Home > Articles > CD38 and NAD+ decline: what enzyme and mouse knockout studies show
CD38 and NAD+ decline: what enzyme and mouse knockout studies show
Prepared by: Research Desk
Prepared by: Research Desk. Sources are linked for verification.
Most explanations of falling NAD+ levels focus on supply: cells make less, so levels drop. A second explanation focuses on demand: something is consuming NAD+ faster. Since 2016, the enzyme most often named in that second explanation is CD38. This explainer covers what CD38 does chemically, what the key mouse studies reported, how its by-products connect back to NAD+ recycling, and where the evidence stops. For the wider picture of NAD+ biochemistry, see our NAD+ guide.
What CD38 is
CD38 was first described as a surface marker on immune cells. Early immunology work studied it as a molecule involved in lymphocyte activation and signalling [1]. Only later did it become clear that CD38 is also an enzyme, and a very active one, whose main substrate is NAD+.
CD38 belongs to a small family of ADP-ribosyl cyclases. In much of its biology, its catalytic domain faces the outside of the cell, so it works as an "ectoenzyme" [2]. That orientation created what researchers called a "topological paradox": the enzyme sits outside the cell, yet the NAD+ it acts on and the calcium-signalling messenger it produces are mostly needed inside the cell [3]. Later work described transport routes and cell-to-cell (paracrine) signalling that partly resolve the puzzle [4][2].
The chemistry: what CD38 does to NAD+
CD38 is multifunctional. Working from NAD+, it can catalyse three related reactions [4][2]:
Hydrolysis (glycohydrolase activity). NAD+ is split into nicotinamide and ADP-ribose (ADPR). This is the dominant reaction and the one that matters most for NAD+ consumption.
Cyclisation (ADP-ribosyl cyclase activity). NAD+ is converted into cyclic ADP-ribose (cADPR), a messenger that releases calcium from internal stores. Only a small fraction of NAD+ takes this route.
cADPR hydrolysis. CD38 can also break cADPR down into ADPR.
Two features make CD38 important for NAD+ levels. First, it is a consuming enzyme: each reaction removes one NAD+ molecule, unlike the redox cycle in which NAD+ and NADH swap back and forth. Second, the hydrolysis reaction is inefficient as a signalling step. Many NAD+ molecules are broken down for each cADPR produced, which is why CD38 has been described as a large NAD+ sink rather than mainly a signalling enzyme.
The 2016 knockout study
The study that put CD38 at the centre of the NAD+ aging debate was Camacho-Pereira and colleagues, published in Cell Metabolism in 2016 [5]. Working in mice, the authors reported that:
CD38 levels and activity rose with age in several mouse tissues.
Mice genetically lacking CD38 (knockout mice) did not show the same age-related fall in tissue NAD+.
Knockout mice kept mitochondrial function closer to young-animal levels, through a mechanism the authors linked to SIRT3, a mitochondrial sirtuin that depends on NAD+.
A commentary published alongside it summarised the argument in its title: NAD+ declines during aging because "it's destroyed" [6]. That framing shifted attention from NAD+ synthesis toward NAD+ consumption.
It helps to be precise about study type. This was animal research using a whole-body genetic deletion. A knockout removes CD38 from every cell for the animal's entire life, which is not the same as a partial or late-life change, and it cannot show what happens in human tissue.
Where the CD38 comes from: macrophages and senescent cells
If CD38 rises with age, which cells are making it? A 2020 Nature Metabolism mouse study by Covarrubias and colleagues pointed to immune cells [7]. The authors reported that senescent cells, which accumulate with age and release inflammatory factors, prompted macrophages to multiply and express CD38. CD38-positive, inflammation-associated macrophages accumulated in tissues such as visceral fat and liver of aged mice, and the authors linked their NAD+-degrading activity to tissue NAD+ decline.
Several later rodent studies extended the macrophage-CD38 idea to specific tissues and conditions:
Cholesterol and neurodegeneration (2024). Terao and colleagues reported that cholesterol activated an LXR/CD38 pathway in macrophages, driving macrophage senescence and NAD+ depletion in their neurodegeneration models [8].
Kidney injury (2025). Yao and colleagues described CD38-high macrophages that promoted a fibrotic transition after acute kidney injury in their models, by modulating NAD+ metabolism [9].
Endometrial aging (2026). Hua and colleagues reported that CD38 from myeloid (bone-marrow-derived immune) cells mediated age-related endometrial changes through NAD+ depletion [10].
Diet and fat tissue (2026). Yang and colleagues reported that long-term high-protein diet intake accelerated adipocyte senescence through macrophage CD38-mediated NAD+ depletion [11].
Together these papers form a consistent pattern in animal and cell work: inflammatory macrophages expressing CD38 appear in aging or stressed tissue, and NAD+ falls in that tissue. A consistent pattern in rodents is still not a demonstration in people.
Nicotinamide: the by-product that gets recycled
CD38's main reaction releases nicotinamide. Nicotinamide is the starting point of the salvage pathway, in which the enzyme NAMPT converts it to NMN, and NMN is then converted back to NAD+. In other words, NAD+ broken down by CD38 is not entirely lost: part of it can be rebuilt.
This matters for interpreting NAD+ measurements. Tissue NAD+ at any moment reflects a balance between consumption (CD38, PARPs, sirtuins) and resynthesis, mainly through salvage. Higher CD38 activity raises consumption, and NAMPT capacity sets how quickly the pool can be refilled. Our explainer on the NAD+ salvage pathway walks through each step. A fall in NAD+ could reflect more consumption, slower salvage, or both, and a single NAD+ number cannot distinguish them.
The gap between mouse knockouts and human aging
Several limits apply before the CD38 model can be read as an account of human aging:
Species. The landmark findings come from mice [5][7]. Mouse and human immune systems and tissue NAD+ turnover differ.
Model design. Lifelong whole-body knockouts are blunt tools. CD38 also has immune-signalling roles [1][4], so removing it changes more than NAD+ consumption.
Measurement. Human studies usually measure NAD+ or its metabolites in blood, while the mouse findings concern tissues such as liver, fat and muscle. Blood values do not necessarily mirror tissue. Our page on reading an NAD+ test report explains why sample type matters.
Human data on decline itself. Whether, where and how much NAD+ falls with age in people is a separate question, covered in Does NAD+ decline with age?.
Intervention evidence. A 2026 PRISMA-guided systematic review of NAD+ supplementation brought together preclinical and clinical studies and illustrates how much of the field still rests on animal work [12]. None of the CD38 studies cited here tested anything in humans.
CD38 research concerns enzyme biology, not any product. For how NAD+-related products are classified, see Is NAD+ legal?.
Frequently asked questions
Is CD38 the only enzyme that consumes NAD+?
No. PARPs and sirtuins also consume NAD+, and other ADP-ribosyl cyclases exist [2]. CD38 drew particular attention because the 2016 mouse knockout study linked it directly to age-related tissue NAD+ decline [5].
Did the 2016 study show CD38 causes NAD+ decline in people?
No. The work was done in mice, comparing normal animals with CD38 knockout animals [5]. It established a strong animal-model hypothesis, not a finding about human aging.
Why are macrophages central to the CD38 story?
Mouse studies reported that senescent cells prompted macrophages to express CD38 and that these CD38-positive macrophages accumulated in aging tissue [7]. Later rodent work tied CD38-expressing macrophages to NAD+ depletion in kidney, endometrial, nervous and fat tissue models [8][9][10][11].
What happens to the nicotinamide CD38 releases?
It can re-enter the salvage pathway, where NAMPT converts it to NMN and then back to NAD+. Overall NAD+ levels therefore reflect a balance between consumption and recycling.
Further reading
Much of the CD38 literature assumes working knowledge of enzyme kinetics, nucleotide metabolism and immune-cell biology. A solid general text on cellular metabolism makes the primary papers far easier to read and evaluate.
When comparing biochemistry textbooks, useful criteria are: a recent edition (NAD+-consuming enzymes have been reassessed substantially since the mid-2010s), dedicated coverage of nucleotide and cofactor metabolism, clear pathway diagrams, chapter references to primary literature, and a format that suits the reader, such as a searchable ebook versus a print reference. Price transparency across editions and formats is also worth checking, since older editions are often much cheaper while covering the core chemistry well.
For a reading list focused on aging and sirtuin science, our evidence-graded book list grades titles by how closely they track published data.
How we choose sources and handle affiliate links is set out on our about page.
Related on this site
NAD+ explained: biochemistry, precursors, testing and human evidence
Reading an NAD+ test report: units, sample types and reference ranges
Last updated: 23 October 2026.
Not medical advice. This page reports published research and regulatory records for information only.
Research-only notice: Content on this site is provided for research and educational purposes only and is not intended for human use. It is not medical advice.
Translation notice: Some content on this site may have been machine-translated using Google Translate and may contain errors.