Mitochondria redistribution organizes the immunosuppressive tumor ecosystem

The tumor microenvironment (TME) imposes severe constraints on oxidative metabolism 1. Hypoxia, abnormal vasculature, metabolic competition, and the accumulation of multiple inhibitory metabolites collectively compromise mitochondrial respiration and are thought to underlie failure of anti-tumor immunity and immune-directed therapies2-6. Yet, paradoxically, tumor cells and immunoregulatory populations within the same hostile TME maintain robust mitochondrial activity, supporting their persistence and function7-9. For instance, despite the canonical Warburg effect10, the vast majority of metabolism in cancer cells is mitochondria- derived8,11. Moreover, regulatory T cells and macrophages rely on mitochondrial metabolism to execute most of their immunosuppressive functions 12,13. Because solid tumors rarely regress spontaneously 14, these observations point to an underlying mechanism by which tumors maintain a mitochondrial metabolic ecosystem that selectively supports cancer cells and their immunosuppressive partners. Precisely how this is achieved remains unresolved. Emerging studies indicate that cancer cells can supplement their metabolic needs by directly harvesting exogenous mitochondria from the host 15-19, a process that has been proposed to boost ATP production in cancer cells15,20,21. However, the fate of these exogenously acquired mitochondria and their broader consequences for the TME are unknown. Moreover, it remains unclear how a relatively small input of transferred exogenous mitochondria elicits dramatic bioenergetic output in recipient cells. Here, leveraging tractable mitochondrial reporter systems across murine models and patient samples, we unexpectedly found that .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint cancer cells fuse acquired exogenous mitochondria with their own and ultimately redistribute them to fuel immunosuppressive cells. In the process, cancer cells themselves benefit by optimizing their metabolic capacity through architectural rearrangement of P5CS, a critical enzyme for ma cromolecular biosynthetic capabilities of cancer cells 22. These findings uncover a previously unrecognized regulatory axis in which cancer cells act not merely as resource competitors, but as active organizers of a mitochondrial economy that fuels immune suppression and tumor progression.

Cancer cells redistribute mitochondria To study intercellular mitochondrial transfer, we developed a flow cytometry and confocal microscopy platform in which cancer cells were co-cultured with leukocytes from PhaMexcised mice, which express a mitochondria -targeted dendra2 (mtD2) reporter23. In agreement with prior reports16,18,24,25, we confirmed that tumors (B16 melanoma or E0771 breast) rapidly acquired mtD2 protein during co-culture, indicating mitochondria transfer from immune cells (Fig S1A and Fig S1B). Three observations raised the possibility that cancer cells do not permanently retain exogenous mitochondria. First, after mitochondria transfer to cancer cells, the mtD2+ signal declined faster than could be explained by proliferative loss from cell division, consistent with elimination or redistribution (Fig S1 C). Second, across tumor: leukocyte co-culture ratios (1:1 to 1:10), the fraction of mtD2+ tumor cells increased, but per-cell mtD2 intensity remained constant ( Fig. S1D ), suggesting a set -point for mitochondrial content. Third, live-cell imaging confirmed early mtD2 uptake by melanoma cells, followed by return of signal to immune cells (Supplementary Video). To formally test whether cancer cells redistribute exogenously acquired mitochondria , we leveraged the CD45.1 and CD45.2 allelic variants mouse systems 26, such that labeled mitochondria originating from CD45.2 cells (CD45.2mtD2+) can be unambiguously tracked in mitochondria-unlabeled CD45.1 cells (CD45.1mtD2-). CD45.1mtD2-:CD45.2 mtD2+ co-cultures

in minimal mtD2 transfer to CD45.1 cells. In triple co -culture with B16 melanoma, CD45.1 cells became mtD2+ (Fig 1A,B). However, when B16 cancer cells are indirectly present (that is, seeded in Boyden chamber with 3 µm filters during the CD45.1mtD2-:CD45.2mtD2+ co-culture to prevent direct contact ), there was minimal mitochondria transfer to CD45. 1 cells (Fig 1C), suggesting that cancer cells act as direct conduits for mitochondria transfer between immune cells. To definitively demonstrate th at cancer cells directly transfer mtD2+ exogenous mitochondria, we first co -cultured CD45.2mtD2+ leukocytes with B16 and sort -purified mtD2+ cancer cells ( Fig 1D ). In subsequent direct co-culture with CD45.1mtD2-, tumor cells transferred some of the previously acquired mtD2 to CD45. 1 cells ( Fig 1 E). Confocal imaging confirmed cancer cell redistribution of mtD2+ mitochondria to CD45.1-derived neutrophils, macrophages, and T cells (Fig 1F). .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint We repeated these co-culture experiments using other cancer cell types, such as E0771 breast cancer. Moreover, this time we tracked mitochondria from C D45.2 (using CD.45.2mtD2+) to CD45.1 (CD45.1mtD2-) to demonstrate that the findings are allele variant-independent. The

were identical to what we described for B16 : in the direct presence of E0771, mitochondria from CD45.2 can be redistributed to CD45.1 cells (Fig S1E). For genetic confirmation of mitochondria redistribution by cancer cells independent of any potential dendra2 artifacts, we co-cultured B16 with leukocytes from NZB/BINJ mice, which are on a CD45.2 background. We subsequently co-cultured the tumor cells with CD45.1 cells (from C57BL6/J). As the NZB exhibits defined mitochondria polymorphisms compared to the C57BL6/J reference mtDNA genome27, ARMS-PCR can be performed on sorted CD45.1 cells which, if positive, indicates the acquisition of mtDNA from NZB 28. In all cases, NZB mtDNA was detected in CD45.1 and B16 cells (Fig 1G). This approach further supported that cancer cells acquire exogenous mitochondria and redistribute them to cells in their microenvironment. Mitochondria redistribution in vivo by mice and human cancers We next tested mitochondria redistribution by tumor cells in vivo using various complementary approaches. First, we established mixed chimeras , in which lethally irradiated C57BL6/J mice were reconstituted with hematopoietic cells from CD45.2mtD2+ and CD45.1mtD2- donors (Fig 1H, and Fig S1F), enabling the tracking of dendra2 from CD45.2 to CD45.1 cells upon tumor implantation. These chimeric mice were then challenged with subcutaneous tumor implantation of B16Tdtomato. After tumors were established (~day 14), they were excised and processed for flow cytometr ic analysis of live cells to examine the acquisition of mitochondria. As expected, tumor cells became positive for mtD2+, suggesting the horizontal transfer of mitochondria from CD45.2 immune cells. Notably, mtD2 was also detected in CD45.1 immune cells in the TME (Fig 1I). Second, we performed parabiosis of CD45.2mtD2+ and CD45.1mtD2- mice. After chimerism was established, tumors were implanted and we confirmed the detection of mtD2 + mitochondria from CD45. 2 cells to CD45. 1 infiltrating the TME ( Fig 1 J). Third, we established a mixed chimera of CD45.2NZB and CD45.1C57BL6/J. Animals were challenged with tumors. NZB mtDNA was detected in tumor cells and in the CD45.1 cells (Fig 1K), paralleling the in vitro results. Collectively, the results indicate that tumors can acquire mitochondria from the host and and disperse them to other immune cells. Cell type –restricted mtD2 expression further demonstrated mitochondria redistribution between immune lineages. In PhAM floxed mice crossed to CD8 Cre, mtD2 was initially confined to CD8+ T cells (Fig S1G). In the presence of B16 tumors, mtD2 was detected in CD11b+ myeloid cells (Fig. S1H). Similarly, we also restricted mtD2 expression in myeloid cells by crossing PhAMfloxed with CD68Cre (PhAM+/+CD68Cre) (Fig S1I). Tumor implantation resulted in mtD2 expression in CD11b-CD3+ lymphoid cells within the TME ( Fig S1 J). .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint Altogether, the results demonstrate that mitochondria may be transferred from one immune cell type to others, a process orchestrated by cancer cells. To demonstrate mitochondria redistribution by human cancer cells, we used mitotracker green dye (MTG) to label the mitochondria of T cells from a donor (APC labeled). Separate T cells from a different donor were differentially -labeled (CD3-PE). These two distinct T cell populations were co-cultured alone or in the presence of the melanoma line SK-MEL-28 directly or in transwell . Only in the direct presence of tumor cells did we observe mitochondria transfer between T cells ( Fig S1 K). Furthermore, we t racked mitochondria redistribution b y cancer cells in human cancer pati ents, leveraging the computational mitochondria transfer inference program, MERCI25. Analysis of previously-validated human tumor scRNA-seq datasets revealed the redistribution of CD8+ T cell–derived mitochondria to multiple immune populations, including macrophages (Fig. 1L), dendritic cells, and B cells across multiple tumor types (Fig S1L). These findings establish mitochondrial redistribution as a conserved feature of both murine and human tumors. Mitochondria fusion in cancer cells prior to redistribution The fate of mitochondria acquired from donor cells is uncertain; it is unclear if they persist as separate entities or integrate into the host cell’s mitochondrial networ k 24,29. To address this in cancer cells, we labeled in situ tumor mitochondria network with mitotracker red (MTR), and co-cultured them with CD45.2mtD2+ leukocytes. Confocal microscopy revealed that the vast majority (>80%) of exogenous mtD2 mitochondria fused seamlessly with in situ endogenous tumor mitochondria (Fig. 2A–C). We frequently observed concatenated hybrid structures containing both red and green signals, indicating physical integration of the two pools (Fig 2B). We next asked whether fusion was mechanistically linked to redistribution. Prior reports suggested that tumor-to-CD8 T cell mitochondria transfer is relatively inefficient compared to immune-to-tumor transfer 16,25,30. Consistent with this, we found that cancer cells exported mitochondria at lower frequency and slower kinetics than they imported them ( Fig. 2D). After MTR+ B16: CD45.2mtD2+ leukocyte co-culture, we sorted mtD2+ cancer cells and co- cultured them with CD45. 1mtD2-. Confocal microscopy revealed transfer of fused mitochondria to CD45.1 cells ( Fig 2E ). The vast majority (~85 %) of mitochondria redistributed to CD45.1 cells were fused, mtD2+MTR+ hybrid mitochondria. About ~12% of cells were only mtD2 positive, and rare CD45.1 cells (~< 3%) were only MTR positive (Fig 2F). These findings indicate that cancer cells redistribute mitochondria not as isolated units, but predominantly as fused composites of endogenous and exogenous origin. To further validate the dispersion of fused mitochondria by cancer cells, we employed a melanoma line engineered by base editing to carry an m.12436G>A mutation 31. We performed triple co -culture with CD45.1 mtD2- and CD45.2mtD2+. We reasoned that if cancer dispersed fused mitochondria, m.12,436G>A mutation would be detected in CD45.1 immune .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint cells that become mtD2+. Indeed, we detected the cancer cell -derived m.12,436G>A point mutation in CD45. 1 cells, but only in the mtD2+ subset (Fig 2 G). To demonstrate the dispersion of fused mitochondria in vivo , we implanted m.1243680% tumors into mixed CD45.2mtD2+/CD45.1mtD2- chimeric mice. After tumor explant, we gated on CD45.1 cells and could detect m.12436 mutation in the mtD2+ subset (Fig 2H), similar to our in vitro studies. Similarly, we subcutaneously implanted m.12,436G>A m tumors into CD45.2NZB/CD45.1C57BL6/J mixed chimeric mice. CD45.1 cells were isolated, and we could confirm m.12436 mutation and NZB mtDNA in those cells (Fig S2A). These multiple lines of evidence indicate that cancer cells disperse fused mitochondria – a mixture of their own endogenous and exogenously-acquired mitochondria – to immune cells. Mechanisms of mitochondria redistribution Among the various mechanisms enabling mitochondrial transfer to cancer cells, tunneling nanotubes (TNTs) play the dominant role, whereas extracellular vesicles account for a modest, though measurable, proportion of transfer events 32 15,16,19. Therefore, we sought to determine whether TNTs similarly mediate dispersion of fused mitochondria from cancer cells to immune cells . Indeed, s uper resolution imaging revealed actin -rich, cellular projections between cancer cells and immune cells, containing fused mitochondria (Fig S2B), consistent with TNTs. Less frequently, we also identified transfer of anuclear, mitochondria- laden vesicular detachments from cancer ce lls ( Fig S 2C), consistent with extracellular vesicles. To determine the relative contribution of these modalities to mitochondria redistribution from cancer, we performed initial tumor: CD45.2 leukocyte co-culture to allow mtD2 transfer to tumor cells. Then, after washing away immune cells, we introduced CD45.1 immune cells as we described previously , or under 0.4 µm filter to block contact, or in the presence of cytochalasin B to block TNTs, or GW 4869 to block EV release . Inhibiting physical contact and TNTs substantially reduced mitochondria redistribution to immune cells by more than 85% By contrast, GW4869 only caused ~15% reduction in mitochondria redistribution to immune cells (Fig S2D). These results indicate that TNT formation between cancer cells and immune cells facilitate the majority of mitochondria redistribution. Redistributed mitochondria reprogram acceptor immune cells Next, we evaluated the function al consequence of mitochondria redistribution to recipient immune cells. Mitochondria are redistributed to many cell types ( Fig 3A). We focused on neutrophils, macrophages, and CD4 T cells because they are the dominant immune infiltrates in tumors that physically interact with tumors and contribute to the immunosuppressive TME33,34. After mtD2+ tumor: CD45.1 immune co -culture, we distinguished recipient CD45.1 cells on the basis of mtD2 and compared their phenotype with mtD2 - counterparts (Fig 3B-D). Neutrophils that acquired mitochondria (CD45.1⁺Ly6G⁺CD11b⁺mtD2⁺) showed increased pro-tumoral features, including elevated CD200R, PD -L1, and NETs formation (Fig 3B). Similarly, mtD2⁺ macrophages displayed higher levels of CD200R, PD-1, MerTK, and CD206 ( Fig. 3C ). In CD4 T cells, mitochondrial uptake promoted the expansion of .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint FoxP3⁺CD25⁺ regulatory T cells (Tregs) ( Fig. 3D ). These Tregs exhibited heightened immunosuppressive activity, with increased PD -1, CD69, and FoxP3 expression ( Fig. 3E), enhanced IL-10 production (Fig. 3F), and stronger suppression of target T cell proliferation (Fig. 3G ). Together, these data indicate that mitochondria redistributed from tumor cells reinforce immunosuppressive programs across multiple immune cell types. To further exa mine the direct role of exogenous fused mitochondria as a causal driver of immunosuppression, we prepared tumor: leukocyte co -cultures and harvested their mitochondria. Then, using mitoception protocol35, we directly introduced the fused cargo into different immune cells. Macrophages differentiated in the presence of exogenous fused mitochondria increased expression of CD206, MerTK, similar to our observation from direct dispersion of mitochondria from cancer cells (Fig S3A). Differentiation of CD4T cells in the presence of fused mitochondria also enhanced FoxP3, CD25, and PD -1 expression (Fig S3B). To characterize the phenotype of immune cells that accept fused mitochondria in vivo, we turned to the mixed CD45.2mtD2/CD45.1mtD2- mice challenged with B16 tumor. We gated on CD45.1 immune cells, and subgated macrophages and CD4 T cells on the basis of mtD2 status (Fig 3H and Fig S3C for gating strategy). F4/80+ CD45.1 cells that acquire exogenous mitochondria upregulated PD-L1, MerTK and CD206 (Fig 3H). CD4+ CD45.1 T cells that acquire exogenous mitochondria upregulated PD -1, FoxP3 and CD25 ( Fig 3 I). They produced more intracellular IL-10 and suppressed the proliferation of target cells with greater potency (Fig 3J). Thus, transfer of mitochondria also enhances immunosuppressive features in vivo. Beyond immunosuppressive cells, CD8+ T cells are also frequent recipients of redistributed mitochondria (Fig 3A ). We found that CD8+ T cell recipients of tumor -redistributed mitochondria showed evidence of exhaustion, including upregulation of PD -1, LAG3, and CD39 (Fig 3K) as well as a higher frequency of PD-1hiLAGhi cells (Fig 3L). Moreover, using SCENITH metabolic analysis 36, we found that these CD8+ T cells demonstrated marked decrease in total metabolic capacity, characterized by decreased mitochondria and glycolytic capacities (Fig 3M). Thus, while mitochondria redistribution enhances suppressive cells, it leads to functional and metabolic exhaustion of CD8+ T cells. To determine the impact of mitochondria dispersion on cancer outcomes, we subcutaneously challenged mice with B16. When tumors were established, we harvested fused mitochondria derived from tumor: immune co-cultured and directly injected them into the tumor bed (Fig 3N). Compared to tumors treated with vehicle, those injected with fused mitochondria showed a more rapid growth rate ( Fig 3O). Direct injection of mitochondria into the tumor bed did not change the frequency of immune cell infiltration into the tumors. However, this expanded the frequencies of Tregs, decreased the frequency CD8+ T cells, but had no impact on the frequency of infiltrating macrophages ( Fig 3P). Conversely, we treated cohorts of .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint B16 tumor-bearing mice with inhibitors of mitochondria transfer using a combination of EV inhibitors (GW4869) and TNT inhibitors ( L-778123)16 (Fig 3S D). Such therapy reduced tumor growth, limited the expansion of Tregs and increased the frequency of CD8 + T cells (Fig S3E). Therefore, mitochondria redistribution can augment tumor growth , whereas inhibition of mitochondria transfer can restrain tumor growth. Mitochondria fusion optimizes metabolic fitness of cancer cells Multiple recent reports indicate that mitochondria transfer can expand regulatory T cells 37- 39. In those studies, the transferred mitochondria were not fused with tumor mitochondria. Therefore, the dispersion of fused mitochondria by cancer cells may not be required to induce the expansion of regulatory cells. On that basis, we speculate d that cancer cells undergo mitochondria fusion for metabolic gains. To explore this possibility, we performed RNAseq of cancer cells with fused mitochondria. Compared to their respective controls , B16 or E0771 cancer cells with fused mitochondria had markedly i ncreased mitochondrial macromolecular biosynthetic capacity (Fig 4A). Membrane potential, measured by TMRM assay, was increased in mtD2+ cancer cells (Fig 4B). SCENITH metabolic analysis revealed that mtD2+ cancer cells substantially increased their total metabolic capacity, largely driven by enhanced mitochondria dependence ( Fig 4C ). These results suggested that exogenous mitochondria augment the mitochondrial macromolecular biosynthetic fitness of cancer cells. Under various bioenergetic demands, mitochondria can segregate into two conformations to enable them execute reductive synthesis and oxidative phosphorylation simultaneously. This is mediated by compartmentalization of mitochondria into two subpopulations: a canonical cristae-rich architecture, which enables complex V docking for ATP production; and a fibrillar architecture, which favors macromolecule biosynthesis. The latter conformation is associated with the sequestration of pyrroline-5-carboxylate synth ase (P5CS) filaments, which drives reductive biosynthe sis22. Therefore, we h ypothesized that transfer red mitochondria trigger P5CS filamentous structural changes in recipient cancer cells. We first tested whether cancer cells utilize mitochondria transfer to import P5CS-rich mitochondria from host immune cells. Measurement of P5CS levels in leukocytes compared with tumor cells showed more than a log -fold less P5CS expression intensity in leukocytes than in cancer cells (Fig 4D). Although tumor cells express abundant P5CS, we failed to colocalize P5CS with any of the transferred mitochondria in mtD2+ cancer cells (Fig 4E, and Fig S4 A). Conversely, when mtD2 + cancer cells redistribute mitochondria to CD45.1 acceptor cells, P5CS is not co-transferred (Fig 4E,F). These results suggest that cancer cells do not i mport P5CS-rich mitochondria and most likely exclude P5CS from the transferred mitochondrial pool . Indeed, n et P5CS increase in cancer cells after they disperse fused mitochondria is minimal (Fig S4B) and P5CS levels do not change in immune cells that accept fused mitochondria ( Fig S 4C). Strikingly, however, we fo und that exogenous mitochondria triggered fibrillar conformation of P5CS in recipient cancer cells (Fig 4G,H), .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint and many cancer cells with exogenous mitochondria showed fibrillar P5CS re-organization in clusters (Fig S4D). P5CS architectural changes are commonly driven by cellular stress response s22,40,41. Therefore, we asked whether the cellular stress associated with mitochondria fusion directly contributes to the filamentous architecture change in cancer cells after acquiring exogenous mitochondria. To that end, we performed B16: CD5.2mtD2+ co-cultures as we have described or in the presence of Myls22, a small molecule inhibitor of OpA1 42,43, which governs mitochondria inner membrane fusion 44. As an additional control for the potential mitochondrial dynamic changes that could be induced by Myls22, we treated separate tumor: immune co -culture cohorts with Mdivi -1, which promotes mitochondria fusion through Drp1.45,46 Disrupting endogenous and exogenous mitochondria fusion in cancer cells prevented the filamentous morphological changes in P5CS, whereas treatment with Mdivi-1 did not ( Fig 4I,J). Importantly, this disruption did not impact membrane potential but reduced the mitochondria-specific metabolic gains in mtD2+ cancer cells ( Fig 4K). These

indicate that the fusion of exogenous mitochondria with endogenous network promotes P5CS architectural organization, contributing to the enhanced metabolic fitness in cancer cells that hijack and redistribute mitochondria. Finally, we asked whether fusion of exogenous mitochondria can enable tumors to utilize dysfunctional, host-derived mitochondria. We created mitochondria reporter mouse on the

of mice lacking Ndufs4 (Nfuds4KO mtD2). Ndufs4 encodes mitochondria complex I and is essential for cellular energy production. Interestingly, the rate of mitochondria transfer from Nduf4 KO and Nduf4WT leukocytes were similar ( FigS4E). The mitochondria transfer from Nduf4 KO leukocytes still improved membrane potential and enhanced the bioenergetics of recipient cancer cells to the same extent as transfer from Nduf4WT leukocytes ( FigS4F, G ). Thus, mitochondria fusion enables cancer cells to repurpose suboptimal mitochondria to promote tumor metabolism.

Vast bioenergetic and biosynthetic resources are required to sustain tumor growth. Mitochondrial metabolism, in particular, fuels cancer cell proliferation and is also required by immunosuppressive cells to direct antitumor immunity7,9,13,47. Yet, the TME is hostile to mitochondrial metabolism. The highly mobile nature of mitochondria48, and the principles of maximum parsimony raised the prospects that a coordinated mechanism might exist to sustain the mitochondrial metabolism of cancer cells and their immunoregulatory partners. Here we uncover such a mechanism: cancer cells import exogenous mitochondria, fuse them with their own networks, and redistribute the fused payload to neighboring immune cell s to fuel the immunosuppressive TME. Mitochondria redistribution has divergent effect on recipient immunosuppressive and cytotoxic CD8 T cells , as it boosts the expansion of the .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint former and exhausts the latter. One possible explanation is that the transferred mitochondria accumulate reactive oxygen species 24. Whereas regulatory cells possess abundant bespoke antioxidant systems and are redox -resilient, cytotoxic effector CD8 T cells are relatively redox-vulnerable49-53. The fusion of exogenous and endogenous mitochondria also reconfigures tumor cell metabolism through P5CS conformational remodeling, thereby allowing cancer cells to couple ecosystem regulation with tumor-intrinsic metabolic gain. The ability of relatively small amounts of transferred mitochondria to dramatically boost the metabolism of recipient cells ha s been met with much skepticism 54,55. Our research shows that, at least in cancer, this can be achieved by P5CS architectural changes in the endogenous mitochondrial network. Whether such mechanisms contribute to the metabolic boost in other cells under various clinical and pathologic settings following transfer remains to be elucidated. The fact that cancer cells acquire and redistribute mitochondria in a fused form has several implications. This axis provides extraordinary metabolic flexibility. As we showed, can cer cells can even incorporate suboptimal mitochondria and still extract metabolic benefit . The maneuver c ould also enable tumors to selectively offload dysfunctional mitochondria to shape their surroundings. An example of the latter has recently been reported, wherein cancer cells transferred in situ mtDNA mutation to recipient CD8 T cells, resulting in impaired immune function30. The incorporation of fresh, host -derived mitochondria along with the expulsion of in situ mitochondria may provide mechanistic basis for how cancer cells, replete with mtDNA mutations31,56, avoid catastrophic collapse of mitochondrial function: they can exploit beneficial mutations while replenishing their network with host-derived mitochondria. Mitochondria transfer is frequently associated with dampening immune response and promoting tolerance in various settings. For instance, it improves inflammation resolution in the lung following pulmonary injury 57,58. In diabetic renal injury, it restricts macrophage inflammation59. Mitochondria transfer from mesenchymal cells also induces potent Tregs and suppressed Th1 proliferation 37-39. Consistent with those reports, cancer cell redistribution of fused mitochondria instructed immune cells in the TME towards immunosuppressive fates. The fundamental underpinning of mitochondria’s effect on immunity remains unclear. We speculate that sharing mtDNA may blur the genetic boundary between tumor and host, increasing camouflage, tolerance, and immune evasion. This perspective of organelle sharing evokes mitochondria symbiosis, the foundational, primordial act of cooperation that beget eukaryotic life, when proto mitochondria integrated with Asgard archaea 60,61. Thus, we propose that redistributed mitochondria from cancer do not only provide metabolic substrate for the expansion of regulatory cells, but that sharing hybrid mitochondria may regulate the TME by communicating cellular co-existence within the tumor ecosystem. In summary, we define mitochondrial redistribution as a mechanism by which tumors simultaneously reinforce their own biosynthetic capacity and construct an .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint immunosuppressive niche. Rather than passive metabolic scavengers, cancer cells are active distributors of organelles, coordinating growth with ecosystem control. Mitochondria redistribution may represent a central organizing principle of tumor biology, opening up new therapeutic levers distinct from current metabolic or checkpoint -based strategies. Given the deleterious impact on CD8 T cells, it may be clinically important to determine whether cancer-redistributed, fused mitochondria contribute to the exha ustion of CAR-T cells upon infiltrating the TME.

We probed numerous publicly available scRNAseq datasets of cancer patients from the CancerSCEM databas e62. Unfortunately, m any datasets were not candidates for computational inference of mitochondria redistribution using MERCI. This is largely attributed to insufficient coverage of mtDNA reads that failed to meet the rigorous threshold necessary for accurate prediction of mitochondria transfer from CD8+ T cells to other cells besides cancer cells (see Methods for details). Moreover, MERCI is only currently trained to predict mitochondria transfer from CD8 + T cells. As such, this limits our ability to extensively infer m itochondria redistribution in real world cases of human cancers. The emergence of mtscATACseq will enable more accurate prediction of mitochondria redistribution in solid cancer to understand the full impact of this mechanism in human disease. .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint

Mice Wildtype C57BL6/J CD45.1 (strain #000664), C57BL6/J CD45.2 (strain # 002014), NZB (strain #:000684), mtDendra2Flox/Flox (PhAM Flox; also referred to as mtD2, strain #18385), PhAMexcised (global mtD2 mitochondria reporter, strain #:018397), CD4Cre (strain # 022071), CD68CreERT2 (strain #:038175), CD8Cre (strain # 008766), Nduf4 -deficient mice (strain #:027058) were purchased from The Jackson Laboratory (Maine, USA) and bred in -house for experimental use. The global mitochondria reporter mice are on the CD45.2 background (CD45.2mtD2). To generate CD45.1 mtD2, mtD2 mice were crossed for more than 10 generations with CD45.1 to eliminate the CD45,2 allele. These mice were a generous gift from the Brestoff Laboratory (Washington University, St Louis). All mice had ad libitum access to food and water and were maintained in a specific -pathogen-free facility with a 12h:12h light: dark cycle (lights on from 0600 to 1800). Experiments were performed in male and female mice between 8 and 12 weeks. Animals were randomly assigned to groups per experiment, ensuring similar distribution of age and gender. No specific blinding method was used. Data represent at least 2 independent experiments that were pooled for analysis. All experiments were executed under the guidelines of the Institutional Animal Care and Use Committee (IACUC # 34210). Bone Marrow Chimera Briefly, recipient mice received two doses of 5 .5 Gy irradiation 3 hours apart. Thereafter, femurs of donor mice (CD45.1 mtD2- and CD45.2mtD2+) were flushed. Red blood cells were lysed and equal numbers of cells from each donor pool were mixed (total 10 x106 cells) and intravenously administered to irradiated recipient mice. Blood, spleen, and marrow were sampled from mixed chimeric mice to confirm successful engraftment prior to experimental use. Parabiosis Parabiosis was performed as previously described63. Parabiosis was performed using 7-week- old female mice, with each pair consisting of a wildtype CD45.1mtD2- and CD45.2mtD2+ mice. Mice were anesthetized with 2 –3% isoflurane and placed on a sterile surgical field. The corresponding lateral aspects (left for one mouse, right for the other) were shaved, cleaned with 70% ethanol and povidone -iodine, and incised from elbow to knee to expose t he underlying musculature. The skin was gently loosened from the subcutaneous tissue along the flank. The olecranon and k nee joints of each mouse were ligated together using nonabsorbable 5-0 silk sutures to ensure musculoskeletal alignment. The dorsal and ventral skin were then sutured together using 6-0 nylon monofilament in a continuous or interrupted .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint fashion. Following surgery, mice were maintained on a heating pad and monitored continuously until fully ambulatory. Analgesia was administered with sustained -release buprenorphine (1.0 mg/kg subcutaneously) and meloxicam (5 mg/kg) as needed. Pairs were housed in large cages with accessible food and water gel. Circulatory cross-engraftment was confirmed by flow cytometry for CD45.1/CD45.2 chimerism. In vivo animal studies (drug and mitochondria treatment) B16 melanoma cancer cells (1 × 10 6 cells) were subcutaneously injected in the flanks of syngeneic mice. The drug therapy was started on day 3 post tumor implantation. Each animal was intraperitoneally injected every alternate day with the vehicle (for the control group), αPD1 (10 mg kg–1) alone, combined L-778123 (80 mg kg–1) and GW4876 ( 3 mg kg–1) or a combination of all three drugs αPD1 (10 mg kg–1, L-778123 (80 mg kg–1), and GW4876 (3 mg kg–1). The tumors were measured at indicated times using a Vernier caliper, and the tumor volume (Vt) was calculated as per the following formula: L × B2/2, where L is the longest dimension and B is the shortest dimension. The total body weight was routinely measured to assess any gross toxicity. All the tumor tissues were harvested for further studies. The maximum permitted tumor volume (2 cm3) was not exceeded in any study. For adoptive mitochondria transfer, mitochondria were isolated from cancer cell and immune cell co-culture using the mitochondria isolation kit (Thermo Fisher Scientific, cat# PI89874). We typically prepared mitochondria from a co-culture of 1x107 and 1x107 blood leukocytes per mouse. Purified mitochondria were injected directly into tumor beginning day 3. Cancer cells, culture and treatments Tumor cells (B16F10, E0771, MC38, SK -MEL-28 were originally obtained from ATCC . Cells were retrovirally transduced to overexpress Tdtomato . Cells were cultured according to manufacturer’s conditions, typically consisting of RPMI 1640 or DMEM, 10% heat inactivated fetal bovine serum (FBS) and 1% penicillin and streptomycin. Cell lines are authenticated through STR profiling, and all lines were subjected to mycoplasma testing every 6 months using the Universal Mycoplasma Detection Kit. For in vitro co -cultures, tumor cells were seeded in 24 -well plates with immune cells , typically using 1 x10 5 tumor cells in1:1 ration unless otherwise specified . After 16 h or indicated time, cells were washed and evaluated for mitochondria transfer to tumor cells. For mitochondria dispersion, cancer cells were sorted from immune cells (CD45.2) and re - cultured with immune cells harvested from CD45.1 mice. For in vivo experiments, cells were implanted subcutaneously on the flanks of recipient mice at 1 x106 in 50 ul PBS. Between 14-21 days, tumors were explanted, digested in collagenase IV, and passed through 70 µm strainer to obtain single cell suspensions. Flow cytometry was then used to identify mtD2 expression by live tumor cells (Tdtomato+CD45-CD31-DAPI-). .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint Mitoception Mitoception was performed as previously described 35. Briefly, mitochondria from tumor: immune co-culture were harvested and prepared in a 500 µl suspension in DMEM/FCS 5%. The suspension was added slowly to the plated recipient cells, close to the bottom of the well, throughout the culture surface. The culture plates were then centrifuged at 1,500 g for 15 min at 4°C. They were then placed in a 37°C cell incubator prior to a second centrifugation in the same conditions, two hours later, and then returned to 37°C cell culture. Confocal Microscopy Cancer cells (1 x104) were seeded on # 1.5 round coverslip prior to the addition isolated mtD2 immune cells in a 1:1 ratio. The cells were co-cultured on the coverslip for indicated times at 37 °C in 5% CO2 incubator. The cells were then fixed with 4% paraformaldehyde and kept at 4˘C until staining. The samples were stained with 300 nM DAPI, Alexa Fluor 647 phalloidin (1:500), and anti -mouse CD45-PE (1:500) for 30 minutes, at room temperatu re, followed by two times washing with PBS. The images were acquired with a Zeiss LSM900 confocal microscope with Plan-Apochromat 63X/1.4 oil objective. Co-cultures of cancer and immune cells were imaged using a Leica Stellaris 5 White Light Laser (WLL) confocal microscope equipped with a motorized stage and a Leica DMi8 inverted microscope base. The system included a 405 nm solid-state laser and a tunable White Light Laser (WLL) providing excitation wavelengths from 485 nm to 790 nm. Fluorescence emission was detected by highly sensitive Power HyD S detectors, with spectral detection ranges optimized for each fluorophore to minimize crosstalk. All images were acquired with a 63x NA 1.4 oil immersion objective. Super-resolution imaging was achieved using the integrated LIGHTNING adaptive deconvolution module within the Leica LAS X software. This algorithm adaptively processes the image data by determining optimal deconvolution parameters for each voxel based on local image properties, thereby increasing image resolution and contrast beyond the diffraction limit. The deconvolution process was applied automatically during or immediately after acquisition to generate the super -resolution images presented throughout the manuscript. Original raw confocal images are also available. All image processing, including brightness, contrast adjustments, and 3D rendering, was performed using Leica LAS X software and ImageJ/Fiji (National Institutes of Health, USA). Antibodies were used at a dilution of 1:1 000 for immunofluorescence unless otherwise indicated. For multiplexed immunostaining, we used h ighly cross -adsorbed secondary antibodies to avoid cross -species contamination. The following antibodies were used: Phalloidin-iFluor™ 647 Conjugate (Cayman Chemicals, Cat # 20555), anti-rabbit IgG alexa fluor 647 (Highly cross -adsorbed, Invitrogen, A -31573; 1:500 for immunofluorescence) , .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint Mitotracker Red ( ThermoFisher), P5CS (Proteintech, 68184 -1-Ig, 1:500 for immunofluorescence). Flow Cytometry For flow cytometric analysis of ex vivo tumors, Tumors were digested with collagenase IV and Dnase I and m ashed through a 70 μm cell strainer (Falcon). After red blood cell lysis, cells were then stained with antibody cocktail in FACS buffer (PBS with 0.5% bovine serum albumin) at 4 degrees for 30 minutes. Thereafter, staining was quenched with washes with washing with FA Cs buffer and immediately analyzed on a flow cytometer. Antibody cocktails included DAPI or 7 -AAD to exclude dead cells. Samples were analyzed with FlowJo software 10.8.2 (BD Biosciences). The following antibodies (BioLegend) were used: anti CD45 ( clone 30-F11), anti-Ly6G (clone 1a8), anti CD11b (clone M1/70), anti-CD69 (clone H1.2F3) anti -CD3 ( clone 145-2C11), anti -CD8a ( clone 53-6.7), anti-PD-L1 (10F.9G2), anti-PD-1 (RMPI -30), were from Biolegend; LSR II or BDFortessa and BD FACSymphony (BD Biosciences) were used for flow cytometry acquisition and a FACSAria Fusion (BD Biosciences) or BD Influx (BD Biosciences) for cell sorting. SCENITH Single cell metabolic analys es of cells w ere performed using SCENITH 36,64 as previously described. In brief, after mitochondria transfer to cancer cells, or after mitochondria redistribution by cancer cells, each sample condition was treated with 10 µL of either wash media, 2-Deoxy-D-Glucose (2-DG, final concentration 100 mM), Oligomycin (Oligo, final concentration 1 μM), or a combination of 2-DG and Oligo. After 15 mins incubation at 37˚, all conditions were treated with 10.5 µ g/mL of puromycin, followed by a 40 mins, 37˚C incubation. Cells were washed in 200 µL cold PBS, centrifuged 500 x g for 3 mins at 4˚C and washed with FACS buffer. Zombie NIR viability dye, Fcblock, and surface staining was performed, followed by fixation, permeabilization and i ntracellular staining of puromycin using anti-puromycin (clone 12D10, Sigma-Aldrich, cat# MABE343-AF488). The samples were incubated 1 hr on ice, in the dark. Samples were washed with PermWash buffer (from the CytoFix/CytoPerm kit) and resuspended in 200 µL FACS buffer and subjected to flow cytometric analysis on a Cytek Aurora spectral flow cytometer configured with 4 lasers (violet, blue, yellow/ green, and red lasers), with 100 μL acquired to enable cell count enumeration. Chemicals Tetramethylrhodamine (TMRE, Invitrogen, T669), Mitotracker deep red (Thermo, M22426), Myls22 (TargetMol, T9127); Mito -Tempol (Medchem express), GW4869 (Selleck .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint Chemicals, S7609), Y27632 (Selleck Chemicals, S1049), cytochalasin B (Medchem Express, HY-16928). RNA seq After tumor-immune co-cultures, cells were resuspended in FACS buffer (PBS in 0.5% BSA) and stained with DAPI and CD45. Live tumor cells (TdTomato+CD45 -DAPI-) were sorted into mtD2+ and mtD2- subpopulations. Total mRNA was isolated using Qiagen RNEasy Plus isolation kit. Bulk RNAseq was performed using the NovaSeq platform. The raw RNA sequencing data was mapped to mouse reference genome hg38 using the STAR aligner, and genes annotated in Gencode v3664 was quantified using featurecounts in the subread package. The differential gene expression analysis was conducted in the DESeq2 package. Gene set enrichment analysis was performed with Gene Set Enrichment Analysis. Significantly different genes were identified by DESeq2 using Wald test. Gene annotation enrich ment analysis was performed using KEGG pathways and GO terms (biological process, cellular component, and molecular function). Functional annotation clustering was performed and terms with p < 0.05 (Benjamini corrected) are shown. Mitochondria DNA PCR After co-culture, mtD2 - and mtD2 - cells were FACS sorted. DNA isolation (Q iagen) following manufacturer’s instructions, qPCR reactions were designed with 100 ng of DNA and PowerUP SYBR Green Master Mix reagent (Thermo, A25742). Primers ARMS22 (5′ - TTATCCACGCTTCCGTTACGTC-3′) and MT20 (5′ -TGGCACTCCCGCTGTAAAAA- 3′) were used to amplify NZB mtDNA as previously described 65. NZB mtDNA (Ct) and the nDNA gene β -actin (Ct) were measured in samples obtained from tumors implanted into C57/NZB or WT/MtD2 chimeras or tumor cells treated with leukocytes isolated from either NZB or C57BL6/J mice. The NZB mtDNA signal was normalized to β -actin. Percentage of NZB mtDNA to the total mtDNA was calculated as NZB mtDNA/ (NZBmtDNA + C57mtDNA). Bio-Rad QX200 Droplet Digital PCR System was used to quantify m34126 mtDNA in immune cells. Primers for m34126 are available as previously reported 31. MERCI Computational inference of MERCI was performed as previously described 25. To validate mitochondria redistribution in human samples, we first analyze scRNAseq datasets and filter samples with two minimal criteria: 1). Each sample contained at least 100 CD8+ T cells and 100 cancer cells (benchmarked for internal control), and 100 of any target cell of interest (example: macrophages). 2) the fraction of transferred mitochondria in cancer cells and the target cell of interest must be significantly associated with the gene expression variations on UMAP (|ρ| > 0.2, p < 0.001). DNA and RNA scores were estimated using mtDNA mutations .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint and gene expression profiles as previously described25. Consistently, RNA and DNA scores with Rcm >1 cutoff was used as a criterion to select samples with mitochondria redistribution. We validate that receivers are sufficiently included in the input mixed cells. Samples that do not meet these criteria are excluded from downstream analysis. We then employ MERCI leave-one-out (LOO) pipeline to estimate the fraction of transferred mitochondria in target cells. Statistics Data are reported as mean ± standard error of the mean. Statistical analyses were performed in Prism v10 or v11 (Graphpad, La Jolla, CA) unless otherwise specified. Paired or unpaired Student’s t-tests were used for two -group comparisons, with Welch’s correction applied when the standard deviations between groups differed. One -way analysis of variance (ANOVA) with Tukey or Fisher’s LSD post -hoc testing was used for three -or-more group comparisons, and two-way ANOVA with Bonferroni post-hoc. Statistical significance was set at p95% purity using BD FACSAria II (BD Biosciences) or Bigfoot Spectral Cell Sorter (Thermo Fisher Scientific). AUTHOR CONTRIBUTIONS Conceptualization: DOD, AT Design & Investigation: AT, ATW, VS, KB, YT, CCR, VS, CK, MM, LW, VRR, PG, TR, JDA, BL, EE, BL, DO-D Formal Analysis: AZ, ATW, VS, YT, BL, DO-D Writing – Original Draft- DO-D, EE, AT Writing – Review & Editing - All Supervision: DO-D, BL, JDA, EE Funding Acquisition: DO-D Funding: This project was supported by NIH R01ES034235 (DO-D), 5U54CA274511 (EE), 5P01CA244114 (EE), NIH R01CA258524 (BL). Conflict: None

1. Hanahan, D., and Weinberg, R.A. (2011). Hallmarks of cancer: the next generation. Cell 144, 646-674. 10.1016/j.cell.2011.02.013. .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint 2. Delgoffe, G.M. (2016). Filling the Tank: Keeping Antitumor T Cells Metabolically Fit for the Long Haul. Cancer Immunol Res 4, 1001-1006. 10.1158/2326-6066.CIR- 16-0244. 3. Delgoffe, G.M., Xu, C., Mackall, C.L., Green, M.R., Gottschalk, S., Speiser, D.E., Zehn, D., and Beavis, P.A. (2021). The role of exhaustion in CAR T cell therapy. Cancer Cell 39, 885-888. 10.1016/j.ccell.2021.06.012. 4. Scharping, N.E., Menk, A.V., Moreci, R.S., Whetstone, R.D., Dadey, R.E., Watkins, S.C., Ferris, R.L., and Delgoffe, G.M. (2016). The Tumor Microenvironment Represses T Cell Mitochondrial Biogenesis to Drive Intratumoral T Cell Metabolic Insufficiency and Dysfunction. Immunity 45, 374 -388. 10.1016/j.immuni.2016.07.009. 5. Simula, L., Fumagalli, M., Vimeux, L., Rajnpreht, I., Icard, P., Birsen, G., An, D., Pendino, F., Rouault, A., Bercovici, N., et al. (2024). Mitochondrial metabolism sustains CD8. Nat Commun 15, 2203. 10.1038/s41467-024-46377-7. 6. Zheng, X., Qian, Y., Fu, B., Jiao, D., Jiang, Y., Chen, P., Shen, Y., Zhang, H., Sun, R., Tian, Z., and Wei, H. (2019). Mitochondrial fragmentation limits NK cell -based tumor immunosurveillance. Nat Immunol 20, 1656-1667. 10.1038/s41590-019-0511- 1. 7. Weinberg, S.E., Sena, L.A., and Chandel, N.S. (2015). Mitochondria in the regulation of innate and adaptive immunity. Immunity 42, 406 -417. 10.1016/j.immuni.2015.02.002. 8. DeBerardinis, R.J., and Chandel, N.S. (2016). Fundamentals of cancer metabolism. Sci Adv 2, e1600200. 10.1126/sciadv.1600200. 9. Vitale, I., Manic, G., Coussens, L.M., Kroemer, G., and Galluzzi, L. (2019). Macrophages and Metabolism in the Tumor Microenvironment. Cell Metab 30, 36 - 50. 10.1016/j.cmet.2019.06.001. 10. Warburg, O., Wind, F., and Negelein, E. (1927). THE METABOLISM OF TUMORS IN THE BODY. J Gen Physiol 8, 519-530. 10.1085/jgp.8.6.519. 11. Bartman, C.R., Weilandt, D.R., Shen, Y., Lee, W.D., Han, Y., TeSlaa, T., Jankowski, C.S.R., Samarah, L., Park, N.R., da Silva -Diz, V., et al. (2023). Slow TCA flux and ATP production in primary solid tumours but not metastases. Nature 614, 349 -357. 10.1038/s41586-022-05661-6. 12. Watson, M.J., Vignali, P.D.A., Mullett, S.J., Overacre-Delgoffe, A.E., Peralta, R.M., Grebinoski, S., Menk, A.V., Rittenhouse, N.L., DePeaux, K., Whetstone, R.D., et al. (2021). Metabolic support of tumour -infiltrating regulatory T cells by lactic acid. Nature 591, 645-651. 10.1038/s41586-020-03045-2. 13. Binnewies, M., Roberts, E.W., Kersten, K., Chan, V., Fearon, D.F., Merad, M., Coussens, L.M., Gabrilovich, D.I., Ostrand -Rosenberg, S., Hedrick, C.C., et al. (2018). Understanding the tumor immune microenvironment (TIME) for effective therapy. Nat Med 24, 541-550. 10.1038/s41591-018-0014-x. 14. Radha, G., and Lopus, M. (2021). The spontaneous remission of cancer: Current insights and therapeutic significance. Transl Oncol 14, 101166. 10.1016/j.tranon.2021.101166. .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint 15. Berridge, M.V., Zobalova, R., Boukalova, S., Caicedo, A., Rushworth, S.A., and Neuzil, J. (2025). Horizontal mitochondrial transfer in cancer biology: Potential clinical relevance. Cancer Cell 43, 803-807. 10.1016/j.ccell.2025.03.002. 16. Saha, T., Dash, C., Jayabalan, R., Khiste, S., Kulkarni, A., Kurmi, K., Mondal, J., Majumder, P.K., Bardia, A., Jang, H.L., and Sengupta, S. (2022). Intercellular nanotubes mediate mitochondrial trafficking between cancer and immune cells. Nat Nanotechnol 17, 98-106. 10.1038/s41565-021-01000-4. 17. Chandel, N.S., Falk, M.J., Santos, J.H., Brestoff, J.R., Lechuga-Vieco, A.V., Sancak, Y.S., Chen, Q., Elorza, A.A., and Quintana-Cabrera, R. (2025). Mitochondria transfer. Nat Metab. 10.1038/s42255-025-01364-0. 18. Hoover, G., Gilbert, S., Curley, O., Obellianne, C., Lin, M.T., Hixson, W., Pierce, T.W., Andrews, J.F., Alexeyev, M.F., Ding, Y., et al. (2025). Nerve -to-cancer transfer of mitochondria during cancer metastasis. Nature. 10.1038/s41586 -025- 09176-8. 19. Watson, D.C., Bayik, D., Storevik, S., Moreino, S.S., Sprowls, S.A., Han, J., Augustsson, M.T., Lauko, A., Sravya, P., Røsland, G.V., et al. (2023). GAP43 - dependent mitochondria transfer from astrocytes enhances glioblastoma tumorigenicity. Nat Cancer 4, 648-664. 10.1038/s43018-023-00556-5. 20. Dong, L.F., Kovarova, J., Bajzikova, M., Bezawork-Geleta, A., Svec, D., Endaya, B., Sachaphibulkij, K., Coelho, A.R., Sebkova, N., Ruzickova, A., et al. (2017). Horizontal transfer of whole mitochondria restores tumorigenic potential in mitochondrial DNA-deficient cancer cells. Elife 6. 10.7554/eLife.22187. 21. Tan, A.S., Baty, J.W., Dong, L.F., Bezawork -Geleta, A., Endaya, B., Goodwin, J., Bajzikova, M., Kovarova, J., Peterka, M., Yan, B., et al. (2015). Mitochondrial genome acquisition restores respiratory function and tumorigenic potential of cancer cells without mitochondrial DNA. Cell Metab 21, 81-94. 10.1016/j.cmet.2014.12.003. 22. Ryu, K.W., Fung, T.S., Baker, D.C., Saoi, M., Park, J., Febres -Aldana, C.A., Aly, R.G., Cui, R., Sharma, A., Fu, Y., et al. (2024). Cellular ATP demand creates metabolically distinct subpopulations of mitochondria. Nature 635, 746 -754. 10.1038/s41586-024-08146-w. 23. Pham, A.H., McCaffery, J.M., and Chan, D.C. (2012). Mouse lines with photo - activatable mitochondria to study mitochondrial dynamics. Genesis 50, 833 -843. 10.1002/dvg.22050. 24. Kidwell, C.U., Casalini, J.R., Pradeep, S., Scherer, S.D., Greiner, D., Bayik, D., Watson, D.C., Olson, G.S., Lathia, J.D., Johnson, J.S., et al. (2023). Transferred mitochondria accumulate reactive oxygen species, promoting proliferation. Elife 12. 10.7554/eLife.85494. 25. Zhang, H., Yu, X., Ye, J., Li, H., Hu, J., Tan, Y., Fang, Y., Akbay, E., Yu, F., Weng, C., et al. (2023). Systematic investigation of mitochondrial transfer between cancer cells and T cells at single -cell resolution. Cancer Cell 41, 1788 -1802.e1710. 10.1016/j.ccell.2023.09.003. 26. Zebedee, S.L., Barritt, D.S., and Raschke, W.C. (1991). Comparison of mouse Ly5a and Ly5b leucocyte common antigen alleles. Dev Immunol 1, 243 -254. 10.1155/1991/52686. .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint 27. Machado, T.S., Macabelli, C.H., Sangalli, J.R., Rodrigues, T.B., Smith, L.C., Meirelles, F.V., and Chiaratti, M.R. (2015). Real -Time PCR Quantification of Heteroplasmy in a Mouse Model with Mitochondrial DNA of C57BL/6 and NZB/BINJ Strains. PLoS One 10, e0133650. 10.1371/journal.pone.0133650. 28. Nakai, R., Varnum, S., Field, R.L., Shi, H., Giwa, R., Jia, W., Krysa, S.J., Cohen, E.F., Borcherding, N., Saneto, R.P., et al. (2024). Mitochondria transfer -based therapies reduce the morbidity and mortality of Leigh syndrome. Nat Metab 6, 1886- 1896. 10.1038/s42255-024-01125-5. 29. Lin, R.Z., Im, G.B., Luo, A.C., Zhu, Y., Hong, X., Neumeyer, J., Tang, H.W., Perrimon, N., and Melero -Martin, J.M. (2024). Mitochondrial transfer mediates endothelial cell engraftment through mitophagy. Nature 629, 660 -668. 10.1038/s41586-024-07340-0. 30. Ikeda, H., Kawase, K., Nishi, T., Watanabe, T., Takenaga, K., Inozume, T., Ishino, T., Aki, S., Lin, J., Kawashima, S., et al. (2025). Immune evasion through mitochondrial transfer in the tumour microenvironment. Nature. 10.1038/s41586 - 024-08439-0. 31. Mahmood, M., Liu, E.M., Shergold, A.L., Tolla, E., Tait -Mulder, J., Huerta -Uribe, A., Shokry, E., Young, A.L., Lilla, S., Kim, M., et al. (2024). Mitochondrial DNA mutations drive aerobic glycolysis to enhance checkpoint blockade response in melanoma. Nat Cancer 5, 659-672. 10.1038/s43018-023-00721-w. 32. Rustom, A., Saffrich, R., Markovic, I., Walther, P., and Gerdes, H.H. (2004). Nanotubular highways for intercellular organelle transport. Science 303, 1007-1010. 10.1126/science.1093133. 33. Camargo, S., Moskowitz, O., Giladi, A., Levinson, M., Balaban, R., Gola, S., Raizman, A., Lipczyc, K., Richter, A., Keren-Khadmy, N., et al. (2025). Neutrophils physically interact with tumor cells to form a signaling niche promoting breast cancer aggressiveness. Nat Cancer 6, 540-558. 10.1038/s43018-025-00924-3. 34. Matusiak, M., Hickey, J.W., van IJzendoorn, D.G.P., Lu, G., Kidziński, L., Zhu, S., Colburg, D.R.C., Luca, B., Phillips, D.J., Brubaker, S.W., et al. (2024). Spatially Segregated Macrophage Populations Predict Distinct Outcomes in Colon Cancer. Cancer Discov 14, 1418-1439. 10.1158/2159-8290.CD-23-1300. 35. Caicedo, A., Fritz, V., Brondello, J.M., Ayala, M., Dennemont, I., Abdellaoui, N., de Fraipont, F., Moisan, A., Prouteau, C.A., Boukhaddaoui, H., et al. (2015). MitoCeption as a new tool to assess the effects of mesenchymal stem/stromal cell mitochondria on cancer cell metabolism and function. Sci Rep 5, 9073. 10.1038/srep09073. 36. Argüello, R.J., Combes, A.J., Char, R., Gigan, J.P., Baaziz, A.I., Bousiquot, E., Camosseto, V., Samad, B., Tsui, J., Yan, P., et al. (2020). SCENITH: A Flow Cytometry-Based Method to Functionally Profile Energy Metabolism with Single - Cell Resolution. Cell Metab 32, 1063-1075.e1067. 10.1016/j.cmet.2020.11.007. 37. Court, A.C., Le-Gatt, A., Luz-Crawford, P., Parra, E., Aliaga-Tobar, V., Bátiz, L.F., Contreras, R.A., Ortúzar, M.I., Kurte, M., Elizondo -Vega, R., et al. (2020). .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint Mitochondrial transfer from MSCs to T cells induces Treg differentiation and restricts inflammatory response. EMBO Rep 21, e48052. 10.15252/embr.201948052. 38. Do, J.S., Zwick, D., Kenyon, J.D., Zhong, F., Askew, D., Huang, A.Y., Van't Hof, W., Finney, M., and Laughlin, M.J. (2021). Mesenchymal stromal cell mitochondrial transfer to human induced T-regulatory cells mediates FOXP3 stability. Sci Rep 11, 10676. 10.1038/s41598-021-90115-8. 39. Piekarska, K., Urban-Wójciuk, Z., Kurkowiak, M., Pelikant-Małecka, I., Schumacher, A., Sakowska, J., Spodnik, J.H., Arcimowicz, Ł., Zielińska, H., Tymoniuk, B., et al. (2022). Mesenchymal stem cells transfer mitochondria to allogeneic Tregs in an HLA-dependent manner improving their immunosuppressive activity. Nat Commun 13, 856. 10.1038/s41467-022-28338-0. 40. Yang, Z., Zhao, X., Shang, W., Liu, Y., Ji, J.F., Liu, J.P., and Tong, C. (2021). Pyrroline-5-carboxylate synthase senses cellular stress and modulates metabolism by regulating mitochondrial respiration. Cell Death Differ 28, 303-319. 10.1038/s41418- 020-0601-5. 41. Pérez-Arellano, I., Carmona -Alvarez, F., Martínez, A.I., Rodríguez -Díaz, J., and Cervera, J. (2010). Pyrroline -5-carboxylate synthase and proline biosynthesis: from osmotolerance to rare metabolic disease. Protein Sci 19, 372-382. 10.1002/pro.340. 42. Noguchi, M., Kohno, S., Pellattiero, A., Machida, Y., Shibata, K., Shintani, N., Kohno, T., Gotoh, N., Takahashi, C., Hirao, A., et al. (2023). Inhibition of the mitochondria-shaping protein Opa1 restores sensitivity to Gefitinib in a lung adenocarcinomaresistant cell line. Cell Death Dis 14, 241. 10.1038/s41419 -023- 05768-2. 43. Herkenne, S., Ek, O., Zamberlan, M., Pellattiero, A., Chergova, M., Chivite, I., Novotná, E., Rigoni, G., Fonseca, T.B., Samardzic, D., et al. (2020). Developmental and Tumor Angiogenesis Requires the Mitochondria -Shaping Protein Opa1. Cell Metab 31, 987-1003.e1008. 10.1016/j.cmet.2020.04.007. 44. Giacomello, M., Pyakurel, A., Glytsou, C., and Scorrano, L. (2020). The cell biology of mitochondrial membrane dynamics. Nat Rev Mol Cell Biol 21, 204 -224. 10.1038/s41580-020-0210-7. 45. Kim, H., Lee, J.Y., Park, K.J., Kim, W.H., and Roh, G.S. (2016). A mitochondrial division inhibitor, Mdivi -1, inhibits mitochondrial fragmentation and attenuates kainic acid-induced hippocampal cell death. BMC Neurosci 17, 33. 10.1186/s12868- 016-0270-y. 46. Bordt, E.A., Clerc, P., Roelofs, B.A., Saladino, A.J., Tretter, L., Adam -Vizi, V., Cherok, E., Khalil, A., Yadava, N., Ge, S.X., et al. (2017). The Putative Drp1 Inhibitor mdivi-1 Is a Reversible Mitochondrial Complex I Inhibitor that Modulates Reactive Oxygen Species. Dev Cell 40, 583-594.e586. 10.1016/j.devcel.2017.02.020. 47. Mills, E.L., Kelly, B., and O'Neill, L.A.J. (2017). Mitochondria are the powerhouses of immunity. Nat Immunol 18, 488-498. 10.1038/ni.3704. 48. Dong, L.F., Rohlena, J., Zobalova, R., Nahacka, Z., Rodriguez, A.M., Berridge, M.V., and Neuzil, J. (2023). Mitochondria on the move: Horizontal mitochondrial transfer in disease and health. J Cell Biol 222. 10.1083/jcb.202211044. .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint 49. Vijayan, V., Wagener, F.A.D.T., and Immenschuh, S. (2018). The macrophage heme- heme oxygenase -1 system and its role in inflammation. Biochem Pharmacol 153, 159-167. 10.1016/j.bcp.2018.02.010. 50. Bea, F., Hudson, F.N., Chait, A., Kavanagh, T.J., and Rosenfeld, M.E. (2003). Induction of glutathione synthesis in macrophages by oxidized low -density lipoproteins is mediated by consensus antioxidant response elements. Circ Res 92, 386-393. 10.1161/01.RES.0000059561.65545.16. 51. Angelin, A., Gil-de-Gómez, L., Dahiya, S., Jiao, J., Guo, L., Levine, M.H., Wang, Z., Quinn, W.J., Kopinski, P.K., Wang, L., et al. (2017). Foxp3 Reprograms T Cell Metabolism to Function in Low-Glucose, High-Lactate Environments. Cell Metab 25, 1282-1293.e1287. 10.1016/j.cmet.2016.12.018. 52. Kishton, R.J., Sukumar, M., and Restifo, N.P. (2017). Metabolic Regulation of T Cell Longevity and Function in Tumor Immunotherapy. Cell Metab 26, 94 -109. 10.1016/j.cmet.2017.06.016. 53. Scharping, N.E., Rivadeneira, D.B., Menk, A.V., Vignali, P.D.A., Ford, B.R., Rittenhouse, N.L., Peralta, R., Wang, Y., DePeaux, K., Poholek, A.C., and Delgoffe, G.M. (2021). Mitochondrial stress induced by continuous stimulation under hypoxia rapidly driv es T cell exhaustion. Nat Immunol 22, 205 -215. 10.1038/s41590 -020- 00834-9. 54. Bertero, E., Maack, C., and O'Rourke, B. (2018). Mitochondrial transplantation in humans: "magical" cure or cause for concern? J Clin Invest 128, 5191 -5194. 10.1172/JCI124944. 55. Lightowlers, R.N., Chrzanowska -Lightowlers, Z.M., and Russell, O.M. (2020). Mitochondrial transplantation-a possible therapeutic for mitochondrial dysfunction?: Mitochondrial transfer is a potential cure for many diseases but proof of efficacy and safety is still lacking. EMBO Rep 21, e50964. 10.15252/embr.202050964. 56. Gorelick, A.N., Kim, M., Chatila, W.K., La, K., Hakimi, A.A., Berger, M.F., Taylor, B.S., Gammage, P.A., and Reznik, E. (2021). Respiratory complex and tissue lineage drive recurrent mutations in tumour mtDNA. Nat Metab 3, 558-570. 10.1038/s42255- 021-00378-8. 57. Morrison, T.J., Jackson, M.V., Cunningham, E.K., Kissenpfennig, A., McAuley, D.F., O'Kane, C.M., and Krasnodembskaya, A.D. (2017). Mesenchymal Stromal Cells Modulate Macrophages in Clinically Relevant Lung Injury Models by Extracellular Vesicle Mitochondr ial Transfer. Am J Respir Crit Care Med 196, 1275 -1286. 10.1164/rccm.201701-0170OC. 58. Islam, M.N., Das, S.R., Emin, M.T., Wei, M., Sun, L., Westphalen, K., Rowlands, D.J., Quadri, S.K., Bhattacharya, S., and Bhattacharya, J. (2012). Mitochondrial transfer from bone -marrow-derived stromal cells to pulmonary alveoli protects against acute lung injury. Nat Med 18, 759-765. 10.1038/nm.2736. 59. Yuan, Y., Yuan, L., Li, L., Liu, F., Liu, J., Chen, Y., Cheng, J., and Lu, Y. (2021). Mitochondrial transfer from mesenchymal stem cells to macrophages restricts inflammation and alleviates kidney injury in diabetic nephropathy mice via PGC -1α activation. Stem Cells 39, 913-928. 10.1002/stem.3375. .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint 60. Eme, L., Spang, A., Lombard, J., Stairs, C.W., and Ettema, T.J.G. (2017). Archaea and the origin of eukaryotes. Nat Rev Microbiol 15, 711 -723. 10.1038/nrmicro.2017.133. 61. Sagan, L. (1967). On the origin of mitosing cells. J Theor Biol 14, 255 -274. 10.1016/0022-5193(67)90079-3. 62. Zeng, J., Nie, Z., Shang, Y., Mai, J., Zhang, Y., Yang, Y., Xu, C., Zhao, J., Fan, Z., and Xiao, J. (2025). CancerSCEM 2.0: an updated data resource of single -cell expression map across various human cancers. Nucleic Acids Res 53, D1278-D1286. 10.1093/nar/gkae954. 63. Kamran, P., Sereti, K.I., Zhao, P., Ali, S.R., Weissman, I.L., and Ardehali, R. (2013). Parabiosis in mice: a detailed protocol. J Vis Exp. 10.3791/50556. 64. Vogel, A., García González, P., and Argüello, R.J. (2024). Measuring the Metabolic State of Tissue-Resident Macrophages via SCENITH. Methods Mol Biol 2713, 363- 376. 10.1007/978-1-0716-3437-0_25. 65. Venegas, V., and Halberg, M.C. (2012). Quantification of mtDNA mutation heteroplasmy (ARMS qPCR). Methods Mol Biol 837, 313 -326. 10.1007/978 -1- 61779-504-6_21. .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint FIGURES AND LEGENDS Figure 1. Cancer cells redistribute exogenously acquired mitochondria A) Gating strategy for the detection of transferred CD45.2 -derived, labeled mitochondria (mtD2) to cancer cells and to CD45.1 immune cells. B) mtD2 mean fluorescence intensity in CD45.1 cells cultured with CD45.2 mtD2+ cells alone, or with CD45.2 mtD2+ cells along with cancer cells (B16 melanoma). n=6 mtD2 mice and CD45.2 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint recipients, 3 independent times. C) Detection of mtD2 in CD45.1 cells cultured alone, or in the presence of CD45.2mtD2+ cells along with the direct or indirect cancer cell co -culture. D) Schematic for the culture of B16 melanoma with CD45.2 mtD2+ cells, followed by sorting of mtD2+ cancer cells and subsequent co -culture with CD45.1 leukocytes. E) mtD2 detection in CD45.1 cells co-cultured with mtD2+ tumor cells. F) Representative immunofluorescence images of CD45.1 immune cells (neutrophils, macrophages and lymphocytes) that acquired mtD2 protein from cancer cells previously co-cultured with CD45.2mtD2+ cells. Scale bar= 5 µm. G) PCR detection rate of NZB mtDNA in CD45.1 cells that were co -cultured with cancer cells that had interacted with NZB leukocytes. H) Schema for the establishment of mixed chimeric mice from CD45.1mtD2- and CD45.2 mtD2+ donors. n=6 chimeras. I) Detection of mtD2 in CD45.1 cells in the mixed chimeric mice challenged with B16 tumors. J) Detection of mtD2 in CD45.1 cells from CD45.2 mtD2+ and CD45.1mtD2- parabionts challenged with tumor. n= 4 parabionts. K) PCR detection of NZB mtDNA heteroplasmy in CD45.1 cells in the tumor microenvironment of NZB/C57 mixed chimeric mice implanted with tumors. n= 6 mixed chimeras per group. L) Computational inference of redistributed mitochondria from CD8 + T cells to macrophages in the tumor microenvironment of human patients with basal cell carcinoma and melanoma from GSE123814. *p<0.05, **p<0.01, ***p<0.001 by unpaired Student’s t test except c (ANOVA). .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint Figure 2. Mitochondria are fused prior to redistribution A) Representative immunofluorescence staining tumor cells without or with exogenous mitochondria . B) 3 - dimensional rendition of cancer cell with fused mtD2 (green) and endogenous mitochondria (red). C)Assessment of exogenous mtD2 mitochondria that are fused (attached to endogenous mitochondria) or alone within cancer cell. D) Rate of endogenous mitochondria transfer from cancer cells to immune cells (blue) and rate of mitochondria transfer from immune cells to cancer cells (black). E) Representative immunofluorescence images of CD45.1 neutrophils, macrophages and lymphocytes with fused exogenous mitochondria obtained from B16 cancer cells. F) Quantification of mitochondria dispersed from cancer cells to immune cells. G) Detection rate of m12,436 in CD45.1 co -cultured with cancer cells harboring exogenous mtD2. H) Detection rate of m12,436 in CD45.1 cells from CD45.1mtD2-/CD45.2mtD2+ mixed chimeric mice challenged with tumor. ***<0.001 by unpaired T test except f (ANOVA). .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint Figure 3. Dispersed mitochondria reprogram acceptor immune cells A) Schematic for B16: CD45.2mtD2 co-culture followed by redistribution of mitochondria to CD45.1 cells . B) Relative expression of PD -L1, NETs and CD200R in CD45.1 neutrophils that accept mtD2 versus neutrophils that do not. C) Relative expression of CD206, MerTK, PD-1 in macrophages that accept exogenous mtD2 versus mtD2- macrophages. D) Expression of Foxp3, CD25+ Tregs in CD4 cells that acquire exogenous mtD2 mitochondria. E) Relative expression of FoxP3, PD -1 and CD69 in Tregs that have accepted exogenous mitochondria (mtD2+) or not (mtD2 -). F) IL-10 secretion and suppressive capacity . G) of mtD2+ and mtD2 - Tregs. H-I) Phenotype of CD45.1 macrophages and Tregs in vivo from CD45.2 mtD2 and CD45.1 mtD2- mixed chimeras inoculated with tumors. Relative expression of CD206, MerTK and PD-L1 by mtD2+ and mtD2- F4/80 macrophages (CD45.1+ cells). J) Relative IL-10 production and proliferative suppression of target cells in mtD2+ .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint and mtD2- Tregs in vivo. K-M) Profile of CD8+ T cells that accept mitochondria from tumors versus those that did not. K shows expression of exhaustion markers on mtD2+ CD8 T cells relative to control. L) shows percentage of CD8+ T cells that express PD -1 and LAG3. M) shows SCENITH metabolic profiling of metabolic capacity, mitochondria and glycolytic capacities. N) Schematic for tumor inoculation and assessment of tumor growth after intratumoral administration of fused mitochondria. O) B16 tumor growth in mice administered with vehicle or fused mitochondria (n=8 per group). P) Frequencies of total immune cells, Tregs, CD8 T cells, and macrophages in control versus mitochondria administered tumors. p<0.05, **p<0.01, ***p<0.005. NS indicates not significant. Statistical analyses were by Student’s test; O by two-way ANOVA followed by Bonferroni correction. .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint Figure 4. Mitochondria dispersion improves metabolic fitness of cancer cells. A) Metabolic pathway enrichment in tumor cells with exogenous mitochondria. n= 3 independent pairs of mtD2+ and mtD2- cancer cells from the same experiment. B) Membrane potential mtD2+ vs mtD2 - cancer cells. n=5 matched cohorts. C) Metabolic profiling of mtD2+ vs mtD2- B16 cancer cells by SCENITH. D) P5CS expression levels in leukocytes and melanoma cells. E) Representative immunofluorescence of P5CS staining of cancer cells that have acquired mtD2 vs mtD2 - cancer cells. P5CS does not colocalize with mtD2. Bottom panel: Immunostaining of P5CS expression in accepting immune cells. P5CS does not colocalize with transferred, fused mitochondria. F) Pearson correlation coefficient analysis for mitochondria areas overlapping with P5CS in tumor cells with or without mtD2, and in immune cells that have recei ved redistributed mtD2 or not. G) Immunofluorescence images of P5CS conformation in mtD2+ vs mtD2 - cancer cell. H) Quantification of the frequency of mtD2+ and mtD2 - cancer cells with filaments P5CS conformation. I) Representative images of P5CS architecture in cancer cells with exogenous mitochondria (mtD2) in the presence of myls22 or mdivi-1. J) Quantification of P5CS filament in mtD2+ cancer cells treated with fusion or fission inhibitors versus control. K) Effect of blocking mitochondria fusion on metabolic capacity of mtD2+ cancer cells. *p<0.05, **p<0.01, ***p<0.005. NS indicates not significant. Statistics were performed using paired t test (B), unpaired t test (C) and ANOVA (F). bar indicates 5 µm. .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted November 13, 2025. ; https://doi.org/10.1101/2025.11.11.687895doi: bioRxiv preprint