Charles Ng, MD, PhD — Physician-Scientist, Memorial Sloan Kettering

Background

Metabolism is usually treated as the cell's power supply. I study it as a source of regulatory information — where a metabolite's stereochemistry, or the route by which ATP reaches the nucleus, determines which genes a cell can open and which fate it takes.

I did my M.D./Ph.D. at NYU, where my thesis work with Dan Littman established how the Cd4 locus is heritably silenced in cytotoxic T cells, using CD4 silencing as a lineage marker for epigenetic memory. I then trained as a physician-scientist at Weill Cornell, in anatomic pathology and the Physician-Scientist Training Program, where I remain a trainee. I work in Craig Thompson’s laboratory at Memorial Sloan Kettering at the interface of metabolism, chromatin, and antitumor immunity, supported by a Tow Foundation Postdoctoral Fellowship.

Research

In my postdoctoral work (Ng et al., Cell, 2026), I showed that D-alpha-hydroxybutyrate — a short-chain alpha-hydroxy acid of microbial origin — acts as a signaling molecule that switches CD8 T cells from glycolysis to fatty-acid-supported oxidative phosphorylation, builds a phosphocreatine reserve, and drives ATP-dependent chromatin remodeling at effector loci, enhancing antitumor function in both mouse and human T cells.

I follow ATP delivery rather than the provision of substrates such as acetyl-CoA for acetylation or SAM for methylation: blocking ATP synthase leaves DAHB-induced histone acetylation intact but abolishes effector gene induction.

Three questions organize the laboratory I am building: how ATP delivery to chromatin is regulated in space and time; how D-2-hydroxyacids set bioenergetics both within a cell and across the whole animal; and how metabolic stress signaling selects which effector genes a T cell expresses. Together they point toward new therapeutic strategies for antitumor immunity and metabolic disease.

Schematic of a CD8 T cell. D-alpha-hydroxybutyrate stimulates the mitochondrial electron transport chain, raising oxidative phosphorylation and fatty acid oxidation. Mitochondrial creatine kinase charges phosphocreatine, which passes into the nucleus where creatine kinase B transfers its phosphate to ADP to regenerate ATP. That ATP supports BAF-dependent chromatin remodeling and increased accessibility at effector loci, raising perforin, interferon gamma and TNF-alpha, driving cytotoxicity toward the tumor and away from exhaustion. — DAHB raises mitochondrial ATP production and expands the phosphocreatine reserve. Creatine kinase B regenerates ATP within the nucleus, supporting BAF-dependent chromatin remodeling at effector loci. From Ng et al., *Cell*, 2026.

Phosphagen systems as spatial ATP delivery

Chromatin remodeling consumes ATP, but ATP does not travel freely through a crowded nucleus. I am testing whether the creatine/phosphocreatine shuttle is what delivers it — mapping creatine kinase B and its partners at chromatin, reading nuclear ATP directly with targeted sensors, and then engineering orthogonal phosphagen circuits into T cells: a cell-engineering strategy that modifies energy supply rather than signaling.

Fluorescence micrographs of CD8 T cell nuclei expressing a nuclear-targeted ATP sensor, in T cell media, with a creatine kinase inhibitor, and with 2-deoxyglucose plus oligomycin. The ATP sensor signal falls progressively across the three conditions while the HaloTag reference is unchanged. Beside them, a scatter plot of the ATP sensor to HaloTag ratio for the same three conditions. — A nuclear-targeted ATP sensor. Depleting ATP with 2-deoxyglucose and oligomycin collapses the GFP channel while the HALO reference is unchanged — the readout reports ATP in the nucleus, not how much sensor is present.

D-2-hydroxyacids as signaling molecules

D-alpha-hydroxybutyrate acts through its stereochemistry rather than through its metabolism. The mirror-image L form does not substitute, and my data show the sensor is not LDHD, the enzyme that consumes it. I am identifying the protein that reads the D configuration — and following the same chemistry outside the immune system, where oral DAHB limits diet-induced weight gain through a mechanism that requires LDHD in the intestinal epithelium.

Three mice photographed side by side against a ruler: normal diet with saline, high-fat diet with saline, and high-fat diet with DAHB. The high-fat saline animal is markedly larger. Beside them, a schematic of DAHB crossing into an intestinal epithelial cell and being oxidised by LDHD to alpha-ketobutyrate. — The same chemistry outside the immune system. A, oral DAHB limits weight gain on a high-fat diet. B, the route under test: DAHB is oxidised by LDHD in the intestinal epithelium, and that signal — not the calories — is what changes energy balance.

Integration of energetic stress

Energetic stress engages several sensors at once — the integrated stress response, AMPK, mTORC1, the heat-shock arm — and my published data show the ISR alone does not account for which effector genes open. I am mapping which stress kinases are actually engaged rather than assuming which ones matter, and doing it cell by cell rather than in bulk, where opposing signals average away.

Schematic of a cell under metabolic stress. AMPK, the integrated stress response, mTORC1 and the heat shock response sit together above ATP, which is linked to them by a double-headed arrow, and an arrow runs from the group into the nucleus to effector genes. — AMPK reads the AMP/ADP to ATP ratio, mTORC1 reads amino acid sufficiency, the ISR converges on eIF2α, and the heat-shock arm acts through HSF1. ATP availability couples them to each other — how that coupling resolves into a single transcriptional decision is the open question.

Selected publications

Mitochondrial ATP promotes T cell differentiation through chromatin accessibility Ng C, Fung TS, Li D, Kropp KN, Somarribas Patterson LF, Markovitz A, Weinberg DN, Jones O, Kim JY, Zhang G, … Thompson CB Cell 2026. 10.1016/j.cell.2026.08.023 · PMID 42727574 · Preprint: 10.64898/2026.03.27.714789
Lactate activates the mitochondrial electron transport chain independently of its metabolism Cai X, Ng CP, Jones O, Fung TS, Ryu KW, Li D, Thompson CB Molecular Cell 2023;83:3904–3920.e7. 10.1016/j.molcel.2023.09.034
The histone chaperone CAF-1 cooperates with the DNA methyltransferases to maintain Cd4 silencing in cytotoxic T cells Ng C, Aichinger M, Nguyen T, Au C, Najar T, Wu L, Mesa KR, Liao W, Quivy JP, Hubert B, Almouzni G, Zuber J, Littman DR Genes & Development 2019;33:669–683. 10.1101/gad.322024.118

Complete list: ORCID · Google Scholar