Sialidase Inhibitors Block HIV-Driven Brain Aging and Memory Loss in Mice

HIV-associated cognitive impairment (HIV-CI) affects roughly 24% of people living with HIV even when viral load is suppressed by antiretroviral therapy, yet the upstream mechanisms driving this decline remain poorly defined. This study identifies pro-inflammatory glycan degradation—specifically loss of sialic acid (hypo-sialylation) and galactose (agalactosylation) from circulating immune glycoproteins—as a previously unrecognized contributor to HIV-CI and accelerated biological aging.

The investigators analyzed longitudinal plasma samples from the ACTG HAILO (A5322) cohort: 40 ART-suppressed PLWH followed over 8 years, matched by sex, age, and ethnicity. IgG and plasma glycomic profiles were quantified by capillary electrophoresis and linked to neuropsychological composite scores (NPZ4). Pro-inflammatory agalactosylated glycan species such as G0F and G0FB were elevated in those with cognitive impairment and inversely correlated with NPZ4 scores. A key finding was that eleven standard inflammatory cytokines—including IL-6, TNFα, and IFNγ—did not differentiate cognitively impaired from unimpaired individuals, whereas glycomic traits did, suggesting superior specificity. These associations were validated in a separate cross-sectional cohort of 80 PLWH, and were especially pronounced in women living with HIV, who exhibited significantly higher agalactosylation and lower sialylation when cognitively impaired.

To test causality and therapeutic potential, the team turned to two animal models. In a humanized mouse model of HIV infection, treatment with sialidase inhibitors (oseltamivir combined with the broad-spectrum compound DANA) preserved surface sialylation on immune cells, reduced virally induced systemic inflammation, and attenuated epigenetic markers of accelerated biological aging. In Eco-HIV—a chimeric virus model enabling detailed behavioral testing in immunocompetent mice—sialidase inhibitor treatment prevented HIV-driven memory deficits as assessed by novel object recognition and related paradigms. In vitro experiments on human PBMCs and monocyte-derived macrophages confirmed that sialidase exposure alone was sufficient to trigger inflammatory cytokine release and immune activation, both reversed by sialidase inhibitor co-treatment. Importantly, neither oseltamivir nor DANA showed direct antiviral effects on HIV infectivity in TZM-bl assays, indicating the benefit was anti-inflammatory rather than antiviral.

These results reposition sialidase inhibitors—drugs already FDA-approved and widely used for influenza—as candidate anti-inflammatory agents for HIV-CI. They also implicate glycomic degradation as a mechanistic bridge linking chronic HIV infection to the hallmarks of premature aging observed in PLWH. The identification of glycomic biomarkers, particularly the IgG Gal-ratio, as more sensitive correlates of HIV-CI than conventional cytokines opens the door to novel prognostic tools.

Caveats include the preprint status of the work, the relatively small human cohort sizes, and the predominance of effects in women, which requires further investigation. Translation from mouse behavioral models and in vitro systems to clinical benefit in humans will require dedicated trials. Nonetheless, the convergence of two independent human cohorts, two complementary animal models, and mechanistic in vitro data provides a compelling case for glycan-targeting strategies in viral cognitive disease.