Mapping Mitochondrial Diversity in NSCLC

Researchers at the University of California, Los Angeles, have uncovered a new wrinkle in non–small cell lung cancer's (NSCLC) complex, heterogeneous landscape: key differences in mitochondrial dynamics and morphology between the two main subtypes, lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC), that regulate tumor bioenergetics and could impact therapeutic targeting of cancer metabolism (Nature 2023;615:712–19).

Often described as a “super engine in cells,” mitochondria “form networks that are highly responsive to tumor environment changes,” says senior author David Shackelford, PhD. “We've known very little about how these organelles behave in vivo, being limited to 2D images until recently. Now there are much better biological profiling tools, so we decided to address this information gap.”

The team devised a workflow that spanned machine learning, PET imaging with [¹⁸F]FBnTP—a tracer they developed to measure mitochondrial oxidative phosphorylation (OXPHOS)—and powerful 3D electron microscopy pioneered by Mark Ellisman, PhD, of the University of California, San Diego. They used this platform to “zoom in on the fine details,” Shackelford says, of mitochondrial networks in mouse models of LUAD and LUSC. The former had considerably higher OXPHOS rates than the latter, a functional distinction that tracked with structural differences: in LUAD, long, fused mitochondria whose cristae, or inner membrane folds, were highly organized, versus smaller, more fragmented mitochondria with disordered cristae in LUSC.

OXPHOS-high mitochondria in LUAD cells utilized diverse nutrient sources, including glucose, glutamine, and fatty acids, the researchers found. They also pinpointed a subpopulation of peridroplet mitochondria (PDM) wrapped around lipid droplets in these cells. By contrast, OXPHOS-low mitochondria relied on glucose and glutamine metabolism. Both PDM and lipids were notably absent in LUSC cells, Shackelford adds, “potentially explaining the lack of fatty acid oxidation, which we'd previously noted.”

Shackelford cautions that these differences “are likely not binary, but on a spectrum—I'd expect there are LUAD cells with OXPHOS-low mitochondria and vice versa.” He views his team's findings as “the first step in generating a more accurate map than we've had so far of what's really going on with mitochondria in the tumor microenvironment.”

“This work is a tour de force and incredibly well done,” says Timothy Pardee, MD, PhD, chief medical officer of Cranbury, NJ–based Cornerstone Pharmaceuticals, which is developing devimistat, a mitochondria-targeting agent. “My take is within [NSCLC], there are more or less mitochondrially robust subtypes, and the latter is where we might do better with a drug like devimistat.”

Pardee, who wasn't involved, points out that “as phenomenal as this study is data-wise, it shows us just the baseline—a picture of complex mitochondrial dynamics when the tumor is happy.” The researchers “have yet to look at how mitochondria remodel themselves in tumors challenged with treatment.” Such metabolic plasticity remains a therapeutic barrier, and “the needle has moved backwards” of late: Devimistat didn't meet its primary endpoint in phase III trials, although subset analyses of the data are ongoing, and development on another drug that showed clinical potential, IACS-010759, recently was halted due to unacceptable neurotoxicity.

Cancer metabolism therapies “haven't fared very well, particularly as single agents,” Shackelford observes. “We need to go after multiple metabolic nodes; hitting just one won't work.”

Combinations will be essential, Pardee agrees, noting that trials are ongoing or planned for devimistat with chemotherapy and immune checkpoint blockade. “We don't think the field is without a future,” he says. “It's a hard nut to crack, but we'll find a way eventually.”

Looking ahead, too, “maybe we find just 5% of [NSCLC] patients have this specific mitochondrial phenotype” that's therapeutically sensitive, Shackelford says. “But 5% of a big number is still a big number.” –Alissa Poh