If you buy NMN or NR to refill your NAD+ tank, a new preclinical study asks an awkward question: what if the tank has a leak? The leak in question is an enzyme called CD38. In rodents, a drug that blocks it raised NAD+ in the brain and eased movement problems in a Parkinson's disease model.
Before anyone clears a shelf of supplements, here is the catch. The compound is CVN14, an experimental drug from Cerevance. It is not a supplement, it is not available to consumers, and the work was done in rats, mice and cell cultures, not people.
What is the 'NAD+ thief'?
NAD+ is a cofactor that cells need for many core processes. It tends to decline with age, and the authors point to two causes: less NAD+ is made, and more is used up.
CD38 sits on the consumption side. It is an NAD+-consuming enzyme (an NADase), and its expression rises with ageing and disease. The researchers argue that this likely contributes to the age-related drop in NAD+. They also note genetic links between CD38 and Parkinson's disease (PD), which gives them a reason to study it in that context.
Using NETSseq datasets from human post-mortem brain tissue, the team found that CD38 is highly enriched in astrocytes, the star-shaped support cells of the brain. Neurons are not the main place this enzyme is found.
The key finding: block CD38, brain NAD+ goes up
CVN14 is a brain-penetrant, selective CD38 inhibitor that the group had described previously. In this paper they characterised it in detail.
- Potency: It inhibited recombinant CD38 with IC50 values of 19.0 nM (human), 5.1 nM (rat) and 2.4 nM (mouse). Lower numbers mean more potent, so it was strongest against the mouse enzyme and weakest, though still potent, against the human one.
- Selectivity: It had no activity at CD157, the closest CD38 relative, or at other NADases.
- Brain NAD+: It increased brain NAD+ in both rats and mice.
- Knockout control: It had no effect in CD38 knockout mice. This supports the idea that the NAD+ rise comes from blocking CD38 and not from some other NAD-chewing enzyme.
The abstract does not give the size of the NAD+ increase, so no percentage claims are made here.
The Parkinson's angle
The researchers then tested whether this NAD+ boost mattered in models of Parkinson's disease, which is marked by the buildup of misfolded α-synuclein protein.
In cell culture: In rat mesencephalic cultures challenged with α-synuclein, CVN14 showed evidence of increasing the survival of dopaminergic neurons. Those are the cells that die in PD. It also reduced the number of neurons positive for pSer129 α-synuclein, a marker of pathological protein. The authors say this supports further investigation of a possible role for CD38 inhibition in clearing pathological protein. The wording is cautious, and it should be read that way.
In rats: In a rat model using AAV1/2-hA53T α-synuclein, CVN14 improved motor function in the cylinder test. That is one behavioural readout, in one animal model.
Read this before you get excited
The limits are substantial:
- Not human data. Everything here comes from rats, mice, and cultured cells, plus human post-mortem tissue used only to map where CD38 is expressed.
- Industry-funded. All authors were employed by Cerevance Ltd, which is developing the compound. That is not a reason to dismiss the work, but it matters for how you weigh it.
- One behavioural test. The motor benefit was shown in the cylinder test. The conclusions stop at supporting "further investigation."
- Correlation within the study. The paper shows brain NAD+ rising and neuroprotective signals appearing in the same system. The abstract does not prove that the NAD+ rise alone drives the benefit.
- No dosing for humans. Nothing here tells you a dose, timing, or safety profile for any person.
What it means for NMN and NR buyers
This paper does not test NMN, NR, or any supplement. It is still useful for thinking about how NAD+ strategies might work in future. Here are the practical takeaways.
1. NAD+ is a supply-and-demand problem. Precursors like NMN and NR address the supply side. CD38 inhibition addresses the demand side. If CD38 is a major drain in an ageing brain, then boosting supply alone may be only half of the equation. That is a hypothesis from this line of research, not a proven fact for supplement users.
2. A combined approach is an idea to watch, not a protocol. The logic of pairing a precursor with a CD38 blocker is appealing. This study did not test it, and there is no human evidence here.
3. Be careful with 'natural CD38 inhibitor' claims. Compounds such as apigenin and quercetin are often described online as CD38 inhibitors. That claim is not from this paper, and their effects on brain NAD+ in people remain unproven. CVN14 was designed to be selective and brain-penetrant. There is no reason to assume a food-derived flavonoid does the same job.
4. Who is this for? Right now, nobody outside a research setting. The stated target is Parkinson's disease and other α-synucleinopathies, and the compound is still preclinical. Anyone with a neurological condition should work with a physician and not try to self-treat based on rodent data.
5. Dose and timing. There are no human dose or timing recommendations to give, because none exist for this compound. Be wary of anyone who claims otherwise.
Bottom line
The study makes a clear case that CD38, concentrated in astrocytes in the human brain, is a credible NAD+ drain worth targeting. A selective inhibitor raised brain NAD+ in rodents, and it did so only when CD38 was present. In Parkinson's models it was linked with better dopaminergic neuron survival in culture, lower pathological α-synuclein signals, and improved motor performance in rats. Whether any of this carries over to humans is the question for future trials.
For now, the main lesson for NAD+ enthusiasts is conceptual: how fast NAD+ is consumed may matter as much as how much precursor you take.