Mitochondrial-cytochrome c oxidase II promotes glutaminolysis to sustain tumor cell survival upon glucose deprivation

Yi, Yong; Wang, Guoqiang; Zhang, Wenhua; Yu, Shuhan; Fei, Junjie; An, Tingting; Yi, Jianqiao; Li, Fengtian; Huang, Ting; Yang, Jian; Niu, Mengmeng; Wang, Yang; Xu, Chuan; Xiao, Zhi-Xiong Jim

doi:10.1038/s41467-024-55768-9

Download PDF

Article
Open access
Published: 02 January 2025

Mitochondrial-cytochrome c oxidase II promotes glutaminolysis to sustain tumor cell survival upon glucose deprivation

Nature Communications volume 16, Article number: 212 (2025) Cite this article

9507 Accesses
12 Citations
11 Altmetric
Metrics details

Subjects

Abstract

Glucose deprivation, a hallmark of the tumor microenvironment, compels tumor cells to seek alternative energy sources for survival and growth. Here, we show that glucose deprivation upregulates the expression of mitochondrial-cytochrome c oxidase II (MT-CO2), a subunit essential for the respiratory chain complex IV, in facilitating glutaminolysis and sustaining tumor cell survival. Mechanistically, glucose deprivation activates Ras signaling to enhance MT-CO2 transcription and inhibits IGF2BP3, an RNA-binding protein, to stabilize MT-CO2 mRNA. Elevated MT-CO2 increases flavin adenosine dinucleotide (FAD) levels in activating lysine-specific demethylase 1 (LSD1) to epigenetically upregulate JUN transcription, consequently promoting glutaminase-1 (GLS1) and glutaminolysis for tumor cell survival. Furthermore, MT-CO2 is indispensable for oncogenic Ras-induced glutaminolysis and tumor growth, and elevated expression of MT-CO2 is associated with poor prognosis in lung cancer patients. Together, these findings reveal a role for MT-CO2 in adapting to metabolic stress and highlight MT-CO2 as a putative therapeutic target for Ras-driven cancers.

MRI assessment of glutamine uptake correlates with the distribution of glutamine transporters and cancer stem cell markers

Article Open access 01 April 2022

Tumor-suppressive function of EZH2 is through inhibiting glutaminase

Article Open access 20 October 2021

Mitochondrial metabolism and cancer therapeutic innovation

Article Open access 04 August 2025

Introduction

Metabolic reprogramming is a hallmark of cancer cells¹. Most cancer cells use glucose as the major energy source and preferentially catabolize glucose to lactate even under abundant oxygen conditions, a phenomenon termed the Warburg effect or aerobic glycolysis¹. Aerobic glycolysis can rapidly but inefficiently generate ATP by consuming large amounts of glucose to meet the needs of rapidly proliferating tumor cells^1,2. It has been well-known that glucose deprivation is one of the characteristics of the tumor microenvironment, which can be attributed to the large consumption of glucose by tumor cells^1,2, and by insufficient blood supply due to poor angiogenesis³. Indeed, it has been reported that the glucose concentrations in the solid tumor tissues are 3- to 10-fold lower than in normal tissues^4,5. Therefore, Tumor cells must sense glucose shortage and seek alternative energy resources for tumor cell survival and proliferation.

Glutamine is an alternative source of energy for cancer cells. Glutamine enters the mitochondria and is catabolized to glutamate by glutaminase-1 (GLS1)⁶. Glutamate can then be converted by glutamate dehydrogenase (GLUD1) to α-ketoglutarate (α-KG), which then enters the TCA cycle and is ultimately oxidized to produce ATP⁷. Furthermore, glutamine-derived α-KG can be reductively carboxylated to generate citrate, which can generate acetyl-CoA for de novo fatty acid synthesis⁸. Moreover, glutamine-derived glutamate also can be converted to non-essential amino acid (NEAA) by alanine or aspartate transaminases (TAs) for protein synthesis⁹. It is reported that glutamine-derived TCA cycle intermediates, including fumarate, malate, and citrate, are significantly increased upon glucose deprivation¹⁰. However, how glucose deprivation triggers metabolic reprogramming to the glutaminolysis pathway for tumor cell survival and growth remains unclear.

Mitochondria are intracellular organelles with the core function to produce the majority of ATP by oxidative phosphorylation¹¹. In mammalian cells, mitochondria consist of 16,569 base pairs of DNA organized in a closed circle, encoding 13 proteins, which are core components of the respiratory chain¹². It has been shown that cancer cells lacking mitochondrial DNA are sensitive to glucose deprivation¹³, underlying a critical role for mitochondria in cellular response to glucose deprivation. However, how mitochondria act in the cellular response to glucose deprivation is unknown.

In this study, we show that glucose deprivation leads to robust upregulation of the expression of MT-CO2, a mitochondrial genome-encoded protein essential in the complex IV of the respiratory chain, which bridges Ras signaling to epigenetic upregulation of JUN gene transcription via activation of the FAD-LSD1-H3K9me2 axis, consequently resulting in elevated GLS1 expression, glutaminolysis, and tumor cell survival. Inhibition of MT-CO2 by shRNA or berberine leads to inhibition of glutaminolysis and suppression of Ras-driven tumor growth.

Results

Mitochondrial MT-CO2 is essential for cancer cell survival upon glucose deprivation and for tumor growth in vivo

Glucose deprivation is often associated with tumor development. However, glucose deprivation can lead to an energy crisis and cell death. Thus, cancer cells must adopt alternative energy sources for cell survival and proliferation. To investigate the underlying molecular changes that lead to cell survival upon glucose deprivation, we selected human lung cancer H292 cells that could survive in DMEM lacking glucose but supplemented with 2 mM glutamine (Glc−/Gln+) and obtained two sets of viable H292 cells (H292-V#1 and H292-V#2) (Supplementary Fig. S1a). As shown in Supplementary Fig. S1b–d, while parental H292 cells were inviable in the Glc-/Gln+ condition, H292-V cells were well survived, as evidenced by either trypan blue exclusion assay or propidium iodide (PI) staining assay. Notably, xenograft tumor mouse models showed that H292-V cells exhibited higher tumorigenicity than parental H292 cells (Fig. 1a–c and Supplementary Fig. S1e, f), indicating that H292-V cells acquired a growth advantage compared to parental H292 cells. Notably, as shown in Supplementary Fig. S1g, h, ethidium bromide (EB)-induced depletion of mitochondrial DNA significantly increased the sensitivity of human lung cancer A549 cells to glucose deprivation, in keeping with a previous report¹³. We, therefore, hypothesized that mitochondrial genome-encoded proteins play a critical role in cancer cell survival upon energy stress. We thus examined the expression of genes encoded by the mitochondrial genome in H292-V and parental H292 cells. As shown in Fig. 1d, compared to parental H292 cells, the steady-state mRNA levels of MT-CO2, which encodes a core subunit of respiratory chain complex IV, were dramatically increased in H292-V cells. Since there was little difference in mitochondrial biogenesis between H292-V and parental H292 cells, as evidenced by the comparable ratio of mitochondrial DNA/nuclear DNA or citrate synthase levels (Supplementary Fig. S1i, j), it suggests that the robust elevation of the MT-CO2 mRNA levels is most likely attributed to the increased gene transcription and/or increased mRNA stability. Consistent with this notion, western blot analyses showed that H292-V cells had higher MT-CO2 protein expression than parental H292 cells (Fig. 1e).

**Fig. 1: MT-CO2 is essential for tumor cell survival and tumor growth upon glucose deprivation.**

Next, we investigated the biological significance of elevated mitochondrial MT-CO2 in sustained tumor cell survival upon glucose starvation. As shown in Supplementary Fig. S1k, l, A549 or H1299 cells with high levels of MT-CO2 survived well upon glucose deprivation. By contrast, H292 or H1975 cells, both of which had low MT-CO2 expression, were unable to survive (Supplementary Fig. S1k, l). To inhibit MT-CO2 expression in A549 and H1299 cells, we designed and verified two different shRNAs that can effectively knockdown MT-CO2 expression (Fig. 1f), in keeping with a previous report that RNAi technology can be used to silence the mitochondrial genome-encoding genes in vitro¹⁴. As shown in Fig. 1g, silencing of MT-CO2 dramatically inhibited A549 and H1299 cell viability upon glucose deprivation. In addition, the knockdown of MT-CO2 also significantly inhibited the proliferation of A549 and H1299 cells (Fig. 1h, i), indicating that MT-CO2 is essential for tumor cell growth even in an environment rich in glucose and glutamine. Consistently, in vivo xenograft tumor mouse models showed that knockdown of MT-CO2 dramatically inhibited tumor growth, concomitant with reduced Ki67 expression (Fig. 1j–l and Supplementary Fig. S1m–o). Importantly, the silencing of MT-CO2 significantly re-sensitized H292-V cells to glucose deprivation (Fig. 1m, n). Moreover, the knockdown of MT-CO2 in H292-V cells markedly inhibited cell proliferation in vitro and tumor growth in vivo (Fig. 1o–r and Supplementary Fig. S1p, q).

Taken together, these results indicate that glucose deprivation leads to upregulation of MT-CO2, which plays a pivotal role in cancer cell survival upon glucose deprivation and in tumor growth in vivo.

MT-CO2 elevates GLS1 expression in promoting glutaminolysis critical for cancer cell survival upon glucose deprivation

Next, we investigated the molecular basis by which MT-CO2 regulates cancer cell survival upon glucose deprivation. Glutamine, fatty acids, or glycogen can serve as alternative energy sources for cancer cells^15,16,17. Notably, inhibition of fatty acid oxidation by etomoxir, a selective inhibitor of carnitine palmitoyl-transferase 1a (CPT1a), or inhibition of glycogen catabolism by CP-91149, a glycogen phosphorylase inhibitor, led to upregulated expression of p53 and p21 (Supplementary Fig. S2a), as previously reported^18,19. However, either etomoxir or CP-91149 had little effect on the viability of A549 or H1299 cells under Glc−/Gln+ condition (DMEM containing 2 mM glutamine in the absence of glucose) (Fig. 2a, b and Supplementary Fig. S2b, c). By sharp contrast, inhibition of glutaminolysis by CB-839, a selective inhibitor of glutaminase-1 (GLS1), dramatically suppressed the survival of A549 and H1299 cells under Glc-/Gln+ condition (Fig. 2a, b and Supplementary Fig. S2b, c), indicating that cancer cells preferentially utilize glutamine as an energy source for survival upon glucose deprivation. Consistently, depletion of glutamine resulted in robust cancer cell death in the absence of glucose (Supplementary Fig. S2d). Furthermore, ectopic expression of GLS1 in H292 cells markedly upregulated cellular oxygen consumption rate (OCR), cell survival under the Glc-/Gln+ condition, and tumor growth in vivo (Supplementary Fig. S2e–i). These results indicate that glutamine is the preferred alternative energy source and that glutaminolysis is essential for cancer cell survival under glucose deprivation.

**Fig. 2: MT-CO2 elevates GLS1 expression in facilitating glutaminolysis to promote tumor cell survival upon glucose deprivation.**

Given that glutaminolysis plays a pivotal role in maintaining tumor cell survival upon glucose deprivation, we hypothesized that MT-CO2 could be involved in regulating glutaminolysis. To address this issue, we performed mass spectrometry analyses on energy metabolites. As shown in Supplementary Fig. S2j, the silencing of MT-CO2 in H1299 cells dramatically increased cellular glutamine levels. Notably, the expression of glutamine transporter, SLC1A5, was not affected by knockdown of MT-CO2 (Supplementary Fig. S2k), suggesting that silencing of MT-CO2 may lead to blockage of glutaminolysis which leads to cellular glutamine accumulation. We then traced the metabolic flux of ¹³C-labeled glutamine (¹³C₅-glutamine). As shown in Fig. 2c, d, there was a reduced enrichment of ¹³C from ¹³C₅-glutamine in downstream metabolites such as glutamate, α-ketoglutarate, succinate, fumarate, and citrate. Additionally, ¹³C₅-glutamine accumulation was observed upon knockdown of MT-CO2 (Fig. 2d), indicating that silencing MT-CO2 inhibits glutaminolysis, blocking tumor cells from utilizing glutamine as a carbon source.

To investigate the molecular basis by which MT-CO2 regulates glutaminolysis, we examined the effects of silencing of MT-CO2 on the expression of the glutaminolysis-related enzymes. As shown in Fig. 2e–g, silencing of MT-CO2 significantly inhibited the mRNA and protein expression of GLS1, while it had little effect on the expression of glutamate dehydrogenase (GLUD1), which catalyzes glutamate to α-ketoglutarate (α-KG). Importantly, silencing of MT-CO2-mediated inhibition of cell viability under the Glc-/Gln+ condition was significantly rescued by supplement of dimethyl α-KG (DM-αKG, a cell-permeable analog of α-KG) (Fig. 2h, i) or ectopic expression of GLS1 (Fig. 2j, k). Moreover, ectopic expression of GLS1 also significantly rescued tumor growth in vivo inhibited by silencing of MT-CO2 (Fig. 2l–r).

Together, these results indicate that silencing of MT-CO2 inhibits glutaminolysis and tumor growth via the downregulation of GLS1 expression.

MT-CO2 facilitates JUN transcription through activation of the FAD-LSD1-H3K9me2 axis to promote GLS1 expression

To explore the molecular mechanism(s) by which MT-CO2 regulates GLS1 expression, we performed RNA-seq analyses. As shown in Fig. 3a, the silencing of MT-CO2 significantly affected the expression of a subset of transcription factors, including c-JUN, KLF2, FOXM1, and SOX2. Notably, c-JUN has been reported as a transcription factor of GLS1²⁰, raising the possibility for MT-CO2 regulating GLS-1 through c-JUN. Indeed, similar to the effects of silencing MT-CO2, the viability of H1299 cells was also dramatically suppressed by the knockdown of c-JUN under Glc−/Gln+ condition (DMEM containing 2 mM glutamine in the absence of glucose) (Supplementary Fig. S3a, b). Furthermore, qPCR and western blot analyses showed that silencing of MT-CO2 significantly inhibited c-JUN mRNA and protein levels (Fig. 3b, c). Importantly, the silencing of MT-CO2-mediated downregulation of GLS1 expression was completely rescued by ectopic expression of c-JUN (Fig. 3d–e). Together, these results indicate that silencing of MT-CO2 leads to inhibition of GLS1 transcription via suppression of c-JUN expression.

**Fig. 3: MT-CO2 facilitates *JUN* transcription through the FAD-LSD1-H3K9me2 axis to promote GLS1 expression.**

We then investigated the molecular basis by which MT-CO2 regulates JUN transcription. Accumulating evidence indicates that disruption of mitochondria homeostasis leads to the alteration of a set of metabolites that can serve as signaling molecules, including AMP/ATP, ROS, NAD+, and acetyl coenzyme A, which are involved in the regulation of gene transcription through various mechanisms²¹. Therefore, we speculated that silencing of MT-CO2 may affect certain signaling molecules to regulate c-JUN expression. Notably, treatment with NAC (N-Acetyl-L-cysteine, a ROS scavenger), nicotinamide mononucleotide (NMN, an NAD+ precursor), ATP, or NaAc (Sodium acetate) had little effect on the silencing of MT-CO2-mediated downregulation of c-JUN and GLS1 expression (Supplementary Fig. S3c–f).

Since MT-CO2 is a key component in the complex IV of the respiratory chain, we examined whether disruption of electron transport chain function would affect c-JUN and GLS1 expression. As shown in Supplementary Fig. S4a, b, inhibition of respiratory chain complex I by small molecule inhibitors, such as rotenone and metformin, did not affect c-JUN expression. In addition, while silencing of MT-ND1 (a subunit of respiratory chain complex I) suppressed the cellular oxygen consumption rate (OCR) as expected, it had little effect on c-JUN and GLS1 expression (Supplementary Fig. S4g, j), suggesting that the disruption of complex I is not engaged in the regulation of c-JUN/GLS1. In contrast, inhibition of respiratory chain complexes II, III, IV, and V by the respective inhibitors (Carboxin for complex II, Antimycin for complex III, NaN₃ for complex IV, and Oligomycin for complex V) significantly suppressed c-JUN expression (Supplementary Fig. S4c–f), suggesting that complex II-V of the respiratory chain is important in the regulation of c-JUN. Notably, the knockdown of MT-CO2 of complex IV not only inhibited expression of c-JUN and GLS1, but also significantly suppressed the OCR levels and the activity of complex II, III, and IV (Fig. 3f, g), in keeping with the observation that glucose-starvation-resistant H292 cells (H292-V) exhibited higher MT-CO2 and OCR levels, along with increased activity of complex II, III, and IV (Figs. 1e, 3h, i). Consistently, the knockdown of MT-CYB (a subunit of respiratory chain complex III) or MT-ATP6 (a subunit of respiratory chain complex V) also led to a marked suppression of OCR levels as well as c-JUN and GLS1 expression (Supplementary Fig. S4h, i, S4k, l). These results suggest that the complex II-V of the respiratory chain plays an important role in regulating the expression of c-JUN and GLS1. It has been documented that the complex II-V of the respiratory chain is essential for FADH₂-mediated ATP generation and flavin adenosine dinucleotide (FAD) regeneration^22,23,24,25. Thus, it would be expected that the disruption of complex II-V of the respiratory chain should result in a reduction of FAD levels. Indeed, the knockdown of MT-CO2, MT-CYB, or MT-ATP6 significantly reduced cellular FAD levels (Fig. 3j and Supplementary Fig. S4m).

To explore whether FAD can rescue the downregulation of c-JUN and GLS1 expression mediated by silencing MT-CO2, we examined the effects of FAD and riboflavin, the precursor of FAD, on intracellular FAD levels. As shown in Supplementary Fig. S4n, o, treatment with FAD or riboflavin significantly increased intracellular FAD levels in a dose-dependent manner, in keeping with a previous report²⁶. Notably, FAD treatment led to a marked upregulation of c-JUN and GLS1 in a time- and dose-dependent manner (Fig. 3k and Supplementary Fig. S4p). Importantly, the silencing of MT-CO2-mediated downregulation of c-JUN and GLS1 protein expression could be effectively rescued by FAD or riboflavin (Fig. 3l–o).

It has been documented that FAD is a key cofactor of lysine-specific histone demethylase 1 (LSD1, also known as KDM1A), which catalyzes the demethylation of histone 3, including H3K4me2 and H3K9me2, to regulate gene expression²⁷. Therefore, we speculated that the MT-CO2-FAD axis might be involved in epigenetically regulating c-JUN expression via LSD1. To address this issue, we examined the effects of LSD1 on the c-JUN protein. As shown in Fig. 3p and Supplementary Fig. S5a, treatment with GSK-LSD1, a specific inhibitor of LSD1, significantly inhibited c-JUN and GLS1 expression, concomitant with markedly upregulated H3K4me2 and H3K9me2 levels. Silencing of MT-CO2 also dramatically increased H3K9me2 and H3K4me2 levels, whereas it had little effect on H3K27me3 expression (Fig. 3q).

Consistent with the knowledge that H3K9me2 serves as a repressive epigenetic mark²⁸, silencing of MT-CO2 upregulated H3K9me2 levels and downregulated c-JUN expression, which was largely rescued by FAD or riboflavin treatment, but not by α-ketoglutarate (Fig. 3m, o, and Supplementary Fig. S5b). We further investigated whether H3K9me2 could bind to the JUN gene promoter to inhibit JUN transcription. To address this issue, we designed 15 primers (P1 to P15) that can cover the -1 to -4000 region of the JUN gene promoter (Supplementary Fig. S5c). Our CHIP-qPCR analyses indicated that H3K9me2 could strongly bind to P10 to P15 (-2405 to -4000) region of the JUN gene promoter (Fig. 3r). Silencing of MT-CO2 significantly increased H3K9me2 binding to the P12 and P13 regions of the JUN gene promoter, both of which could be largely rescued by FAD treatment (Fig. 3s). Parallel experiments showed that H3K4me2 could also bind to the JUN gene promoter, which was unaffected by the silencing of MT-CO2 (Supplementary Fig. S5d–f).

Together, these results indicate that the knockdown of MT-CO2 leads to reduced electron flows of complex II-IV and FAD levels, resulting in reduced LSD1 enzymatic activities, which in turn downregulates JUN transcription via the elevation of H3K9me2 levels, consequently resulting in suppressing GLS1 expression.

Glucose deprivation activates Ras signaling to promote the MT-CO2-GLS1-glutaminolysis axis required for cancer cell survival and tumor growth

Our aforementioned data indicate that glucose deprivation promotes MT-CO2 expression to facilitate glutaminolysis and tumor cell survival. Yet, how glucose starvation upregulates MT-CO2 is unknown. To explore the underlying molecular mechanism, we performed RNA-seq analyses to profile gene expression in glucose-starvation-resistant H292 cells (H292-V) versus parental H292 cells. As shown in Supplementary Fig. S6a, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses of DEGs (Differentially expressed genes) showed that multiple signaling pathways, including Ras signaling, MAPK signaling, Hippo signaling, and Wnt signaling were significantly enriched in H292-V cells. In addition, whole exon sequencing (WES) analyses showed that compared to H292 cells, H292-V cells had no acquisition of additional gene mutations in the Ras pathway (Supplementary Table 1). However, Gene Ontology (GO) enrichment analyses showed that, in H292-V cells, there was the acquisition of gene mutations enriched in multiple GO terms, which can affect the Ras signaling, such as Ras GTPase binding, Rab GTPase binding, and GTPase activator activity (Supplementary Fig. S6b). Therefore, we speculated that glucose deprivation could promote MT-CO2 expression via activating the Ras signaling. Indeed, western blot analyses showed that the Ras downstream effector, ERK, was significantly activated in H292-V cells (Fig. 4a).

**Fig. 4: Glucose deprivation activates Ras signaling to promote MT-CO2 expression critical for cancer cell survival and tumor growth.**

We next examined the effects of activated Ras on MT-CO2 expression and tumor cell survival upon glucose starvation. As shown in Supplementary Fig. S1k, l, A549 or H1299 cells harboring activated Ras (KRas^G12S or NRas^Q61K, respectively) could survive well under Glc-/Gln+ (DMEM containing 2 mM glutamine in the absence of glucose) condition, concomitant with higher MT-CO2 protein expression. By contrast, H292 or H1975 cells harboring wild-type Ras and expressing lower levels of MT-CO2 were unable to survive (Supplementary Fig. S1k, l). Notably, ectopic expression of KRas^G12V in H292 or H1975 cells promoted cell survival under the Glc-/Gln+ condition, concomitant with dramatically elevated mRNA and protein levels of MT-CO2 (Fig. 4b–d). Furthermore, not only KRas^G12V but also HRas^G12V could significantly elevate MT-CO2 expression (Supplementary Fig. S6c). By contrast, silencing of KRas in A549 or H292-V cells significantly downregulated MT-CO2 protein expression (Fig. 4e, g), concomitant with inhibited A549 or H292-V cell viability under Glc-/Gln+ condition (Fig. 4f, h). Importantly, the knockdown of MT-CO2 completely inhibited KRas^G12V-induced survival of H292 or H1975 cells under the Glc-/Gln+ condition (Fig. 4i, j). Together, these results indicate that activation of Ras signaling promotes the upregulation of MT-CO2 expression and that MT-CO2 is indispensable to Ras-mediated cell viability in response to glucose deprivation.

It has been established that activation of Ras is a driving force of glutaminolysis through promoting the expression of glutamine transporter, SLC1A5, or glutamate-oxaloacetate transaminase (GOT)^8,29. Whether Ras-mediated upregulation of MT-CO2 plays a role in Ras-dependent glutaminolysis is unknown. As shown in Supplementary Fig. S7a, b, consistent with glucose-starvation-resistant H292 cells, ectopic expression of KRas^G12V significantly upregulated the activity of respiratory chain complex II-IV and oxygen consumption rate (OCR). Moreover, ectopic expression of KRas^G12V also significantly upregulated c-JUN and GLS1 protein expression, FAD levels as well as dramatically downregulated H3K9me2 levels (Supplementary Fig. S7c–e), all of which could be completely rescued by either silencing of MT-CO2 or treatment with LSD1 inhibitor GSK-LSD1 (Supplementary Fig. S7f–i). Importantly, the knockdown of MT-CO2 inhibited KRas^G12V-induced cell survival under the Glc-/Gln+ condition, which could be markedly rescued by supplement of dimethyl α-KG (DM-αKG) (Supplementary Fig. S7j). Together, these results indicate that activation of Ras promotes glutaminolysis via activating the MT-CO2-FAD-LSD1-H3K9me2-c-JUN-GLS1 axis.

Next, we examined the role of MT-CO2 in KRas^G12V-mediated cell proliferation and tumor growth. As shown in Fig. 4k–q, ectopic expression of KRas^G12V promoted H292 cell proliferation in vitro and tumor growth in vivo, which was significantly inhibited by the simultaneous silencing of MT-CO2. In the Rosa26-LSL-KRas^G12D lung cancer mouse model, activated Ras also dramatically elevated MT-CO2 expression in KRas^G12D-driven lung tumors (Supplementary Fig. S8a, b). To further investigate the effects of MT-CO2 on tumor growth in the KRas^G12D genetic mouse model, we designed 4 specific shRNAs for mouse MT-CO2 (mMT-CO2). As shown in Supplementary Fig. S8c, shmMT-CO2-#2 and shmMT-CO2-#4 could effectively silence MT-CO2 expression in MEF cells. To verify that these shRNAs can be used to silence MT-CO2 in vivo, we examined the effects of lentivirus-shmMT-CO2-#2 on the MT-CO2 protein expression in the mouse lung by nasal drip. As shown in Supplementary Fig. S8d, shmMT-CO2 effectively silenced MT-CO2 expression in the mouse lung. Importantly, the silencing of MT-CO2 dramatically inhibited KRas^G12D-induced lung tumor growth (Fig. 4r, s).

Taken together, these results demonstrate that glucose starvation-mediated activation of Ras signaling stimulates the MT-CO2-GLS1 axis to promote glutaminolysis, tumor cell survival, and tumor growth.

Activated Ras promotes MT-CO2 transcription via facilitating TFAM expression to promote glutaminolysis and cancer cell survival in response to glucose deprivation

We next investigated the molecular basis by which activated Ras promotes MT-CO2 expression. As shown in Supplementary Fig. S9a, b, ectopic expression of KRas^G12V had little impact on mitochondrial biogenesis, as evidenced by the comparable ratio of mitochondrial DNA to nuclear DNA and citrate synthase levels. Additionally, ectopic expression of KRas^G12V had little effect on MT-CO2 mRNA stability (Supplementary Fig. S9c), indicating that the elevated MT-CO2 mRNA levels induced by activated Ras are most likely due to increased gene transcription. Since TFAM is a key mitochondrial transcription factor responsible for recruiting mitochondrial RNA polymerase to activate mitochondrial gene transcription, including MT-CO2^30,31, we hypothesized that TFAM may act as a critical transcription factor mediating oncogenic Ras signaling and mitochondrial MT-CO2 transcription. Indeed, activated Ras significantly upregulated TFAM mRNA and protein levels (Supplementary Fig. S9d, e). Silencing of TFAM dramatically inhibited activated Ras-induced upregulation of MT-CO2 expression and FAD levels (Supplementary Fig. S9f–h). Importantly, KRas^G12V-induced cell survival in response to glucose deprivation was significantly inhibited by the knockdown of TFAM (Supplementary Fig. S9i). Furthermore, inhibition of ERK by pharmacological inhibitor PD98059 not only effectively blocked KRas^G12V-induced upregulated of TFAM and MT-CO2 expression and FAD levels (Supplementary Fig. S9j–m), but also significantly inhibited cell viability under glucose deprivation (Supplementary Fig. S9n). Together, these results indicate that oncogenic Ras activates ERK signaling to upregulate the expression of TFAM and consequently promote MT-CO2 transcription.

Glucose deprivation specifically upregulates MT-CO2 expression via suppressing IGF2BP3

The human mitochondrial genome is a circular double-stranded DNA, comprising heavy (H) and light (L) strands³². The H-strand encodes 12 subunits of the respiratory chain complex, derived from the H-strand promoter, including MT-CO2, MT-ND1, MT-CYB, and MT-ATP6³². In contrast, the L-strand only encodes MT-ND6³². The transcription of mtDNA results in the formation of long pre-MT-mRNA, which is processed into mature MT-mRNA³². Therefore, the 12 subunits encoded by the H-strand should theoretically have equal mRNA levels. However, it has been reported that the levels of MT-mRNA encoded by the H-strand are variable^33,34, suggesting additional mechanisms in modulating MT-mRNA stability. Interestingly, glucose starvation led to robust elevation of MT-CO2 mRNA levels, but not other MT-mRNAs, including MT-ND1, MT-ND5, and MT-ATP6 (Fig. 1d). Given that activation of Ras had little effect on MT-CO2 mRNA stability (Supplementary Fig. S9c), these results suggest that glucose starvation specifically upregulates MT-CO2 mRNA stability in a Ras pathway-independent manner.

Accumulating evidence indicates that N6-methyladenosine (m6A) modification plays a critical role in regulating mRNA stability³⁵. Yet, there are currently no reports on m6A modification of mitochondrial mRNA. Notably, as shown in Supplementary Fig. S10a, b, m6A could be found in mitochondria. RNA immunoprecipitation (RIP) analyses revealed that MT-CO2 mRNA had higher m6A modification levels than other MT-mRNAs (Fig. 5a). In addition, the analyses using the SRAMP database (http://www.cuilab.cn/sramp) also indicated that MT-CO2 has a high confidence for m6A modification (Supplementary Table 2).

**Fig. 5: Glucose starvation specifically facilitates MT-CO2 expression via downregulation of IGF2BP3.**

m6A modification is mediated by m6A writers and erasers and is recognized by m6A readers³⁵. We observed that, compared to parental H292 cells, the mRNA and protein levels of the m6A reader, IGF2BP3, were markedly reduced in glucose-starvation-resistant H292 cells (H292-V) (Fig. 5b, c). Furthermore, both the endogenous and exogenous IGF2BP3 could be found in mitochondria (Supplementary Fig. S10c–f). These results suggest that IGF2BP3 could specifically regulate MT-CO2 expression. Indeed, silencing IGF2BP3 specifically upregulated MT-CO2 mRNA and protein levels in H1299 cells (Fig. 5d, e). Conversely, ectopic expression of IGF2BP3 specifically suppressed MT-CO2 mRNA and protein levels (Fig. 5f, g).

We then examined the effects of IGF2BP3 on MT-CO2 mRNA stability. As shown in Fig. 5h, the knockdown of IGF2BP3 significantly upregulated MT-CO2 mRNA stability. In contrast, ectopic expression of IGF2BP3 markedly reduced MT-CO2 mRNA stability (Fig. 5i). Moreover, RIP analyses showed that IGF2BP3 could bind to MT-CO2 mRNA, but not to other MT-mRNAs, such as MT-CYB, MT-ND1, MT-CO1, MT-ND3, or MT-ND4L (Fig. 5j). Importantly, restoration of IGF2BP3 in H292-V cells markedly reduced MT-CO2 expression to levels similar to those in parental H292 cells (Fig. 5k).

Together, these results indicate that IGF2BP3 specifically inhibits MT-CO2 mRNA stability. Glucose starvation upregulates MT-CO2 expression via downregulation of IGF2BP3.

MT-CO2 expression is linked to activated Ras and the levels of c-JUN and GLS1 in human lung cancer samples

Next, we investigated the clinical relevance of this study. We first examined MT-CO2 and GLS1 protein expression in human lung cancer samples. As shown in Fig. 6a–c, both MT-CO2 and GLS1 protein expression were significantly upregulated in lung cancer samples compared to adjacent tissues. Importantly, the protein expression of MT-CO2 and GLS1 exhibited a significant and positive correlation in clinical lung cancer samples (r = 0.69, p < 0.0001; Fig. 6d). In addition, lung cancer samples with KRAS mutations showed higher mRNA levels of MT-CO2, c-JUN, or GLS1 (Fig. 6e–g). Notably, a positive correlation was observed regarding mRNA expression between MT-CO2 and GLS1 (r = 0.66, p = 0.0005), or between MT-CO2 and c-JUN (r = 0.55, p < 0.0001) (Fig. 6h, i). Furthermore, lung cancer patients with higher mRNA levels of MT-CO2 exhibited lower overall survival (OS) (Fig. 6j). Consistently, Kaplan–Meier database analyses showed that lung cancer patients with higher mRNA levels of GLS1 had lower overall survival (OS) (Fig. 6k).

**Fig. 6: Elevated MT-CO2 expression is linked to oncogenic Ras and is positively correlated with the expression of c-JUN and GLS1, and with poor clinical outcomes in human lung cancer.**

Berberine inhibits the MT-CO2-GLS1 axis to suppress oncogenic Ras-induced glutaminolysis and lung tumor growth

Our abovementioned results show that MT-CO2 plays a critical role in oncogenic Ras-induced glutaminolysis and tumor growth. We thus rationalized that cancer-associated elevation of MT-CO2 may be a putative therapeutic target for Ras-driven cancers. To search for potential small chemical MT-CO2 inhibitors, we examined the effects of several known mitochondrial respiratory chain inhibitors (metformin³⁶, oligomycin, rotenone, and berberine³⁷), ROS inducers (elesclomol³⁸, tolperisone³⁹, and piperlongumine⁴⁰) or glycolysis inhibitor (2-Deoxy-D-glucose, 2-DG) on MT-CO2 expression. As shown in Fig. 7a and Supplementary Fig. S11a, berberine, a clinically approved drug used to treat diarrhea and diabetes⁴¹, remarkably inhibited MT-CO2 expression in Ras-activated A549 and H1299 cells, concomitant with reduced GLS1 expression in a dose-dependent manner. Notably, berberine inhibited MT-CO2 mRNA expression without affecting mitochondrial mass (Supplementary Fig. S11b, c), suggesting that berberine inhibits MT-CO2 transcription. In keeping with our aforementioned observation that TFAM upregulates MT-CO2 gene transcription (Supplementary Fig. S9f, g), berberine significantly inhibited the expression of TFAM, accompanied by the suppression of MT-CO2, and GLS1 expression (Supplementary Fig. S11d). Importantly, ectopic expression of TFAM effectively rescued berberine-induced downregulation of MT-CO2 expression (Supplementary Fig. S11e, f). Together, these results indicate that berberine inhibits MT-CO2 transcription via suppression of TFAM.

**Fig. 7: Berberine inhibits the MT-CO2-GLS1 axis to suppress oncogenic Ras-induced glutaminolysis and tumor growth.**

We further investigated the effects of berberine on oncogenic Ras-induced glutaminolysis and cell proliferation. As shown in Supplementary Fig. S11g, h, berberine dramatically suppressed A549 or H1299 cell growth, which was markedly rescued by a supplement of α-KG, suggesting that berberine inhibits glutaminolysis. In supporting this notion, berberine-mediated inhibition of cell viability under the Glc-/Gln+ (DMEM containing 2 mM glutamine in the absence of glucose) condition was largely rescued by a supplement of α-KG (Fig. 7b). Furthermore, berberine strongly inhibited KRas^G12V-induced upregulated MT-CO2 and GLS1 expression, resulting in suppression of cell proliferation and cell viability, both of which could be markedly rescued by supplement of α-KG (Fig. 7c, d and Supplementary Fig. S11i). Notably, the oncogenic Ras-induced upregulation of MT-CO2 and GLS1 expression as well as lung tumor growth were significantly inhibited by berberine treatment in vivo (Fig. 7e–h and Supplementary Fig. S11j, k). Moreover, berberine significantly inhibited A549-derived xenograft tumor growth, concomitant with a marked reduction in MT-CO2 and GLS1 expression (Fig. 7i–l and Supplementary Fig. S11l, m). We further investigated whether berberine can also inhibit the growth of Ras mutant tumors in mice with a healthy immune system. As shown in Fig. 7m–o and Supplementary Fig. S11n, berberine not only significantly suppressed tumor growth derived from mouse LLC cells (Lewis lung carcinoma cell line harboring a KRAS^G12C allele) but also prolonged the overall survival of the C57BL/6 J mice. Importantly, a low dose of berberine treatment significantly suppressed the proliferation of RAS mutant tumor cells, while having little effect on the growth of tumor cells bearing wild-type RAS alleles (Supplementary Fig. S12).

Together, these results indicate that berberine is an effective MT-CO2 inhibitor and highlight that MT-CO2 is a valuable therapeutic target for oncogenic Ras-driven cancers.

Discussion

Glucose deprivation is often associated with tumor development, forcing tumor cells to seek alternative energy sources for survival and growth. This study demonstrates that MT-CO2 promotes metabolic reprogramming by switching the energy source from glucose to glutamine in response to glucose deprivation. We show that glucose deprivation activates Ras signaling while suppressing IGF2BP3, leading to the upregulation of MT-CO2 expression, which in turn facilitates glutaminolysis. We provide compelling evidence that MT-CO2 is not only essential for tumor cell survival upon glucose deprivation but is also indispensable for KRas-driven glutaminolysis and tumor growth. Mechanistically, we show that increased MT-CO2 upregulates flavin adenosine dinucleotide (FAD) levels to activate lysine-specific histone demethylase 1 (LSD1), which epigenetically upregulates JUN transcription via suppressing H3K9me2 levels, consequently resulting in increased GLS1 expression and glutaminolysis (Fig. 7p).

It is well-established that glucose is a major energy source for tumor cell survival and growth. However, it remains a central issue how tumor cells sense glucose shortage and adapt to alternative energy sources. Tumor cells can facilitate fatty acid oxidation or glycogen catabolism to maintain survival in response to glucose deprivation^15,17. These processes are regulated by highly conserved energy sensors and signaling hubs, such as AMP-activated protein kinase (AMPK) and mTORC1, both critical for cell survival under glucose deprivation⁴². Glucose starvation can activate AMPK to facilitate lipid droplet lipolysis and NADPH homeostasis, thereby promoting tumor cell survival^43,44. Inhibition of mTORC1 can suppress oxidative stress by maintaining NADPH homeostasis, aiding tumor cell survival under glucose deprivation⁴². A recent study reports that glucose shortage leads to AMPK-mediated phosphorylation of PDZD8, a scaffold protein binding to and activating GLS1 enzymatic activity to promote glutaminolysis ahead of utilization of fatty acid⁴⁵. Here, we demonstrate that glucose starvation enhances TFAM-mediated MT-CO2 expression by activating Ras signaling, which promotes GLS1 expression, glutaminolysis, and tumor cell survival, revealing that the Ras-MT-CO2-GLS1 axis as a pathway that triggers a shift in carbon source from glucose to glutamine. These observations point to the pleiotropic effects in the cellular response to glucose deprivation. It is possible that the Ras-MT-CO2 axis promotes GLS1 expression, whereas AMPK signaling enhances GLS1 activity upon glucose shortage, both promoting glutaminolysis in supporting tumor cell viability and growth.

These findings position MT-CO2 as a crucial regulator in orchestrating metabolic reprogramming to sustain tumor cell survival under glucose deprivation. Notably, MT-CO2 is not only essential in promoting glutaminolysis for cancer cell survival but also is indispensable for tumor cell growth in vitro and in vivo. Given that glutamine addiction is a key feature in metabolic reprogramming for tumor growth⁴⁶, it is plausible that MT-CO2 also plays a crucial role in glutamine addiction through the stimulation of glutaminolysis.

Our results show that glucose starvation activates Ras signaling to upregulate the transcription of MT-CO2. It is known that MT-CO2, along with 11 other MT-mRNAs, is encoded by the heavy (H) strand of the mitochondrial genome through a single H-strand promoter³² so that the H-strand promoter would drive the transcription to yield an equal amount of pre-mRNAs for each gene. Then, a critical issue is how glucose starvation specifically upregulates MT-CO2 expression. Our results show that glucose starvation reduces the expression of IGF2BP3, a well-known N6-methyladenosine (m6A) modification reader. Importantly, IGF2BP3 can specifically bind to and destabilize MT-CO2 mRNA, revealing that m6A modification can also applied to mitochondrial DNA-encoded mRNA. Therefore, this study reveals that glucose deprivation-mediated robust upregulation of MT-CO2 expression is achieved by Ras-mediated transcription upregulation and IGF2BP3-mediated modulation of MT-CO2 mRNA stability.

Abundant evidence demonstrates that glutaminolysis is crucial in promoting tumor growth⁴⁷. Genetic or pharmacological inhibition of GLS1 or SLC1A5 significantly suppresses tumor growth in vivo^48,49,50. Despite previous reports that the GLS1 inhibitor CB-839 failed to suppress KRas-driven pancreatic ductal adenocarcinoma growth⁵¹ and showed no advantage over standard immunotherapy in stage IV non-small cell lung cancer⁵², a phase 1b/2 trial found that CB-839 combined with azacytidine is a safe and effective treatment for myelodysplastic syndrome⁵³. In addition, a phase II trial showed that its combination with everolimus significantly improved progression-free survival in renal cell carcinoma patients⁵⁴. These developments highlight the therapeutical potential of targeting glutaminolysis in cancer treatment.

It is well-known that complexes I-IV of the mitochondrial respiratory chain are responsible for oxidative phosphorylation. Our findings show that inhibition of respiratory chain complexes II-IV by either genetic maneuvers or by small molecule inhibitors significantly suppresses c-JUN/GLS1 expression via inactivation of the FAD-LSD1-H3K9me2 axis. In contrast, inhibition of respiratory chain complex I has little effect on FAD levels and does not affect c-JUN/GLS1 expression. Together, these results uncover a critical function for respiratory chain complexes II-IV in regulating glutaminolysis and provide strong evidence for a mitochondrial retrograde signaling pathway through MT-CO2-mediated epigenetic regulation of GLS1.

Given glutaminolysis is essential in supporting various biological processes, it is plausible that defective MT-CO2 may be involved in human diseases. Indeed, It has been reported that supplementing glutamine can reduce the risk of stroke and coronary artery disease (CAD)⁵⁵, suggesting that glutaminolysis is involved in the development of these diseases. Notably, a clinical study reported that mitochondrial MT-CO2 gene 8231C>A (L216M) missense mutation is significantly associated with CAD (p = 0.007)⁵⁶, the molecular basis of which is not yet known. It might be possible that MT-CO2^L216M is defective in the regulation of GLS1 expression, leading to the inhibition of glutaminolysis.

Targeting Ras remains a vast challenge for cancer therapy, despite the recent clinical approval of AMG-510⁵⁷, a specific small molecular inhibitor of KRas^G12C, which is found in approximately 13% of lung adenocarcinomas and 1–3% of other solid tumors⁵⁷. A recent study has reported that tumor cells rapidly acquire resistance to small molecule inhibitors of KRas^G12C⁵⁸. Therefore, it is urgent to develop therapeutic strategies for Ras-activated cancers. Here, we show that either knockdown of MT-CO2 or berberine significantly suppresses Ras-induced tumor growth in vivo, highlighting MT-CO2 as a putative therapeutic target for treating Ras-activated cancers.

Methods

Ethics statement and mouse models

All animal care and animal experiments in this study were performed in accordance with China’s National Legislation and the institutional ethical guidelines and were approved by the institutional review board of Sichuan University (Approval Number: SCU23030314). Mice were fed a standard chow diet and were maintained in individual cages at a room temperature of 22 ± 2 °C on a 12: 12 light–dark cycle (lights on at 09:00 h).

For in vivo tumor growth, six-week-old female BALB/c nude mice (BALB/cNj-Foxn1nu/Gpt) were purchased from GemPharmatech (Chengdu, China). Indicated cells were subcutaneously inoculated into the right scruff of each nude mouse. Mice were monitored for tumor size and sacrificed on the indicated day. Tumor weight, volume, and photos were taken. The xenograft tumor samples were subjected to IHC staining analyses. Tumor size was measured with a caliper and tumor volume was calculated by width² × length × 1/2. The maximal tumor size permitted by the IACUC of Sichuan University is 20 mm at the largest diameter in mice. The maximal tumor size in our animal experiments did not exceed the permitted maximal tumor size.

For transgenic mice, eight-week-old LSL-Kras^G12D knock-in mice (C57BL/6 background, provided by Dr. Chong Chen of Sichuan University) were infected with 2 × 10⁷ PFU adenovirus-based Cre by nasal drip. 28 days post-infection, mice were infected with 2 × 10⁷ PFU lentivirus-based shGFP or shMT-CO2 by nasal drip. On day 56 after the lentivirus infection, lungs were dissected, fixed in 4% polyformaldehyde, embedded in paraffin, and sectioned onto microscope slides for hematoxylin and eosin (H&E) staining before histological analysis. The number of nodules in the lungs per mouse was counted.

Cell culture and drug treatment

H1975 (CRL-5908), H292 (CRL-1848), and LLC (Lewis lung carcinoma, CRL-1642) were obtained from ATCC (Manassas, VA, USA). H1299 (CL-0165) were obtained from Procell Life Science&Technology (Wuhan, China). HEK-293FT (R70007) was obtained from Thermo Fisher Scientific (Waltham, MA, USA). A549 (BNCC337696) was obtained from the BeNa Culture Collection (Beijing, China). MEF (mouse embryonic fibroblast) cells were isolated from day 13.5 mouse embryos. All cell lines used in this study were routinely tested to be negative for mycoplasma contamination and were kept at low passages to maintain their identity and were authenticated by morphology check and growth curve analysis.

A549, H1299, HEK 293FT, and MEF cells were cultured in DMEM medium (GIBCO, Rockville, MD, USA) supplemented with 10% fetal bovine serum (FBS; Hyclone, Logan, UT, USA). Human lung mucoepidermoid pulmonary carcinoma H292 and human NSCLC H1975 cells were cultured in RPMI-1640 medium (GIBCO) supplemented with 10% FBS (Hyclone). All cells were grown in a medium supplemented with 2 mM glutamine, 100 units/mL penicillin (GIBCO), and 100 μg/mL streptomycin (GIBCO) with or without 4.5 mg/mL glucose as indicated. Cells were maintained in a humidified 37 °C incubator under a 5% CO₂ atmosphere. Cells at 75–85% confluence were treated with an indicated chemical compound(s). PD98059 (S1177), oligomycin (S1478), rotenone (S2348), elesclomol (S1052), tolperisone (S4200), 2-Deoxyglucose (2-DG, S4701), ATP (S1985), and nicotinamide mononucleotide (NMN, S5259) were purchased from Selleck (Houston, TX, USA). Berberine (HY-18258), carboxin (HY-B2064), antimycin A (HY-107406), flavin adenine dinucleotide (FAD+, HY-B1654A), riboflavin (HY-B0456), CB-839 (HY-12248), CP-91149 (HY-13525), etomoxir (HY-50202), and GSK-LSD1 (HY-100546A) were purchased from MCE (Monmouth Junction, NJ, USA). Metformin (PHR1084), dimethyl α-ketoglutarate (DM-αKG, 349631), and piperlongumine (SML0221) were purchased from Sigma (St. Louis, MO, USA). Sodium acetate (NaAc, ST351) was purchased from Beyotime (Beijing, China).

Plasmid transfection, lentiviral infection, and RNA interference

Cells at 70–80% confluence were transfected using Lipofectamine 2000 transfection reagent (Invitrogen, USA). Expression plasmids were used in this study including KRas^G12V, HRas^G12V, TFAM, GLS1, c-JUN, or IGF2BP3. Recombinant lentiviruses were amplified by transfecting HEK-293FT cells with psPAX2 and pMD2.G packaging plasmids and lentiviral expression plasmid using Lipofectamine 2000. Viruses were collected at 60 h after transfection. Cells at 50% confluence were infected with a recombinant lentivirus in the presence of 10 μg/mL polybrene, followed by 12 h of incubation at 37 °C with 5% CO₂. Lentiviral-based shRNAs specific for green fluorescent protein (GFP), human MT-CO2, TFAM, MT-ND1, MT-CYB, MT-ATP6, IGF2BP3, or mouse MT-CO2 were constructed into a pLKO.1-puromycin lentiviral vector. The shRNA sequences were listed in Supplementary Table 3.

Western blot, immunohistochemistry, and immunofluorescence staining analyses

For Western blot, cells were collected, washed with cold PBS, and resuspended in EBC₂₅₀ lysis buffer (250 mM NaCl, 50 mM Tris pH 8.0, 0.5% Nonidet P-40, 50 mM NaF, 1 mM phenylmethylsulfonyl fluoride, 2 μg/mL aprotinin, and 2 μg/mL leupeptin). Equal amounts of total protein were loaded, separated by SDS-PAGE, transferred to PVDF membranes (Millipore, Darmstadt, Germany), and hybridized to an appropriate primary antibody and horseradish peroxidase (HRP)-conjugated secondary antibody for subsequent detection by enhanced chemiluminescence (Bio-Rad). Western blot images were analyzed using Image Lab Software 5.0. Antibodies for Ras (3965, 1:1000), pERK (Thr202/Tyr204, 9101, 1:1000), ERK (9102, 1:1000), EZH2 (5246, 1:1000), c-JUN (9165, 1:1000), HA-Tag (5017, 1:1000), H3K27me3 (9733, 1:1000), H3K27ac (8173, 1:1000), pAMPK (Thr172, 2535, 1:1000), and AMPK (2532, 1:1000) were purchased from Cell Signaling Technology (CST, Danvers, MA, USA). Antibodies for GLS1 (CY5719, 1:1000), MT-CO2 (human, CY5717, 1:1000), MT-ND1 (CY8176, 1:1000), H3K4me2 (CY5123, 1:1000), H3K9me2 (CY5761, 1:1000), LSD1 (CY5517, 1:1000), p21 (CY5543, 1:1000), and GAPDH (AB0037, 1:5000) were purchased from Abways (Shanghai, China). Antibodies for MT-CO2 (mouse, ab198286, 1:1000) and TFAM (ab131607, 1:2000) were purchased from Abcam (Cambridge, UK). Antibody for MT-CO1 (DF8920, 1:1000) was purchased from Affinity (Jiangsu, China). Antibodies for MT-ATP6 (A8193, 1:1000), Citrate synthase (A23371, 1:1000), and SLC1A5 (A6981, 1:1000) were purchased from ABclonal (Wuhan, China). Antibody for p53 (SC-126, 1:300) was purchased from Santa Cruz (CA, USA). Antibody for IGF2BP3 (14642-1-AP, 1:1000) was purchased from Proteintech (Wuhan, China).

For immunohistochemistry (IHC) analyses, human lung adenocarcinoma tissue microarrays (TMA) (HLugA060PG02) were obtained from Outdo Biotech (Shanghai, China). Paraffin-embedded tumors or lung samples were sliced into 5 µm thickness. Tissue sections were rehydrated through a decreasing ethanol gradient, and treated by boiling in citrate buffer (pH 6.0) or Tris-EDTA (pH 9.0) for antigen retrieval. Endogenous peroxidases were blocked using 0.3% H₂O₂. After blocking with 5% BSA, the sections were incubated with primary antibody and followed by HRP-conjugated secondary antibody. The sections were subsequently stained with a DAB Detection Kit (ZLI-9018, ZSGB-BIO, China). Antibodies used in IHC for GLS1 (CY5719, 1:200) and MT-CO2 (human, CY5717, 1:200) were purchased from Abways (Shanghai, China). Antibody for human Ki67 (9449, 1:800) was purchased from CST (Danvers, MA, USA). The antibody for mouse Ki67 (ab15580, 1:200) was purchased from Abcam (Cambridge, UK). Slides were scanned through NanoZoomer (Hamamatsu, Japan). Scanned images were then subjected to integrated optical density (IOD) measurements using Image-Pro Plus 6.0 to calculate the average optical density (AOD) using the formula: AOD = IOD/Area⁵⁹.

For Immunofluorescence staining, cells grown on coverslips were fixed with 4% polyformaldehyde in PBS, permeabilized with 0.1% Triton X-100 in PBS, blocked with 4% bovine serum albumin in PBS, hybridized to an appropriate primary antibody anti-m6A (m6A: 1:500, 68055-1-Ig, Proteintech; MT-CO2: 1: 200, CY5717, Abways; IGF2BP3: 1:200, 14642-1-AP, Proteintech; Flag-Tag: 1:200, 8146, CST; Hsp60: 1:200, 66041-1-Ig, Proteintech), followed by incubation with a second antibody (Goat anti-Mouse Alexa Fluor 488, A-11029 or Goat anti- Rabbit Alexa Fluor 514, A-31558, ThermoFisher). The cells were counterstained with ProLong® Gold Antifade Reagent with DAPI (#8961, CST) prior to visualization and photographed using a Leica TCS SP5II confocal laser scanning microscope. LAS X (V3.3.0) was used to analyze fluorescent images. To determine the IGF2BP3 or m6A co-localization with Hsp60 or MT-CO2, the free software Image J. Fiji coupled with the Coloc 2 plugin and Pearson’s correlation coefficient were used to calculate double fluorescence correlation coefficients⁶⁰, and co-localized fluorescence quantifications were presented by scatter diagrams.

Quantitative PCR (qPCR) and measurement of the ratio of mitochondrial DNA/nuclear DNA

Total RNA was extracted from cells using an RNA extraction kit (QIAGEN, Germany) followed by reverse transcription using ReverTra Ace qPCR RT Master Mix (FSQ-201, TOYOBO, Japan). qPCR was carried out for MT-CO2, TFAM, GLS1, GLUD1, c-JUN, and GAPDH (The primer sequences used in the reactions were listed in Supplementary Table 3). The qPCR reactions were performed in CFX-960 Real-time PCR System (Bio-Rad) and using Bio-Rad SoFast Eva-Green Supermix Kit (Bio-Rad) according to the manufacturer’s instructions. Relative quantitation values were calculated using the ΔΔCt method. For measurement of the ratio of mitochondrial DNA/nuclear DNA (mtDNA/nDNA), nuclear DNA and mitochondrial DNA were extracted from cells using the TIANamp Genomic DNA Kit (TIANGEN BIOTECH, China). qPCR was carried out for CYTB and H2B (The primer sequences used in the reactions were listed in Supplementary Table 3). The ratio of mtDNA/nDNA = CYTB copies/H2B copies.

Chromatin immunoprecipitation (ChIP) assay

ChIP assays were performed using ChIP-IT Kit (Active Motif, USA) using antibodies specific for H3K9me2 (CY5761, Abways), H3K4me2 (CY5123, Abways) or normal rabbit IgG (2729, Cell Signaling Technology). To examine the strength of H3K9me2 or H3K4me2 for binding to JUN gene promoter elements, ChIP samples were subjected to qPCR using primers as listed in Supplementary Table 3. The value of each ChIP sample was normalized to its corresponding input.

RNA immunoprecipitation (RIP) assay

The RIP assay was carried out using a BeyoRIP™ RIP Assay Kit (P1801S, Beyotime, China), according to the manufacturer’s instructions. Briefly, 1 × 10⁷ cells were lysed in complete RIP lysis buffer containing protein A/G Agarose conjugated with anti-m6A (68055-1-Ig, Proteintech) or anti-IGF2BP3 (14642-1-AP, Proteintech) at 4 °C for 4 h. RNA was then purified using TRIzol™ Reagent (15596026, Invitrogen) and used for qPCR analysis.

RNA sequencing and whole exon sequencing

For RNA sequencing analyses, total RNA was extracted using TRIzol reagent (Thermo Fisher Scientific, USA) according to the manufacturer’s protocol. RNA purity and quantification were evaluated using the NanoDrop 2000 spectrophotometer (Thermo Scientific, USA). RNA integrity was assessed using the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). Then the libraries were constructed using TruSeq Stranded mRNA LT Sample Prep Kit (Illumina, San Diego, CA, USA) according to the manufacturer’s instructions. The libraries were sequenced on an Illumina HiSeq X Ten platform and 150 bp paired-end reads were generated and analyzed (OE Biotech Co., Ltd, Shanghai, China). Differential expression analysis was performed using the DESeq (2012) R package. P value < 0.05 and foldchange >2 or foldchange <0.5 was set as the threshold for significantly differential expression. KEGG pathway enrichment analysis of differentially expressed genes was performed respectively using R based on the hypergeometric distribution.

Whole exon sequencing and data analysis were conducted by Seqhealth Technology Co., LTD (Wuhan, China). DNA was extracted from cells using a commercial kit (TIANGEN Biotech, Beijing, Chin). Genomic DNA integrity was checked by 0.5% agarose gel electrophoresis. Qualified DNA was finally quantified by Qubit3.0 with QubitTM RNA Broad Range Assay kit (Life Technologies, Q10210). Libraries of DNA were prepared using 50~200 ng DNA, captured by xGenTM Exome Hyb Panel v2 (Integrated DNA Technologies) following the manufacturer’s instructions. PCR products corresponding to 200–500 bps were enriched, quantified, and finally sequenced on a Novaseq 6000 sequencer (Illumina) with PE150 mode. The raw FastQ files were first filtered by fastp and reads were mapped to the hg19 reference genome. Both single nucleotide polymorphism (SNP) and insertion/deletion (INDEL) were called, filtered, and annotated using GATK HaplotypeCaller, GATK VariantFiltration, and ANNOVAR, respectively. The sequencing depth of the chromosomes and the distribution of the variants were visualized using Circos.

Detection of targeted metabolites

1 × 10⁷ cells were frozen in liquid nitrogen for 10 min. Extraction and analysis of metabolites were carried out by Metware Biotechnology Co. Ltd. (Wuhan, China) based on the AB Sciex QTRAP 6500 LC-MS/MS platform. Briefly, the sample was thawed on ice, then mixed with 500 µL of 80% methanol/water (precooled at −20 °C), and then frozen in liquid nitrogen for 5 min, removed on ice for 5 min, after that, the sample was vortexed for 2 min. The previous step was repeated 3 times. The sample was centrifuged at 13,000 × g for 10 min at 4 °C. Take 300 μL of supernatant into a fresh centrifuge tube and place the supernatant in a −20 °C refrigerator for 30 min. Then the supernatant was centrifuged at 13,000 × g for 10 min at 4 °C. After centrifugation, 200 μL of supernatant were analyzed using an LC-ESI-MS/MS system (UPLC, ExionLC AD, https://sciex.com.cn/; MS, QTRAP® 6500+ System, https://sciex.com/). All metabolites were detected by MetWare (http://www.metware.cn/) based on the AB Sciex QTRAP 6500 LC-MS/MS platform. Significantly regulated metabolites between groups were determined by VIP and absolute Log2FC (foldchange).

¹³C₅-glutamine metabolic flux analysis

Cells were grown in DMEM containing 2 mM ¹³C₅-glutamine for 8 h. After incubation, the cells were ground in liquid nitrogen and re-suspended in 1 mL of cold (−40 °C) 50% aqueous methanol. The suspension was placed on dry ice for 30 min, then thawed on ice. Subsequently, 0.4 mL of chloroform was added, and the mixture was vortexed for 30 s. The samples were centrifuged at 14,000 × g for 15 min at 4 °C, and the resulting supernatant was transferred to a fresh 1.5 mL tube and stored at −80 °C until further analysis. Water-soluble metabolites were analyzed using a Q Exactive PLUS hybrid quadrupole-orbitrap mass spectrometer (Thermo Scientific) coupled to hydrophilic interaction chromatography (HILIC). Separation was performed on an XBridge BEH Amide column (150 mm × 2.1 mm, 2.5 μM particle size, Waters, Milford, MA) using a solvent gradient of A (95% H₂O:5% acetonitrile with 20 mM ammonium acetate and 20 mM ammonium hydroxide, pH 9.4) and B (100% acetonitrile). The flow rate was maintained at 150 μL/min, with a 5 μL injection volume and a column temperature of 25 °C. Mass spectrometry was conducted in negative ion mode with a resolution of 140,000 at m/z 200, an automatic gain control (AGC) target of 1 × 10⁶, and a scan range of m/z 75–1000. Data from isotope labeling experiments were processed using El-MAVEN, applying natural abundance correction.

Mitochondrial electron transport chain complex activity assays

Complex II-IV activities were determined by Mitochondrial Complex Activity Assay Kit (complex II: BC3230, Solarbio; complex III: BC3240, Solarbio; complex IV: BC0945, Solarbio) followed the manufacturer’s instructions using Visible spectrophotometer (GENESYS 30, Thermo Scientific).

Oxygen consumption rate assay

The oxygen consumption rate (OCR) was measured using the Seahorse XFp Extracellular Flux Analyzer (Seahorse Bioscience). Experiments were performed according to the manufacturer’s instructions. The Seahorse XF Mito stress test kit (Agilent Technologies) was used to test OCR by sequential injection of 50 μM oligomycin, 10 μM FCCP, and 2.5 μM rotenone/antimycin A.

Measurement of cellular FAD levels

The cellular FAD levels were measured using the Flavin Adenine Dinucleotide (FAD) Assay Kit (Abcam, ab204710). Cells (1 × 10⁶) were harvested and resuspended in 400 µL ice-cold FAD Assay Buffer, then homogenized by pipetting. The homogenate was centrifuged at 10,000 × g for 5 min at 4 °C to remove insoluble material. The supernatant was collected, and ice-cold 4 M Perchloric Acid (PCA) was added to a final concentration of 1 M, mixed briefly, and incubated on ice for 5 min. After centrifugation at 13,000 × g for 2 min at 4 °C, the supernatant was transferred to a fresh tube, and the volume was measured. Ice-cold 2 M Potassium Hydroxide (KOH) was added to equal 34% of the supernatant volume, followed by vortexing and final centrifugation at 13,000 × g for 15 min at 4 °C. The supernatant was then collected for FAD measurement.

Cell proliferation and cell viability analyses

Cell proliferation assay (MTS) was performed using CellTiter 96_ kit (Promega, USA) as described in the manufacturer’s instructions. The trypan blue exclusion assay (C0011, Beyotime) was performed according to the manufacturer’s instructions.

Collection of patient samples

To evaluate the clinical relevance of MT-CO2, primary human lung tumor specimens were collected from 23 male lung cancer patients (aged 48–72 years) who underwent surgical resection at Sichuan Provincial People’s Hospital (Ethics Approval No. 2024-79-1). Samples were collected according to the institutional regulation of the Hospital Clinic Ethical Committee and were examined by pathologists. There were 11 lung tumor specimens harboring wild-type KRAS alleles (KRAS^WT) and 11 lung tumor specimens harboring KRAS mutant alleles (KRAS^MT). Tumor samples were subjected to qPCR analyses for MT-CO2, c-JUN, or GLS1 mRNA levels. No sex- or gender-based analyses were conducted in this study. All procedures were performed with signed informed consent from the patients, and the donation of lung tumor tissues was entirely voluntary and uncompensated.

Statistics and reproducibility

GraphPad Prism 8.0 (GraphPad Software Inc., USA) was used for data recording, collection, processing, and calculation. Data from three independent experiments in vitro were presented as mean ± standard deviation (SD), and data from animal experiments were presented as mean ± standard error of the mean (SEM). Unpaired two-tailed Student’s t test was used for comparing two groups of data. One/two-way ANOVA with Tukey’s test or Bonferroni’s test was used to compare multiple groups of data. P values ≤ 0.05 were considered significant. No statistical method was used to predetermine the sample size. No data were excluded from the analyses. Mice were randomly assigned to different experimental groups. Investigators measuring the mice were blinded to the treatment groups. Other experiments, such as the western blot assay, were conducted in a non-blinded manner.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The raw RNA sequencing and whole-exon sequencing data have been deposited in the NCBI Sequence Read Archive (SRA) under accession number PRJNA1002470. The mass spectrometry data for targeted metabolites and ¹³C₅-glutamine metabolic flux generated in this study are available in the Open Archive for Miscellaneous Data (OMIX) at the China National Center for Bioinformation / Beijing Institute of Genomics, Chinese Academy of Sciences, under accession codes OMIX008195 and OMIX008035. All data generated or analyzed during this study are included in this article and its Supplementary Information files. The uncropped gel or blot figures and original data underlying Figs. 1–7 and Supplementary Figs. S1–S12 are provided as a Source Data file. Source data are provided with this paper.

References

Hanahan, D. & Weinberg, R. A. Hallmarks of cancer: the next generation. Cell 144, 646–674 (2011).
Article CAS PubMed MATH Google Scholar
Lin, S. Y., Zhang, C. S. & Lin, S. C. Carbohydrates: Not All that Bad? Cell Metab. 28, 671–672 (2018).
Article CAS PubMed MATH Google Scholar
Ferreira, L. M., Hebrant, A. & Dumont, J. E. Metabolic reprogramming of the tumor. Oncogene 31, 3999–4011 (2012).
Article CAS PubMed MATH Google Scholar
Hirayama, A. et al. Quantitative metabolome profiling of colon and stomach cancer microenvironment by capillary electrophoresis time-of-flight mass spectrometry. Cancer Res. 69, 4918–4925 (2009).
Article CAS PubMed MATH Google Scholar
Birsoy, K. et al. Metabolic determinants of cancer cell sensitivity to glucose limitation and biguanides. Nature 508, 108–112 (2014).
Article ADS CAS PubMed PubMed Central Google Scholar
Jin, H. et al. A powerful drug combination strategy targeting glutamine addiction for the treatment of human liver cancer. Elife 9, https://doi.org/10.7554/eLife.56749 (2020).
Saqcena, M. et al. Blocking anaplerotic entry of glutamine into the TCA cycle sensitizes K-Ras mutant cancer cells to cytotoxic drugs. Oncogene 34, 2672–2680 (2015).
Article CAS PubMed MATH Google Scholar
Bernfeld, E. & Foster, D. A. Glutamine as an Essential Amino Acid for KRas-Driven Cancer Cells. Trends Endocrinol. Metab. 30, 357–368 (2019).
Article CAS PubMed Google Scholar
Choi, B. H. & Coloff, J. L. The Diverse Functions of Non-Essential Amino Acids in Cancer. Cancers 11, https://doi.org/10.3390/cancers11050675 (2019).
Le, A. et al. Glucose-independent glutamine metabolism via TCA cycling for proliferation and survival in B cells. Cell Metab. 15, 110–121 (2012).
Article CAS PubMed PubMed Central MATH Google Scholar
Vega, R. B., Horton, J. L. & Kelly, D. P. Maintaining ancient organelles: mitochondrial biogenesis and maturation. Circulation Res. 116, 1820–1834 (2015).
Article CAS PubMed Google Scholar
Copeland, W. C. & Longley, M. J. Mitochondrial genome maintenance in health and disease. DNA Repair 19, 190–198 (2014).
Article CAS PubMed PubMed Central MATH Google Scholar
Haga, N. et al. Mitochondria regulate the unfolded protein response leading to cancer cell survival under glucose deprivation conditions. Cancer Sci. 101, 1125–1132 (2010).
Article CAS PubMed PubMed Central MATH Google Scholar
Gao, K. et al. Active RNA interference in mitochondria. Cell Res., https://doi.org/10.1038/s41422-020-00394-5 (2020).
Li, X. X. et al. Nuclear Receptor Nur77 Facilitates Melanoma Cell Survival under Metabolic Stress by Protecting Fatty Acid Oxidation. Mol. Cell 69, 480–492.e487 (2018).
Article CAS PubMed MATH Google Scholar
Wang, Y. et al. Coordinative metabolism of glutamine carbon and nitrogen in proliferating cancer cells under hypoxia. Nat. Commun. 10, 201 (2019).
Article ADS PubMed PubMed Central MATH Google Scholar
Favaro, E. et al. Glucose utilization via glycogen phosphorylase sustains proliferation and prevents premature senescence in cancer cells. Cell Metab. 16, 751–764 (2012).
Article CAS PubMed Google Scholar
Schnier, J. B., Nishi, K., Monks, A., Gorin, F. A. & Bradbury, E. M. Inhibition of glycogen phosphorylase (GP) by CP-91,149 induces growth inhibition correlating with brain GP expression. Biochem. Biophys. Res. Commun. 309, 126–134 (2003).
Article CAS PubMed Google Scholar
Lin, T. et al. Disturbance of Fatty Acid Metabolism Promoted Vascular Endothelial Cell Senescence via Acetyl-CoA-Induced Protein Acetylation Modification. Oxid. Med. Cell Longev. 2022, 1198607 (2022).
Article PubMed PubMed Central Google Scholar
Lukey, M. J., Greene, K. S., Erickson, J. W., Wilson, K. F. & Cerione, R. A. The oncogenic transcription factor c-Jun regulates glutaminase expression and sensitizes cells to glutaminase-targeted therapy. Nat. Commun. 7, ARTN 11321, https://doi.org/10.1038/ncomms11321 (2016).
Mottis, A., Herzig, S. & Auwerx, J. Mitocellular communication: Shaping health and disease. Science 366, 827–832 (2019).
Article ADS CAS PubMed Google Scholar
Gnaiger, E. Complex II ambiguities-FADH(2) in the electron transfer system. J. Biol. Chem. 300, 105470 (2024).
Article CAS PubMed MATH Google Scholar
Hou, Y. et al. Co-production of 7-chloro-tryptophan and indole pyruvic acid based on an efficient FAD/FADH(2) regeneration system. Appl. Microbiol. Biotechnol. 107, 4873–4885 (2023).
Article CAS PubMed MATH Google Scholar
Hargreaves, I. P. et al. Inhibition of mitochondrial complex IV leads to secondary loss complex II-III activity: implications for the pathogenesis and treatment of mitochondrial encephalomyopathies. Mitochondrion 7, 284–287 (2007).
Article CAS PubMed MATH Google Scholar
LaMoia, T. E. et al. Metformin, phenformin, and galegine inhibit complex IV activity and reduce glycerol-derived gluconeogenesis. Proc. Natl Acad. Sci. USA 119, e2122287119 (2022).
Article CAS PubMed PubMed Central Google Scholar
Hirano, K. & Namihira, M. FAD influx enhances neuronal differentiation of human neural stem cells by facilitating nuclear localization of LSD1. FEBS Open Bio 7, 1932–1942 (2017).
Article CAS PubMed PubMed Central MATH Google Scholar
Hino, S. et al. FAD-dependent lysine-specific demethylase-1 regulates cellular energy expenditure. Nat. Commun. 3, 758 (2012).
Article ADS PubMed MATH Google Scholar
Fukuda, K. et al. Regulation of mammalian 3D genome organization and histone H3K9 dimethylation by H3K9 methyltransferases. Commun. Biol. 4, 571 (2021).
Article CAS PubMed PubMed Central MATH Google Scholar
Mukhopadhyay, S., Vander Heiden, M. G. & McCormick, F. The Metabolic Landscape of RAS-Driven Cancers from biology to therapy. Nat. Cancer 2, 271–283 (2021).
Article CAS PubMed PubMed Central MATH Google Scholar
Ekstrand, M. I. et al. Mitochondrial transcription factor A regulates mtDNA copy number in mammals. Hum. Mol. Genet 13, 935–944 (2004).
Article CAS PubMed MATH Google Scholar
Kozhukhar, N. & Alexeyev, M. F. Limited predictive value of TFAM in mitochondrial biogenesis. Mitochondrion 49, 156–165 (2019).
Article CAS PubMed PubMed Central MATH Google Scholar
Jedynak-Slyvka, M., Jabczynska, A. & Szczesny, R. J. Human Mitochondrial RNA Processing and Modifications: Overview. Int. J. Mol. Sci. 22, https://doi.org/10.3390/ijms22157999 (2021).
Psarra, A. M. & Sekeris, C. E. Glucocorticoids induce mitochondrial gene transcription in HepG2 cells: role of the mitochondrial glucocorticoid receptor. Biochim. Biophys. Acta 1813, 1814–1821 (2011).
Article CAS PubMed Google Scholar
Bhattacharya, D., Shah, V., Oresajo, O. & Scime, A. p107 mediated mitochondrial function controls muscle stem cell proliferative fates. Nat. Commun. 12, 5977 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Jiang, X. et al. The role of m6A modification in the biological functions and diseases. Signal Transduct. Target Ther. 6, 74 (2021).
Article CAS PubMed PubMed Central MATH Google Scholar
Wheaton, W. W. et al. Metformin inhibits mitochondrial complex I of cancer cells to reduce tumorigenesis. eLife 3, e02242 (2014).
Article PubMed PubMed Central MATH Google Scholar
Turner, N. et al. Berberine and its more biologically available derivative, dihydroberberine, inhibit mitochondrial respiratory complex I: a mechanism for the action of berberine to activate AMP-activated protein kinase and improve insulin action. Diabetes 57, 1414–1418 (2008).
Article CAS PubMed Google Scholar
Kirshner, J. R. et al. Elesclomol induces cancer cell apoptosis through oxidative stress. Mol. Cancer Therapeutics 7, 2319–2327 (2008).
Article CAS MATH Google Scholar
Shaw, A. T. et al. Selective killing of K-ras mutant cancer cells by small molecule inducers of oxidative stress. Proc. Natl Acad. Sci. USA 108, 8773–8778 (2011).
Article ADS CAS PubMed PubMed Central MATH Google Scholar
Karki, K., Hedrick, E., Kasiappan, R., Jin, U. H. & Safe, S. Piperlongumine Induces Reactive Oxygen Species (ROS)-Dependent Downregulation of Specificity Protein Transcription Factors. Cancer Prev. Res. 10, 467–477 (2017).
Article CAS Google Scholar
Zhang, Y. et al. Treatment of type 2 diabetes and dyslipidemia with the natural plant alkaloid berberine. J. Clin. Endocrinol. Metab. 93, 2559–2565 (2008).
Article CAS PubMed MATH Google Scholar
Levy, T. et al. mTORC1 regulates cell survival under glucose starvation through 4EBP1/2-mediated translational reprogramming of fatty acid metabolism. Nat. Commun. 15, 4083 (2024).
Article ADS CAS PubMed PubMed Central MATH Google Scholar
Jeon, S. M., Chandel, N. S. & Hay, N. AMPK regulates NADPH homeostasis to promote tumour cell survival during energy stress. Nature 485, 661–665 (2012).
Article ADS CAS PubMed PubMed Central Google Scholar
Liu, R. et al. Choline kinase alpha 2 acts as a protein kinase to promote lipolysis of lipid droplets. Mol. Cell 81, 2722–2735.e2729 (2021).
Article CAS PubMed MATH Google Scholar
Li, M. et al. AMPK targets PDZD8 to trigger carbon source shift from glucose to glutamine. Cell Res., https://doi.org/10.1038/s41422-024-00985-6 (2024).
Li, S. et al. CENPA promotes glutamine metabolism and tumor progression by upregulating SLC38A1 in endometrial cancer. Cell. Signal. 117, 111110 (2024).
Article CAS PubMed MATH Google Scholar
Altman, B. J., Stine, Z. E. & Dang, C. V. From Krebs to clinic: glutamine metabolism to cancer therapy. Nat. Rev. Cancer 16, 749 (2016).
Article CAS PubMed Google Scholar
Xiang, Y. et al. Targeted inhibition of tumor-specific glutaminase diminishes cell-autonomous tumorigenesis. J. Clin. Investig. 125, 2293–2306 (2015).
Article PubMed PubMed Central MATH Google Scholar
Wu, S. et al. Targeting glutamine dependence through GLS1 inhibition suppresses ARID1A-inactivated clear cell ovarian carcinoma. Nat. Cancer 2, 189–200 (2021).
Article CAS PubMed PubMed Central Google Scholar
Zhang, Z. et al. ASCT2 (SLC1A5)-dependent glutamine uptake is involved in the progression of head and neck squamous cell carcinoma. Br. J. Cancer 122, 82–93 (2020).
Article CAS PubMed Google Scholar
Biancur, D. E. et al. Compensatory metabolic networks in pancreatic cancers upon perturbation of glutamine metabolism. Nat. Commun. 8, 15965 (2017).
Article ADS CAS PubMed PubMed Central Google Scholar
Vasan, K. & Chandel, N. S. Molecular and cellular mechanisms underlying the failure of mitochondrial metabolism drugs in cancer clinical trials. J. Clin. Invest. 134, https://doi.org/10.1172/jci176736 (2024).
DiNardo, C. D. et al. Glutaminase inhibition in combination with azacytidine in myelodysplastic syndromes: a phase 1b/2 clinical trial and correlative analyses. Nat. Cancer, https://doi.org/10.1038/s43018-024-00811-3 (2024).
Lee, C. H. et al. Telaglenastat plus Everolimus in Advanced Renal Cell Carcinoma: A Randomized, Double-Blinded, Placebo-Controlled, Phase II ENTRATA Trial. Clin. Cancer Res. 28, 3248–3255 (2022).
Article CAS PubMed PubMed Central MATH Google Scholar
Jauhiainen, R. et al. The association of nine amino acids with cardiovascular events in Finnish men in a 12-year follow-up study. J. Clin. Endocrinol. Metabol., https://doi.org/10.1210/clinem/dgab562 (2021).
Heidari, M. M. et al. Novel Point Mutations in Mitochondrial MT-CO2 Gene May Be Risk Factors for Coronary Artery Disease. Appl. Biochem. Biotechnol. 191, 1326–1339 (2020).
Article CAS PubMed MATH Google Scholar
Canon, J. et al. The clinical KRAS(G12C) inhibitor AMG 510 drives anti-tumour immunity. Nature 575, 217–223 (2019).
Article ADS CAS PubMed MATH Google Scholar
Xue, J. Y. et al. Rapid non-uniform adaptation to conformation-specific KRAS(G12C) inhibition. Nature 577, 421–425 (2020).
Article ADS CAS PubMed PubMed Central MATH Google Scholar
Li, J. et al. USP5-Beclin 1 axis overrides p53-dependent senescence and drives Kras-induced tumorigenicity. Nat. Commun. 13, 7799 (2022).
Article ADS CAS PubMed PubMed Central MATH Google Scholar
Yan, G. et al. Genome-wide CRISPR screens identify ILF3 as a mediator of mTORC1-dependent amino acid sensing. Nat. Cell Biol. 25, 754–764 (2023).
Article CAS PubMed MATH Google Scholar

Download references

Acknowledgements

We are grateful to Dr. Chong Chen (State Key Laboratory of Biotherapy, West China Hospital, Sichuan University, China) for the LSL-KRAS^G12D knock-in mice. We thank members of the Z.-X.J.X. laboratory for stimulating discussions during the study. This work was supported in part by the National Natural Science Foundation of China (82073248) to Y.Y. and the National Key R&D Program of China (2022YFA1103700) to Z.-X.J.X.

Author information

Authors and Affiliations

Center of Growth, Metabolism and Aging, Key Laboratory of Bio-Resource and Eco-Environment, Ministry of Education, College of Life Sciences, Sichuan University, Chengdu, China
Yong Yi, Guoqiang Wang, Wenhua Zhang, Shuhan Yu, Junjie Fei, Tingting An, Jianqiao Yi, Jian Yang, Mengmeng Niu, Yang Wang & Zhi-Xiong Jim Xiao
Department of Oncology & Cancer Institute, Department of Laboratory Medicine and Sichuan Provincial Key Laboratory for Human Disease Gene Study, Sichuan Academy of Medical Sciences, Sichuan Provincial People’s Hospital, University of Electronic Science and Technology of China, Chengdu, China
Shuhan Yu, Ting Huang, Chuan Xu & Zhi-Xiong Jim Xiao
School of Biosciences and Technology, Chengdu Medical College, Chengdu, China
Fengtian Li
State Key Laboratory of Biotherapy, West China Hospital, Sichuan University, Chengdu, China
Zhi-Xiong Jim Xiao

Authors

Yong Yi
View author publications
Search author on:PubMed Google Scholar
Guoqiang Wang
View author publications
Search author on:PubMed Google Scholar
Wenhua Zhang
View author publications
Search author on:PubMed Google Scholar
Shuhan Yu
View author publications
Search author on:PubMed Google Scholar
Junjie Fei
View author publications
Search author on:PubMed Google Scholar
Tingting An
View author publications
Search author on:PubMed Google Scholar
Jianqiao Yi
View author publications
Search author on:PubMed Google Scholar
Fengtian Li
View author publications
Search author on:PubMed Google Scholar
Ting Huang
View author publications
Search author on:PubMed Google Scholar
Jian Yang
View author publications
Search author on:PubMed Google Scholar
Mengmeng Niu
View author publications
Search author on:PubMed Google Scholar
Yang Wang
View author publications
Search author on:PubMed Google Scholar
Chuan Xu
View author publications
Search author on:PubMed Google Scholar
Zhi-Xiong Jim Xiao
View author publications
Search author on:PubMed Google Scholar

Contributions

Y.Y., Z.-X.J.X., C.X., and Y.W. conceived and designed the research. Y.Y. and G.Q.W. performed most of the experiments with assistance from W.H.Z., S.H.Y., J.J.F., T.T.A., J.Q.Y, and J.Y. T.H., F.T.L., M.M.N. contributed to the data discussion. Y.Y. and Z.-X.J.X. wrote the manuscript. All the authors read and approved the manuscript.

Corresponding authors

Correspondence to Yong Yi, Yang Wang, Chuan Xu or Zhi-Xiong Jim Xiao.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks Alessandro Carrer, Evangelos Michelakis, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Description of Supplementary Information

Supplementary Data 1

Supplementary Data 2

Supplementary Data 3

Supplementary Data 4

Supplementary Data 5

Supplementary Data 6

Reporting Summary

Peer Review file

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Yi, Y., Wang, G., Zhang, W. et al. Mitochondrial-cytochrome c oxidase II promotes glutaminolysis to sustain tumor cell survival upon glucose deprivation. Nat Commun 16, 212 (2025). https://doi.org/10.1038/s41467-024-55768-9

Download citation

Received: 02 August 2023
Accepted: 20 December 2024
Published: 02 January 2025
DOI: https://doi.org/10.1038/s41467-024-55768-9

This article is cited by

HADHA-mediated regulation of JAK/STAT3 signaling in glioblastoma: a metabolic-epigenetic axis
- Kan Wang
- Yifei Xiao
- Yu Cheng
Cell Death Discovery (2025)
Dexmedetomidine preserves neuronal function by promoting mitochondrial biogenesis through the AMPK/PGC-1α pathway
- Li Wang
- Meng Zhang
- Yanlin Chen
In Vitro Cellular & Developmental Biology - Animal (2025)