Monday, October 5, 2026

The Oncometabolite That Needs No Mutation

The Oncometabolite That Needs No Mutation

Succinate, fumarate and 2-hydroxyglutarate earned the name by disabling the same family of enzymes — and each one requires a mutation to accumulate. Ammonia reaches comparable endpoints with no mutation at all, and disables natural killer cells, dendritic cells and T cells by a mechanism so plain it needs nothing but a pH gradient.

This series began by asking where a tumour's surplus nitrogen goes, and spent four articles finding that interventions aimed at it keep failing for one reason: the tumour has more ways out than the immune cell does. That conclusion held across alanine, glutamine, methionine and arginine.

What it did not address is the thing left behind when the disposal routes are working — the fraction of nitrogen that stays as free ammonia. The literature on that has moved quickly, and it is worse than the series assumed.

Ammonia is a weak base. It diffuses through membranes uncharged and raises the pH of any acidic compartment it reaches. That single physical property disables perforin maturation in NK cells, antigen processing in dendritic cells, and lysosomal integrity in T cells — three separate arms of anti-tumour immunity, failing by one mechanism that requires no receptor, no transporter and no mutation. It also depletes the molecule that oncometabolites compete for. We have been classifying it as waste.

Part I / One Mechanism, Three Failures

What a Weak Base Does to an Immune Cell

The clearest result came out of a question nobody in cancer metabolism was asking: why do NK cells in tumours conjugate with their targets, degranulate normally, and then fail to kill?

The answer is perforin. Perforin is made as a 70 kDa precursor and matured to its active 60 kDa form inside secretory lysosomes, and that processing requires an acidic compartment. Ammonia is lysosomotropic — it enters, becomes protonated, and raises the pH. At 2 to 5 mM, concentrations found in real tumour microenvironments, it impairs NK cytotoxicity dose-dependently; at 4 to 5 mM it abolishes killing entirely.

And it blunts the antibodies too. The same work found ammonia suppressed antibody-dependent cellular cytotoxicity with rituximab, daratumumab and trastuzumab — three drugs whose mechanism depends on NK cells doing the killing — and suppressed CAR-NK cell activity as well. The cells arrive, engage, and discharge granules that cannot lyse anything.

Dendritic cells fail by the same physics in a different compartment. Antigen processing, peptide loading and MHC assembly all happen in acidic endosomes and lysosomes. Ammonia at 5 mM drives dendritic cells into dysfunction — diminished phagocytosis, cell swelling, excessive reactive oxygen species, mitochondrial damage, and reduced capacity to stimulate lymphocytes. Dendritic cells taken from patients with cirrhosis showed the same phagocytic defect as healthy cells treated with ammonia in vitro.

T cells complete the set. Ammonia infiltrates mitochondria and lysosomes, disrupts mitochondrial integrity and raises lysosomal pH, which is the proposed basis of what the field has begun calling ammonia-induced cell death — a selective death of immune cells rather than tumour cells.

One weak base, three acidic compartments, three immune failures Ammonia diffuses across membranes as uncharged NH3 and becomes protonated inside acidic compartments, raising their pH. In NK cells this raises the pH of secretory lysosomes and blocks the maturation of perforin from its 70 kilodalton precursor to the active 60 kilodalton form, so the cells engage and degranulate but cannot kill. In dendritic cells it alkalinizes endosomes and lysosomes, impairing antigen processing and peptide loading, so phagocytosis and lymphocyte stimulation fall. In T cells it disrupts mitochondrial integrity and raises lysosomal pH, leading to cell death. No receptor, transporter or mutation is involved. One weak base, three acidic compartments NH₃ crosses membranes uncharged, is protonated inside, and raises the pH of whatever it enters Ammonia, 2–5 mM lysosomotropic; no receptor needed NK cell secretory lysosome Perforin stays at 70 kDa instead of maturing to 60 engages and degranulates, then fails to kill Dendritic cell endosome · lysosome Antigen processing and peptide loading impaired phagocytosis down, T cell priming fails T cell lysosome · mitochondria Mitochondrial integrity lost, lysosomal pH rises exhaustion, and a selective cell death No mutation is required at any point in this diagram One weak base, three acidic compartments, three immune failures A vertical version. Ammonia at 2 to 5 millimolar diffuses in uncharged and raises the pH of acidic compartments. In NK cells perforin stays at 70 kilodaltons instead of maturing to 60, so the cells degranulate but cannot kill. In dendritic cells antigen processing and peptide loading are impaired, so phagocytosis and T cell priming fail. In T cells mitochondrial integrity is lost and lysosomal pH rises, causing exhaustion and selective death. No mutation is required. One weak base, three compartments NH₃ crosses membranes uncharged and raises the pH inside Ammonia, 2–5 mM lysosomotropic; no receptor needed NK cell secretory lysosome Perforin stays at 70 kDa, never matures to 60 degranulates, cannot kill Dendritic cell endosome · lysosome Antigen processing and peptide loading impaired priming fails T cell lysosome · mitochondria Mitochondrial integrity lost, lysosomal pH rises exhaustion and death No mutation required anywhere above
Three arms of anti-tumour immunity, three different compartments, one piece of physical chemistry.

The series had already shown the tumour holds the enzyme the T cell lacks in exchange after exchange. This is something else. It is not a competition the immune cell loses on the margin — it is a chemical property of the microenvironment that switches off the killing machinery of three cell types at once.

With one encouraging detail. Both the NK and dendritic cell defects are reversible. Perforin returned to control levels after sixteen hours of ammonia washout, and dendritic cell dysfunction reversed on ammonia elimination. These are not damaged cells. They are working cells in a disabling medium — which is precisely the kind of problem that clearance, rather than blockade, is suited to.

Part II / The Cycle That Sustains It

Ammonia and Lactate Hold Each Other in Place

Two waste products, usually discussed in separate literatures, turn out to be mutually reinforcing — and the direction of the reinforcement is not what I expected when I started looking.

Ammonia pushes cells toward glycolysis. Fixing ammonia onto α-ketoglutarate consumes a TCA cycle intermediate, and draining it stalls the cycle. In the clearest demonstration of this, mitochondrial respiration fell within hours and the effect was not attributable to pH. A cell in that state has oxygen it cannot use, and glycolysis is the only ATP source left — which produces lactate.

And lactate's acidity protects the tumour from the ammonia. This is the part that completes the circuit. Ammonia and ammonium interconvert with a pKa near 9.25, so at the acidic extracellular pH that lactate efflux creates — 6.5 to 6.9 against a normal 7.4 — ammonia sits predominantly as charged NH4+, which does not cross membranes. The acidic rind is a shield. The recent literature states it directly: the acidic microenvironment protects cancer cells from ammonia toxicity.

ammonia → α-KG drained → TCA stalls → glycolysis → lactate
lactate → acidosis → NH3 held as NH4+ outside → tumour protected

So the tumour generates the conditions that make its own nitrogen waste tolerable to itself. Whether immune cells in the same space are equally protected is the question nobody has answered — but they are the cells whose function depends on maintaining acidic compartments against a weak base, and they are the ones demonstrably failing at these concentrations.

Which is an unresolved problem for buffer therapy. Neutralizing tumour acidity improves responses to checkpoint blockade and adoptive transfer — that is well demonstrated. But raising extracellular pH also converts NH4+ back to diffusible NH3, and diffusible NH3 is what disables perforin maturation and antigen processing. Bicarbonate could relieve one immunosuppressive mechanism while amplifying another. It could also strip the tumour's own protection. Nobody has measured both at once in the same experiment, and until somebody does, combining buffer therapy with a high ammonia burden is a genuinely open question rather than an obviously additive one.

Part III / The Classification Problem

Why This Belongs in the Oncometabolite Category

Succinate, fumarate and 2-hydroxyglutarate are called oncometabolites for a specific reason: they inhibit the α-ketoglutarate-dependent dioxygenases, producing a pseudohypoxic state and widespread epigenetic change in a cell with adequate oxygen. Each of them accumulates because of a mutation — in SDH, in FH, in IDH1 or IDH2.

Ammonia arrives at a comparable endpoint by a different route. It does not compete with α-ketoglutarate at the enzyme's active site. It removes α-ketoglutarate, by being fixed onto it. The dioxygenases lose their co-substrate rather than being outcompeted for it, and the prolyl hydroxylases that act as the cell's oxygen sensor cannot function without it.

The recognized oncometabolites

Succinate, fumarate, 2-HG compete at the α-KG site of the dioxygenases.

Each requires a mutation in a TCA or IDH enzyme to accumulate.

Ammonia

Consumes α-KG outright when fixed onto it, removing the co-substrate.

Requires no mutation. Any cell with a heavy nitrogen load and an inadequate exit generates it.

That difference is the argument. The oncometabolite concept is built around mutations that cause a metabolite to accumulate. Ammonia accumulates from ordinary metabolic flux — glutaminolysis, amino acid catabolism, the microbiota — in any tumour whose disposal capacity has not kept pace with its appetite. It needs no enabling lesion.

And the Framing That Misses It

Cancer is usually introduced as a genetic disease: changes in DNA that disrupt normal growth and division. That framing has been extraordinarily productive and it is not wrong. But it has a known incompleteness, and the incompleteness is not a fringe position — it is in the field's own canonical review.

Hanahan's 2022 update to the hallmarks added nonmutational epigenetic reprogramming as an enabling characteristic, explicitly distinct from genomic instability, and unlocking phenotypic plasticity as a hallmark capability. He notes that certain pediatric tumours lack recurrent mutations yet show aberrant growth governed by a gene regulatory program induced by hypoxia, and that the physical properties of the microenvironment can cause broad epigenomic change. He singles out α-ketoglutarate and oncometabolites as regulators of differentiation state without mutation.

So the refinement is already sanctioned; what is missing is an obvious candidate. If α-ketoglutarate availability governs differentiation state non-mutationally, then anything that depletes α-ketoglutarate is a non-mutational driver. Ammonia depletes it, is present at millimolar concentrations in tumours, needs no enabling mutation, and separately disables three arms of immune surveillance by a mechanism requiring nothing but a pH gradient. On the existing criteria it qualifies. It has mostly been classified as waste because waste is what nitrogen metabolism was assumed to produce.

The honest version of the claim is narrower than "cancer is a metabolic disease," and I would not make that broader one — the genetic account explains too much to discard, and metabolic reprogramming in most tumours is downstream of oncogenic signalling. The narrower claim is this: a genuinely causal layer of cancer biology operates below the level of mutation, the field has formally acknowledged it, and the most abundant candidate in that layer has been sitting in the discard pile.

Part IV / What Follows Practically

The Rule the Series Arrived At

If the problem is a metabolite rather than a mutation, the intervention is removal of the metabolite rather than inhibition of an enzyme. That is also what four articles of failed enzyme inhibition independently suggested.

Every subtractive intervention in this series damaged the immune compartment first, because the tumour has a stroma, macropinocytosis, isoform redundancy and transporters, while the activated T cell has none of those and needs more of the nutrient in question, not less. Glutaminase inhibition impaired the CD8 response to anti-PD-1. Chronic methionine restriction abolished it. Meanwhile every intervention that added capacity worked.

The urea cycle with each intervention marked at the step it acts on A four-stage cycle running clockwise. Stage one, fixation: CPS1 combines ammonia and bicarbonate into carbamoyl phosphate, requiring N-acetylglutamate as an activator; carglumic acid and bicarbonate act here. Stage two, carry: OTC combines carbamoyl phosphate with ornithine to make citrulline; supplemental ornithine acts here. Stage three, convert: ASS1 and ASL turn citrulline and aspartate into arginine; supplemental citrulline acts here, and only cells retaining ASS1 can use it. Stage four, release: arginase cleaves arginine into urea, which leaves the body, and ornithine, which returns to stage two. A side branch shows ornithine decarboxylase diverting ornithine into polyamines, which DFMO blocks. Phenylbutyrate is shown outside the cycle, conjugating glutamine for urinary excretion. Where each agent acts on the nitrogen exit Green: adds capacity at that step. Red: a leak, and the agent that plugs it. STAGE 1 · FIX CPS1 NH₃ + HCO₃⁻ → carbamoyl phosphate needs N-acetylglutamate to run at all ▶ carglumic acid · bicarbonate STAGE 2 · CARRY OTC + ornithine → citrulline ornithine is the limiting carrier ▶ ornithine STAGE 3 · CONVERT ASS1 · ASL + aspartate → arginine only cells keeping ASS1 can use citrulline ▶ citrulline STAGE 4 · RELEASE Arginase → urea, which leaves the body → ornithine, which returns to stage 2 ornithine recycles Leak: ODC diverts ornithine to polyamines — DFMO blocks it Outside the cycle: phenylbutyrate consumes glutamine to carry N out The urea cycle with each intervention marked at the step it acts on A vertical version of the same four stages. Fix: CPS1 combines ammonia and bicarbonate into carbamoyl phosphate and needs N-acetylglutamate; carglumic acid and bicarbonate act here. Carry: OTC adds ornithine to make citrulline; supplemental ornithine acts here. Convert: ASS1 and ASL add aspartate to make arginine, and only cells keeping ASS1 can use citrulline; supplemental citrulline acts here. Release: arginase produces urea which leaves the body and ornithine which recycles. A leak diverts ornithine to polyamines via ODC, which DFMO blocks. Phenylbutyrate sits outside the cycle and consumes glutamine. Where each agent acts Green adds capacity at that step. Red is a leak and what plugs it. STAGE 1 · FIX CPS1 NH₃ + HCO₃⁻ → carbamoyl P needs N-acetylglutamate ▶ carglumic acid · bicarbonate STAGE 2 · CARRY OTC + ornithine → citrulline ▶ ornithine STAGE 3 · CONVERT ASS1 · ASL + aspartate → arginine needs ASS1, which tumours silence ▶ citrulline STAGE 4 · RELEASE Arginase → urea out · ornithine recycles back to stage 2 Leak: ODC → polyamines; DFMO Outside: phenyl- butyrate eats Gln
Four stages, each with its own limiting factor and its own agent. Only one entry on this map is a blockade.
Agent Step Adds or removes Evidence Main caveat
Carglumic acid CPS1 activation Adds — activates the fixation step Approved for hyperammonemia; no cancer data Entirely untested here; mechanism only
Bicarbonate CPS1 substrate; TME pH Adds — but acts mainly on acidity Mouse: improved anti-PD-1, anti-CTLA-4, adoptive transfer Raises diffusible NH3; see Part II
Ornithine OTC substrate Adds — the limiting carrier Mouse: +72% survival with anti-PD-L1 Diverts to polyamines without ODC blockade
DFMO Blocks ODC leak Adds — protects the carrier, strips polyamines Mouse: T-cell-dependent control; approved in neuroblastoma Not tested with any of the above
Citrulline ASS1 substrate Adds — selectively, where ASS1 is silenced Mouse: 427 → 198 mm³ added to anti-PD-1 No citrulline-only arm; no ASS1 stratification
Phenylbutyrate Outside the cycle Removes — consumes glutamine Mouse: reduced tumour size as monotherapy Depletes a nutrient T cells cannot replace; pan-HDAC inhibitor
Five agents add capacity. One removes a shared nutrient, which is the thing this series says not to do.

The Thing I Keep Returning To

An NK cell arrives at a tumour, recognizes its target, forms a conjugate, and releases its granules. Everything works. The cell does its job completely and the target survives, because a weak base got into the compartment where perforin is cut to size, and raised the pH by a fraction of a unit.

No mutation was involved. No receptor was engaged. Nothing was signalled. A small molecule — NH₃ — that the field has spent decades describing as metabolic waste walked through a membrane and switched off the killing machinery of the immune system's most direct weapon — reversibly, at concentrations that are ordinary inside a tumour.

That is not a detail at the edge of cancer biology. If it is right, it is close to the middle of it, and we have been looking past it because we classified it before we understood it.

Selected References

  1. Ammonia suppresses the antitumor activity of natural killer cells and T cells by decreasing mature perforin.
    Domagala J, Winiarska M, et al. Cancer Research. 85(13):2448 (2025) — Cancer Res; preprint. The 70 to 60 kDa processing block, the ADCC and CAR-NK suppression, and the sixteen-hour washout reversal.
  2. Ammonia drives dendritic cells into dysfunction.
    Luo C, Shen G, Liu N, et al. The Journal of Immunology. 193(3):1080–1089 (2014) — J Immunol. 5 mM NH4Cl; reduced phagocytosis and lymphocyte stimulation, reversed on ammonia elimination; cirrhosis patients' dendritic cells showed the same defect.
  3. Ammonia metabolism and ammonia-induced cell death: role in cancer therapy.
    Xu Y, Wang J, Wu A, Wang M. Cell Communication and Signaling. (2025) — Full text. Lysosomal pH and mitochondrial disruption in T cells; the acidic microenvironment protecting cancer cells from ammonia.
  4. Microenvironmental ammonia enhances T cell exhaustion in colorectal cancer.
    Bell HN, Shah YM, et al. Cell Metabolism. (2023) — PMC. Ornithine plus anti-PD-L1 raised survival 72%; lactulose did not work, locating the source in the epithelium.
  5. Hallmarks of cancer: new dimensions.
    Hanahan D. Cancer Discovery. 12(1):31–46 (2022) — DOI. Nonmutational epigenetic reprogramming as an enabling characteristic; α-ketoglutarate and oncometabolites regulating differentiation without mutation.
  6. Ammonia inhibits energy metabolism in astrocytes in a rapid and glutamate dehydrogenase 2-dependent manner.
    Drews L, Zimmermann M, Westhoff P, et al. Disease Models & Mechanisms. (2020) — PMC. Ammonia fixed onto α-ketoglutarate; respiration falling within hours, independent of pH.
  7. Neutralization of tumor acidity improves antitumor responses to immunotherapy.
    Pilon-Thomas S, et al. Cancer Research. 76(6):1381–1390 (2016) — PubMed. See also the published correction, PMID 28461565.
  8. Polyamine depletion inhibits tumor growth and reverses immunosuppression in syngeneic mouse tumor models.
    Keough MP, et al. Cancer Immunology Research. (2013) — Cancer Immunol Res
  9. Glutaminase inhibition impairs CD8 T cell activation in STK11-/Lkb1-deficient lung cancer.
    Best SA, Gubser PM, et al. Cell Metabolism. 34(6):874–887 (2022) — Full text
  10. Enhancement of anti-PD-1 immunotherapy in non-small cell lung cancer using arginine and citrulline supplementation.
    Miyamoto N, Yoshida M, Tsukumo S, et al. Journal of Thoracic Disease. 17(7):4814–4825 (2025) — Full text
at No comments:
Subscribe to: Posts (Atom)