Longevity & AgingResearch PaperOpen Access

CD38 Drives Brain NAD+ Collapse During Zika Virus Infection

New mouse study reveals CD38, not viral replication itself, depletes brain NAD+ during Zika infection — pointing to a druggable target.

Thursday, June 4, 2026 0 views
Published in iScience
Glowing NAD+ molecules fragmenting inside a translucent mouse brain cross-section surrounded by immune cells bearing CD38 surface markers

Summary

Researchers at UFRJ investigated why NAD+ levels fall in the brains of Zika-infected neonatal mice — a model mimicking third-trimester human infection. Using a detailed time-course analysis, they found that NAD+ depletion is a late-stage event, emerging one week after peak viral replication. Early induction of antiviral PARPs (PARP10, PARP12) tracked viral load but not NAD+ decline. Instead, NAD+ loss coincided precisely with rising CD38 expression and enzymatic activity, driven by infiltration of CD38-positive immune cells — particularly lymphocytes — into the brain. Blocking CD38 pharmacologically prevented NAD+ loss, identifying CD38 as the primary NAD+-depleting enzyme in this context and a potential therapeutic target for Congenital Zika Syndrome.

Detailed Summary

Zika virus (ZIKV) infection during pregnancy can devastate fetal brain development, causing microcephaly and a constellation of abnormalities known as Congenital Zika Syndrome (CZS). Prior work showed that NAD+ metabolism is disrupted in fetal brains of infected mice and that supplementing with the NAD+ precursor nicotinamide riboside reduces neuronal death, but the mechanisms driving NAD+ depletion were unknown. This study was designed to fill that gap.

Researchers infected postnatal day 3 (P3) mice subcutaneously with ZIKV and harvested brains every three days up to 30 days post-infection (dpi). NAD+ levels were stable through the first two weeks but dropped significantly at 18–30 dpi, long after viral replication peaked at 12 dpi. The temporal dissociation between viral burden and NAD+ decline suggested an indirect mechanism.

The team systematically tested candidate NAD+-consuming enzymes. PARP10 and PARP12 — antiviral mono-ADP-ribosyl transferases — were powerfully induced during early infection, tracking viral RNA levels with strong statistical correlation, but their expression peaked before NAD+ fell and declined before the nadir of NAD+. PARP1 showed only minor, transient induction. Simultaneously, NAMPT (the rate-limiting salvage enzyme) was upregulated during active replication, suggesting the cell attempted to compensate for early NAD+ consumption. None of these patterns aligned with the later NAD+ decline.

CD38, a multifunctional ectoenzyme and major mammalian NADase, emerged as the critical driver. Both CD38 mRNA and protein rose at the same time points when NAD+ fell (18–30 dpi). CD38 enzymatic activity — measured by cyclic ADP-ribose hydrolase assay — mirrored this pattern. Crucially, pharmacological inhibition of CD38 with the specific inhibitor 78c prevented NAD+ loss in infected brains, directly implicating CD38 activity in the depletion. Flow cytometry revealed that the rise in CD38 expression coincided with infiltration of CD38-positive immune cells, predominantly lymphocytes, into the brain parenchyma, preceded by induction of inflammatory cytokines IL-6, TNF, and CCL5/RANTES.

Taken together, the data support a model in which ZIKV replication triggers early antiviral PARP induction and cytokine release, which then recruits CD38-high immune cells into the brain. These infiltrating cells drive sustained NAD+ hydrolysis at a scale that overwhelms NAMPT-dependent salvage, ultimately collapsing NAD+ homeostasis. This mechanistic clarity opens two therapeutic avenues: CD38 inhibitors (several already exist clinically) and NAD+ precursor supplementation, potentially in combination, to protect the developing brain from CZS-related damage.

Key Findings

  • Brain NAD+ levels fell significantly at 18–30 dpi, at least one week after peak ZIKV replication at 12 dpi.
  • PARP10 and PARP12 tracked viral load closely but peaked before NAD+ declined, ruling them out as primary drivers.
  • CD38 mRNA, protein, and enzymatic activity rose in parallel with NAD+ depletion at late infection stages.
  • Pharmacological CD38 inhibition (compound 78c) prevented NAD+ loss in infected brains.
  • Neuroinflammation and infiltration of CD38+ lymphocytes preceded and accompanied the NAD+ decline.

Methodology

P3 neonatal mice were subcutaneously infected with ZIKV and brains harvested every 3 days through 30 dpi. Measurements included total NAD+ (enzymatic cycling assay), absolute RT-qPCR for viral RNA and gene expression, CD38 enzymatic activity (cADPR hydrolase), flow cytometry for immune cell infiltrates, and pharmacological CD38 inhibition with compound 78c.

Study Limitations

The study uses neonatal mouse infection as a proxy for third-trimester human fetal exposure, which may not fully recapitulate transplacental ZIKV pathogenesis. The causal role of immune cell-derived CD38 versus resident brain cell CD38 is not yet resolved. All pharmacological inhibition data are from mouse models, and translational relevance to human CZS requires further validation.

Enjoyed this summary?

Get the latest longevity research delivered to your inbox every week.