There is still no effective vaccine or cure for HIV. Scientists are considering options ranging from longer-term antiretroviral therapy (ART) that space out injections by several years to long-lasting pre-exposure prophylaxis (PrEP) that acts as a vaccine while immunization is achieved. What else can be done? The “Innovations in HIV virology: Translating discoveries into novel therapies” symposium in basic science at the 13th IAS Conference on HIV Science (IAS 2025), which took place from July 13 to 17, 2025, in Kigali, Rwanda, showcased some of the new ideas that the scientific community are developing.

While people living with HIV can lead virtually normal lives thanks to antiretroviral therapy (ART), HIV persists in a latent state within cellular reservoirs that scientists do not know how to eliminate. “Transcription is a critical step in the viral life cycle. … But there are currently no drugs suppressing HIV transcription, and that may be one of the reasons why current antiretroviral therapy is not curative,” Melanie Ott told the audience at the 13th IAS Conference on HIV Science this week in Kigali, Rwanda.

Degradation is a therapeutic strategy that could offer possibilities to get at currently undruggable target proteins. In targeted degradation, compounds induce interactions between a target protein and a protein that can tag the target for degradation. In principle, there are several pathways that could be used for such tagging; the most attention has gone to ubiquitin ligases, in particular cereblon, a protein that is part of a ubiquitin ligase complex and the target of several approved drugs.

Degradation is a therapeutic strategy that could offer possibilities to get at currently undruggable target proteins. In targeted degradation, compounds induce interactions between a target protein and a protein that can tag the target for degradation. In principle, there are several pathways that could be used for such tagging; the most attention has gone to ubiquitin ligases, in particular cereblon, a protein that is part of a ubiquitin ligase complex and the target of several approved drugs.

Aussie researchers have used CRISPR gene editing tools to “armor” chimeric antigen receptor (CAR) T cells to activate additional cancer-fighting proteins at the tumor site, enabling them to target cancer cells in solid tumors.

Degradation is a therapeutic strategy that could offer possibilities to get at currently undruggable target proteins. In targeted degradation, compounds induce interactions between a target protein and a protein that can tag the target for degradation. In principle, there are several pathways that could be used for such tagging; the most attention has gone to ubiquitin ligases, in particular cereblon, a protein that is part of a ubiquitin ligase complex and the target of several approved drugs.

Degradation is a therapeutic strategy that could offer possibilities to get at currently undruggable target proteins. In targeted degradation, compounds induce interactions between a target protein and a protein that can tag the target for degradation. In principle, there are several pathways that could be used for such tagging; the most attention has gone to ubiquitin ligases, in particular cereblon, a protein that is part of a ubiquitin ligase complex and the target of several approved drugs.

Aussie researchers have used CRISPR gene editing tools to “armor” chimeric antigen receptor (CAR) T cells to activate additional cancer-fighting proteins at the tumor site, enabling them to target cancer cells in solid tumors.

An experimental drug for treating diabetes and obesity has been shown to lower blood sugar levels and increase fat burning. It is a β2-adrenergic receptor (β2AR) agonist that mimics the effects of physical exercise by activating skeletal muscle metabolism. Unlike GLP-1-based treatments such as semaglutide and tirzepatide, this new compound, developed by researchers at the Karolinska Institute, Stockholm University, and the biotech company Atrogi AB, does not suppress appetite or cause muscle loss. 

The Alphafold machine learning system for predicting a protein’s structure from its amino acid sequence has been adapted to make it possible to design de novo proteins that fold in a particular way and bind to prespecified target proteins. The sister system, called Alphadesign, works by generating random strings of amino acids, using Alphafold to predict their structure, and then iteratively optimizing the design.