Respiratory complex I is essential to induce a Warburg profile in mitochondria-defective tumor cells - PubMed (original) (raw)
doi: 10.1186/2049-3002-1-11.
Luisa Iommarini, Ivana Kurelac, Maria Antonietta Calvaruso, Mariantonietta Capristo, Pier-Luigi Lollini, Patrizia Nanni, Christian Bergamini, Giordano Nicoletti, Carla De Giovanni, Anna Ghelli, Valentina Giorgio, Mariano Francesco Caratozzolo, Flaviana Marzano, Caterina Manzari, Christine M Betts, Valerio Carelli, Claudio Ceccarelli, Marcella Attimonelli, Giovanni Romeo, Romana Fato, Michela Rugolo, Apollonia Tullo, Giuseppe Gasparre, Anna Maria Porcelli
Affiliations
- PMID: 24280190
- PMCID: PMC4178211
- DOI: 10.1186/2049-3002-1-11
Respiratory complex I is essential to induce a Warburg profile in mitochondria-defective tumor cells
Claudia Calabrese et al. Cancer Metab. 2013.
Abstract
Background: Aerobic glycolysis, namely the Warburg effect, is the main hallmark of cancer cells. Mitochondrial respiratory dysfunction has been proposed to be one of the major causes for such glycolytic shift. This hypothesis has been revisited as tumors appear to undergo waves of gene regulation during progression, some of which rely on functional mitochondria. In this framework, the role of mitochondrial complex I is still debated, in particular with respect to the effect of mitochondrial DNA mutations in cancer metabolism. The aim of this work is to provide the proof of concept that functional complex I is necessary to sustain tumor progression.
Methods: Complex I-null osteosarcoma cells were complemented with allotopically expressed complex I subunit 1 (MT-ND1). Complex I re-assembly and function recovery, also in terms of NADH consumption, were assessed. Clones were tested for their ability to grow in soft agar and to generate tumor masses in nude mice. Hypoxia levels were evaluated via pimonidazole staining and hypoxia-inducible factor-1α (HIF-1α) immunoblotting and histochemical staining. 454-pyrosequencing was implemented to obtain global transcriptomic profiling of allotopic and non-allotopic xenografts.
Results: Complementation of a truncative mutation in the gene encoding MT-ND1, showed that a functional enzyme was required to perform the glycolytic shift during the hypoxia response and to induce a Warburg profile in vitro and in vivo, fostering cancer progression. Such trigger was mediated by HIF-1α, whose stabilization was regulated after recovery of the balance between α-ketoglutarate and succinate due to a recuperation of NADH consumption that followed complex I rescue.
Conclusion: Respiratory complex I is essential for the induction of Warburg effect and adaptation to hypoxia of cancer cells, allowing them to sustain tumor growth. Differently from other mitochondrial tumor suppressor genes, therefore, a complex I severe mutation such as the one here reported may confer anti-tumorigenic properties, highlighting the prognostic values of such genetic markers in cancer.
Figures
Figure 1
Expression and functional effects of allotopic nND1. (A) Western blot analysis for ND1 in control (CC), OS-93 and OS-93ND1 representative clones. Voltage-dependent anion channel (VDAC) was used as a loading control. One representative experiment of three is shown. (B) m.3571insC mutation load evaluation by fluorescent (F)-PCR. Wild-type and mutant fragments are distinguished based on the length of the homopolymeric stretch, where 7C corresponds to the C insertion. (C) Complex I in-gel activity (CI-IGA) assay in isolated mitochondria. CI-IGA band is indicated with an arrow. One representative experiment of four is shown**.** (D-E) NAD+/NADH ratio and NADH levels were measured in cell lysates. Data (mean ± SD) are expressed as pmoles of NADH and normalized for protein content (n = 3; *P <0.05; **P <0.01). (F) Oxygen consumption rate (OCR). Measurements were performed upon injection of 1 μM oligomycin (O), 0.1 μM trifluorocarbonylcyanide phenylhydrazone (FCCP) (F), 1 μM rotenone (R) and 1 μM antimycin A (AA). The inset shows the OCR control cell line (CC) profile, analogous to that of OS-93ND1 cells. Data (mean ± standard error of the mean (SEM)) are expressed as pmoles of O2 per minute per 3 × 104 cells (n = 3). (G) Mitochondrial ATP synthesis driven by pyruvate/malate and succinate, CI and CII substrates, respectively. CS, citrate synthase. (n = 4; #undetectable value; **OS-93 v_s_ CC, P <0.01; **OS-93ND1 vs OS-93, P <0.01; *CC vs OS-93ND1, P <0.05). (H) Mitochondrial membrane potential evaluation in CC, OS-93 and OS-93ND1 cells. Arrows indicate the addition of 6 μM oligomycin (O) and 4 μM FCCP (F). Data are mean ± SEM (n = 6). Fluorescence readings following the addition of oligomycin and preceding that of FCCP revealed a statistically significant difference (P <0.05) for all time points between OS-93 and both CC and OS-93ND1 cells.
Figure 2
Complex I (CI) function is required for recovery of tumorigenic potential in vitro and in vivo. (A) Representative images of anchorage-independent colony growth in soft agar of control (CC), OS-93 and OS-93ND1 cell lines. (B) Colony count on soft agar plate after 7 days; data are mean ± SD (n = 3, *P <0.05). (**C**) Tumor growth induced upon injection of CC, OS-93 and OS-93ND1 cell lines in nude mice. Data are mean ± standard error of the mean (SEM) (n = 3, 5 to 10 animals inoculated in each experiment; *_P <_0.05; *CC and OS-93ND1 versus OS-93). (**D**) CI-in gel activity (IGA) assay in tumor homogenates from CC, OST-93 and OST-93ND1 tumors. One representative experiment of three is shown. (**E**) Representative electron micrographs of CC, OST-93 and OST-93ND1 tumors. Asterisks indicate diverse mitochondrial morphology. (**F**) Fluorescent-PCR analysis of the m.3571insC in OS-93 and OS-93ND1 cell lines and xenografts (OST). All xenografts maintained the same m.3571insC mutant load as their corresponding cell lines (>90%).
Figure 3
Transcriptional profile of OST-93 and OST-93 ND1 xenografts. (A) Heatmap displaying expression levels of the 521 differentially expressed (DE) genes in the samples analyzed. Dark red = upregulated genes; cyan = downregulated genes. (B) The transcriptional profile of the Warburg phenotype**.** Heatmap showing gene expression profile of 21 DE genes from the hypoxia inducible factor-1α (HIF-1α) activation pathway. Genes are ordered by decreasing log2 fold change. Glucose transporters and glycolytic genes are marked in red. (C) Glucose uptake and glycolytic reactions. Genes overexpressed in OST-93ND1 xenografts are labeled in red. Red circles represent glucose molecules. (D) Quantitative real-time PCR validation performed on biological replicates of OST-93 (n = 2) and OST-93ND1 (n = 7) for 9/21 HIF-1α-responsive genes found DE in RNA-Seq (*P <0.05).
Figure 4
Complex I (CI) rescue correlates with decrease in α-ketoglutarate (α-KG)/ succinate (SA) ratio and recovery of hypoxia inducible factor-1α (HIF-1α) stabilization. (A) Representative immunohistochemical (IHC) analysis of CI (NDUFB8) and HIF-1α in OST-93 and OST-93ND1 xenografts. Positive NDUFB8 (b) and HIF-1α staining (d) is observed in OST-93ND1 but not in OST-93 xenografts (a, c); magnification 100×. Pimonidazole staining of representative OST-93 and OST-93ND1 xenografts (e, f); magnification 63×. (B) Correlation between NDUFB8 and HIF-1α IHC staining scores. X-axis represents NDUFB8 staining scores obtained considering the percentage of positive cells and staining intensity. Y-axis represents the percentage of HIF-1α-positive nuclei. (C) Western blot analysis of HIF-1α and lactate dehydrogenase A (LDHA) protein levels in OST-93 and OST-93ND1 tumors. Coomassie staining was used as loading control (input). (D) The ratio of α-KG and SA levels was calculated by measurements for each metabolite in OST-93- and OST-93ND1-derived cell lines. Data are mean ± SD (n = 3, *P <0.05). (E) Scheme of the metabolic changes in the absence/recovery of functional CI in cancer cells. Non-functional CI leading to an increase of the α-KG/SA ratio (left panel), which may foster activity of prolyl-hydroxylases (PHDs) with subsequent HIF-1α degradation even at low oxygen. Inactivation of HIF-1α leads to the downregulation of glycolysis needed to compensate for the defective mitochondrial respiration. This scenario may not allow the metabolic adaptation of tumor cells, possibly inducing a short-circuited mitochondrial compensatory proliferation. Due to a recuperated NADH consumption, CI rescue restores the α-KG/SA balance (right panel), hence not preventing HIF-1α stabilization. HIF-1α may therefore translocate into the nucleus with HIF-1β and activate transcription of target genes (red ovals and rectangles), among which those contributing to increase the glycolytic flux, hence conferring a Warburg phenotype and allowing tumor adaptation and growth. Red elements indicate activation or overexpression.
References
- Koppenol WH, Bounds PL, Dang CV. Otto Warburg’s contributions to current concepts of cancer metabolism. Nat Rev Cancer. 2010;11:325–337. - PubMed
- Warburg O, Posener K, Negelein E. Uber den Stoffwechsel der Carcinomzelle. Biochem Zeitschr. 1924;152:309–344.
LinkOut - more resources
Full Text Sources
Other Literature Sources