Cells face a continual metabolic decision: should fatty acids be stored for later, or mobilized and burned for energy? Mitochondria sit at the heart of both processes. Although they are best known for oxidizing nutrients to produce energy, distinct populations of mitochondria can also support the accumulation of cellular fat. Brown adipose tissue (BAT), which can rapidly shift between storing and consuming lipids in response to physiological demand, provides a particularly useful system for understanding how mitochondria help control this balance. A new study published in The EMBO Journal identifies mitochondrial calcium as an important signal at this metabolic decision point1.

The current study builds upon previous work from the Shirihai laboratory at the University of California, Los Angeles, demonstrating that unique subpopulations of mitochondria coexist within brown adipocytes2. One population, called peridroplet mitochondria (PDM), physically attaches to lipid droplets and preferentially oxidizes pyruvate, helping support lipid-droplet expansion. A second population, cytosolic mitochondria (CM), is spatially separated from lipid droplets and has a greater capacity to oxidize fatty acids. Thus, within the same cell, mitochondria can participate in opposing sides of lipid metabolism: supporting the build-up of stored fat or facilitating its utilization.

A pivotal question addressed by the new study is how cells transition between these states. To tackle this problem, the authors developed cell-free assays in which mitochondria and lipid droplets could be separated and then brought back together under controlled conditions. Remarkably, isolated PDM retained a greater intrinsic capacity than CM to bind lipid droplets. Their association was also influenced by the fuel available to mitochondria: pyruvate plus malate favored PDM attachment, whereas palmitoyl-carnitine, a fatty acid-derived substrate, favored detachment. The authors found evidence that these different metabolic signals converge on a common regulatory axis involving calcium within the mitochondrial matrix.

At the center of this mechanism is a transient rise in mitochondrial matrix calcium. During adrenergic stimulation, calcium enters mitochondria, and the authors found that elevated matrix calcium promotes a striking CypD-dependent architectural transition of PDM from their characteristic elongated, lipid-droplet-associated morphology toward a rounded, swollen form that favors detachment. Importantly, this represents a reversible mitochondrial shape transition rather than the irreversible mitochondrial damage associated with sustained calcium overload. Removing PDM from the lipid-droplet surface also increased the ability of recombinant lipases to liberate fatty acids, while blocking triglyceride breakdown did not prevent PDM detachment. Together, these findings place mitochondrial detachment upstream of lipid mobilization and support the idea that PDM can act as both a metabolic and a physical barrier to lipid utilization.

The duration of the calcium signal is tightly regulated. Following adrenergic stimulation, PDM and CM take up calcium at similar initial rates, but PDM clear it more slowly, exposing them to a larger and more prolonged calcium signal. The mitochondrial sodium/calcium exchanger NCLX counteracts this process by extruding calcium from the matrix and favoring mitochondrial association with lipid droplets. The authors further identified phosphodiesterase 2A (PDE2A), which is enriched in PDM, as an important regulator of NCLX activity. Inhibiting PDE2A accelerated NCLX-dependent calcium extrusion, preserved PDM association with lipid droplets, reduced fatty-acid mobilization, and increased glucose utilization.

The pathway was also manipulable in vivo. In obese ob/ob mice, three weeks of PDE2A inhibition produced a greater than fourfold increase in PDM abundance in BAT and increased glucose uptake by more than 50%, together with metabolic and morphological features consistent with BAT remodeling. Importantly, the study does not establish whether increasing PDM is itself responsible for this remodeling or whether PDM accumulation occurs as part of a broader change in the tissue. Nevertheless, the findings demonstrate that mitochondrial–lipid droplet interactions can be pharmacologically altered in a metabolically challenged animal.

Altogether, the study positions mitochondrial calcium not simply as a regulator of mitochondrial activity, but as a signal capable of changing mitochondrial positioning and, in turn, influencing how cells handle fat. Brown adipose tissue offers a particularly vivid system in which to study this principle because lipid storage and utilization can be rapidly switched in response to physiological demand. Whether analogous mechanisms operate in other cell types remains an intriguing question, with potential implications for understanding altered lipid metabolism in both physiological and disease states.

A Statement of Significance from Dr. Shirihai:

Brown adipose tissue provides a unique system for understanding how mitochondria can regulate both lipid storage and lipid utilization within the same cell. Our study identifies mitochondrial matrix calcium as a key decision point in this process: a transient rise in calcium within peridroplet mitochondria promotes their detachment from lipid droplets and shifts the cell toward fatty acid mobilization. This translates into a shift in fuel preference, toward lipid utilization. Beyond fuel, detachment of mitochondria may facilitate the release of lipids for signaling. A byproduct of this study was the investigation into the role of metabolites in inducing attachment and detachment through calcium. Investigation into this question revealed that mitochondrial matrix calcium is altered by specific mitochondrial fuels. Remarkably, acyl-carnitines elevate free matrix calcium, while pyruvate reduces it.  This can explain various observations, including the previously reported induction of lipolysis by pyruvate carrier inhibitors.

But my favorite part of the study is the cell-free system that allowed us for the first time, to ask if PDM know they are PDM, and indeed they remember! If you separate them from a beloved lipid droplet, they come back and attach, and if you show them a lipid droplet from a white adipocyte they remain naïve.

Next is a manuscript about a PDM-detacher small molecule that induces lipolysis, shrinks lipid droplets, and inhibits viral proliferation.

A conversation with the authors:

MitoWorld: Brown adipose tissue provides a striking system for studying how mitochondria participate in both lipid storage and lipid utilization. As this story unfolded in the lab, which aspects of this metabolic switch surprised you most?

Authors: That metabolites are the actual signaling molecules for detachment and attachment, and that mitochondrial fragmentation has a new role: detachment via architectural change. Fragmentation also facilitates lipid import to mitochondria, as shown by by Ngo et al. EMBO 20233. Therefore, fragmentation is critical for BAT to activate thermogenesis.

MitoWorld: Given that peridroplet mitochondria undergo cycles of attachment to and detachment from lipid droplets, how do you suppose that PDM maintain a unique identity alongside cytosolic mitochondria?

Authors: PDM remain faithful to lipid droplets. If separated, PDM have a higher affinity for binding to lipid droplets. The mechanism is likely to be through tethering proteins such as PLIN5, DGAT2, MIGA2.

MitoWorld: To what extent do you think distinct mitochondrial subpopulations exist in other cell types and are tailored to different physiological responses?

Authors: We already know that neurons have distinct mitochondrial populations in the neurites vs the soma, and other papers have shown that mitochondria isolated from purified nuclei have distinct functional features. The important implication is to consider that our research and therapeutic targets may be a subpopulation of mitochondria in a given cell type and as such we should design our studies to zoom in on them, exactly as we do in a mixed group of cells.

MitoWorld: Your findings suggest that mitochondrial calcium can influence whether cells favor fatty-acid storage or utilization. How broadly do you think this principle might apply to pathological alterations in lipid metabolism, such as those occurring in diabetes or cancer?

Authors: I think the principle could be much broader than brown adipose tissue. The important concept is that mitochondrial subpopulations can sense the metabolic environment and make a local decision between lipid storage and lipid utilization. In PDM, metabolites change matrix calcium and calcium becomes part of the switch that controls attachment to lipid droplets and access to stored fat. We do not yet know whether the same mechanism operates in other tissues, but many diseases, including diabetes, fatty liver disease and cancer, involve abnormal lipid accumulation or altered lipid utilization. It will be very interesting to ask whether similar mitochondrial decision points exist in these settings and whether they can be manipulated therapeutically.

MitoWorld: In a rodent model of obesity, pharmacological modulation of this pathway produced substantial metabolic remodeling in brown adipose tissue. With the advent of GLP-1 receptor agonists for treating obesity in humans, do you think targeting mitochondrial control of lipid metabolism could eventually provide an alternative or complementary strategy?

Authors: I see it more as a complementary strategy. GLP-1 therapies are extremely effective at reducing energy intake and body weight, while what we are studying is a different level of metabolic control: what happens to fat once it is inside the cell. The exciting possibility is that we may be able to control whether lipids are stored or mobilized by targeting specific mitochondrial populations and their interaction with lipid droplets. This could be useful not only for obesity, but also for diseases in which ectopic fat accumulation or abnormal lipid utilization contributes to pathology, such as diabetes, fatty liver disease, and potentially some cancers. The direction of the intervention may also depend on the disease: we may want to promote lipid utilization in one setting and limit it in another

MitoWorld: Can you elaborate about future directions for this research?

Authors: One direction is to understand the molecular machinery that gives PDM their identity and allows them to attach and detach from lipid droplets. Another is to understand how metabolites communicate with this machinery through mitochondrial calcium. But I am particularly interested in whether we can pharmacologically control this decision. We are now studying small molecules that promote PDM detachment, increase lipolysis and reduce lipid-droplet size. The larger question is whether manipulating mitochondrial–lipid droplet interactions can be used in diseases where abnormal lipid storage or utilization is part of the pathology. I think this will also teach us whether the PDM concept is unique to adipocytes or represents a more general principle of mitochondrial specialization.

References:

  1. Acin-Perez R, Assali EA, Veliova M, Ngo J, Brownstein AJ, Villalobos F, Petcherski A, Hernansanz-Agustin P, Kim-Vasquez D, Xu S, Tamboline M, Silva RM, Upcher A, Shu C, Ferriss DE, Liesa M, Enriquez JA, Sekler I, Shirihai OS. Mitochondrial calcium regulates lipid metabolism by modulating tethering of mitochondria to lipid droplets. The EMBO Journal. 2026;45(14):4820–4848. Published online July 3, 2026.
  2. Benador IY, Veliova M, Mahdaviani K, Petcherski A, Wikstrom JD, Assali EA, Acín-Pérez R, Shum M, Oliveira MF, Cinti S, Sztalryd C, Barshop WD, Wohlschlegel JA, Corkey BE, Liesa M, Shirihai OS. Mitochondria Bound to Lipid Droplets Have Unique Bioenergetics, Composition, and Dynamics that Support Lipid Droplet Expansion. Cell Metabolism. 2018;27(4):869–885.e6.
  3. Ngo J, Choi DW, Stanley IA, Stiles L, Molina AJA, Chen PH, Lako A, Sung ICH, Goswami R, Kim MY, Miller N, Baghdasarian S, Kim-Vasquez D, Jones AE, Roach B, Gutierrez V, Erion K, Divakaruni AS, Liesa M, Danial NN, Shirihai OS. Mitochondrial morphology controls fatty acid utilization by changing CPT1 sensitivity to malonyl-CoA. EMBO J. 2023 Jun 1;42(11):e111901. doi: 10.15252/embj.2022111901. Epub 2023 Mar 14. PMID: 36917141; PMCID: PMC10233380.