BGPT: Paper Review: Oncogenic KRAS supports pancreatic cancer through regulation of nucleotide synthesis

Explore by Goal

Quick Explanation Copied

Core claim (from the paper): Oncogenic KRAS–driven pancreatic cancer survival correlates with metabolomic programs that preserve nucleotide pools—via a MAPK→MYC→RPIA axis controlling non‑oxidative PPP—so KRAS/MEK-inhibition resistance is linked to maintained nucleotide synthesis; targeting PPP (RPIA) or pyrimidine synthesis (DHODH) disrupts growth.

Long Explanation

Paper Review (visual-first): KRAS supports PDAC via nucleotide synthesis

Target paper: 10.1038/s41467-018-07472-8

1) One-screen mechanistic map (what they measured → what they conclude)

All nodes/edges summarize the paper’s experimental chain of reasoning and measured readouts.

2) What the authors actually did (and why it matters)

KRAS “dependence” stratification: They classify PDAC cell lines into KRAS-sensitive vs KRAS-resistant based on differential loss of clonogenic growth/viability after lentiviral KRAS shRNA, with KRAS-resistant lines showing less apoptosis and more survival.
Metabolomic readouts: After KRAS extinction they perform targeted LC-MS/MS to track glycolytic intermediates, PPP intermediates, and pyrimidine nucleotides; they also measure ECAR as a glycolysis proxy.
Rescue via nucleosides: They test whether nucleotide depletion is causally linked to death by supplementing nucleosides (uridine and inosine and related nucleoside mixes). Rescue supports a nucleotide-pool mediation model for at least part of the lethal phenotype.
Signaling-to-transcription mechanism: They test whether MAPK→MYC→RPIA explains the stratification by measuring ERK phosphorylation, MYC, and RPIA expression after KRAS loss and after MEK inhibition.
Genetic and pharmacologic targeting of nucleotide synthesis: They inhibit PPP via RPIA knockdown/CRISPR deletion and inhibit pyrimidine synthesis via DHODH knockdown and two inhibitors (brequinar; leflunomide), with additional 3D spheroid and in vivo xenograft experiments.

3) Central results (evidence, then what’s still uncertain)

3.1 KRAS extinction rewires metabolism—but nucleotide/pyrimidine loss is larger in sensitive lines

Glycolysis drops broadly: After KRAS knockdown, multiple glycolytic intermediates and ECAR decrease in all tested lines (consistent with KRAS influencing glycolytic gene expression).
PPP and pyrimidines split sensitive vs resistant: PPP intermediates and pyrimidine nucleotides (e.g., UDP/UTP) decrease more strongly in sensitive than resistant cells.
Nucleosides restore survival: Nucleoside supplementation suppresses the cell-death phenotype caused by KRAS knockdown in sensitive lines, supporting nucleotide pool involvement in KRAS dependency.

Uncertainty to keep skeptical about: Nucleoside rescue strengthens the “nucleotide pools matter” argument, but it does not fully exclude other KRAS-dependent vulnerabilities (e.g., redox/mitochondrial effects) that might also contribute to the sensitive phenotype. The authors also state that KRAS loss affects proliferation and metabolism broadly, and they discuss differences between genetic KRAS ablation and MEK inhibition that may reflect KRAS functions beyond MAPK.

3.2 The MAPK–MYC–RPIA axis tracks KRAS/MEK dependence patterns

MEK inhibition phenocopies KRAS dependence: KRAS-resistant lines show higher IC50s for AZD8330 and maintain the expected metabolic phenotype under MEK inhibition.
MYC and RPIA suppression is larger in sensitive lines: After KRAS knockdown and MEK inhibition, MYC and RPIA decrease mainly in sensitive cells, while resistant cells maintain RPIA (including a resistant subset with maintained RPIA despite reduced MYC).
MYC knockdown kills broadly: MYC depletion reduces survival/proliferation across PDAC lines, and reduces RPIA levels; this supports MYC’s role as an upstream mediator for RPIA-linked survival.

Key blind spot: The paper argues resistance is largely via maintained nucleotide synthesis rather than global KRAS reactivation, but “maintained nucleotide synthesis” could itself be downstream of additional pathways not fully mapped here (for example compensatory PI3K/AKT signaling under MEK inhibition). The authors do report PI3K/AKT compensatory activation after MEK inhibition in sensitive lines but not after KRAS knockdown.

3.3 RPIA and DHODH targeting disrupts nucleotide-linked survival

RPIA depletion inhibits growth and can induce death: RPIA knockdown reduces clonogenic survival across PDAC cell lines; death is attenuated by nucleoside supplementation, consistent with on-target nucleotide dependence.
Non-transformed cells are less affected: RPIA inhibition is comparatively weak in non-transformed lines, suggesting a possible therapeutic window.
DHODH inhibition blocks pyrimidines and respiration-related phenotypes: DHODH knockdown and inhibitors (brequinar, leflunomide) reduce growth; metabolomics show accumulation of upstream pyrimidine precursors and depletion of pyrimidine nucleotides, consistent with pathway blockade.
In vivo: early tumor suppression, later adaptation: Brequinar treatment slows xenograft growth with some regressions, but the effect diminishes by study end, and endpoint tumor metabolomics shows persistent on-target pyrimidine suppression with adaptive changes.

4) Critical appraisal (skeptical, mechanistic, and falsification-focused)

4.1 Strong points

Mechanism is triangulated: They connect (i) KRAS dependence phenotypes, (ii) metabolomics of glycolysis/PPP and nucleotide pools, (iii) signaling/transcription readouts (ERK phosphorylation, MYC, RPIA), and (iv) rescue/functional perturbations (nucleosides; genetic and pharmacologic inhibition).
Rescue supports causal direction (within limits): Nucleoside supplementation attenuates cell death from KRAS loss and from RPIA loss, aligning with nucleotide pool depletion as a mediator rather than a mere correlate.
Resistance framing is metabolically explicit: They emphasize differential metabolic rewiring between KRAS-sensitive and -resistant cells and validate that MEK inhibition can induce a cytostatic effect distinct from KRAS knockdown, highlighting biological context differences rather than assuming equivalence.

4.2 Weak spots / what could mislead

Cell-line panel generalization: The model relies on a panel of PDAC cell lines; KRAS “dependence” and metabolic phenotypes can be cell-line specific. The authors explicitly use a panel of human PDAC lines and murine PDAC-derived lines, but the mapping from in vitro dependence to patient tumors remains uncertain.
Genetic KRAS ablation vs pharmacologic MEK inhibition: They note differences—KRAS knockdown induces death in sensitive lines while MEK inhibition yields cytostasis and can trigger compensatory AKT signaling. Any resistance inference based solely on MEK inhibition may not fully represent KRAS-driven biology.
Adaptive metabolic escape is real: DHODH inhibition shows early tumor regression but diminished effect later, implying adaptive mechanisms. That limits how “universal” nucleotide synthesis blockade might be as a standalone strategy.
Off-target/pharmacology caveats (general): The paper uses pharmacological inhibitors (brequinar, leflunomide). While metabolomics supports pathway blockade, pharmacology can still have off-target effects; the authors partially address this via genetic knockdown and metabolomic pathway readouts.

4.3 What would most disprove the paper’s thesis?

If KRAS/MEK-resistant PDAC cells showed no ability to maintain non‑oxidative PPP and nucleotide pools despite MYC/RPIA maintenance, then the mechanistic model (nucleotide pools as mediator of resistance) would be weakened.
If RPIA or DHODH inhibition reduced growth but failed to produce pathway-consistent metabolomic signatures (e.g., no pyrimidine depletion for DHODH blockade), then causality from nucleotide synthesis disruption would be undermined.
If nucleoside rescue did not attenuate cell death after KRAS or RPIA perturbation, the “nucleotide pool mediation” causal link would weaken.

5) Conflict of interest transparency (important for skepticism)

The paper discloses that A.C. Kimmelman has financial interests in Vescor Therapeutics, and serves on advisory boards, and is an inventor on patents related to KRAS-regulated metabolic pathways and related targets.

Skeptical implication: this does not invalidate the mechanistic data presented (especially metabolomics + genetic perturbations), but it increases the value of replication, independent datasets, and careful evaluation of specificity and adaptive escape.

6) Bespoke BGPT follow-ups (author reviews + deeper asks)

7) Author reviews (all full names found in the provided paper text)

Feedback:

Updated: April 06, 2026