap values ≥ 70% were considered statistically significant. Results are shown in Fig. 1: the gag sequence of A01 formed a monophyletic cluster with the subtype A1 reference strain (bootstrap value = 100%), indicating a close genetic relationship between the two. The env region sequence formed monophyletic groups with subtype G (bootstrap value = 100%) and subtype A1 (bootstrap value = 99%) in different segments. The pol region sequence formed a monophyletic group with the subtype G reference strain (bootstrap value = 79%), a statistically significant result.

Fig. 1.

NJ phylogenetic tree analysis of each functional region of sample A01. Sample (green), Group M (blue); a phylogenetic analysis of the gag region of sample A01; b phylogenetic analysis of the env region of sample A01. c phylogenetic analysis of the pol region of sample A01. Sample A01 is highlighted in green in all panels; reference sequences of HIV-1 Group M are shown in blue

Given that the three major functional regions (gag, pol, env) of A01 form monophyletic groups with subtypes A1 and G, respectively, it is speculated that A01 may belong to the A1/G recombinant strains. To validate this hypothesis, we downloaded A1/G recombinant strains from the HIV BLAST database and constructed a phylogenetic tree using the neighbor-joining method based on the pol region. The results, shown in Fig. 2, reveal that A01 forms a monophyletic cluster with CRF02_AG (bootstrap value = 94%).

Fig. 2.

NJ Phylogenetic tree analysis of A01 pol region (green) and A1/G recombinant strain (blue)

NFLG recombination pattern analysis

Preliminary phylogenetic analysis revealed that A01 forms a monophyletic clade with the CRF02_AG strain. To further characterize its recombination structure, we first compared the CRF02_AG reference strain with the A01 near-full-length genome (NFLG) using jpHMM (Fig. 3a, b). This comparison delineated the recombinant breakpoints, establishing the final genomic mosaic structure of A01 (Fig. 3c). Comparative analysis revealed highly consistent recombination patterns in the gag and pol regions between the two strains. However, differences emerged in the env region: at positions 6194–7137 and 7788–8532 of the HXB2 (GenBank accession K03455) reference genome, the CRF02_AG reference strain exhibited subtype A1 characteristics, whereas A01 displayed subtype G features in these segments. A01 largely matches the CRF02_AG mosaic structure except for two subtype G insertions in env (6194–7137 and 7788–8532 in HXB2). Based on these findings, we concluded that A01 represents a unique HIV-1 recombinant strain formed by a secondary recombination event in the env region between CRF02_AG and subtype G. The cross-subtype clustering of different gene fragments in this strain aligns with this complex secondary recombination pattern.

Fig. 3.

JPHMM Recombinant pattern analysis of the new recombinant strain of sample A01. a CRF02_AG NFLG Reference jpHMM Recombination Pattern Analysis. b NFLG jpHMM Recombination Combination Pattern Analysis for Sample A01. c NFLG Mosaic Structure and Gene Recombination Breakpoint Analysis for Sample A01

To further determine the location of the recombination breakpoint, this study utilized the Bootscan analysis function in Simplot v3.5.1 software to analyze the nucleotide genetic similarity between the near-full-length genome (NFLG) of A01 and the CRF02_AG standard strain, as well as the G subtype reference strain. Figure 4 shows that the A01 genome exhibits secondary recombination characteristics, featuring a CRF02_AG (A1/G recombinant) backbone with two G subtype gene fragments chimeric within the CRF02_AG genomic sequence. Based on the specific locations of recombination breakpoints, this strain was confirmed to be a unique recombinant strain resulting from secondary recombination between CRF02_AG and subtype G, rather than a direct recombinant product of CRF02_AG and subtype G.

Fig. 4.

Bootscan similarity analysis of the A01 near-full-length genome (NFLG). Color Legend: CRF02_AG subtype (dark green), G subtype (light green)

Phylogenetic analysis

A BLAST search against the HIV database yielded 27 CRF02_AG reference sequences from seven countries. All sequences were sourced from international databases due to the limited availability of domestic data. Based on the time-resolved phylogenetic analysis using BEAST v10.5.0, the identified strain in Ningxia was found to share a genetic relationship with strains from multiple countries. Bayesian factor analysis indicated (Fig. 5) that the UK strain had the strongest genetic relationship with the Ningxia strain (log BF = 9.6), followed by South Korea (log BF = 4.1), while Belgium and Burkina Faso had slightly weaker associations with the Ningxia strain (1.0 < log BF < 2.0). Additionally, although the association was weak (log BF ≥ 0.5), our results suggest a potential genetic relationship between strains from Senegal, Ghana, and Cameroon and the Ningxia strain. These log BF values indicate that the target sequence shares the closest genetic relationship with sequences in these national datasets. These results reflect phylogenetic proximity within the available reference dataset and do not establish directionality or epidemiological linkage.

Fig. 5.

Shows the transmission pathway of the HIV-1 CRF02_AG recombinant virus. Each geographical location is colored according to HIV-1 prevalence data (UNAIDS), and the lines connecting the locations represent the migration paths between CRF02_AG subpopulations. The line colors reflect the Bayesian factor test support for epidemiological associations between regions

Discussion

Since its discovery in the 1980s, HIV/AIDS has remained a significant challenge for global public health. Despite global efforts in HIV prevention, treatment, and education, the epidemic remains a major challenge, as evidenced by the 2024 UNAIDS estimate of 40.8 million people living with HIV, including 1.3 million new infections. According to China's national molecular epidemiological survey, as the infected population has shifted from primarily high-risk groups to a mix of high-risk and general populations, the number of CRF and URF strains has increased rapidly, and the distribution of HIV-1 subtype strains has also undergone significant changes [9, 10]. China has a complex HIV-1 epidemic strain profile and is a high-risk region for inter-subtype recombination. To date, at least 59 typical recombinant strains have been identified in China, with CRF01_AE, CRF07_BC, CRF08_BC, and the B subtype being the predominant recombinant strains. In recent years, the CRF55_01B subtype has also begun to circulate and spread in China [10]. Although CRF02_AG is globally prevalent, its overall prevalence in China remains low, significantly lower than in Europe, America, and Africa. In Ningxia, the prevalence of CRF01_AE and CRF07_BC has been gradually increasing, making them the dominant strains in the region. In this context, we identified and characterized a unique URF in Ningxia—a second-generation recombinant derived from CRF02_AG and subtype G. This suggests possible introduction into the region; additional surveillance is needed to determine whether onward transmission has occurred. The emergence of novel recombinant strains significantly impacts HIV prevention and control strategies, as they may alter viral transmissibility, pathogenicity, and treatment response.

Obtaining the full-genome sequence of HIV-1 is a crucial foundation for conducting molecular virology research on HIV-1. It plays a significant role in studying viral evolution within specific populations or regions, identifying new subtypes or prevalent recombinant types, understanding viral transmission and distribution patterns, and developing diagnostic reagents and vaccines. Next-generation sequencing (NGS) offers advantages over first-generation sequencing, including higher throughput, faster sequencing speeds, and greater sensitivity. The application of next-generation sequencing technology for deep sequencing of HIV-1 facilitates accurate analysis of viral strains [11, 12]. The HIV-1 genome is approximately 9.7 kb in length, with the pol region spanning approximately 1.1 kb. HIV-1 subtypes are typically determined based on the pol region sequence. In the initial phase of this study, the three major gene regions (gag, pol, env) of sample A01 formed monophyletic groups with either subtype A1 or G in the phylogenetic tree constructed based on HIV-1 group M reference sequences, suggesting that A01 may be an A1/G recombinant strain. To determine its specific recombination pattern, we conducted a phylogenetic analysis comparing the pol sequence of A01 with known reference sequences of the A1/G circulating recombinant. The analysis results indicated that the pol region of A01 formed a monophyletic group with CRF02_AG.

The identification of the CRF02_AG/G second-generation recombinant in this study highlights the increased genetic diversity of the virus in the Ningxia region, which may be associated with an accelerating trend in viral transmission. The global distribution of the HIV-1 G subtype is highly uneven, primarily concentrated in Central and West Africa as well as Europe, accounting for only 5.0% [13] of global infections; however, the G subtype has been present for a longer period, and through viral mutation and recombination, it has widely circulated in combination with other HIV-1 subtypes [14]. For example, the recombinant strain CRF02_AG is a result of recombination between HIV-1 subtype A and subtype G, with CRF02_AG infections accounting for 7.7% of global HIV-1 infections [15]. CRF02_AG is in a low-prevalence state among HIV participants in China and is classified as a non-dominant strain. The first known cases of the CRF02_AG strain in China were identified in Shenzhen in 1992, with the affected individuals belonging to a cross-border population. [4]. This suggests that the emergence of the CRF02_AG recombinant strain in Ningxia may have been facilitated by interprovincial population mobility or local high-risk behaviors.

Our research group has conducted long-term monitoring of HIV-1 genetic diversity in Ningxia. In 2024, through pol region genotyping screening, we detected signals of the non-local dominant strain CRF02_AG in one sample. Subsequent near-full-length genome sequencing and in-depth analysis revealed that this strain was not a typical CRF02_AG, but rather a CRF02_AG/G second-generation recombinant formed by a CRF02_AG genomic backbone with an additional G subtype fragment inserted into the env region. Such novel re-emerging forms (URFs), due to their unique genomic structure, may compromise the accuracy of diagnostic reagents designed for specific gene regions and potentially alter viral susceptibility to antiviral drugs, thereby posing challenges to existing treatment regimens. The findings of this study further underscore the necessity of maintaining molecular epidemiological surveillance, particularly among highly mobile populations, and dynamically adjusting prevention and control strategies to address the ongoing evolution of HIV-1.

Bayesian methods, as a core technique in molecular epidemiology research, integrate molecular clock models, geographic diffusion analysis, and Bayesian factor calculations to reconstruct potential virus transmission pathways and trace infection sources [16, 17]. This study employed the method to analyze the newly emerging CRF02_AG/G second-generation recombinant strain in Ningxia. The results showed that this strain was associated with strains prevalent in multiple countries, with the strongest association observed with the UK strain, followed by the South Korean strain. This finding suggests that the recombinant strain shares the closest genetic relationship with sequences from the UK and South Korean datasets. The study results not only confirm that international population mobility is an important risk factor for HIV transmission but also reveal its potential role in facilitating the formation of new recombinant strains, providing important scientific evidence for improving regional prevention and control strategies.

This study utilized blood samples from one HIV-1 participant in Ningxia to determine recombination breakpoints through near-full-length genome (NFLG) amplification and phylogenetic analysis, combined with jpHMM and Simplot recombination analysis. The strain was genetically subtyped as a CRF02_AG/G second-generation recombinant, and molecular tracing of this recombinant strain was conducted. Several limitations remain in this study. First, with only a single sample and a limited retrospective analysis timeframe, we could not precisely trace the strain's initial emergence locally, limiting the assessment of transmission dynamics. Second, while phylogenetic analysis confirmed its recombinant nature, the potential biological implications of this mosaic structure remain unverified due to a lack of functional experimental data.

Conclusions

This study identifies an HIV-1 CRF02_AG/G second-generation recombinant strain in Ningxia. Phylogenetic analysis and Bayesian geospatial analysis suggest this strain exhibits a relatively close genetic relationship with strains found overseas. This finding further underscores the escalating risks of viral importation and local transmission in settings with high population mobility. Consequently, the urgency of establishing a sustainable, high-resolution molecular surveillance system for early warning is particularly pronounced. Despite retrospective design limitations and sample size constraints, this study provides robust evidence for enhanced surveillance. Future efforts should integrate multi-source data to develop more sensitive monitoring systems capable of addressing the challenges posed by ongoing viral evolution within dynamic transmission environments.

Institutional review board

The study was conducted in accordance with the Declaration of Helsinki and approved by the Ningxia CDC Institutional Review Board (Application No. 2025-LLSC-220).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Abbreviations

AIDS

Acquired Immune Deficiency Syndrome

HIV

Human Immunodeficiency Virus

CRFs

Circulating recombinant forms

URFs

Unique recombinant forms

ART

Antiretroviral therapy

NFLG

Near full-length genome

HXB2

HIV-1 IIIB/BRU (GenBank accession K03455)

MSM

Homosexual transmission

Author contributions

Conceptualization, M.S. and Z.W.; Methodology, J.P. and Z.W.; Validation, D.Y. and Z.W.; Formal Analysis, M.S.; Investigation, Z.L., Y.D., Y.L., and X.Z.; Data curation, X.Z., Z.L., Y.D., and D.Y.; Writing—original draft, M.S. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by the Ningxia Natural Science Foundation of China (No: 2025AAC020047).

Data availability

The original data presented in the study are openly available in the National Center for Biotechnology Information, https://www.ncbi.nlm.nih.gov/

Declarations

Ethics approval and consent to participate

Informed consent was obtained from all subjects involved in the study.