The coenzyme A precursor pantethine restrains sarcoma growth through promotion of type 1 immunity

doi:10.21203/rs.3.rs-2345803/v1

Download PDF

Article

The coenzyme A precursor pantethine restrains sarcoma growth through promotion of type 1 immunity

https://doi.org/10.21203/rs.3.rs-2345803/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

The tumor microenvironment is a dynamic network of stromal, cancer and immune cells that interact and compete for resources. Mitochondria play an essential role in the control of metabolic plasticity and contribute to tumor progression and immune cell functionality. We previously identified the Vanin1 pathway as a tumor suppressor of sarcoma development via vitamin B5 and coenzyme A regeneration. Using an aggressive sarcoma cell line that lacks Vnn1 expression, we showed that administration of pantethine, a vitamin B5 precursor, impairs tumor growth in immunocompetent mice. Pantethine boosts anti-tumor type 1 immunity including polarization of myeloid and dendritic cells towards enhanced IFNγ-driven antigen presentation pathways and improved development of hypermetabolic effector CD8⁺ T cells endowed with potential anti-tumor activity. At later stages of treatment, the effect of pantethine is limited by the development of immune cell exhaustion. Nevertheless, its activity is comparable to that of anti-PD1 treatment in sensitive tumors. In humans, VNN1 expression correlates with improved survival and immune cell infiltration in soft tissue sarcomas but not osteosarcomas. Pantethine could be a potential therapeutic immunoadjuvant for the development of anti-tumor immunity.

Biological sciences/Immunology/Tumour immunology

Biological sciences/Cancer

Metabolic rewiring plays an essential role to maximize the exploitation of metabolic resources by tumors. Genetic alterations or environmental cues such as hypoxia drive changes in mitochondrial activity that limit reactive oxygen species (ROS) production while preserving anabolic pathways. Tumor-driven metabolic changes impose competition for oxygen, glucose or amino acids and lead to tissue reorganization and impairment of immune responses (1–3). Hypoxic conditions can transiently boost glycolysis and the development of effector T cell responses (4, 5). In contrast, under chronic stimulation, hypoxia pushes CD8⁺ T cells towards exhaustion (6). Exhausted CD8⁺ lymphocytes display abnormal mitochondria biogenesis (7), with a reduction of cristae structure and mitochondria/endoplasmic reticulum (ER) contacts (8). Accordingly, PGC1α-transfection in CD8⁺ T cells (9) or pharmacologically supplementation by NAD⁺ (8) boosted the development of anti-tumor CD8⁺ tumor-infiltrating lymphocytes (TILs). Since the loss of mitochondrial fitness contributes to exhaustion (10), pharmacological strategies modulating mitochondrial activity may induce distinct and possibly exploitable effects on both tumor and immune cells.

Vanin1 (Vnn1) acts as a negative regulator of sarcoma development in mouse by antagonizing the Warburg effect through increased mitochondrial respiration (11). Indeed, both serum and tissue VNN1 pantetheinase isoforms contribute to the maintenance of coenzyme A (CoA) pools by regenerating pantothenate / vitamin B5 (PanSH / vitB5) (FigS1A) (12, 13). Therefore, the lack of Vnn1 expression in an energy-demanding tissue such as a tumor (11) or an inflamed tissue (14) leads to a VitB5 and consequently CoA deficit. This defect could be rescued by Vnn1 overexpression or administration of pantethine, a stable dimeric form of pantetheine, leading to enhanced tissue tolerance to stress (15). Interestingly, VNN1 overexpression by cancer cells was associated with signatures of immune cell activation, in spite of the lack of VNN1 expression by mouse immunocytes (11). This suggested that VNN1-mediated restoration of vitB5 levels in a tumor microenvironment could paracrinally activate immune responses. Interestingly, it was recently found that vitB5 or CoA could boost in vitro macrophage or CD8⁺ T lymphocyte maturation, respectively (16, 17). We therefore tested whether, independently of VNN1 expression, enhancing vitB5 levels via pantethine administration to a tumor-bearing mouse might simultaneously limit tumor growth while boosting anti-tumor immunity.

Pantethine therapy reduces MCA205 growth in immunocompetent mice

We implanted the MCA205 (MCA) sarcoma cells in immunocompetent mice to test whether a systemic administration of pantethine (Pant) would compensate for the lack of Vnn1 in the tumor mass (18). Daily Pant therapy was started on established tumors at day 10 post cell engraftment (Fig. 1A). Growth reduction occurred within a few days post injection and persisted during the treatment period without leading to a complete tumor regression (Fig. 1B). A similar result was obtained by co-injecting PanSH and cysteamine (CEA), the products of pantetheinase activity (FigS1A-B). Importantly, when Pant was administered to a Nude mouse bearing MCA tumor, no growth inhibition was detected (Fig. 1C). This result suggested that Pant effect might directly or indirectly regulate immune responses. Since Pant enhances mitochondrial activity (11), we compared its effect to that of dichloroacetate (DCA), a compound that increases pyruvate uptake in mitochondria via PDK inhibition (19) and has previously been shown to exert anti-tumor activity (20, 21). As shown in Fig. 1D, Pant and DCA comparably reduced MCA tumor growth without synergy. We then performed a global metabolomics analysis of tumors from mice treated with vehicle PBS, Pant, DCA or combined drugs. Tumors harvested on day 28 (D28) showed that the global metabolic changes were not superimposable (Fig. 1E). Pant enhanced the level of metabolites associated with stress response, amino acids and nucleotide metabolism whereas DCA enhanced carbohydrate metabolism and mitochondrial activity. Interestingly, their association had a significant effect on the production of mitochondrial energetic metabolites and biogenic amines, a phenotype confirmed by nuclear magnetic resonance analysis (Fig S1C). We then focused our metabolic analysis on in vitro grown or ex vivo isolated cancer cells. In vitro, addition of Pant slightly enhanced the oxygen consumption (OCR) and extracellular acidification rates (ECAR) measured by a Seahorse analysis (Fig. 1F), and more significantly the production of mitochondrial ROS (mtROS) (Fig. 1G). In vivo, from D24 to D28, the mitochondrial polarization over mass ratio, scored by flow cytometry using the MitoTracker Deep Red (MDR) and MitoTracker Green (MG) probes, respectively, increased similarly between PBS and Pant samples (Fig. 1H-S1D) (22). Since alterations in mitochondrial metabolism can contribute to the reinforcement of a Warburg effect, we tested lactate production in tumor masses (Fig. 1I) and quantified by qRT-PCR the expression level of transcripts associated with an increased glycolytic flow on isolated cancer cells (Fig. 1J). The results showed that neither lactate nor transcript levels were modified by Pant. Altogether, Pant administration modulates metabolic cues in the tumor but does not fully rescue mitochondrial activity nor reduce the Warburg phenotype associated to cancer cell growth. These findings suggest that, in the MCA model, Pant might rather impact the activity of tumor-associated cells such as immunocytes.

Pantethine enhances myeloid signatures associated with antigen presentation and type 1 immunity

We performed a follow-up of immune cell infiltration at different stages of MCA tumor development using flow cytometry. As shown in Fig. 2A, CD11b⁺ myeloid cells predominated, with a myeloid/lymphoid cell ratio evolving progressively towards an enrichment in lymphocytes after one week of Pant administration. In Pant-treated mice, the proportion of tumor-infiltrating monocytes on day 20 (D20) was reduced whereas that of tumor-associated macrophages (TAMs) was increased (Fig. 2B, C). The latter subset was significantly enriched in MHCII^high cells (Fig. 2D), previously identified as owing potential anti-tumor functions (23).

To further dissect myeloid cell heterogeneity and maturation stages, we performed a single-cell RNA sequencing analysis of CD45⁺ tumor-infiltrating cells at D20 and D28 of tumor progression. We focused on the three main metaclusters corresponding to myeloid, lymphoid and antigen-presenting cells (APC), the cluster “others” being invariant in our experimental conditions (FigS2A). The myeloid metacluster was further split in 11 clusters (Fig. 2E) based on the preferential expression of cell-specific markers (Fig. 2F) defining neutrophils (Ly6g, cluster 8), monocytes (Ly6c2, cluster 7), TAM subsets (Adgre1, Cd64, clusters 2, 4–6, 10, 11), proliferating cells (mKi67, cluster 9) or cell signatures linked with activation or polarization pathways (cluster 11, TNF/NF-κB; cluster 10 and 2 for M1- versus M2-like phenotypes) (Fig. 2F, S2C,D). Transitional stages of TAM maturation were annotated mono-TAM A to C (clusters 4–6) as described (24). Interestingly, the density plots clearly distinguished untreated from Pant-treated samples (FigS2B) and showed an enhanced transition of mono-TAM A into mono-TAM C and M1-like TAMs in D20 to D28 Pant samples (Fig. 2G and S2B). Furthermore, Pant samples expressed higher levels of MhcII, Cd74, Cxcl9 transcripts linked to antigen presentation and lymphocyte chemoattraction pathways. In contrast, control tumors overexpressed Arg1 and Nnip3, Pgam1, Ldha, Ddit4 associated with M2 polarization and tumor hypoxia (25) (Volcano plot in Fig. 2H for D20 and maintained till D28 as shown in FigS2C). Within the APC metacluster (FigS3A), 7 clusters were defined and further characterized by differential gene expression (FigS3B). Among them, the proliferating (D20) and IFNγ-responsive monocyte-derived (D28) DC subsets were enriched in Pant samples whereas control PBS samples contained more cDC2 cells (Fig. 2I-S3C). Membrane MHCII expression by various APC was higher in Pant samples (FigS3D). This observation was confirmed by pathway enrichment analysis using Geneset Enrichment Analysis (GSEA) (Table 1) and Kyoto Encyclopedia of Genes and Genomes pathways (KEGG, FigS2D). As shown in Table 1, from D20 to D28 Pant samples overexpress transcripts related to the IFNγ response and antigen presentation process in myeloid and dendritic cells. On D28, these cells show reduced expression of genes linked with oxidative phosphorylation (OXPHOS) and aerobic respiration. Interestingly, an immunofluorescence analysis of tumor sections stained for MHCII to detect antigen presenting cells and CD8 documented 2 to 3-fold more MHCII⁺ / CD8⁺ cell contacts (Fig. 2J, S3E). In some discrete myeloid subsets, chemokine signaling was highlighted (Table 1). To identify dominant chemokine profiles at the protein level, we performed a proteome analysis (Fig. 2K) and a cytometry beads assay (CBA) (FigS3F). Control PBS samples showed an overrepresentation of proteins associated with tissue clearance (pentraxin), extra cellular matrix (MMP9, Serpin) and endothelial cell reorganization (angiopoietin, VEGF, endostatin). Pant-treated samples contained higher levels of cytokines or chemokines associated with type 1 immunity. This included CCL2 or CXCL9 (26, 27) (Fig. 2J) involved in monocyte (via CCR2) or effector lymphocyte (via CXCR3) recruitment, IL2p40 (28, 29) or Flt3-L (30), associated with M1 or DC maturation (Fig. 2J). Within each cluster, we did not observe major changes in the overall expression of different genes previously associated with anti or pro-tumor function (FigS3G). Altogether, Pant samples are enriched in APC subsets highlighting IFNγ responsiveness, enhanced antigen presentation and chemoattraction potential.

Pantethine enhances the development of effectors within adaptive lymphocytes

We used the CD8⁺/Treg cell ratio as a global indicator of anti-tumor lymphoid responses and observed a progressive enrichment in CD8⁺ over Treg CD4⁺ T cells after D17 till D28 (Fig. 3A). We took advantage of MCA cells expressing the OVA antigen to track tumor-specific CD8⁺ cell responses using a SIINFEKL tetramer. On D20, quantification of lymphocytes showed that Pant therapy led to a 2-fold increase in the number of NK/ILC, CD4⁺ and effector CD8⁺ T cells, including OVA-specific CD8⁺ T cells in the tumor and the draining lymph nodes (TDLNs) (Fig. 3B,C). In contrast, the proportion of immune cells remained globally unchanged by Pant administration in lymph nodes or spleen, at steady state or in a tumor context (FigS4A). This phenotype was comforted as CD8⁺ T cells tended to express higher levels of effector molecules such as GZMB, PRF1 or TNFα (Fig. 3D). The analysis of the lymphoid metacluster derived from the single-cell RNA sequencing analysis (FigS5A-C) confirmed the cytometry results. On D20, total CD4⁺ and more dramatically CD8⁺ T cells were increased to the expense of Tregs and activated memory CD8⁺ cells (Fig. 3E and S5A-C). Furthermore, CD8⁺ T cells showed an enriched representation of pathways associated with IFNγ response, cytotoxic function, chemokine production and to a lesser extent PD1 signaling (Fig. 3F). On D28, NK/ILC or Treg signatures predominated in Pant or PBS samples, respectively. We completed this analysis by quantifying cell surface expression of checkpoint inhibitors on various T lymphocyte subsets on D28 (Fig. 3G). In Pant treated samples, two-fold less CD4⁺ but not CD8⁺ T cells expressed CTLA4, PD1 or TIM3 molecules. Since exhaustion is associated with mitochondrial alterations, we monitored various parameters reflecting mitochondrial respiration and depolarization among lymphocytes. In D20 Pant tumors, a Seahorse analysis performed on purified CD8⁺ lymphocytes showed increased levels of OCR and ECAR witnessing a higher energetic state (Fig. 3H). We then quantified mitochondrial depolarization defined by the MDR/MG ratio on CD4⁺ and CD8⁺ T cells at different time points. Whereas no difference was observed in both cell subsets on D20, this ratio was lowered in D28 CD8⁺ T cells from Pant samples (Fig. 3I). Altogether, our results show that Pant therapy induces a more robust lymphocyte response affecting preferentially effector cell subsets. We thus scored the number of genes co-expressed by each cell and associated with a cytotoxic program. Although this analysis was limited by the total amount of cells, CD8⁺ T cells derived from Pant samples expressed a higher number of cytokine or cytotoxicity-associated genes, an argument in favor of their polyfunctionality (FigS5D). Furthermore, we used a global score recapitulating the expression level of genesets involved in activation or exhaustion programs in CD8⁺ lymphocytes. PBS and Pant samples were comparable for each cell subset, suggesting that Pant administration modifies the proportion and not the transcriptional state of these cells (FigS5E). While the relative number of total NK/ILC1 was increased in the cytometry analysis, their proportion within the lymphoid metacluster was reduced in Pant samples. Interestingly, by projecting NK versus ILC1 signatures (FigS5F) and applying single cell velocity analysis of the NK/ILC cluster (Fig. 3J), we observed an enhanced dynamic conversion of NK into ILC1 cells in control samples, whereas Pant therapy tended to reinforce NK cell signatures (31, 32), showing by GSEA an enrichment in IFNγ response (Fig. 3K).

Pantethine therapeutic efficacy depends on IFNγ, cDC1 and tumor-infiltrating CD8 ⁺ T cells

Efficacy of cytotoxic CD8⁺ T lymphocytes requires antigen cross presentation by cDC1 cells and their ability to infiltrate a tumor, a program driven by type 1 cytokines. We performed an immunohistochemistry analysis of CD3 and CD8⁺ TILs on tumor sections. As shown in Fig. 4A (quantified in Fig. 4B), in control tumors were either in reduced numbers or confined to the periphery of the tumor. In contrast, TILs were more abundant and infiltrating in Pant samples, in agreement with the IFNγ-driven increased CXCL9 expression (33). Accordingly, MCA tumors grew more rapidly in IFNGR1-deficient mice where Pant rather enhanced tumor growth (Fig. 4A). This result reinforced the notion that Pant efficacy relies on IFNγ-driven programs in APC and lymphocytes (Table 1). To dissect Pant effect at the cellular level, we transferred OT1 cells in mice, treated them or not with Pant prior to immunization with OVA. We then scored the expansion of CTV-labeled CD8 OT1 lymphocytes in draining lymph nodes. In this context, no difference between PBS and Pant samples was detected (FigS6A,B), suggesting that a chronic inflammatory might be required for Pant effect. We thus injected OVA-expressing MCA cells, extracted CD11c⁺ DC from TDLNs (FigS3D) and co-cultured them in vitro with CFSE-labeled OT-1 CD8⁺ T cells in the presence or absence of OVA (Fig. 6SC,D). Again, both cell numbers and CD8⁺ T cell proliferation index were comparable under the two conditions. To evaluate the contribution of cDC1 cells in vivo, we tested Pant therapy on MCA tumors grafted in control versus Xcr1-deficient mice that lack cDC1. As shown in Fig. 4B, tumor progression was uncontrollable in the absence of cDC1 cells, confirming the importance of this cell subset in the development of anti-tumor immunity (34). Furthermore, although Pant could reduce tumor growth in control mice, it slightly enhanced tumor growth in the absence of cDC1 cells. This showed that the therapeutic effect of Pant depended on the presence of cDC1 cells.

To prove the contribution of CD8⁺ T cells to Pant efficacy, we injected tumor-bearing mice with a depleting anti-CD8 mAb from D10 every 3 days onwards. As shown in Fig. 4C (and depletion controls in Fig. 6E,F), the depletion of CD8⁺ T cells boosted tumor growth in control mice as expected. In Pant-treated mice, this effect was also detected but tumor growth did not reach the level observed in control mice. This suggested that CD8⁺ T cells mediated anti-tumor effect might have been initiated before D10 or that Pant might have affected the function of other anti-tumor cells. Indeed, we observed an augmentation of NK1.1⁺ cells at an early stage (D20) of tumor progression and these cells may participate to tumor control (Fig. 3E) (32). We thus tested whether NK1.1⁺ cells might also contribute to Pant effect. Administration of a depleting anti-NK1.1 mAb had a modest enhancing effect on tumor growth after D22 (Fig. 4D). However, anti-tumor Pant effect persisted at the same level under both conditions. Altogether, these results show that Pant boosts type 1 immunity and that its anti-tumor effect depends critically on the presence of an IFNγ / cDC1 / CD8⁺ T cell pathway. Since this pathway is also required for the control of viral infection, we used a model of cutaneous HSV-1 infection in which the virus load and the proportion of virus-specific T cells could be monitored. The virus load was equivalent under both conditions showing that Pant does not have a significant impact on its replication in vivo (Fig. 4E). As seen in Fig. 4F, the proportion of central memory and effector CD8⁺ T cells in DLNs was enhanced 2-fold by Pant treatment, including virus-specific CD8⁺ T cells. This confirmed that Pant exerted its beneficial effect in the context of type 1-driven tissue inflammation.

Potential and limits of Pant therapy

Sarcoma cell lines show variable levels of susceptibility to immune checkpoint inhibitors (ICI) (35). As shown in Fig. 5A-B, administration of an anti-CTLA4 or anti-PD1 mAbs inhibited MCA growth with variable efficacy and combination with Pant modestly accentuated the inhibition. We then tested an osteosarcoma cell line in which ICI therapy alone or combined with doxorubicine had no effect. Addition of Pant treatment did not change the behavior of this highly resistant tumor (Fig. 5C) or the organization of the immune infiltrate (FigS7A-B). Nevertheless, the combined therapy enhanced the extent of necrosis (FigS7A-B). In the B16F10 melanoma model, known to be sensitive to anti PD1 therapy (36), Pant inhibited B16F10 growth to a level comparable to that obtained with MCA tumors and its effect was equivalent to that of an anti-PD1 mAb or their combination (Fig. 5D). Given the role of IFNγ in the development of resistance to immunotherapy (37), we administered an inhibitory anti-IFNγ antibody during the late phase where resistance to immunotherapy might develop. As seen in Fig. 5E, neutralization of IFNγ boosted tumor growth in control animals suggesting that it participated to the development of anti-tumor immunity. In Pant-treated mice, this boosting effect was absent suggesting that Pant therapy might have imprinted earlier so that it could not be reversed by IFNγ neutralization at later time points. In agreement with an early imprinting effect, we found that administration of Pant from D1 to D10 also inhibited growth (Fig. 5F).

High VNN1 expression in STS correlates with higher immune response and better prognosis

Since Pant therapy mimics VNN1 function (11), we first queried the TCGA database (including 206 soft tissue sarcoma STS samples) searching for VNN1-coexpressed gene modules. VNN1 expression matched with MSigDB hallmarks associated with IFNγ / inflammatory responses (FigS8A). Using a Cibersort deconvolution method to link VNN1 expression with the presence of immunocyte subsets, we found a significant correlation between high VNN1 expression and the presence of immune cells that do not express VNN1 such as M1/M2 macrophages and T cells (including CD8⁺) transcripts (FigS8B). This suggested that VNN1 expression within the tumor mass positively correlated with the presence of immunocytes with anti-tumor potential. To broaden this analysis, we searched for correlations between VNN1 mRNA expression and clinicopathological and immune variables in our large cohort of 1377 clinical STS samples (Fig S6C) collected during the surgery of non-metastatic primary tumors. Their characteristics are summarized in FigS6D. VNN1 expression was heterogeneous across the cohort with a range of intensities over 20 units in log2 scale (Fig. 7A), allowing the search for correlations with tumor variables. VNN1 expression, assessed as discrete variable (high vs. low), correlated (FigS8D) with patient’s age at diagnosis pathological tumor grade, and tumor site: as compared to the “VNN1-low” class, the “VNN1-high” class included older patients (p = 3.78E-08), more grade 1 tumors (p = 5.22E-06), and more extremity and less superficial trunk tumor sites (p = 5.20E-03). No significant correlation was found with patient’s sex, pathological tumor size, and tumor depth. Then, we analyzed the correlation of VNN1 classes with immune variables (Fig. 7B). First, the probability of activation of IFNα, IFNγ and STAT3 immune pathways was higher in the “VNN1-high” class than in the “VNN1-low” class (p < 0.05). Second, we compared the composition and functional orientation of tumor-infiltrated immune cells between VNN1 classes using the 24 Bindea’s immune cell types defined as the immunome. “VNN1-high” tumors displayed a higher infiltrate concerning 18 immune cell types (p < 0.05) including B cells, T cells, Tem cells, Th1 cells, TFH cells, Th17 cells, CD8⁺ T cells, γδ T cells, cytotoxic cells, CD56dim NK cells, dendritic cells (DC, interstitial DC, activated DC and plasmacytoid DC), eosinophils, macrophages, mast cells, and neutrophils. By contrast, “VNN1-low” tumors displayed a higher infiltrate in two immune cell types, NK cells and CD56bright NK cells (p < 0.05). Third, “VNN1-high” samples displayed higher ICR score (p = 1.81E-05) and immune cytolytic activity score (p = 3.13E-29), reflects of an antitumor cytotoxic immune response, than “VNN1-low” samples, and higher scores for signatures associated with response to ICI: tumor inflammation signature (TIS) (p = 1.58E-33) and tertiary lymphoid structure (TLS) (p = 1.56E-36). Finally, “VNN1-high” samples displayed higher TILs scores using lymphoid-alone, myeloid-alone and combined signatures (p < 0.05), and higher Antigen-Processing Molecules (APMs) score (p = 7.55E-18). Next, we searched for correlation of VNN1 classes with the clinical outcome. The patients with “VNN1-high” samples displayed longer MFS than those with “VNN1-low” samples (Fig. 7C with respective 5-year MFS equal to 66% (95%CI 60–73) vs. 61% (95%CI 54–68) (p = 3.01E-02, Fig. 7D). In univariate prognostic analysis (Fig. 7D), high VNN1 expression was associated with longer metastasis-free survival (p = 3.08E-02), as was the pathological grade 1 (p = 8.46E-03). In multivariate prognostic analysis, the VNN1 status tended to remain significant (Hazards Ratio HR = 0.64, 95%CI 0.41-1.00, p = 0.051) (Fig. 7D). Altogether, these results showed heterogeneous VNN1 expression levels in STS, and correlations with immune variables and clinical outcome. Of note, such prognostic value of VNN1 expression was not observed in a cohort of 326 patients with bone sarcoma (p = 0.397), nor in the subcohort of 94 patients with osteosarcoma (p = 0.229; FigS9).

Pantethine administration aiming at restoring vitB5 and mitochondrial homeostasis lead to the reduction of the growth of an aggressive sarcoma cell line in immunocompetent mice. Pant acts predominantly via cysteamine-dependent redox modulation and CoA regeneration. An intrinsic tumor suppressive effect via metabolic rewiring was unlikely. Indeed, Pant did not inhibit tumor growth in Nude mice and, as many aggressive tumors displaying mitochondrial alterations (38), MCA cells showed reduced basal or stimulated respiratory capacity. Rather, Pant administration enhanced mtROS generation, a phenotype observed in cells with electron transport chain (ETC) deficiencies (39). Increasing oxidative stress has been envisaged as a therapeutic strategy in embryonal rhabdomyosarcomas (ERMS) (38, 40) to induce cell death or immunostimulate via damage-associated molecular pattern (DAMP) release (41, 42). Therefore, we suspected that metabolic changes induced by Pant could boost anti-tumor immunity. In Pant-treated mice, tumor-associated myeloid and lymphoid cells displayed signatures of IFNγ-driven antigen presentation pathways. The contribution of mitochondrial activity to antigen recognition and efficacy of immunotherapy is still debated (43) (44, 45) (46). In macrophages, CoA administration enhances acetyl-CoA-mediated epigenetic activation of pro-inflammatory gene transcription, and boosting TAM anti-tumor activity in breast cancer (17). Similarly, in vitro, dendritic cells require enhanced mitochondrial activity for the acquisition of antigen-presenting capacity (47). Furthermore, Pant enhanced tumor growth upon suppression of IFNγ signaling and cDC1 cells suggesting that it might also boost the activity of pro-tumor cells. Indeed, CoA- and its acylated forms have a strong impact on fatty acid oxidation and mitochondrial activity that participate to pro-tumorigenic type 2 immunity though the regulation of macrophage (48) and Treg functions (49).

Pant therapeutic efficacy required the presence of cDC1 and tumor infiltrating CD8⁺ T cells. Indeed, cDC1 cells cross present antigens (50–52) and, through cooperation with Th1 cytokine-producing NK or CD4⁺ Th1 cells (53), prime anti-tumor CD8⁺ T cell responses. Interestingly, perturbation of fatty acid oxidation or mitochondrial clearance was shown to increase the development of cDC2 to the expense of cDC1 cells (54). Furthermore, CD8⁺ dendritic cells use the Hippo pathway to boost mitochondrial energy production and maintain cross presentation or IL-12 production (55, 56). Mitochondrial activity is essential for T lymphocyte activation, cytokine production and differentiation (57–59), in part through mitochondrial-derived ROS (58, 60–63). Interestingly, tumors from Pant-treated mice featured an increased IL12p40 and CXCL9 production involved in Th1 effector cell priming and recruitment via the CXCR3 axis (33). Accordingly, they were two-three-fold more infiltrated by CD4⁺ or CD8⁺ effector T cells, in which the simultaneous expression of Th1 cytokines and markers of cytolytic activity was enhanced, an argument in favor of their polyfunctionality. In D20 Pant-treated samples, CD8⁺ T cells showed a hypermetabolic profile with simultaneous enhancement of ECAR and OCR. Accordingly, TLR-induced IFNγ production by CD8⁺ T cells depends on mitochondrial ATP (64). ETC complexes 1 and 2 are differentially required for Th1 cell proliferation and epigenetic remodelling versus maintenance of terminal effector functions, respectively (65). Interestingly, velocity index analyses suggested that the identity of effector innate or adaptive lymphocytes was reinforced upon Pant therapy. In vitro, addition of CoA was shown to enhance the production of effector memory CD8⁺ T cells that upon in vivo injection rejected a tumor (16). These results indicate that Pant, through its effect on mitochondrial fitness, enhances the functionality and/or stability of anti-tumor cells.

Chronic activation of lymphocytes commonly leads to cell exhaustion (6, 8, 66, 67) and expression of inhibitory PD1 or TIM3 molecules. This phenotype is correlated to long-lived effector cells that survived the contraction phase of the immune response but develop various dysfunctions and mitochondrial alterations (7, 68), in part induced by inhibitory molecules such as PD1 (69). Our results indicated that PD1 / TIM3 expression was lower in CD4⁺ but not CD8⁺ T cells from Pant-treated tumors compared to the PBS situation. Furthermore, mitochondrial depolarization was more important in CD8⁺ T cells from Pant-treated mice, maybe because of a higher activation level at earlier stages of tumor development. Although the difference between PBS and Pant samples was not detected at the transcriptional level for each cell subset, the progressive loss of functionality might limit Pant efficiency and explain the lack of synergy with anti-PD1 therapy. Indeed, Pant and anti-PD1 therapies were equally effective or ineffective on the B16 melanoma or osteosarcoma cell lines, respectively, possibly because they converge on similar metabolic pathways (70). To optimize the anti-tumor effect of Pant therapy, one may require additional signals reinforcing transcriptional programs as those induced by bezafibrate (71).

Pant administration mimicks the biological effects linked to the surexpression of the Vnn1 pantetheinase in vivo (11, 14). We therefore tested whether VNN1 expression in human sarcomas might be correlated with enhanced survival and/or development of anti-tumor responses. VNN1 expression was variable in a heterogeneous series of complex genomics sarcomas and inversely correlated with their severity. However, when detected, its expression positively correlated with improved prognostic and, furthermore, with the presence of immune signatures recapitulating the phenotypes observed in the mouse model. Inversely, in human bone sarcomas, VNN1 expression did not correlate with prognosis and Pant did not control tumor growth. This phenotype should be put in relation with the fact that loss of mtDNA is often found associated with aggressive osteosarcomas (72).

In conclusion, our results extend the former observation that vitB5 is a limiting metabolite in many tumors, and that this defect participates to the impaired development of anti-tumor immunity. Pantetheine might present a therapeutic interest in a context of immune enhancement prior to surgery or in vaccine programs.

Animals

Female 8–10 weeks old C57BL/6 and nude mice were purchased from Janvier Laboratories. Xcr1^DTA and OT-I (ovalbumin-specific TCR-transgenic mice) mice obtained from B. Malissen’s laboratory were previously described (34, 73, 74). Ifngr1^KO mice were purchased from CNRS - INEM - UMR 7355. Mice were housed under a standard 12-h:12-h light–dark cycle with ad libitum access to food and non-acid water, 22°C +/− 1°C, 45–60% humidity and maintained under specific pathogen-free conditions at animal facility of the CIML (F13-055-10). For osteosarcoma murine model, 5 weeks old BALB/cByj mice were purchased from Charles River Laboratories and housed at the CRCL animal facilities under same conditions as mentioned previously. All experiments were conducted in accordance with institutional committees and French and European guidelines for animal care. Experimentation was approved by the Ethical Committee for Animal Experimentation.

Cell lines

The MCA205 mouse fibrosarcoma cell line (SCC173), the OVA-expressing cells (a kind gift of Laurence Zitvogel, Institut Gustave Roussy, Villejuif, France) and the B16F10 melanoma cell line (ATCC-CRL-6475) were cultured in Dulbecco’s modified Eagle’s medium F-12 (Thermo Fisher) supplemented with 10% fetal bovine serum (FBS, Gibco), 100 units/mL penicillin-streptomycin (Gibco), 2 mM of L-glutamine and 1 mM of sodium pyruvate at 37°C with 10% CO2. Murine OsA K7M2 (ATCC – CRl-2836) were cultured in complete DMEM medium and maintained 2 weeks in culture before mice cell engraftment. Mycoplasma status was evaluated using MycoAlert Lonza Detection kit and cells were used at low passage.

Treatments

For drug treatment mice were intraperitoneally injected starting at day 10 post cell engraftment. D-pantethine (1g/Kg) was administered every day and dichloroacetate (1g/Kg) every 2 days. For all experiments, control PBS vehicle was used. All reagents were purchased from Sigma-Aldrich. For osteosarcoma experiment, mice were treated with doxorubicine (0.75mg/Kg - Baxter) or NaCl vehicle two days a week.

Tumor experiments

10⁵ MCA 205 cells or OVA-expressing MCA205 were subcutaneously and bilaterally grafted in the flanks of C57BL/6, nude mice, XCR1^DTA mice and Ifngr1^KO mice. 3.10⁵ B16F10 were intradermally and bilaterally injected to C57BL/6. Murine OsA K7M2 were injected in the tibia of BALB/cByj mice using syringe and catheter from adjacent soft tissue. Mice were anesthetized with 2.5% isoflurane every two days and tumor growth was scored using a caliper by measuring the length (L) and width (W) of the tumor and volumes were calculated using the following formula: (L × W)² /2. A limit point was settled when tumor volume was above 1500 mm³ or symptoms of poor health were detected (weight loss, prostration, bleeding, scars). Tumors were harvested between D10-D28 post cell engraftment, each tumor being considered as an experimental unit referred as n. Tumors were mechanically and enzymatically digested using the Miltenyi Biotech GentleMACS Octo Dissociator technology. Samples were filtered through a 70µm Cell Strainer (Becton Dickinson) and submitted to red blood cells lysis (eBioscience). Cells suspensions were analyzed by flow cytometry. Alternatively, CD45⁻ cells, CD8a⁺ naïve T cells or CD8⁺ TILs were isolated using microbeads from Miltenyi Biotech and the MultiMACS™ Cell24 separation on LS columns according to the manufacturer’s protocols. The inguinal tumor-draining lymph nodes were collected between D14 and D28 post cell engraftment, scratched on 70µm Cell Strainer and analyzed by flow cytometry. Spleen samples were processed as described for lymph nodes. For osteosarcoma experiments, tumors were harvested, fixed in PFA 4%, included in paraffin for subsequent histology analysis.

LC-MS

All the dried polar extracts from samples were first reconstituted with 200µL acetonitrile/water (50:50; v:v). The samples were separated using a high-performance liquid chromatography (UPLC) ultimate 3000 (Thermo Scientific), coupled to a high-resolution mass spectrometer (HRMS). Q-Exactive Plus quadrupole-orbitrap hybrid equipped with electrospray ionization source (H-ESI II). The chromatographic separation was performed on a binary solvent system using a reverse phase C18 column (Hypersil Gold, Thermo Scientific, 100mm X 2.1mm, 1.9µm) at 40°C with a flow rate of 0.4 mL min-1 and a HILIC column (Merk, SeQuant® ZIC®-HILIC, 150 mm x 2.1 mm, 5µm, 200 A) at 25°C with a flow rate of 0.25 mL min-1. The injection volume for both columns was 5µL. The chromatographic separation was performed on a binary solvent system at a flow rate of 0.4 mL min-1 using a reverse phase C18 column (Hypersil Gold, Thermo Scientific, 100mm X 2.1mm, 1.9µm) at 40°C. The mobile phase consisted of a combination of solvent A (0.1% formic acid in water, v/v) and solvent B (0.1% formic acid in acetonitrile, v/v). The injection volume was 5µL. The following gradient conditions were used: 0 to 1 minute, isocratic 100% A. 1 to 11 minutes, linear from 0 to 100% B. 11 minutes to 13 minutes, isocratic 100% B. 13 minutes to 14 minutes, linear from 100–0% B. 14 to 16 minutes, isocratic 100% A. The separated molecules were analyzed in both positive and negative ionization modes in the same run. The mass spectra were collected using resolving power 35 000 Full Width at Half Maximum (FWHM) for the theoretical mass to charge ratio (m/z) 200. Full scan mass spectra were acquired in the 80 − 1000 m/z range. The ionization source parameters for positive and negative ion mode were as follows: capillary temperature 320°C, spray voltage 3.5 kV, sheath gas 30 (arbitrary units), auxiliary gas 8 (arbitrary units), probe heater temperature 310°C and S-Lens RF level was set at 55 v. except MS/MS experiments were performed using Higher energy Collision induced Dissociation (HCD) and the normalized collision energy (NCE) applied was ramped from 10 to 40%. To ensure a good repeatability of the analysis, a quality control sample (QC) was formed by pooling a small aliquot of each biological sample. The QC sample is analyzed intermittently (because of the small number of samples, the QC was analysed 1 out of every 3 samples) for the duration of the analytical study to assess the variance observed in the data throughout the sample preparation, data acquisition and data pre-processing steps. The mobile phase for the HILIC column separation consisted of a combination of solvent A (100% water, 16mM Ammonium formate) and solvent B (100% Acetonitrile 0.1% formic acid). The following gradient conditions were used: 0 to 2 minute, isocratic 97% B. 2 to 10 minutes, linear from 97 to 70% B. 10 to 15 minutes, linear 70 to 10% B. 15 to 17 minutes, isocratic 10% B. 17 to 18 minutes linear from 10 to 97% B. from 18 to 22 minutes isocratic 97% B. The separated molecules were analyzed in both positive and negative ionization modes in the same run. The mass acquisition parameters were the same used for the C18 column. The repeatability of the analysis was insured by analyzing intermittently (1 out of every 5 samples) the quality control sample (QC) as described in section II-3-1.

NMR

Samples of thawed tissue (about 15 mg) were placed into a 30 µl disposable insert in which 10 µl of D2O was added to provide field-lock signal. The disposable insert was then inserted into a 4-mm ZrO2 HRMAS rotor for the HRMAS NMR spectral acquisition. All NMR experiments were carried out on a Bruker Avance III spectrometer operating at 400MHz for the 1H frequency equipped with a 1H/13C/31P HRMAS probe. Spectra were obtained at 278 K with a spin rate of 4 kHz. In order to attenuate NMR signals of macromolecules, the Carr-Purcell-Meiboom-Gil (CPMG) NMR sequence with an overall spin echo time of 150 ms, preceded by a water presaturation pulse during a relaxation delay of 2 s was employed. For each sample, 256 free induction decays (FID) of 32 k data points were collected using a spectral width of 6000Hz. The FIDs were multiplied by an exponential weighting function corresponding to a line broadening of 0.3Hz and zero-filled once prior to Fourier transformation. Subsequently, the spectra were phased and baseline corrected manually and calibrated to the alanine methyl signal (δ = 1.47 ppm). To facilitate NMR signal assignments, 2D NMR experiments using 1H − 1H TOCSY, 1H − 13C HSQC and online databases (Human Metabolome Database, Wishart et al, 2013) were employed. The 1H 1D NMR spectra were directly exported to AMIX 3.8 software (Bruker Biospin GmbH, Karlshure, Germany) and divided into 0.001 ppm-width buckets. In order to remove the effects of possible variations in the water suppression efficiency, the region between 4.70 and 5.20 ppm was discarded. The obtained NMR dataset X-matrix (40 observations x 8001 buckets) was then normalised to the total spectrum intensity and then subjected to multivariate statistical analysis using the software SIMCA-P + v.16 (Umetrics, Umea, Sweden). Initially, the principal component analysis (PCA) of the 1H NMR spectral data was carried out to check the homogeneity of the dataset, group clustering and detect potential outliers. Following PCA, a supervised orthogonal partial least squares discriminant analysis (OPLSDA) was then applied to the X-matrix, in which we defined a Y-matrix as the matrix of sample classes in order to target metabolic differences between the groups of interest. The resulting score and loading plots were used to visualize the discriminant features. A leave-one out internal cross-validation was performed in order to calculate R2Y and Q2 values representing, respectively, the explained variance of the Y matrix and the predictive ability of the model. Model validation was performed by random permutation of the Y matrix with n = 999 times and by the use of a CV-ANOVA p value from SIMCA-P + v.16 (analysis of variance in the cross-validated residuals of a Y variable).

Seahorse

Oxygen consumption rate of in vitro grown MCA205 cells were measured on a Seahorse XF-24 Metabolic Flux Analyzer. 1.10⁵ cells were seeded on XF-24 V7 multi-well plates, then treated or not with 100µM of Pant for 16 h at 37°C. OCR was evaluatedusing the XF cell Mito Stress Kit (Agilent). 5.10⁴ CD8 TILS isolated from control or Pant-treated murine tumors were analyzed using T cell Metabolic Profiling kit on Seahorse HS Mini Xfp.

MitoSox

1.10⁵ MCA205 cells were seeded on 24-well plate in RPMI1 medium supplemented with 10% FBS and stimulated with 1mM Pant for 4h. A 20 min incubation with Antimycin A (5µg.mL^− 1) was used as a positive control for mitochondrial ROS production. The MitoSox probes were used at 5µM 20 min in pre-warmed HBSS1X. Cells were harvested and MitoSox fluorescence analyzed by flow cytometry.

Lactate Assay

Lactate concentrations were quantified according to the manufacturer’s instructions (Sigma-Aldrich). Briefly, tissues were homogenized in four volumes of lactate assay buffer and centrifuged at 13000 g for 10 min to remove insoluble material. Samples were deproteinized with a 10-kD cut-off spin filter. A master reaction mix containing 46 µL lactate assay buffer, 2 µL lactate enzyme mix, and 2 µL lactate probe was added to a 50 µL sample solution. Reactions were incubated at RT for 30 min and sample absorbance measured at 570 nm on a microplate reader.

Flow Cytometry and antibodies

For cytometry analysis, tumors were harvested at the indicated time points in DMEM F12 media, mechanistically dissociated before dissociation using GENTLEMACS OctoDissociator with heaters from Miltenyi Biotech according to the manufacturing protocols. Tumor draining lymph nodes were scratched and mechanistically dissociated. After filtration and red blood cell lysis using eBiosciences reagents, tumor and TDLN cell suspensions were labeled with LiveDead Fixable Blue and CD16/CD32 in PBS EDTA for 30 minutes. Cell surface labelling was performed in FACS Buffer during 1 hour at 4°C with the indicated antibodies. If needed, streptavidin was incubated after cell surface labelling during 15 min at 4°C. For intracellular staining, cells were permeabilized using the FoxP3 staining kit from eBiosciences. Intracellular antibodies were incubated 45 minutes at 4°C. Cells were resuspended in FACS Buffer before analysis on BD LSR Symphony or Fortessa. Mitotracker Deep Red and MitroTracker Green were used following manufactured protocols to assess the membrane mitochondrial potential and mitochondrial mass, respectively. The OVA-tetramer (SIINFEKL – H2-K^b) was purchased from the NIH tetramer core facility and incubated with cell suspension prior to cell surface labeling during 1 hour at 4°C. Beads were added to cell suspension for calibration and data normalization. Flow cytometry data were analysis using FlowJo 10.8. The gating strategies for the analysis of myeloid and lymphoid cells are indicated in FigS1E and S4C, respectively.

Single Cell RNA sequencing

Tumors from MCA-205 bearing mice treated with PBS or Pant were harvested at D20 or D28 in DMEM F12 medium. After dissociation and RBC lysis, cells suspension from the four experimental conditions were tagged using TotalSeq B0301 and pooled. CD45 positive Live Dead negative cells were sorted on ARIA SORP, collected in SVF-coated tube and pooled in equivalent ratio. Then, cells suspension was immediately processed according to the Chromium Next GEM Single Cell v3.1 GEM protocols. HTO and DNA libraries were sequenced at the CIML genomic facility using a P2 flowcell in Illumina NextSeq 1000/2000. 2 independent experiments were performed and analyzed separately. FASTQ raw files were aligned to the mouse genome reference (mm10) using 10x Genomics Cell Ranger 6.0.1 (75), which performs filtering, barcode counting, and unique molecular identifier (UMI) counting. After individual quality control, all count matrices were loaded and processed together using R package Seurat (6.1.0) (76). A second Quality control (QC) was performed to evaluate potential batch-related biases, perform HTO demultiplexing, and identify poor quality cells to be excluded before downstream analyses. Cells with less than 1000 counted UMIs or more than 5000 detected genes (potential doublets), and cells with more than 10% mitochondrial gene expression (apoptotic cells) or less than 5% ribosomal genes were excluded from analyses. Then, datasets from two independent experiments were pooled and analyzed. 5 metaclusters were identified : Myeloid cells, Lymphoid cells, APC cells including DC cells and B cells and an undefined metacluster call “Others” which include tumor cell, fibroblasts and stromal cells. For subsequent analysis, the four immune cells metacluster were recomputed separately. After normalization (lognormalize) and centering, the 2000 most variable features were identified and used for dimensionality reduction by PCA (50 PCs). Further dimensionality reduction was performed using Uniform manifold approximation and projection (UMAP) embedding, and cell clustering using Louvain algorithm (77). Marker genes for each cluster were extracted with Wilcoxon method using Seurat function ‘FindMarkers’. Functional enrichment analyses were performed using the R packages clusterprofiler (78) and fgsea (79), and the external database KEGG (80). GSEA enrichment analyses and figures were performed with R package ‘fgsea’ against selected reference gmt files from MSigDB (81, 82) Heatmaps were generated using the R package complex heatmap (83), other custom figures were generated using ggplot2 (84). RNA velocity analysis was performed with scVeloAll (85) detailed parameters for analyses are available in technical reports). For further characterization of NK cells, results of a differential expression analysis from a previous study (32) comparing ILC and NK cells were downloaded. The top 20 up- and down-regulated genes (sorted by pvalue) from this study were used to create gene module scores in Seurat, projected on previously computed uMAP for NK cells population.

Cytokine array

Frozen tumors were directly homogenized in PBS with protease inhibitors and 1% Triton. After debris removal, protein concentrations were assessed with BCA Assay. 200 µg of tissue lysates were processed according to the Mouse XL Cytokine Array Kit procedure (R&D systems). Membrane were scanned with Samsung Digital Presenter with 720P HD document camera with a 14x optical zoom and 3x digital zoom picture and quantify with Image J.

CBA

Tumor cell suspensions were centrifuged and cytokine or chemokine content measured in the supernatant using the Cytometric bead array (CBA). TME cytokines were determined with mouse CCL2, CXCL9 cytometric bead array (CBA) flex set kits (BD Biosciences), in accordance with the manufacturer’s instructions. Samples were analyzed on a BD LSR II flow cytometer with FCAP Array TM Software V3 (BD Bioscience) to determine the cytokine concentrations of the experimental samples.

In vivo Depletion

CD8 mAb-based depletion (Clone 53 − 5.8) was obtained after intraperitoneal injections of 200µg and 150µg per mouse from day 10 and every three-day onwards, respectively. A rat IgG1 anti-horseradish mAb (Clone HRPN) was used as an isotype CTRL. NK1.1/ILC depletion was obtained by two injections of anti-NK1.1 mAb (Clone PK136) at 100 µg per mouse two days prior to MCA205 cell injection and maintained by injections at days 12 and 24. The mouse IgG2a mAb (Clone C1.18.4) was used as isotype control. The neutralizing IFNγ Ultra-LEAF™ Purified antibody was intraperitoneally injected at days 20 and 23 at 200 µg per mouse.

HSV-1 infection

For HSV-1 infection, the KOS strain of HSV-1, previously described in (86) was used. Briefly, HSV-1 was grown and titrated on confluent Vero cells in minimal essential medium containing 10% FCS, 50µM 2-mercaptoethanol, 2 mM L-glutamine, 100 U/mL penicillin and 100 µg/mL streptomycin. Mice were inoculated with HSV-1 by flank scarification, as previously described (van Lint et al, 2004; Gebhardt et al, 2011). Female mice, 8 weeks old, were anaesthetized by the i.p. injection (10 µL/g of body weight) of ketamine (2%)/Rompun (5%) in saline solution. The left flank of each mouse was clipped and depilated with Veet hair remover cream. A small area of skin was abraded with a MultiPro power tool (Dremel, Racine, WI) and 10 µL volume of viral suspension containing 10⁶ PFU were applied to the skin lesion. Skin samples were harvested using a 12x12mm skin punch biopsy and immune infiltrate analyzed at day 8 post HSV-1 infection by flow cytometry. To study the virus-specific CD8⁺ T cell response, we used a tetramer H-2K(b) composed of HSV-1-derived peptide of SSIEFARL sequence encoding glycoprotein B (NIH tetramer core facilities).

OVA immunization

CD8⁺ OTI were isolated with Untouched Mouse CD8 dynabeads, labelled with Cell Trace Violet (10µM) and enriched from spleen and LNs of Rag2KOxTgOTIxB6. Ly5.1 mice. Adoptive transfer of 10⁶ CD8⁺ OT-I T cell was performed seven days after starting Pant therapy by retro-orbital intravenous injections. The day after adoptive transfer, mice received 10 µl of OVA (5 and 10 µg) in a mix of Alum Inject (240µg) and cholera toxin (0,5 µg) by intradermal injection in the ear pinna using a 25-µl syringe. The auricular draining LNs were harvested, pooled, digested and labeled for flow cytometry analysis in BD Fortessa X20. Beads were added to cell suspension for data normalization.

DC-OT-1 co-culture

Tumor-draining lymph nodes from control PBS and Pant mice were collected and CD11c⁺ cells were sorted with the CD11c Microbeads UltraPure mouse kit from Miltenyi Biotec, according to the manufacturer’s instructions. CD11c⁺ cells were incubated in a 96-well plate for 1 h at 37°C with or without the OVA peptide (SIINFEKL). T cells were isolated and purified from spleens of transgenic OT-1 mice by negative selection with the CD8α⁺ cell Isolation Kit from Miltenyi Biotec, according to the manufacturer’s instructions, and stained with the Cell Trace Violet Cell Proliferation Kit (Invitrogen), according to the manufacturer’s instructions. CD11c⁺ cells and T cells were co-cultured (ratio of 1:10) for three days. Cells were stained with a fixable blue dead-cell staining kit (Invitrogen) and an anti-mouse CD8 antibody (PE-Cy5, 53 − 6.7, 55304 BD). T cell proliferation was assessed by flow cytometry.

Immunofluorescence

For immunofluorescence experiments, tumors were harvested, washed and fixed in Antigenfix solution (Diapath) for 2h under agitation. After several steps of PBS 1X washing, samples were placed in PBS 30% sucrose overnight and included in OCT. After saturation with PBS 2% BSA, samples were stained by overnight incubation at 4°C with anti-CD8b (BioXcell - Clone 53 − 5.8) and MHC-II (BD-Pharmingen – 2G9) mAbs. After incubation with a Donkey anti-Rabbit Cy3 (Polyclonal) and strepatividin Alexa488 (Biolegend), samples were mounted in mounting in mounting medium. Images were acquired with a Zeiss LSM780 confocal microscope and analyzed with ZEN 2.3 software.

Immunohistology

Histology analysis of osteosarcoma, and MCA tumors were realized on 5µm-thick slice mounted in SuperFrost-Plus (VWR, Radnor, Pennsylvanie USA). After deparaffinization and rehydration, hematoxylin-phlowin-safran (HPS) staining was performed to discriminate tumor and surrounding tissues. CD8, CD3, CD31 were stained by immunohistochemistry at the Lyon-Est Research Anatomopathology platform on an automated staining system Ventana Discovery XT (Ventana Medical Systems, Roche, Tucson, Arizona, USA). Slides are digitized by the Histech 3D slide scanner. The slide scans are analyzed by QuPath v0.3.0 software (87). After manual annotation of the tumor area, a specific script for each IHC and HPS staining detection is created to quantify the labeled immune cells and the rate of necrotic tissue present in the tumors. The immune infiltration score displayed in Fig. 4B was calculated for each section as follows : 0: cold tumors, 1 : peri-infiltrated tumors, 2–3 : weak or strong infiltrations by TILs in the tumor mass.

Analysis of VNN1 expression in TCGA Database and Cibersort

TCGA-SARC dataset was retrieved using TCGA biolinks package. Raw HTSeq counts were normalized using DESeq2 variance stabilizing transformation. Co-expression unsigned network was produced using WGCNA with biweight midcorrelation and soft Power = 13. Each module was further 2-means clustered to separate correlated and anticorrelated patterns, and their gene sets were queried for functional enrichment against MSigDB Hallmark (78, 81).

Analysis of VNN1 expression in soft tissue sarcoma clinical samples

We analyzed our database (88) including 1.377 cases of STS clinical samples clinically annotated and with available normalized gene expression profiles. Data had been gathered from 15 public data sets (Fig.S8C), all samples were from operative specimen of previously untreated primary tumors, and the gene expression profiles had been generated using DNA microarrays or RNA-sequencing. VNN1 mRNA expression was analyzed as discrete variable (high vs. low) by using its mean expression level of the whole series as cut-off. Because we found a link between VNN1 and immunity in our mouse model, we searched for correlations between VNN1 tumor expression and several immunity-related variables. These latter included the following multigene classifiers/scores: the 3 Gatza’s immune pathway activation scores (89), the 24 Bindea’s innate and adaptive immune cell subpopulations (90), two signatures associated with antitumor cytotoxic immune response (the Immunologic Constant of Rejection (ICR) classifier (88), and the Rooney’s cytolytic activity score (91)), two metagenes associated with response to immune checkpoint inhibitors ICI (the Ayers’ T-cell-inflamed signature (TIS) (92), and the Coppola’s tertiary lymphoid structures (TLS) signature (93), the Ballot’s Tumor Infiltrating Lymphocytes Signature (TILS) along with both of its inner lymphoid and myeloid signatures (94) and the Thompson’s Antigen Processing and presenting Machinery Score (APMS) (95). The correlations between VNN1 expression-based classes and clinicopathological variables were measured using the Fisher’s exact test or Student’s t-test when appropriate. The correlations of immune variables with VNN1-based classes were assessed by logistic regression analysis with the glm function (R statistical package; significance estimated by specifying a binomial family for models with a logit link). The endpoint of prognostic analysis was the metastasis-free survival (MFS), calculated from the date of diagnosis until the date of metastatic relapse or death from any cause, whichever occurred first. The follow-up was measured from the date of diagnosis to the date of last news for event-free patients. Survivals were estimated using the Kaplan-Meier method and curves compared with the log-rank test. Uni- and multivariate prognostic analyses were done using Cox regression analysis (Wald test). The variables tested in univariate analysis were the patients’ age and sex, pathological grade (2–3, 1), pathological size (pT1, pT2), tumor depth (deep, superficial), tumor site (extremities, head and neck, internal trunk, superficial trunk) and the VNN1-based classification (high, low). Multivariate analysis incorporated all variables with a p-value inferior to 5% in univariate analysis. All statistical tests were two-sided, and the significance threshold was 5%. Analyses were done with the survival package (version 2.43) from R software (version 3.5.2).

Statistical analysis

Sample size were designed to minimize number of individual experimental units (mice or samples) and obtain informative results and appropriate material for downstream analysis. This represents 5 mice per group and experiments were typically performed twice, unless otherwise stated in figure legends. GraphPad Prism 7 software was used for statistical significance assessment. Gaussian distribution was tested using D’Agostino–Pearson omnibus normality test. When passing the normality test, a Student t test was used. Otherwise, a Mann–Whitney U test was used. Differences were considered to be statistically significant when ****p-value < 0.0001, ***p-value < 0.001, **p-value < 0.01, * p-value < 0.05.

Acknowledgments

The ANR PLBIO19-015 grant supported the work realized in PN, JC, LS and JYB teams and RM funding. We thank the core facilities involved in this project: CIML Flow cytometry, CIML Genomics Platform, CIML Animal facilities, Imagimm. BM and SH were supported by DCBIOL LabEx (grants ANR-11-LABEX-0043 and ANR-10-IDEX-0001-02 PSL). AR and SU were supported by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program, under grant agreement No. 648768, from the Agence Nationale de la Recherche (ANR) (No. ANR-14-CE14-00 09-01), and the Fondation pour la Recherche Médicale (FRM). This work was also supported by institutional grants from INSERM, CNRS, Aix-Marseille University and Marseille-Immunopole to the CIML. Guillaume Charbonnier performed the bioinformatics analysis of the TCGA database. We thank Thomas Vannier from the CENTURI Multi-Engineering Platform for the preliminary work on the datamining of the cancer databases. The project leading to this publication has received funding from France 2030, the French Government program managed by the French National Research Agency (ANR-16-CONV-0001) and from Excellence Initiative of Aix-Marseille University - A*MIDEX “. JYB is supported by funds from NetSARC, LYRIC (INCA-DGOS 4664), LYon Recherche Innovation contre le CANcer, European Clinical trials in Rare Sarcomas (FP7-278742), and European network for Rare Adult solid Cancer. We thank A. Carrier and T. Nguyen for assistance with the use of the Cell Culture Platform facility (Centre de Recherche contre le Cancer de Marseille, U1068) in charge of the Seahorse platform.

Availability of data and materials

Data used for the bioinformatics analysis come from the publicly available TCGA database. The source code for all the analyses is available at https://github.com/guillaumecharbonnier/mw-miallot2022 and a compiled version of the Bookdown report is available at https://guillaumecharbonnier.github.io/mw-miallot2022

Author contribution

RM, VM, AR, CT, PB, JSL performed experiments. JC, LS, BM, SH, SU provided technical help, expertise and mice. Single cell analysis was performed by RF. AD and PF contributed to the bioinformatic analysis of data from sarcoma patients, collected and supervised by JYB and FB. RM, VM and PN analyzed data and prepared the figures. FG contributed to manuscript corrections and coaching. PN conceptualized and supervised the project and RM and PN prepared the manuscript.

Conflict of interest

The authors declare that they have no conflict of interest.

Pearce EL, Walsh MC, Cejas PJ, Harms GM, Shen H, Wang L-S, et al. Enhancing CD8 T-cell memory by modulating fatty acid metabolism. Nature. Nature Publishing Group; 2009;460:103–7.
Chang C-H, Qiu J, O’Sullivan D, Buck MD, Noguchi T, Curtis JD, et al. Metabolic Competition in the Tumor Microenvironment Is a Driver of Cancer Progression. Cell. 2015;162:1229–41.
Reinfeld BI, Madden MZ, Wolf MM, Chytil A, Bader JE, Patterson AR, et al. Cell-programmed nutrient partitioning in the tumour microenvironment. Nature. Nature Publishing Group; 2021;593:282–8.
Doedens AL, Phan AT, Stradner MH, Fujimoto JK, Nguyen JV, Yang E, et al. Hypoxia-inducible factors enhance the effector responses of CD8(+) T cells to persistent antigen. Nat Immunol. 2013;14:1173–82.
Gropper Y, Feferman T, Shalit T, Salame T-M, Porat Z, Shakhar G. Culturing CTLs under Hypoxic Conditions Enhances Their Cytolysis and Improves Their Anti-tumor Function. Cell Rep. 2017;20:2547–55.
Scharping NE, Rivadeneira DB, Menk AV, Vignali PDA, Ford BR, Rittenhouse NL, et al. Mitochondrial stress induced by continuous stimulation under hypoxia rapidly drives T cell exhaustion. Nat Immunol. 2021;22:205–15.
Scharping NE, Menk AV, Moreci RS, Whetstone RD, Dadey RE, Watkins SC, et al. The Tumor Microenvironment Represses T Cell Mitochondrial Biogenesis to Drive Intratumoral T Cell Metabolic Insufficiency and Dysfunction. Immunity. 2016;45:374–88.
Yu Y-R, Imrichova H, Wang H, Chao T, Xiao Z, Gao M, et al. Disturbed mitochondrial dynamics in CD8 + TILs reinforce T cell exhaustion. Nat Immunol. 2020;21:1540–51.
LeBleu VS, O’Connell JT, Gonzalez Herrera KN, Wikman H, Pantel K, Haigis MC, et al. PGC-1α mediates mitochondrial biogenesis and oxidative phosphorylation in cancer cells to promote metastasis. Nat Cell Biol. 2014;16:992–1003.
Soto-Heredero G, Desdín-Micó G, Mittelbrunn M. Mitochondrial dysfunction defines T cell exhaustion. Cell Metabolism. 2021;33:470–2.
Giessner C, Millet V, Mostert KJ, Gensollen T, Vu Manh T-P, Garibal M, et al. Vnn1 pantetheinase limits the Warburg effect and sarcoma growth by rescuing mitochondrial activity. Life Sci Alliance. 2018;1:e201800073.
Rommelaere S, Millet V, Gensollen T, Bourges C, Eeckhoute J, Hennuyer N, et al. PPARalpha regulates the production of serum Vanin-1 by liver. FEBS Letters. 2013;587:3742–8.
Naquet P, Kerr EW, Vickers SD, Leonardi R. Regulation of coenzyme A levels by degradation: the ‘Ins and Outs.’ Progress in Lipid Research. 2020;78:101028.
Millet V, Gensollen T, Maltese M, Serrero M, Lesavre N, Bourges C, et al. Harnessing the Vnn1 pantetheinase pathway boosts short chain fatty acids production and mucosal protection in colitis. Gut [Internet]. BMJ Publishing Group; 2022 [cited 2022 Oct 24]; Available from: https://gut-bmj-com.proxy.insermbiblio.inist.fr/content/early/2022/09/29/gutjnl-2021-325792
Naquet P, Giessner C, Galland F. Metabolic adaptation of tissues to stress releases metabolites influencing innate immunity. Current Opinion in Immunology. 2016;38:30–8.
Coenzyme A fuels T cell anti-tumor immunity. Cell Metabolism. Cell Press; 2021;33:2415–2427.e6.
Timblin GA, Tharp KM, Hoeve J ten, Baydemir I, Khantwal C, Farahzad JN, et al. Coenzyme A governs proinflammatory macrophage metabolism [Internet]. bioRxiv; 2022 [cited 2022 Oct 29]. page 2022.08.30.505732. Available from: https://www.biorxiv.org/content/10.1101/2022.08.30.505732v1
Burnet M. IMMUNOLOGICAL FACTORS IN THE PROCESS OF CARCINOGENESIS. Br Med Bull. 1964;20:154–8.
Michelakis ED, Webster L, Mackey JR. Dichloroacetate (DCA) as a potential metabolic-targeting therapy for cancer. Br J Cancer. Nature Publishing Group; 2008;99:989–94.
Whitehouse S, Cooper RH, Randle PJ. Mechanism of activation of pyruvate dehydrogenase by dichloroacetate and other halogenated carboxylic acids. Biochemical Journal. 1974;141:761–74.
Tataranni T, Piccoli C. Dichloroacetate (DCA) and Cancer: An Overview towards Clinical Applications. Oxid Med Cell Longev. 2019;2019:8201079.
Xiao B, Deng X, Zhou W, Tan E-K. Flow Cytometry-Based Assessment of Mitophagy Using MitoTracker. Frontiers in Cellular Neuroscience [Internet]. 2016 [cited 2022 Sep 27];10. Available from: https://www.frontiersin.org/articles/10.3389/fncel.2016.00076
Wang B, Li Q, Qin L, Zhao S, Wang J, Chen X. Transition of tumor-associated macrophages from MHC class IIhi to MHC class IIlow mediates tumor progression in mice. BMC Immunol. 2011;12:43.
Liu Y, Cao X. The origin and function of tumor-associated macrophages. Cell Mol Immunol. 2015;12:1–4.
Semenza GL. HIF-1 mediates metabolic responses to intratumoral hypoxia and oncogenic mutations. J Clin Invest. American Society for Clinical Investigation; 2013;123:3664–71.
Marcovecchio PM, Thomas G, Salek-Ardakani S. CXCL9-expressing tumor-associated macrophages: new players in the fight against cancer. J Immunother Cancer. BMJ Specialist Journals; 2021;9:e002045.
Dangaj D, Bruand M, Grimm AJ, Ronet C, Barras D, Duttagupta PA, et al. Cooperation between Constitutive and Inducible Chemokines Enables T Cell Engraftment and Immune Attack in Solid Tumors. Cancer Cell. 2019;35:885–900.e10.
Cooper AM, Khader SA. IL-12p40: an inherently agonistic cytokine. Trends Immunol. 2007;28:33–8.
Abdi K, Singh NJ. Making many from few: IL-12p40 as a model for the combinatorial assembly of heterodimeric cytokines. Cytokine. 2015;76:53–7.
Cueto FJ, Sancho D. The Flt3L/Flt3 Axis in Dendritic Cell Biology and Cancer Immunotherapy. Cancers. Multidisciplinary Digital Publishing Institute; 2021;13:1525.
Gao Y, Souza-Fonseca-Guimaraes F, Bald T, Ng SS, Young A, Ngiow SF, et al. Tumor immunoevasion by the conversion of effector NK cells into type 1 innate lymphoid cells. Nat Immunol. Nature Publishing Group; 2017;18:1004–15.
Vienne M, Etiennot M, Escalière B, Galluso J, Spinelli L, Guia S, et al. Type 1 Innate Lymphoid Cells Limit the Antitumoral Immune Response. Frontiers in Immunology [Internet]. 2021 [cited 2022 Oct 24];12. Available from: https://www.frontiersin.org/articles/10.3389/fimmu.2021.768989
Chow MT, Ozga AJ, Servis RL, Frederick DT, Lo JA, Fisher DE, et al. Intratumoral Activity of the CXCR3 Chemokine System Is Required for the Efficacy of Anti-PD-1 Therapy. Immunity. 2019;50:1498–1512.e5.
Wohn C, Le Guen V, Voluzan O, Fiore F, Henri S, Malissen B. Absence of MHC class II on cDC1 dendritic cells triggers fatal autoimmunity to a cross-presented self-antigen. Sci Immunol. 2020;5:eaba1896.
Eulo V, Van Tine BA. Immune checkpoint inhibitor resistance in soft tissue sarcoma. Cancer Drug Resistance. OAE Publishing Inc.; 2022;5:328–38.
Kleffel S, Posch C, Barthel SR, Mueller H, Schlapbach C, Guenova E, et al. Melanoma cell-intrinsic PD-1 receptor functions promote tumor growth. Cell. 2015;162:1242–56.
Benci JL, Xu B, Qiu Y, Wu T, Dada H, Victor CT-S, et al. Tumor Interferon Signaling Regulates a Multigenic Resistance Program to Immune Checkpoint Blockade. Cell. 2016;167:1540–1554.e12.
Miallot R, Galland F, Millet V, Blay J-Y, Naquet P. Metabolic landscapes in sarcomas. Journal of Hematology & Oncology. 2021;14:114.
Zorov DB, Juhaszova M, Sollott SJ. Mitochondrial Reactive Oxygen Species (ROS) and ROS-Induced ROS Release. Physiol Rev. 2014;94:909–50.
Chen X, Stewart E, Shelat AA, Qu C, Bahrami A, Hatley M, et al. Targeting Oxidative Stress in Embryonal Rhabdomyosarcoma. Cancer Cell. 2013;24:710–24.
Rongvaux A. Innate immunity and tolerance toward mitochondria. Mitochondrion. 2018;41:14–20.
Porporato PE. Mitochondrial metabolism and cancer. Cell Research. 2018;28:16.
Jones AE, Divakaruni AS. Macrophage activation as an archetype of mitochondrial repurposing. Molecular Aspects of Medicine. 2020;71:100838.
Harel M, Ortenberg R, Varanasi SK, Mangalhara KC, Mardamshina M, Markovits E, et al. Proteomics of Melanoma Response to Immunotherapy Reveals Mitochondrial Dependence. Cell. 2019;179:236–250.e18.
Yin M, O’Neill LAJ. The role of the electron transport chain in immunity. The FASEB Journal. 2021;35:e21974.
Bonifaz L, Cervantes-Silva M, Ontiveros-Dotor E, López-Villegas E, Sánchez-García F. A Role For Mitochondria In Antigen Processing And Presentation. Immunology. 2014;
Bykovskaia SN, Shurin GV, Graner S, Bunker ML, Olson W, Thomas R, et al. Differentiation of Immunostimulatory Stem-Cell- and Monocyte-Derived Dendritic Cells Involves Maturation of Intracellular Compartments Responsible for Antigen Presentation and Secretion. STEM CELLS. 2002;20:380–93.
Jha AK, Huang SC-C, Sergushichev A, Lampropoulou V, Ivanova Y, Loginicheva E, et al. Network integration of parallel metabolic and transcriptional data reveals metabolic modules that regulate macrophage polarization. Immunity. 2015;42:419–30.
Field CS, Baixauli F, Kyle RL, Puleston DJ, Cameron AM, Sanin DE, et al. Mitochondrial Integrity Regulated by Lipid Metabolism Is a Cell-Intrinsic Checkpoint for Treg Suppressive Function. Cell Metab. 2020;31:422–437.e5.
Broz ML, Binnewies M, Boldajipour B, Nelson AE, Pollack JL, Erle DJ, et al. Dissecting the tumor myeloid compartment reveals rare activating antigen-presenting cells critical for T cell immunity. Cancer Cell. 2014;26:638–52.
Brewitz A, Eickhoff S, Dähling S, Quast T, Bedoui S, Kroczek RA, et al. CD8 + T Cells Orchestrate pDC-XCR1 + Dendritic Cell Spatial and Functional Cooperativity to Optimize Priming. Immunity. 2017;46:205–19.
Dorner BG, Dorner MB, Zhou X, Opitz C, Mora A, Güttler S, et al. Selective expression of the chemokine receptor XCR1 on cross-presenting dendritic cells determines cooperation with CD8 + T cells. Immunity. 2009;31:823–33.
Barry KC, Hsu J, Broz ML, Cueto FJ, Binnewies M, Combes AJ, et al. A natural killer-dendritic cell axis defines checkpoint therapy-responsive tumor microenvironments. Nat Med. 2018;24:1178–91.
Kratchmarov R, Viragova S, Kim MJ, Rothman NJ, Liu K, Reizis B, et al. Metabolic control of cell fate bifurcations in a hematopoietic progenitor population. Immunology & Cell Biology. 2018;96:863–71.
Du X, Wen J, Wang Y, Karmaus PWF, Khatamian A, Tan H, et al. Hippo/Mst signalling couples metabolic state and immune function of CD8α + dendritic cells. Nature. Nature Publishing Group; 2018;558:141–5.
Farrand KJ, Dickgreber N, Stoitzner P, Ronchese F, Petersen TR, Hermans IF. Langerin + CD8α + Dendritic Cells Are Critical for Cross-Priming and IL-12 Production in Response to Systemic Antigens. The Journal of Immunology. American Association of Immunologists; 2009;183:7732–42.
Chang C-H, Curtis JD, Maggi LB, Faubert B, Villarino AV, O’Sullivan D, et al. Posttranscriptional control of T cell effector function by aerobic glycolysis. Cell. 2013;153:1239–51.
Sena LA, Li S, Jairaman A, Prakriya M, Ezponda T, Hildeman DA, et al. Mitochondria are required for antigen-specific T cell activation through reactive oxygen species signaling. Immunity. 2013;38:225–36.
Almeida L, Dhillon-LaBrooy A, Carriche G, Berod L, Sparwasser T. CD4 + T-cell differentiation and function: Unifying glycolysis, fatty acid oxidation, polyamines NAD mitochondria. Journal of Allergy and Clinical Immunology. 2021;148:16–32.
van der Windt GJW, Pearce EL. Metabolic switching and fuel choice during T-cell differentiation and memory development. Immunol Rev. 2012;249:27–42.
Phan AT, Doedens AL, Palazon A, Tyrakis PA, Cheung KP, Johnson RS, et al. Constitutive Glycolytic Metabolism Supports CD8 + T Cell Effector Memory Differentiation during Viral Infection. Immunity. 2016;45:1024–37.
Kamiński MM, Sauer SW, Kamiński M, Opp S, Ruppert T, Grigaravičius P, et al. T cell activation is driven by an ADP-dependent glucokinase linking enhanced glycolysis with mitochondrial reactive oxygen species generation. Cell Rep. 2012;2:1300–15.
Ma EH, Verway MJ, Johnson RM, Roy DG, Steadman M, Hayes S, et al. Metabolic Profiling Using Stable Isotope Tracing Reveals Distinct Patterns of Glucose Utilization by Physiologically Activated CD8 + T Cells. Immunity. 2019;51:856–870.e5.
Salerno F, Guislain A, Cansever D, Wolkers MC. TLR-Mediated Innate Production of IFN-γ by CD8 + T Cells Is Independent of Glycolysis. The Journal of Immunology. American Association of Immunologists; 2016;196:3695–705.
Bailis W, Shyer JA, Zhao J, Canaveras JCG, Al Khazal FJ, Qu R, et al. Distinct modes of mitochondrial metabolism uncouple T cell differentiation and function. Nature. Nature Publishing Group; 2019;571:403–7.
Yang M, Du W, Yi L, Wu S, He C, Zhai W, et al. Checkpoint molecules coordinately restrain hyperactivated effector T cells in the tumor microenvironment. OncoImmunology. Taylor & Francis; 2020;9:1708064.
Bengsch B, Johnson AL, Kurachi M, Odorizzi PM, Pauken KE, Attanasio J, et al. Bioenergetic Insufficiencies Due to Metabolic Alterations Regulated by the Inhibitory Receptor PD-1 Are an Early Driver of CD8(+) T Cell Exhaustion. Immunity. 2016;45:358–73.
Blank CU, Haining WN, Held W, Hogan PG, Kallies A, Lugli E, et al. Defining “T cell exhaustion.” Nat Rev Immunol. 2019;19:665–74.
Wei J, Zheng W, Chapman NM, Geiger TL, Chi H. T cell metabolism in homeostasis and cancer immunity. Current Opinion in Biotechnology. 2021;68:240–50.
Chamoto K, Chowdhury PS, Kumar A, Sonomura K, Matsuda F, Fagarasan S, et al. Mitochondrial activation chemicals synergize with surface receptor PD-1 blockade for T cell-dependent antitumor activity. Proc Natl Acad Sci USA [Internet]. 2017 [cited 2022 Oct 29];114. Available from: https://pnas.org/doi/full/10.1073/pnas.1620433114
Chowdhury PS, Chamoto K, Kumar A, Honjo T. PPAR-Induced Fatty Acid Oxidation in T Cells Increases the Number of Tumor-Reactive CD8 + T Cells and Facilitates Anti-PD-1 Therapy. Cancer Immunol Res. 2018;6:1375–87.
Jackson M, Serada N, Sheehan M, Srinivasan S, Mason N, Guha M, et al. Mitochondrial genome and functional defects in osteosarcoma are associated with their aggressive phenotype. Bai Y, editor. PLoS ONE. 2018;13:e0209489.
Voehringer D, Liang H-E, Locksley RM. Homeostasis and effector function of lymphopenia-induced “memory-like” T cells in constitutively T cell-depleted mice. J Immunol. 2008;180:4742–53.
Hogquist KA, Jameson SC, Heath WR, Howard JL, Bevan MJ, Carbone FR. T cell receptor antagonist peptides induce positive selection. Cell. 1994;76:17–27.
Zheng GXY, Terry JM, Belgrader P, Ryvkin P, Bent ZW, Wilson R, et al. Massively parallel digital transcriptional profiling of single cells. Nat Commun. 2017;8:14049.
Hao Y, Hao S, Andersen-Nissen E, Mauck WM, Zheng S, Butler A, et al. Integrated analysis of multimodal single-cell data. Cell. 2021;184:3573–3587.e29.
McInnes L, Healy J, Melville J. UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction [Internet]. arXiv; 2020 [cited 2022 Nov 13]. Available from: http://arxiv.org/abs/1802.03426
Yu G, Wang L-G, Han Y, He Q-Y. clusterProfiler: an R Package for Comparing Biological Themes Among Gene Clusters. OMICS: A Journal of Integrative Biology. Mary Ann Liebert, Inc., publishers; 2012;16:284–7.
Korotkevich G, Sukhov V, Budin N, Shpak B, Artyomov MN, Sergushichev A. Fast gene set enrichment analysis [Internet]. bioRxiv; 2021 [cited 2022 Dec 2]. page 060012. Available from: https://www.biorxiv.org/content/10.1101/060012v3
Kanehisa M, Furumichi M, Sato Y, Kawashima M, Ishiguro-Watanabe M. KEGG for taxonomy-based analysis of pathways and genomes. Nucleic Acids Res. 2022;gkac963.
Liberzon A, Birger C, Thorvaldsdóttir H, Ghandi M, Mesirov JP, Tamayo P. The Molecular Signatures Database Hallmark Gene Set Collection. cels. Elsevier; 2015;1:417–25.
Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102:15545–50.
Gu Z, Eils R, Schlesner M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics. 2016;32:2847–9.
ggplot2 [Internet]. [cited 2022 Nov 13]. Available from: https://link.springer.com/book/10.1007/978-3-319-24277-4
Bergen V, Lange M, Peidli S, Wolf FA, Theis FJ. Generalizing RNA velocity to transient cell states through dynamical modeling. Nat Biotechnol. Nature Publishing Group; 2020;38:1408–14.
Filtjens J, Roger A, Quatrini L, Wieduwild E, Gouilly J, Hoeffel G, et al. Nociceptive sensory neurons promote CD8 T cell responses to HSV-1 infection. Nat Commun. 2021;12:2936.
Bankhead P, Loughrey MB, Fernández JA, Dombrowski Y, McArt DG, Dunne PD, et al. QuPath: Open source software for digital pathology image analysis. Sci Rep. Nature Publishing Group; 2017;7:16878.
Bertucci F, Niziers V, de Nonneville A, Finetti P, Mescam L, Mir O, et al. Immunologic constant of rejection signature is prognostic in soft-tissue sarcoma and refines the CINSARC signature. J Immunother Cancer. 2022;10:e003687.
Gatza ML, Lucas JE, Barry WT, Kim JW, Wang Q, Crawford MD, et al. A pathway-based classification of human breast cancer. Proc Natl Acad Sci U S A. 2010;107:6994–9.
Bindea G, Mlecnik B, Tosolini M, Kirilovsky A, Waldner M, Obenauf AC, et al. Spatiotemporal Dynamics of Intratumoral Immune Cells Reveal the Immune Landscape in Human Cancer. Immunity. 2013;39:782–95.
Rooney MS, Shukla SA, Wu CJ, Getz G, Hacohen N. Molecular and genetic properties of tumors associated with local immune cytolytic activity. Cell. 2015;160:48–61.
Ayers M, Lunceford J, Nebozhyn M, Murphy E, Loboda A, Kaufman DR, et al. IFN-γ-related mRNA profile predicts clinical response to PD-1 blockade. J Clin Invest. 2017;127:2930–40.
Coppola D, Nebozhyn M, Khalil F, Dai H, Yeatman T, Loboda A, et al. Unique ectopic lymph node-like structures present in human primary colorectal carcinoma are identified by immune gene array profiling. Am J Pathol. 2011;179:37–45.
Ballot E, Ladoire S, Routy B, Truntzer C, Ghiringhelli F. Tumor Infiltrating Lymphocytes Signature as a New Pan-Cancer Predictive Biomarker of Anti PD-1/PD-L1 Efficacy. Cancers (Basel). 2020;12:E2418.
Thompson JC, Davis C, Deshpande C, Hwang W-T, Jeffries S, Huang A, et al. Gene signature of antigen processing and presentation machinery predicts response to checkpoint blockade in non-small cell lung cancer (NSCLC) and melanoma. J Immunother Cancer. 2020;8:e000974.

Table 1 is available in the Supplementary Files section

There is NO Competing Interest.

SuppFigures.pdf
Figures 1 to 9 Figure S1 A. Enzymatic reaction catalyzed by the Vanin1 pantetheinase: pantetheine is hydrolyzed by Vanin1 into cysteamine and vitamin B5 or pantothenic acid, the latter being the Coenzyme A biosynthetic precursor. B. Effect of administration of pantetheinase products CEA and PanSH on MCA growth. C. NMR PCA analysis of PBS, Pant, DCA, DCA + Pant samples showing the top 5 enriched metabolites detected in samples. Correlation coefficients were calculated for each metabolite (n=10 per condition). D. Plot representation of MDR versus MG staining of CD45^- cells from PBS- or Pant-tumors harvested at day 20 and 28 post engraftment. E. Gating strategy for the analysis of CD45+ cells by flow cytometry shown in Fig2A. Figure S2 A. uMAP representation of the 4 metaclusters identified after analysis of the pooled single-cell experimental data: myeloid, APC, lymphoid and non-immune cells called others. B. Density representation of the myeloid metacluster for each experimental condition C. Volcano plot highlighting the results of differential expression analysis in PBS versus Pant sample at day 28. The name of the top 10 up- and down-regulated genes sorted by Pvalue are highlighted. D. KEGG graph representation of the pathway “Antigen processing and presentation” for myeloid cells at D20. Colored boxes show the relative logFC ratio computed during differential expression analysis (red = enriched in Pant; blue = enriched in PBS). Figure S3 A, B. Projection on an individual uMAP (A) and identification of dendritic cell clusters (B): cDC1, cDC1, regulatory-like DC, proliferating DC, IFN-responsive monoDC, pDC and B cells. C. Density plot representation of the APC metacluster for each experimental condition D. Geometric MFI of MHC II staining on tumor infiltrating APC at day 20 from PBS and Pant-treated mice evaluated by flow cytometry. (n=5-7 per condition). Mann-Whitney test; **p-value < 0.01. E. Immunofluorescence analysis of CD8β and MHC II staining on PBS and Pant tumors sections (n=10) at day 20. The contact points between CD8β⁺ cells and MHC II⁺ cells were quantified and normalized by the number of CD8β⁺ cells present in the field. F. CBA quantification of CCL2 and CXCL9 concentrations in tumor lysates at day 20 from PBS and Pant-treated mice (n=7 per condition). Mann-Whitney test; **p-value < 0.01. * p-value<0.05. G. Normalized expression values as modulescore (Seurat) of previously anti-tumoral and protumoral genesets in different clusters from Day 20 and 28 from PBS and Pant-treated tumors. Values are centered-scaled for each row. Figure S4 A,B. Quantification of immune cells in lymph nodes and spleen at steady state or in the presence of a tumor, with or without Pant administration at D24. C. Gating strategy for the flow cytometry analysis of lymphoid populations. Figure S5 A. Identification on individual uMAP projection of the 8 clusters segregating the lymphoid metacluster: NK1.1/ILC1, CD8⁺, Activated CD8⁺, CD4⁺, Treg and Ki67 CD4⁺ T cell, Tregs and NK cells, represented in the uMAP. B, C. Density plot representation of the lymphoid metacluster for each experimental condition (B) and cell-specific marker-based identification of cell subset (C). D. Representation of the number of co-expressed activation-associated genes (defined as positive) in CD8⁺ or CD4⁺ T cells under various experimental conditions. E. Module score representation of normalized expression values averaged across cells for activation and exhaustion genesets in CD4⁺ and CD8⁺ T cells from D20-D28 tumors. Values are centered-scaled for each row. F. Relative expression of top 20 ILC (left) and NK (right) cells markers genes as described in the method section. Figure S6 A, B. Quantification (A) and proportion of CTV-labelled CD8⁺ OT-1 T cells for each cycle of division from G1 to G7 (B) harvested from ear draining lymph nodes after id OVA immunization; n=5. C,D. Absolute numbers (C) and CFSE quantification (D) of OT1+ cells present in the culture with DC extracted for TDLNs in the presence or absence of OVA SIINFEKL peptide (n=5 per condition). E, F. Flow cytometry control of CD8 (E) and NK1.1+ (F) cell depletion in peripheral blood lymphocytes. Figure S7: A, B. Pathway Immunochemistry analysis (A) and quantification (B) of paraffin-embedded OsA tumor samples using CD8, CD31 labeling and necrosis scoring; n=5. Figure S8: VNN1 expression in human soft tissue sarcomas and correlation with immune signatures and clinicopathological characteristics. A. Pathway analysis using the WGCNA tool of Vanin1-coexpressed genes based on the analysis of the TCGA Database highlighting overexpressed MSigDB hallmarks. B. Cibersort analysis of deconvoluting immune cell subsets enriched in VNN1-high tumors. C. List of soft tissue sarcoma data sets included in our analysis. D. Clinicopathological characteristics of patients and samples in the whole population and according to VNN1-based classification. Figure S9: Absence of prognostic value for MFS of VNN1 expression in human bone sarcomas and osteosarcomas. A. List of bone sarcoma and osteosarcoma data sets included in analysis. B. Kaplan-Meier MFS curves according to VNN1 mRNA expression in a cohort of 326 patients with bone sarcoma (left), and a subcohort of 94 patients with osteosarcoma (right). The p-values are for the log-rank test.
Table1.pdf

Download PDF

Version 1

posted

You are reading this latest preprint version

The coenzyme A precursor pantethine restrains sarcoma growth through promotion of type 1 immunity

Status:

Version 1

Abstract

Figures

Introduction

Results

Pantethine therapy reduces MCA205 growth in immunocompetent mice

Pantethine enhances myeloid signatures associated with antigen presentation and type 1 immunity

Pantethine enhances the development of effectors within adaptive lymphocytes

Potential and limits of Pant therapy

High VNN1 expression in STS correlates with higher immune response and better prognosis

Discussion

Material & Methods

Animals

Cell lines

Treatments

Tumor experiments

LC-MS

NMR

Seahorse

MitoSox

Lactate Assay

Flow Cytometry and antibodies

Single Cell RNA sequencing

Cytokine array

CBA

In vivo Depletion

HSV-1 infection

OVA immunization

DC-OT-1 co-culture

Immunofluorescence

Immunohistology

Analysis of VNN1 expression in TCGA Database and Cibersort

Analysis of VNN1 expression in soft tissue sarcoma clinical samples

Statistical analysis

Declarations

References

Table

Additional Declarations

Supplementary Files

Status:

Version 1