In a previous 2023 study published in PNAS (see below and featured in CRISPR Medicine News), the same group showed that simultaneous CRISPR targeting of the HIV-1 LTR-gag region and the host CCR5 co-receptor, combined with long-acting antiretroviral therapy (LA-ART), could excise integrated provirus and achieve viral clearance in some mice.
“No major CRISPR-specific mutations are observed. The molecular viral signatures demonstrate an accelerated HIV-1 drug resistance escape from ART rather than from the generation of CRISPR mutants”Zhang et al.
However, viral persistence and rebound remained unresolved in a subset of mice. Here, the authors profile the genetic features of rebound viruses and determine that drug resistance, not CRISPR-induced mutagenesis, accounts for viral escape in these cases.
In the current study, humanised mice were infected with either HIV-1_NL4-3 or HIV-1_ADA and assigned to one of four groups: untreated, LA-ART only, CRISPR-Cas9 only, or sequential LA-ART followed by CRISPR (see Figure 1). LA-ART consisted of nanoformulated dolutegravir, lamivudine, abacavir, and rilpivirine. The CRISPR-Cas9 system, delivered via AAV9 vectors, targeted the HIV-1 LTR-gag region.
details, plasma collection and RNA isolation leading up to subsequent sequencing assays and data analysis. Reproduced under the Creative Commons License from Figure 1 in Zhang et al. (2025) Communications Biology, https://doi.org/10.1038/s42003-025-08499-6
Plasma viral RNA was collected at multiple stages—pre-treatment, post-LA-ART, and study endpoint—and gag, pol, and env regions were sequenced using both Sanger and high-sensitivity next-generation methods. Sequencing data revealed significant viral evolution in all treatment groups, with the env gene showing the highest mutation rates (see Figure 2). Rebound viruses displayed multiple drug resistance-associated mutations (DRAMs) in reverse transcriptase (e.g., M41L, V75I), integrase (R263K, V151I), and protease (I54T, F53L), particularly in LA-ART and dual-treated animals.
These mutations were absent in untreated and CRISPR-only animals. Notably, no indels or substitutions were found at or near the CRISPR target site in gag, and CRISPR-treated rebound viruses lacked signs of editing-induced escape. Longitudinal analysis confirmed that most resistance mutations arose during the ART phase, not following CRISPR exposure.
Additional unique mutations – some undocumented in resistance databases – were identified in pol, especially in the reverse transcriptase region. These included shared substitutions (e.g., E29G and Q151R) in the LA-ART and dual-treated groups, potentially representing compensatory adaptations under drug pressure.
Interestingly, while LA-ART-only samples showed a broader range of high-frequency novel mutations, dual-treated rebound viruses harboured fewer such changes, suggesting a possible limiting effect of CRISPR on viral diversity. Two of the five dual-treated samples also exhibited non-synonymous substitutions in the env CD4-binding loop, potentially affecting host-cell binding affinity; however, the functional impact requires further study.
“The data underscores the limited role of CRISPR excision in generating these rebound HIV-1 mutants from dual-treated hu-mice”Zhang et al.
These findings suggest that viral rebound in this model is primarily driven by the selection of resistant variants induced by ART. On the other hand, CRISPR-Cas9 exerts minimal off-target pressure and does not contribute to the emergence of escape mutations. The high specificity of CRISPR was supported by the absence of mutations at its target site and the preservation of the highly conserved gag region across rebound samples. However, indirect selection effects may arise through disruption of viral replication capacity, contributing to adaptive changes in untargeted regions such as pol and env. Host restriction factors, such as APOBEC3G, were also implicated, based on mutational patterns in treated animals.
While CRISPR-Cas9 effectively excises provirus and does not provoke resistance, the study demonstrates that viral escape can still occur under ART pressure. These results reinforce the need for optimised combinatorial strategies to suppress replication-competent variants and address latent reservoirs. This work also validates the use of NGS in detecting low-frequency DRAMs that are often missed by standard sequencing methods, which may be critical for anticipating rebound risk in clinical settings.
The study was led by Chen Zhang and Prasanta K. Dash from the University of Nebraska Medical Center. It was published in Communications Biology on 19 July 2025. The group’s previous study was published in PNAS in 2023.
To get more CRISPR Medicine News delivered to your inbox, sign up to the free weekly CMN Newsletter here.
ArticleCMN BriefsCMN HighlightsNewsHuman Immunodeficiency Virus Infection, HIVNext gen sequencing