The Nutrient That Helps Ovarian Cancer Survive Chemotherapy

Interview and article by Pamela Pont

A five-year study led by OCRA grantee Dr. Apoorva Uboveja found that a common nutrient primarily in animal-based foods lets a hard-to-treat subset of ovarian cancer repair itself faster than chemotherapy can damage it. A drug that blocks it is already in human trials for other conditions.

• The paper: Uboveja A, Yang B, Buj R, et al. "αKG-mediated carnitine synthesis drives DNA repair via histone acetylation." Nature, May 2026. DOI: 10.1038/s41586-026-10584-7
• The atlas: Generated from αKG, Carnitine Synthesis and DNA Repair. Explore here.
• The researcher: Dr. Apoorva Uboveja, Staff Scientist in Dr. Katherine Aird’s Lab, Molecular and Cellular Oncogenesis Program, The Wistar Institute
• Funded in part by: an OCRA grant awarded to Dr. Uboveja in her second year.

The short version

Chemotherapy works by damaging a cancer cell's DNA badly enough that the cell cannot recover. Most cancers are bad at repairing that damage, which is exactly why the treatment works.

One subset of ovarian cancer does the opposite. These tumors, marked by high levels of expression from a gene called cyclin E, are unusually good at DNA repair. They patch the damage and keep growing. Chemotherapy doesn’t have such a dire impact.

Dr. Aird's and Dr. Synder’s team spent five years finding out what powers that repair. The answer turned out to be carnitine, a nutrient your body makes on its own and also absorbs from food, particularly red meat. In these tumors, carnitine production is running high, and that makes the DNA more accessible via chemical modifications, which in turn makes the repair machinery more efficient.

When the team blocked carnitine production in mice and gave chemotherapy alongside it, tumor burden dropped by about 30 percent.

Why this particular group of patients matters

This is the part that makes the finding urgent rather than merely interesting.

When a cancer's main DNA repair pathway is broken, doctors have a treatment for that. PARP inhibitors work by closing off the backup repair routes the cell falls back on, and they are standard of care for those patients.

Cyclin E-high cell ovarian cancers do not fit that picture. Their main repair pathway is not broken. It is working well, which means PARP inhibitors have nothing to exploit, and the tumors resist platinum chemotherapy at the same time.

"Because they are not sensitive to chemotherapy, there are really no therapies for patients with such a cancer," Dr. Uboveja says.

That is the gap this work is aimed at.

How they found it

The trail started with an observation from the lab of Dr. Katherine Aird, Dr. Uboveja's mentor : cyclin E-high cells carry unusually high levels of a metabolite called alpha-ketoglutarate. Nobody knew what it was doing there.

Alpha-ketoglutarate is known to feed a family of enzymes involved in removing methyl tags from proteins, so methylation was the obvious place to look.

To test it properly, the team built a targeted CRISPR screen covering 64 of those enzymes, knocking each out one at a time to see which ones cyclin E-high cells could not survive under chemotherapy. The screen returned an unexpected enzyme: TMLHE, the first and rate-limiting step in the body's own carnitine production.

"I had never imagined that this particular enzyme could be involved," Dr. Uboveja says. "I couldn't imagine it until we got the results from the screen."

And the expected answer was wrong. Methylation was affected, but it did not matter. Acetylation did. The chain runs: alpha-ketoglutarate feeds TMLHE, TMLHE drives carnitine synthesis; carnitine supplies the acetyl units for histone (a special type of protein that organizes DNA and helps it fit inside a cell's nucleus) acetylation, and that acetylation, at specific sites, makes homologous recombination repair more efficient.

"Once we were able to link all of this, the mechanism was so clear," she says. "Then the mechanism was so clear that this is what is happening."

What it could mean for patients

A possible way to identify these patients. Patients treated with DNA-damaging chemotherapy, high TMLHE, and therefore high acetylcarnitine, in the blood correlated with worse progression-free survival. If that holds up, a blood measurement might one day help flag the patients least likely to respond to standard chemotherapy, before treatment starts rather than after it fails.

A drug that already exists. The team tested mildronate, a carnitine synthesis inhibitor already used in human trials for other conditions. Combined with chemotherapy in mouse models, it reduced tumor burden by roughly 30 percent. A compound with existing human safety data is a considerably shorter path than starting from a new molecule.

What this does not mean yet. Dr. Uboveja is careful not to overpromise, and when asked what a patient reading this could do right now, she did not reach for a comforting answer.

"That's a really difficult question to answer, because our data has not gone through human trials," she says. "This is one of the limitations of my study. I would agree that we really do need to extend it to patient level."

Nothing here is a treatment today. Nothing here is dietary advice. The dietary carnitine question is a research question she is actively working on, not a recommendation, and patients should not change how they eat based on this paper. What the study offers is a mechanism nobody had described before and two plausible routes forward.

"For any medicine to form, or any cure to form, basic biology plays a very big role," she says.

The OCRA grant

Dr. Uboveja applied for OCRA funding in her second year of the project, well before anyone knew where the project would land.

"When I got the grant, I wasn't at a point where I could have imagined this would become such a big paper," she says. "We started working on it in the direction I had proposed, and it turned out amazing."

She is clear about what the funding did beyond the budget line. "The grant I got from OCRA was really beneficial and really helpful, and motivating me that I was on the right track."

What's next

Dr. Uboveja is pursuing several threads that the published paper does not cover.

Does diet matter? Mildronate blocks only the carnitine your body makes itself. Carnitine also comes from food. How much dietary carnitine contributes to chemotherapy resistance is still an open question, and one she considers central to making the finding useful to patients.

Is the gut microbiome involved? In pilot experiments, gut bacteria metabolize dietary carnitine into another compound, TMAO, which may be the actual driver of resistance. When she used antibiotics to clear the microbiome, chemotherapy response improve.

Why did the immune system change the answer? The published preclinical work used murine models without immune systems. When she repeated the dietary experiments in both immune-competent and immune-deficient models, the antibiotic effect appeared only in the models that had an immune system. Diet, microbes, and immunity appear to be interacting in a way nobody has mapped yet.

Does this extend beyond ovarian cancer? Cyclin E is not the only gene that produces the DNA replication stress that is the root of this mechanism. MYC does it too, and the paper includes preliminary results suggesting the finding reaches past cyclin E. Whether the alpha-ketoglutarate to carnitine to acetylation axis operates in other cancers with the same resistance profile is, in her words, "yet to be seen."

"We have just published one arm of this mechanism," she says.

In her own words

Did you always know you would end up studying metabolism and cancer?

My PhD was oncogenic pathways in colorectal cancer, and I had not yet studied anything about cancer metabolism. But through the experience I gained during my PhD, I became more interested in understanding why different cancers behave the way they do. I read that it is actually the metabolism of the cell that gets rewired once it becomes cancerous, and that can lead to so many of the things responsible for how cancer cells behave.

So for my postdoc I really wanted to study cancer cell metabolism, and it was by chance that I got into this great lab studying that topic in ovarian cancer. I really wanted to study ovarian cancer, because it is very hard to treat. Patients relapse after the first round of chemotherapy and I wanted to understand why.

Can you explain in plain language why this discovery matters?

Cancer cells need to constantly adapt to survive and grow. We discovered that, in a group of these aggressive cancers, there is a certain way they use a nutrient called carnitine, and this nutrient is actually helping the cancer cells repair their DNA damage, which makes them harder to kill with cancer treatment.

Through our work, we showed that if we can interfere with this metabolic adaptation-the way they use carnitine- we might be able to make these cancers more vulnerable to treatment.

Was there a result that made you think this is really going to matter?

The most amazing observation was that carnitine is involved at all. We found acetylcarnitine, the form that carries the acetyl units needed for histone acetylation, is high in patient serum, and it correlates with worse progression-free survival. So high carnitine in the serum could act as a biomarker for these patients.

What would you want a patient or family member to take away?

The motive of this group is not just to do basic science, but to do science that could extend to translational significance. Even though this is basic science in preclinical models, it is a hope that this mechanism can be targeted to help patients with this cancer - patients for whom current treatments are not applicable and who are not getting any relief. Maybe this could be a future thing that could help, even as a biomarker.

Read the paper

Uboveja A, Yang B, Buj R, Amalric A, Wang H, Tangudu NK, et al. αKG-mediated carnitine synthesis drives DNA repair via histone acetylation. Nature. 2026 May;655(8123):790-800. DOI: 10.1038/s41586-026-10584-7

Senior authors: Nathaniel W. Snyder (Temple University, Philadelphia) and Katherine M. Aird (The Wistar Institute,Philadelphia).

Article citation data retrieved from PubMed.