Udutha et al. · Neuro-Oncology · 2026

1p/19q codeletion induces targetable and imageable vulnerabilities in glucose metabolism in oligodendrogliomas

Udutha S, Batsios G, Taglang C, Gillespie AM, Viswanath P

The study

What was asked, and what was found

Udutha et al., Neuro-Oncology, 2026 took the genetic lesion that defines oligodendroglioma and asked whether it leaves the tumour with a weakness worth attacking. Losing one arm of chromosome 1 removes a copy of ENO1, the ubiquitous enolase of glycolysis, and the authors found the remaining copy is silenced further by histone methylation at its promoter. The tumour compensates by raising the brain isoform ENO2, which makes it dependent on a single enzyme that normal brain cells and the other kind of glioma do not need in the same way.

AUMsilence sdASOs were used to test that dependence. Two nonoverlapping sequences against ENO2, at 5 μM in patient-derived oligodendroglioma cells with a nontargeting oligonucleotide as the control, abolished enolase activity, stopped proliferation and arrested the cells in S phase. Astrocytoma cells and neural progenitor cells given the same treatment carried on unaffected. Carbon-13 glucose tracing then showed something more interesting than a simple slowdown: with ENO2 gone, glucose stopped flowing to lactate and the tricarboxylic acid cycle and went instead into serine, one-carbon metabolism and purine nucleotides. The escape route ran through PHGDH, which rose in message and protein after the knockdown, and silencing AMPK returned PHGDH to control levels, placing AMPK in between.

Closing the escape route was lethal. Silencing ENO2 and PHGDH together, again with AUMsilence sdASOs, killed the tumour cells by apoptosis, while either knockdown alone did not, and the same double knockdown left astrocytoma cells and neural progenitor cells alive. Pyruvate rescued the ENO2 arrest and serine partly rescued the double-knockdown death, tying each phenotype to the enzyme that was removed. The authors then reproduced the pair with two brain-penetrant small molecules, which is what carried the work into animals: in intracranial patient-derived xenografts the drug pair caused tumour regression and longer survival, and deuterium imaging of lactate production reported the response within 5 plus or minus 2 days, ahead of any volume change visible on MRI. The oligonucleotide experiments in this paper are all in cells; every animal experiment used the small molecules.

Key findings

Two nonoverlapping AUMsilence sdASOs against ENO2, at 5 μM in patient-derived oligodendroglioma cells, abolished enolase activity, stopped proliferation and arrested the cells in S phase, while astrocytoma cells and neural progenitor cells given the same treatment were unaffected.(Results, Genetic or Pharmacological Inhibition of ENO2 Inhibits ODG Proliferation and Cell Cycle Progression, Fig. 3A to 3C)
Carbon-13 tracing showed the silencing did not only slow glucose use, it redirected it: less label reached lactate and the tricarboxylic acid cycle, and more reached serine, one-carbon metabolism and purine nucleotides.(Results, Targeting ENO2 Abrogates [U-13C]-Glucose Metabolism via Glycolysis and the TCA Cycle in ODGs, Fig. 3H and 3I)
The rerouting was traced to PHGDH, which rose in message and protein after the ENO2 knockdown, and silencing AMPK brought PHGDH back to control levels, which places AMPK between the two.(Results, ENO2 Loss Upregulates Serine Biosynthesis via PHGDH in ODGs, Fig. 4G to 4J)
Silencing ENO2 and PHGDH together killed the tumour cells, while silencing either one alone did not, and the same double knockdown did not kill astrocytoma cells or neural progenitor cells. That is a synthetic lethal pair demonstrated with oligonucleotides.(Results, Targeting PHGDH Is Synthetically Lethal in Combination with ENO2 in ODGs, Fig. 5B)
Adding pyruvate back restored cell cycle progression in cells given the ENO2 oligonucleotides, which shows the arrest came from losing enolase activity rather than from anything else the treatment did.(Results, Targeting ENO2 Abrogates [U-13C]-Glucose Metabolism via Glycolysis and the TCA Cycle in ODGs, Fig. 3J and 3K)
Supplying serine partly rescued the death caused by the double knockdown, which ties the killing to the collapse of serine synthesis rather than to a general toxicity.(Results, Targeting PHGDH Is Synthetically Lethal in Combination with ENO2 in ODGs, Fig. 5I and 5J)
The knockdown results were reproduced with two small-molecule inhibitors, which then carried the work into animals, where the pair caused tumour regression and longer survival and could be watched by deuterium imaging within days.(Results, Combination of POMHEX and D8 Abrogates [U-13C]-Glucose Metabolism and Induces Tumor Regression in ODGs In Vivo, Fig. 6G to 6J)

Product: AUMsilence sdASO
Targets: ENO2 · human · mRNAPHGDH · human · mRNAENO2, using two independent sequences · human · mRNA, specificity control by second sequence
Models: patient-derived oligodendroglioma cells SF10417, BT88, BT54, TS603 and NCH612, cultured in serum-free medium · primary cellspatient-derived astrocytoma cells SF10602 and BT142 and neural progenitor cells, treated with the same oligonucleotides as specificity controls · primary cells, comparison armsoligodendroglioma and astrocytoma patient biopsies from the UCSF Brain Tumor Center Biorepository · human tissue, no oligonucleotide usedintracranial patient-derived xenografts in SCID mice, treated with small-molecule inhibitors rather than oligonucleotides · in vivo, no oligonucleotide used
Delivery: added to patient-derived cells at 5 μM. No transfection reagent is named for the oligonucleotides, and the separate small interfering RNA reagents used for other genes are named as a different product from a different supplier.
Readouts: enolase activity assay, cell cycle analysis by flow cytometry, Ki67 proliferation assay, RT-qPCR, immunoblot, stable isotope tracing with uniformly labelled carbon-13 glucose and mass spectrometry, apoptosis by annexin V, chromatin immunoprecipitation for histone H3K27 trimethylation, deuterium metabolic imaging and MRI, bioluminescence imaging and survival analysis
Measured: oligonucleotide concentration on patient-derived cells: 5 μM (ENO2 and PHGDH silencing in oligodendroglioma, astrocytoma and neural progenitor cells)number of independent sequences against ENO2: 2 nonoverlapping (enolase activity in three oligodendroglioma lines, two astrocytoma lines and neural progenitor cells)potency of the small-molecule ENO2 inhibitor, for comparison with the knockdown: 35-50 nM in oligodendroglioma cells against greater than 5-20 μM in astrocytoma cells and neural progenitor cells (viability, no oligonucleotide involved)replication of each in vitro comparison: n = 3 (stated for every panel involving the oligonucleotides)synergy of the matching drug combination, no oligonucleotide involved: 16.59, 15.01 and 15.95 Bliss synergy score, above 10 indicates synergy (SF10417, BT88 and NCH612 cells)how quickly deuterium metabolic imaging reported the response, and how that compared with MRI: 5 plus or minus 2 days, ahead of any volume change visible on MRI (intracranial NCH612 and BT88 xenografts on the drug combination, no oligonucleotide involved)
Figures: Figure 3 The ENO2 knockdown: enolase activity, cell cycle arrest and the glucose tracing, with astrocytoma and neural progenitor cells as unaffected comparators, and the pyruvate rescue.Figure 4 PHGDH rising after ENO2 is silenced, AMPK phosphorylation, and PHGDH returning to control levels when AMPK is silenced too.Figure 5 The synthetic lethal pair: apoptosis only when both ENO2 and PHGDH are silenced, absent in astrocytoma and neural progenitor cells, with the serine rescue.
Citation: Udutha et al. 1p/19q codeletion induces targetable and imageable vulnerabilities in glucose metabolism in oligodendrogliomas. Neuro-Oncology 2026;28:1711-1726. doi:10.1093/neuonc/noag063

For research use only. Not for use in diagnostic or therapeutic procedures.