Tumor Metabolic Dependency and Target Validation

A dependency is not a sensitivity, and neither is a target.

Four things often reported as one

TermDefinitionEvidence required
ExpressionThe enzyme is present, often higher in tumour than normalTranscript or protein, with a reference tissue
SensitivityInhibiting it reduces viabilityDose–response, one culture condition
DependencyThe cell requires the activity to grow or surviveGenetic and pharmacological concordance; rescue
TargetInhibiting it produces a therapeutic marginEffect in vivo at achievable exposure, with normal-tissue tolerance

Most published "metabolic targets" sit at row two. The distance from row two to row four is where programmes fail, and each row requires a different experiment.

Gate 1 — Is this the right node?

QuestionWhy it decides the outcome
Which isoform?Isoform redundancy buffers single-enzyme inhibition; pan-isoform tools obscure it
Which compartment?Cytosolic and mitochondrial pools of the same metabolite are separate
Is the enzyme a control point?Control is distributed; the enzyme with the largest expression change is often not the one that sets flux
Is there a bypass?Substrate-level alternatives (e.g. exogenous serine, salvage nucleotides) can make the node irrelevant

Important distinction. High expression is a hypothesis about a node, not evidence that the node controls flux. Control analysis in central metabolism repeatedly shows that no single step holds most of the control, which is why an enzyme can be 90% inhibited with a small flux change.

Caveat on isoform coverage. A tool compound that inhibits one isoform of a redundant pair will look inactive, and the conclusion "this pathway is dispensable" will be wrong. The genetic experiment — knocking out both — is what distinguishes redundancy from dispensability.

Gate 2 — Does inhibition change flux, not just the pool?

A metabolite pool is the integral of production minus consumption. It can fall because production fell, or because consumption rose.

MeasurementWhat it supportsWhat it does not
Steady-state poolThe concentration changedDirection of the flux change
Labelled precursor incorporationRoute and rate of entryAbsolute flux without a model
Product secretion rateNet pathway outputWhich step changed
Rescue with the downstream productThe pathway, not the enzyme, is requiredThat the enzyme is the target
Rescue with the upstream substrateThe block is at or before that stepWhere exactly

Readout note. For a drug programme the useful measurement is the product, not the substrate. Inhibiting a biosynthetic enzyme raises its substrate and lowers its product; only the product change maps onto the biological consequence. A programme that reports substrate accumulation alone has shown target engagement, not pathway inhibition.

Gate 3 — Does the tumour depend on it where the tumour lives?

Dependency is a property of the environment as much as of the genome.

EnvironmentWhat changesWhy it matters
Culture mediumGlucose, glutamine, serine, lipids at supraphysiological concentrationsRemoves the restriction that creates dependency
Two-dimensional versus spheroidNutrient gradients, oxygen, proliferation gradientAverages over heterogeneity
XenograftMouse plasma composition, host metabolism, deliveryThe relevant comparison for a drug
Genetic modelIntact immune system, native stromaDependency may be immune-mediated

Important distinction. Many metabolic dependencies are medium artefacts in the specific sense that they exist only when the relevant nutrient is limiting. Serine synthesis, for example, is dispensable in standard medium that contains serine and becomes essential when serine is withdrawn. Both statements are true; only one is relevant to a patient, and which one depends on the tumour's vascular supply — which is the measurement nobody makes.

Gate 4 — Is there a window?

Normal tissueWhy it is at riskHow it limits the target
Bone marrowHigh proliferative flux; nucleotide and one-carbon demandAntifolates and nucleotide inhibitors
Intestinal epitheliumRapid turnover, high glutamine useGlutaminase inhibitors
Activated immune cellsGlycolytic and glutamine-dependent on activationImmunosuppression as on-target toxicity
LiverCentral carbon and nitrogen hubHepatotoxicity at effective exposure

Important distinction. The most attractive metabolic targets are those required by the tumour and not required by the adult organism. Nucleotide salvage versus de novo synthesis is the clearest example: adults can salvage, many tumours lose the salvage route, so the de novo pathway becomes tumour-selective. This is a stronger argument than differential expression, because it is a difference in requirement rather than in abundance.

Genotype-selected vulnerability

The strongest version of a metabolic target is one where the tumour's genotype creates the dependency, because the genotype is also the biomarker.

LesionConsequenceTherapeutic logic
MTAP deletion (with CDKN2A)MTA accumulates; MAT2A and PRMT5 axis stressedMAT2A inhibition collapses the axis further
Somatic SDH or FH lossOncometabolite accumulationDownstream dependencies, not the enzyme itself
Isocitrate dehydrogenase mutation2-HG productionThe mutant enzyme, not the pathway
VHL lossConstitutive HIF signallingContext-dependent; usually not a metabolic node

Caveat on the biomarker. Genotype-selected programmes require the biomarker assay to be validated in the same tissue and platform as the trial. A deletion call from a targeted panel and a call from whole-genome sequencing can disagree at the boundary, and the disagreement falls exactly on the patients who would be enrolled.

The validation sequence, in the order that saves money

  1. Isoform and compartment resolved genetically — knockout both isoforms, or use compartment-specific perturbation.
  2. Flux, not pool — product secretion or labelled incorporation.
  3. Rescue with the product — establishes that the pathway, not a downstream side effect, is required.
  4. Two culture conditions — permissive and restricted medium; report both.
  5. In vivo at achievable exposure — with plasma concentration measured, not assumed.
  6. Normal-tissue tolerance — at least one proliferating normal tissue.

Steps 1 and 3 are cheap and eliminate most candidates. Step 6 is expensive and is the one most often deferred to a point where it cannot change the decision.

Failures that recur

  • • Sensitivity called dependency. One cell line, one medium, one compound.
  • • Compound-only evidence. No genetic arm, so off-target effects are indistinguishable.
  • • Rescue with a downstream metabolite that also feeds other pathways. Succinate, pyruvate and aspartate all do this.
  • • Supraphysiological inhibitor concentrations used to define the phenotype, then a nanomolar compound advanced.
  • • Dependency measured in medium that contains the nutrient in question.
  • • Window asserted from expression data. Expression is not requirement.

Where this connects

  • • Glucose, lipid, amino acid and mitochondrial pillars supply the node-level mechanisms and the flux rules this gate sequence assumes.
  • • Disease pillar supplies the host-state frame: a dependency measured in a lean, young xenograft host may not exist in the metabolic state of the patients who have the disease.
  • • Synthetic lethality — the genetic-interaction logic behind the genotype-selected row — is treated in the companion articles on metabolic synthetic lethality and on designing the screen.

Scientific interpretation and experimental design

  • • Expression, sensitivity, dependency and target are four different claims.
  • • Resolve isoforms and compartments. Redundancy is the default.
  • • Measure the product. Substrate accumulation is engagement, not inhibition.
  • • Report the medium. Dependency is often a property of the medium.
  • • Genetic and pharmacological arms together. Neither alone excludes off-target effects.
  • • Prefer requirement differences to abundance differences. Salvage loss beats overexpression.
  • • Measure exposure in vivo. Achievable concentration is part of the claim.

Selected references for further reading

  • • Vander Heiden MG. Targeting cancer metabolism: a therapeutic window opens. Nature Reviews Drug Discovery. 2011;10:671–684.
  • • Vander Heiden MG, DeBerardinis RJ. Understanding the intersections between metabolism and cancer biology. Cell. 2017;168:657–669.
  • • Kryukov GV, et al. MTAP deletion confers enhanced dependency on the PRMT5 arginine methyltransferase in cancer cells. Science. 2016;351:1214–1218.
  • • Mavrakis KJ, et al. Disordered methionine metabolism in MTAP/CDKN2A-deleted cancers leads to dependence on PRMT5. Science. 2016;351:1208–1213.
  • • Sullivan LB, Gui DY, Vander Heiden MG. Altered metabolite levels in cancer: implications for tumour biology and cancer therapy. Nature Reviews Cancer. 2016;16:680–693.

Related Products

Product name Cat.No. Species Gene ID
PBRM1 Knockout A-549 Cell Line EDJ0001-K01 Human 55193 Details Get a Quote
YTHDC1 Knockout A-549 Cell Line EDC07652 Human 91746 Details Get a Quote
TP53 Knockout HCT 116 Cell Line EDC07854 Human 7157 Details Get a Quote
Sirt2 Knockout MH-S Cell Line EDJ-KQ78173 Mouse 64383 Details Get a Quote
TFE3 Knockout KGN Cell Line EDJ-KQ48 Human 7030 Details Get a Quote
Stub1 Knockout MB49 Cell Line EDJ-KQ53 Mouse 56424 Details Get a Quote
ARID1A Knockout SNK-6 Cell Line EDJ-KQ64 Human 8289 Details Get a Quote
Tet1 Knockout Kupffer Cell Line EDJ-KQ88 Mouse 52463 Details Get a Quote
NR1H3 Knockout HEK293T Cell Line EDJ-KQ109 Human 10062 Details Get a Quote
NR1H2 Knockout HEK293T Cell Line EDJ-KQ110 Human 7376 Details Get a Quote
RPS6KB2 Knockout HEK293 Cell Line EDJ-KQ125 Human 6199 Details Get a Quote
SMAD6 Knockout HEK293 Cell Line EDJ-KQ126 Human 4091 Details Get a Quote
JUN Knockout HEK293 Cell Line EDJ-KQ176 Human 3725 Details Get a Quote
JUN Knockout HEK293T Cell Line EDJ-KQ184 Human 3725 Details Get a Quote
ALKBH5 Knockout HEK293 Cell Line EDJ-KQ186 Human 54890 Details Get a Quote
Contact Us
*
*
*
How did you hear about us: