Down-regulation of β-F1-ATPase (the catalytic subunit of the mitochondrial H+-ATP synthase) is a hallmark of many human tumours. The expression level of β-F1-ATPase provides a marker of the prognosis of cancer patients, as well as of the tumour response to chemotherapy. However, the mechanisms that participate in down-regulating its expression in human tumours remain unknown. In the present study, we have investigated the expression of β-F1-ATPase mRNA (termed β-mRNA) in breast, colon and lung adenocarcinomas and squamous carcinomas of the lung. Despite the down-regulation of the protein, tumour β-mRNA levels remained either unchanged (breast and lung adenocarcinomas) or significantly increased (colon and squamous lung carcinomas) when compared with paired normal tissues, suggesting a specific translation-masking event for β-mRNA in human cancer. Consistently, we show using cell-free translation assays that a large fraction (~70%) of protein extracts derived from breast and lung adenocarcinomas specifically repress the translation of β-mRNA. We show that the 3′UTR (3′ untranslated region) of human β-mRNA is a relevant cis-acting element required for efficient translation of the transcript. However, an RNA chimaera bearing the 3′UTR of human β-mRNA does not recapitulate the inhibitory effect of tumour extracts on β-mRNA translation. Overall, the findings of the present study support the hypothesis that down-regulation of the bioenergetic activity of mitochondria in human tumours is exerted by translation silencing of β-mRNA.

INTRODUCTION

Mitochondria play a central role in the physiology of eukaryotic cells. The biogenesis of mitochondria is a complex genetic programme that requires the concerted transcriptional response of nuclear and mitochondrial genes [1]. However, mechanisms that control the localization, stability and translation of mRNAs also contribute to defining the mitochondrial phenotype of the cell [2]. In oxidative phosphorylation, ATP is synthesized by the H+-ATP synthase, a rotatory engine of the inner mitochondrial membrane that utilizes as a driving force the proton electrochemical gradient generated by the respiratory chain. β-F1-ATPase (the β-subunit of the mitochondrial H+-ATP synthase) is the catalytic subunit of the complex and rate-limiting component for the production of ATP [3]. The expression of β-F1-ATPase in the liver [46], in brown adipose tissue [7] and during progression through the cell cycle [8] indicates that the mitochondrial localization and translation of β-F1-ATPase mRNA (termed β-mRNA) are primary sites for regulating the spatial and temporal expression of the protein. Similar findings have been obtained in yeast, where the sorting of β-mRNA to the vicinity of mitochondria is mediated by its 3′UTR (3′ untranslated region), and deletion of this element leads to deficient protein import, reduced ATP synthesis, mtDNA (mitochondrial DNA) depletion and respiratory dysfunction [9,10]. In rat liver, the 3′UTR of β-mRNA has a functional internal ribosome entry sequence [11,12] which controls the synthesis of the protein at the G2/M-phase of the cell cycle [8]. The binding of developmentally regulated RNA-binding proteins to the 3′UTR of β-mRNA has been shown to inhibit its translation during foetal stages of liver development [4]. The re-installment of the same mechanism of translational repression seems to control the expression of β-F1-ATPase in rat hepatomas [13].

Mitochondrial malfunctioning is implicated in the pathogenesis of various human disorders including cancer [14,15]. Cancer cells are characterized by a predominant glycolytic metabolism even under aerobic conditions [16,17]. The distinctive glycolytic shift of cancer cells is accompanied by silencing of the bioenergetic activity of mitochondria [3]. In this regard, we and others have reported that the expression level of β-F1-ATPase is consistently down-regulated in different human tumours when compared with normal tissue [3,14,1824]. Moreover, the relative cellular expression level of the protein [β-F1-ATPase/GAPDH (glyceraldehyde-3-phosphate dehydrogenase) ratio] provides a bioenergetic signature of the tumour with potential clinical utility [16]. In fact, a diminished bioenergetic signature of the tumour is associated with poor prognosis as assessed in large cohorts of colon [14,25], lung [3,19] and breast [20] cancer patients. Interestingly, the bioenergetic signature of cancer cells and tumours also affords a promising predictive marker of the response to therapy [2528]. Even though the expression of β-F1-ATPase is compromised in cancer and its down-regulation involved in progression of the disease, our knowledge of the mechanisms that control β-F1-ATPase expression in human tumours is virtually nil. In the present study we show that post-transcriptional expression of β-mRNA plays a key role in defining the bioenergetic phenotype of colon, lung and breast carcinomas. Consistently, we illustrate in cell-free translation assays that a large fraction of breast and lung carcinoma extracts trigger the specific repression of β-mRNA translation when compared with extracts derived from normal tissue of the same patients. In agreement with previous findings in the rat [4,12], we document that the human 3′UTR of β-mRNA is a cis-acting element required for efficient translation. However, in contrast with findings in foetal rat liver [4] and rat hepatocarcinomas [13], human tumour extracts did not recapitulate the translation inhibitory effect on a reporter bearing the 3′UTR of human β-mRNA, indicating that control of β-mRNA translation in human cells requires additional elements.

EXPERIMENTAL

Patient specimens and protein extraction

Non-embedded frozen tissue obtained from surgical specimens of untreated cancer patients with primary ductal infiltrating carcinomas of the breast, adenocarcinomas of the colon and lung, and squamous cell lung carcinomas were obtained from the Banco de Tejidos y Tumores, IDIBAPS (Instituto de Investigaciones Biomédicas Pi y Suñer), Hospital Clinic, Barcelona, Spain. Routine histopathological study of all the cases had been previously performed by an experienced pathologist and the histological type, grade and size of the tumour, as well as regional lymph-node involvement, was recorded (Table 1). All tissue samples were anonymized and received in a coded form to protect patient confidentiality. The Institutional Review Board approved the project and samples were provided with informed consent from the patients. Tissue sections of the tumour and normal tissue of each patient were analysed and the regions of the tumours that did not contain significant areas of fibrosis, inflammation or necrosis were chosen [18]. For protein extraction, the samples were homogenized with a glass potter in a buffer containing 20 mM Hepes (pH 7.9), 100 mM KCl, 3 mM MgCl2, 1 mM DTT (dithiothreitol), 0.5 mM PMSF and protease inhibitors (Roche) in a 1:5 (w/v) ratio, and further freeze-thawed three times in liquid nitrogen. After protein extraction, the samples were centrifuged (15000 g) at 4 °C for 25 min. For in vitro translation assays, 200 μg of protein extracts were concentrated using centricon-10 filters (Amicon, Millipore) and the buffer exchanged to 20 mM Hepes (pH 7.9), 20 mM KCl, 1 mM DTT, 0.5 mM PMSF and protease inhibitors. The protein concentration was determined using Bradford reagent (Bio-Rad) using BSA as a standard. Aliquots were stored at −80 °C until use.

Table 1
Summary of clinicopathological characteristics of the cohorts of patients studied

AC, adenocarcinoma; DIC, ductal infiltrating carcinoma; N/R, non-registered; SCC, squamous cell carcinoma.

Breast cancer Lung cancer Colon cancer 
Characteristic 
Histology n Histology n Histology n 
 Normal breast 14  Normal lung 35  Normal colon 27 
 DIC 50  AC 19  AC 27 
   SCC 16   
Stage  Stage  Stage  
 I 10  IA  I 
 IIA 19  IB  IIA 10 
 IIB  IIA  IIB 
 III 12  IIB  IIIA 
   IIIA  IIIB 
   IIIB  IIIC 
     N/R 
Grade  Grade  Grade  
 1  1  1 
 2 18  2 17  2 17 
 3 28  3 12  3 
   N/R  N/R 
Lymph node metastases  Lymph node metastases  Lymph node metastases  
 No 24  No 13  No 17 
 Yes 24  Yes 22  Yes 
 N/R  N/R  N/R 
Breast cancer Lung cancer Colon cancer 
Characteristic 
Histology n Histology n Histology n 
 Normal breast 14  Normal lung 35  Normal colon 27 
 DIC 50  AC 19  AC 27 
   SCC 16   
Stage  Stage  Stage  
 I 10  IA  I 
 IIA 19  IB  IIA 10 
 IIB  IIA  IIB 
 III 12  IIB  IIIA 
   IIIA  IIIB 
   IIIB  IIIC 
     N/R 
Grade  Grade  Grade  
 1  1  1 
 2 18  2 17  2 17 
 3 28  3 12  3 
   N/R  N/R 
Lymph node metastases  Lymph node metastases  Lymph node metastases  
 No 24  No 13  No 17 
 Yes 24  Yes 22  Yes 
 N/R  N/R  N/R 

Western blot analysis

Protein extracts were fractionated by SDS/PAGE (10% gels) and transferred on to PVDF membranes. The antibodies used for Western blot analysis were: anti-(β-F1-ATPase) (used at 1:20000) and anti-GAPDH (used at 1:20000) [14]. Peroxidase conjugated anti-rabbit or anti-mouse IgGs (1:5000) (Nordic Immunology) were used as secondary antibodies for detection by ECL (enhanced chemiluminescence; Amersham Bioscience). Membranes were exposed to X-ray films and the quantification of the immunoreactive bands (in arbitrary units) was accomplished using a Kodak DC120 Zoom digital camera and the Kodak 1D Image Analysis Software for Windows.

Quantification of mRNA expression

Human β-mRNA levels in normal and tumour tissues were determined by qPCR (quantitative PCR) using TissueScan Tissue qPCR Arrays from OriGene Technologies. The breast cancer panel (BCRT101) contains cDNA from seven normal breast and 41 ductal breast adenocarcinoma biopsies. The lung cancer panel (HLRT104) includes cDNA from paired normal and tumour tissue derived from ten adenocarcinomas, eight squamous and six other lung carcinoma patients. The colon cancer panel (HCRT103) includes cDNA from paired normal and tumour tissue of 24 adenocarcinoma patients. Only samples from clinic stages I, II and III patients were considered in the analysis. For detailed pathological information of these patients see Table 1. Real-time qPCR was performed using an ABI PRISM 7900 SDS thermocycler and the Power SYBR Green Master Mix (Applied Biosystems) following the manufacturer's instructions. The following forward (F) and reverse (R) primers were used to amplify human β-F1-ATPase and β-actin cDNAs: F, 5′-CAGCAGATTTTGGCAGGTG-3′ and R, 5′-CTTCAATGGGTCCCACCATA-3′; and F, 5′-CAGCCATGTACGTTGCTATCCAGG-3′ and R, 5′-AGGTCCAGACGCAGGATGGCATG-3′ respectively. The relative expression level of β-mRNA was determined using the comparative ΔΔCt method [29] using β-actin as a control and relative to the ΔCt mean value of the normal samples.

Cloning strategies

To obtain the plasmid pBS-hβF1-fl, encoding the full-length human β-F1-ATPase cDNA, we followed the same bipartite strategy as for the rat cDNA [4] using human RNA (30 μg) as a template for the RT (reverse transcription) (AMV reverse transcriptase; Boehringer Ingelheim). The primers and restriction enzymes used were as follows: for the 5′-end fragment, hβcDNA-ss (5′-GCGCGGCTCGAGAGTCTCCACCCGGACTACGCCA3′) and hβcDNA-as (5′-TGGTCTCCTTCAGGGGTACCAGCTT-3′), digested with XhoI and PstI and placed in the pBluescript-SK+ vector; for the 3′-end fragment, hβ500-ss (5′-GCGCGGGGGCCCGGATCCCAACATTGTTGGCAG-3′) and hβ3′β-as (5′-GCGCGGGAATTCTTTTTTTTTTTTT-3′), digested with PstI. The plasmid encoding ARF (ADP-ribosylation factor) fused to the 3′UTR of human β-mRNA was obtained by subcloning the ApaI/EcoRI cDNA fragment from pBS-hβF1-fl into pARF [4]. Plasmids expressing a hybrid GFP (green fluorescent protein) that contains the mitochondrial targeting sequence of β-F1-ATPase (pβ) with the 3′UTR of human β-mRNA (pβGFP-3′β) or α-F1-ATPase mRNA (pβGFP-3′α) or without (pβGFP-3′Δ) were developed in the pcDNA3.1 vector (Invitrogen) using the pβ-GFP-β-3′UTR [8] and human cDNA as a template. The following primers were used: F-pβ-GFP-KpnI, 5′-GCGGGTACCCGAATCCAGTCTC-3′; R-GFP-EcoRI, 5′-GCGGAATTCTTCACTTGTACAGCTCGTCCATG-3′; F-3′β-h-EcoRI, 5′-GCGGAATTCGGGGTCTTTGTCCTCTGTA-3′; R-3′β-h-XhoI, 5′-GCGCTCGAGTTTTTTTTTTTTTTGAGGGGTGTA-3′; F-3′α-h-EcoRI, 5′-GCGGAATTCACTCCTGTGGATTCACATCAA-3′; and R-3′α-h-XhoI, 5′-GCGCTCGAGTTTTTTTTTTTTAACTATGCATTATG-3′. PCR amplifications were performed with the Expand High Fidelity PCR System (Roche).

Transcription reactions and UV cross-linking experiments

In vitro transcriptions were carried out for 2 h at 37 °C using the T7 MEGAscript kit (Ambion/Applied Biosystems), the corresponding linearized plasmid DNA (1 μg) and 6 mM m7G(5′)ppp(5′)G cap analogue (Ambion/Applied Biosystems) [12]. The restriction enzymes used to generate the following RNAs are indicated in brackets: hβF1 (EcoRI), hβ-3′UTR (EcoRI) and hβF1-3′Δ (HindIII), ARF (ApaI), ARF-3′β (EcoRI), pβGFP-3′β and pβGFP-3′α (XhoI) and pβGFP-3′Δ (EcoRI). RNAs were purified and the quality assessed by agarose gel electrophoresis. UV cross-linking experiments were performed as previously described [4]. In brief, 10 μg of breast or lung tissue extracts were incubated with the radiolabelled 3′UTR of human β-mRNA (5×105 c.p.m.) at 30 °C for 30 min. The reaction mixtures were exposed to UV light (254 nm) (Stratalinker 1800; Stratagene) for 6 min on ice, before the addition of 20 units of RNase T1 (Boehringer). For competition experiments, an excess of unlabelled 3′UTR of β-mRNA was added 10 min before the addition of the labelled probe. The RNA–protein complexes were resolved by SDS/PAGE.

Translation assays

In vitro translations of capped RNAs were carried out using nuclease-treated rabbit reticulocyte lysates (GE Healthcare) [4]. The reactions were performed in the presence of 1.2 μCi/μl of [35S]methionine labelling mix (Redivue Promix; GE Healthcare), 100 mM K+ and 1.5 mM Mg2+ ions for 30 min at 30 °C in the presence or absence of human extracts (0–10 μg of protein) derived from normal or tumour tissues. Tissue extracts were diluted in translation buffer (20 mM Hepes, pH 7.9, 20 mM KCl and 1 mM DTT supplemented with protease inhibitors). The translation products were further analysed by SDS/PAGE and fluorography [4]. The input RNA was recovered after translation and the content of the remaining β-mRNA determined by RT-qPCR [30].

Statistical analysis

Statistical analysis was performed using a Student's t test for paired samples. One-way ANOVA was used to detect differences in the bioenergetic signature within the normal and tumour biopsies. A standard F test was used to assess significance. Statistical tests were two-sided at the 5% level of significance.

RESULTS

Evidence of post-transcriptional regulation of β-F1-ATPase expression in human cancer

The expression of mitochondrial β-F1-ATPase is down-regulated in a wide variety of carcinomas concurrent with the up-regulation of the glycolytic GAPDH (Figure 1). In fact, a significant inverse correlation between these two protein markers of metabolism have been reported in the analysis of large cohorts of breast, colon and lung cancer patients [3,14,20]. The results in Figure 1 illustrate the significant down-regulation of the bioenergetic signature (β-F1-ATPase/GAPDH ratio) of breast, colon and lung adenocarcinomas, as well as in squamous carcinomas of the lung, when compared with the bioenergetic signature in paired normal tissues from the same patients. These findings raised the question of the mechanism(s) that could control β-F1-ATPase expression in human tumours. Therefore we determined the expression level of β-mRNA in normal and tumour tissue of breast, colon and lung cancer patients (Figure 1). The expression of β-mRNA in cancer varied significantly depending upon the tissue and histological type being considered (Figure 1). Expression of β-mRNA in breast (Figure 1A) and lung (Figure 1C) adenocarcinomas was not significantly different from normal tissues. In contrast, expression of β-mRNA was significantly augmented in adenocarcinomas of the colon (Figure 1B) and squamous carcinomas of the lung (Figure 1C) when compared with paired normal samples. These findings indicate that the decreased expression of β-F1-ATPase in human tumours (Figure 1) could not be ascribed to a limitation in the availability of β-mRNA due to an impaired transcription and/or increased degradation of the transcript. Therefore the results suggest that the lower content of β-F1-ATPase could originate from translational silencing of β-mRNA as it has been previously documented in rat liver during development [4], in brown adipose tissue [7] and in rat hepatocarcinomas [13].

Expression of β-mRNA in human tumours

Figure 1
Expression of β-mRNA in human tumours

The bioenergetic signature (β-F1-ATPase/GAPDH ratio) was determined by Western blot analysis in human breast (A), colon (B) and lung adenocarcinomas (Adeno) and squamous carcinomas (C) (T, closed bars) and in paired normal (N, open bars) tissue of the same patients. For illustration purposes, the result of a patient is shown for each type of carcinoma. The expression of β-mRNA was assessed by qPCR in normal and tumour tissue of breast (A), colon (B) and lung (C). The number of patients analysed is shown in brackets. The results shown are means±S.E.M. *P<0.05, when compared with normal tissue (as measured using a Student's t test). a.u., arbitrary units; β-F1, β-F1-ATPase.

Figure 1
Expression of β-mRNA in human tumours

The bioenergetic signature (β-F1-ATPase/GAPDH ratio) was determined by Western blot analysis in human breast (A), colon (B) and lung adenocarcinomas (Adeno) and squamous carcinomas (C) (T, closed bars) and in paired normal (N, open bars) tissue of the same patients. For illustration purposes, the result of a patient is shown for each type of carcinoma. The expression of β-mRNA was assessed by qPCR in normal and tumour tissue of breast (A), colon (B) and lung (C). The number of patients analysed is shown in brackets. The results shown are means±S.E.M. *P<0.05, when compared with normal tissue (as measured using a Student's t test). a.u., arbitrary units; β-F1, β-F1-ATPase.

The 3′UTR of human β-mRNA is a relevant cis-acting element involved in the control of its translation

Previous studies have demonstrated a relevant role for the 3′UTR of rat liver β-mRNA in the control of its translation [4,11,12]. Therefore a first question was to verify whether the human 3′UTR is also necessary for efficient translation of the transcript. Translation in vitro of the human β-mRNA (hβF1) and of a truncated version of the mRNA that lacks the 3′UTR (hβF1-3′Δ) revealed a 2-fold decrease in the amount of synthesized pβ-F1-ATPase when the 3′UTR is missing (Figure 2A). To further confirm the need of the human 3′UTR in translation, we fused the human 3′UTR of β-mRNA to ARF and GFP mRNAs and tested the translational efficiency of the chimaeric reporters (Figure 2B). The ARF transcript was selected as a reporter to maintain uniformity with previous findings where the translational activity of the 3′UTR of rat β-mRNA has been described [4,13]. The results revealed an autonomous translation-enhancing activity of the human 3′UTR because the amount of ARF derived from the ARF-3′β chimaera was ~10-fold higher when compared with that of the ARF mRNA (Figure 2B). Likewise, the amount of GFP derived from the GFP-3′β reporter was ~2-fold higher than that derived from GFP chimaeras that lacked any 3′UTR (GFP-3′Δ) or contained the 3′UTR of α-F1-ATPase (GFP-3′α), a partner subunit of β-F1-ATPase in the H+-ATP synthase complex. The rather large differences in the translational activity of the 3′UTR as assessed with the ARF-3′β and GFP-3′β chimaeras could result from intrinsic translational differences of the two open reading frames. In any case, these results demonstrate that the human 3′UTR of β-mRNA is also necessary for efficient translation of the transcript, and that it might act as a relevant element in the control of β-mRNA translation [4,13].

The 3′UTR of human β-mRNA is required for efficient translation

Figure 2
The 3′UTR of human β-mRNA is required for efficient translation

(A) The human in vitro synthesized and capped β-mRNA containing (hβF1) or not containing (hβF1-3Δ) the 3′UTR of the transcript were translated in a cell-free system. The fluorograms show the synthesized protein product (pβF1). (B) The in vitro synthesized and capped ARF and GFP mRNA containing (ARF3′β, pβGFP-3′β) or not containing (ARF, pβGFP-3′Δ) the 3′UTR of β-mRNA, or containing the 3′UTR of α-F1-ATPase mRNA (pβGFP-3′α), were translated in a cell-free system. The fluorograms of the synthesized protein products are shown. The histograms illustrate the relative amount of synthesized protein. The results shown are means±S.E.M. *P<0.05, when compared with hβF1, ARF and pβGFP-3′Δ respectively (as measured using a Student's t test). a.u., arbitrary units.

Figure 2
The 3′UTR of human β-mRNA is required for efficient translation

(A) The human in vitro synthesized and capped β-mRNA containing (hβF1) or not containing (hβF1-3Δ) the 3′UTR of the transcript were translated in a cell-free system. The fluorograms show the synthesized protein product (pβF1). (B) The in vitro synthesized and capped ARF and GFP mRNA containing (ARF3′β, pβGFP-3′β) or not containing (ARF, pβGFP-3′Δ) the 3′UTR of β-mRNA, or containing the 3′UTR of α-F1-ATPase mRNA (pβGFP-3′α), were translated in a cell-free system. The fluorograms of the synthesized protein products are shown. The histograms illustrate the relative amount of synthesized protein. The results shown are means±S.E.M. *P<0.05, when compared with hβF1, ARF and pβGFP-3′Δ respectively (as measured using a Student's t test). a.u., arbitrary units.

Breast and lung tumour extracts specifically repress β-mRNA translation

The lower content of β-F1-ATPase in tumours, despite an unchanged or increased availability of β-mRNA, suggested that its translation could be hampered in human cancer. Therefore we next studied whether protein extracts from paired normal and tumour tissue derived from the same patient have a different effect on the synthesis of pβ-F1-ATPase (pβF1) using cell-free translation assays [4,13]. It was observed that in seven out of the 11 (64%) breast cancer patients studied, the addition of carcinoma extracts promoted a dose-dependent reduction in the amount of pβF1 synthesized when compared with assays in which normal breast extract was added (Figures 3A and 3B). The higher inhibitory activity of breast tumour extracts on the synthesis of pβF1 compared with paired normal extracts is shown in the histogram (Figure 3B). Similar findings were obtained in six out of the eight (75%) lung cancer patients analysed when using extracts of lung adenocarcinomas and of its paired normal tissue (Figures 4A and 4B). The higher inhibitory activity of lung tumour extracts on the synthesis of pβF1 is shown in the histogram (Figure 4B). It should be noted that differences in the level of pβF1 synthesized are not the result of an increased degradation of β-mRNA in the presence of tumour extracts, because the recovery of the transcript at the end of translation assays was the same, irrespective of the extract added (see the histograms in Figures 3A and 4A).

Human breast tumour extracts specifically repress β-mRNA translation

Figure 3
Human breast tumour extracts specifically repress β-mRNA translation

(A) The human in vitro synthesized and capped β-mRNA (hβF1) was translated in a cell-free system in the presence of different amounts of human breast tissue extracts derived from normal (N, open circles and continuous line) and tumour (T, closed circles and discontinuous line) biopsies. The fluorogram and plot show the synthesized precursor of β-F1-ATPase (pβF1) in the presence of different protein inputs (μg of protein) of the extracts in a representative experiment derived from one cancer patient. The quantification of the inhibitory activity of normal (open bars) and tumour (closed bars) extracts for hβF1 translation in seven patients is summarized in the bottom panel of (B). The histogram in (A) shows the quantity of β-mRNA recovered after translation assays with normal (N) and tumour (T) extracts. (B) The in vitro synthesized and capped ARF and ARF-3′β RNAs were translated in a cell-free system in the presence of different amounts of human breast tissue extracts derived from normal and tumour biopsies. A representative experiment (same patient as in A) illustrates the synthesized ARF protein in the presence of different protein inputs (μg of protein) of the extracts. The histograms show the translation inhibitory activity of tumour (closed bars) extracts on the different RNAs tested when compared with the activity of normal extracts (open bars). The results shown are means±S.E.M. of seven different patients. *P<0.05, when compared with normal breast extracts (as measured using a Student's t test). a.u., arbitrary units.

Figure 3
Human breast tumour extracts specifically repress β-mRNA translation

(A) The human in vitro synthesized and capped β-mRNA (hβF1) was translated in a cell-free system in the presence of different amounts of human breast tissue extracts derived from normal (N, open circles and continuous line) and tumour (T, closed circles and discontinuous line) biopsies. The fluorogram and plot show the synthesized precursor of β-F1-ATPase (pβF1) in the presence of different protein inputs (μg of protein) of the extracts in a representative experiment derived from one cancer patient. The quantification of the inhibitory activity of normal (open bars) and tumour (closed bars) extracts for hβF1 translation in seven patients is summarized in the bottom panel of (B). The histogram in (A) shows the quantity of β-mRNA recovered after translation assays with normal (N) and tumour (T) extracts. (B) The in vitro synthesized and capped ARF and ARF-3′β RNAs were translated in a cell-free system in the presence of different amounts of human breast tissue extracts derived from normal and tumour biopsies. A representative experiment (same patient as in A) illustrates the synthesized ARF protein in the presence of different protein inputs (μg of protein) of the extracts. The histograms show the translation inhibitory activity of tumour (closed bars) extracts on the different RNAs tested when compared with the activity of normal extracts (open bars). The results shown are means±S.E.M. of seven different patients. *P<0.05, when compared with normal breast extracts (as measured using a Student's t test). a.u., arbitrary units.

Human lung adenocarcinoma extracts specifically repress β-mRNA translation

Figure 4
Human lung adenocarcinoma extracts specifically repress β-mRNA translation

(A) The human in vitro synthesized and capped β-mRNA (hβF1) was translated in a cell-free system in the presence of different amounts of human lung tissue extracts derived from normal (N, open circles and continuous line) and tumour (T, closed circles and discontinuous line) biopsies. The fluorogram and plot show the synthesized precursor of β-F1-ATPase (pβF1) in the presence of different protein inputs (μg of protein) of the extracts in a representative experiment derived from one cancer patient. The quantification of the inhibitory activity of normal (open bars) and tumour (closed bars) extracts for hβF1 translation in six patients is summarized in the bottom panel of (B). The histogram in (A) shows the quantity of β-mRNA recovered after translation assays with normal (N) and tumour (T) extracts. (B) The in vitro synthesized and capped ARF and ARF-3′β RNAs were translated in a cell-free system in the presence of different amounts of human lung tissue extracts derived from normal and tumour biopsies. A representative experiment (same patient as in A) illustrates the synthesized ARF protein in the presence of different protein inputs (μg of protein) of the extracts. The histograms show the translation inhibitory activity of tumour (closed bars) extracts on the different RNAs tested when compared with the activity of normal extracts (open bars). The results shown are means±S.E.M. of six different patients. *P<0.05, when compared with normal lung extracts (as measured using a Student's t test). a.u., arbitrary units.

Figure 4
Human lung adenocarcinoma extracts specifically repress β-mRNA translation

(A) The human in vitro synthesized and capped β-mRNA (hβF1) was translated in a cell-free system in the presence of different amounts of human lung tissue extracts derived from normal (N, open circles and continuous line) and tumour (T, closed circles and discontinuous line) biopsies. The fluorogram and plot show the synthesized precursor of β-F1-ATPase (pβF1) in the presence of different protein inputs (μg of protein) of the extracts in a representative experiment derived from one cancer patient. The quantification of the inhibitory activity of normal (open bars) and tumour (closed bars) extracts for hβF1 translation in six patients is summarized in the bottom panel of (B). The histogram in (A) shows the quantity of β-mRNA recovered after translation assays with normal (N) and tumour (T) extracts. (B) The in vitro synthesized and capped ARF and ARF-3′β RNAs were translated in a cell-free system in the presence of different amounts of human lung tissue extracts derived from normal and tumour biopsies. A representative experiment (same patient as in A) illustrates the synthesized ARF protein in the presence of different protein inputs (μg of protein) of the extracts. The histograms show the translation inhibitory activity of tumour (closed bars) extracts on the different RNAs tested when compared with the activity of normal extracts (open bars). The results shown are means±S.E.M. of six different patients. *P<0.05, when compared with normal lung extracts (as measured using a Student's t test). a.u., arbitrary units.

To verify the specificity of the inhibitory effect of tumour extracts on human β-mRNA translation, we studied the effect of the extracts in the translation of an unrelated mRNA such as ARF, which was previously used as a reporter to document similar findings with the rat transcript [4,13]. The results showed no relevant effect of breast (Figure 3B) and lung (Figure 4B) tumour extracts on the synthesis of ARF when compared with paired normal extracts (compare the open and closed bars in Figures 3B and 4B respectively). These results illustrate the specific inhibition of β-mRNA translation exerted by tumour extracts.

Contrary to our expectations, given the relevant functional role that the 3′UTR of human β-mRNA has in translation (Figure 2) and the findings obtained with the rat transcript [4,13], we observed that human breast (Figure 3B) and lung (Figure 4B) tumour extracts were unable to recapitulate the translation inhibitory effect on RNA chimaeras that contained the 3′UTR of human β-mRNA (compare the open and closed bars in Figures 3B and 4B respectively). These results indicate the existence of mechanistic differences among mammals for controlling the expression of β-mRNA and suggest that, in human tissues, in addition to the 3′UTR, other elements of the transcript are required for its appropriate translation masking in cancer. Taken together, the results indicate that in a large fraction of human breast and lung adenocarcinomas the decreased expression of β-F1-ATPase could originate from a specific translation repression event of β-mRNA.

DISCUSSION

A characteristic of normal proliferating cells is the reprogramming of its energetic metabolism, switching from a prevalent oxidative phosphorylation in G1-phase to glycolysis in the rest of the cell cycle [16,31,32]. On top of these changes, a large number of studies support the malfunctioning of the bioenergetic activity of mitochondria in cancer cells and tumours (for reviews see [16,33]), a condition that might be superimposed to the acquisition of the metabolic phenotype of proliferation. Several mechanisms have been proposed to explain the Warburg effect observed in tumours (for reviews see [16,17]). However, irrespective of the decisive molecular alteration that could explain the Warburg phenotype, there are large biochemical similarities between tumours and embryonic tissues, especially with regard to the expression of isoforms of glycolytic enzymes and with the control of mitochondrial biogenesis [2]. Indeed, conserved mechanisms for regulating the expression of β-F1-ATPase have been described between foetal rat liver and rat hepatomas [2,13]. In agreement with these observations, we have shown that the expression of β-F1-ATPase is decreased in various human tumours, and its altered expression was linked to the glycolytic switch experienced by the tumours [3,14,20]. However, to date, the molecular mechanisms underlying the down-regulation of β-F1-ATPase expression in human cancers have remained largely unexplored. In the present study we have shown that the control of β-mRNA translation is a molecular mechanism that could participate in defining the bioenergetic phenotype of human tumours, supporting a role for the misregulation of the translation of a specific transcript in the aetiology of the disease [34].

Consistent with the complexity of cancer as a disease [35], and with the cell-type-specific mechanisms that regulate the biogenesis of mitochondria [36], we observed, with regard to the expression of β-mRNA, that different tissues and histological types respond differently in oncogenesis to the drop in its bioenergetic signature. Whereas in breast and lung adenocarcinomas there are no relevant changes in β-mRNA expression, colon adenocarcinomas and squamous carcinomas of the lung revealed a significant increase in β-mRNA expression. Changes in the cellular availability of β-mRNA in these tissues could result from an increase in the transcription rate of the gene and/or in the stability of the transcript, since both mechanisms have been shown to control the expression of the mammalian ATP5B gene in order to compensate for the energetic imbalance [5,13].

In agreement with previous findings with rat β-mRNA [4], we have shown that the 3′UTR of human β-mRNA is required for efficient translation of the transcript, acting also as a translational enhancer of reporter constructs. Following from these similarities, we have demonstrated that human β-mRNA translation is specifically inhibited in the presence of lung and breast adenocarcinoma extracts, as was previously shown for the translation of rat β-mRNA in the presence of foetal rat liver [4] and hepatocarcinoma [13] extracts, two other conditions where translation masking of β-mRNA has been described. However, in total contrast with these situations [4,13], a reporter with the 3′UTR of human β-mRNA is unable to recapitulate the translation inhibitory activity of tumour extracts (Figures 3B and 4B). It is assumed that the binding of proteins to the 3′UTR sterically hinder the initiation of translation by preventing 43S ribosomal recruitment on to the ‘closed loop’ pseudo-circular mRNA that is competent in translation [37]. In contrast with the rat 3′UTR, where silencing of β-mRNA translation and of 3′UTR-containing chimaeras is due to the specific binding of regulatory proteins to the 3′UTR of the transcript [4,13], the 3′UTR of human β-mRNA does not interact specifically with breast and lung tumour proteins (see Supplementary Figure S1 at http://www.BiochemJ.org/bj/426/bj4260319add.htm). In fact, and in contrast with rat data, the RNA–protein complexes that could be visualized in UV cross-linking experiments are not competed by an excess of the unlabelled human 3′UTR (see Supplementary Figure S1). This result is consistent with the absence of a specific effect of these extracts on the translation of the ARF-3′β chimaera and suggest that, although the 3′UTR is a sequence element required for efficient translation of the mRNA (Figure 2), in the human transcript additional cis-acting elements are required to recapitulate translational inhibition, thus indicating that the control of β-mRNA translation in human tissues is more complex than anticipated.

Translation masking of β-mRNA has been observed in foetal rat liver [4], brown adipose tissue [7], rat hepatomas [13,38] and in the human tumours analysed in the present study. Translation silencing is usually mediated by 3′UTR-mediated sequestration of the mRNA into RNPs (ribonucleoproteins) [39,40] and/or by miRNA (microRNA)-mediated inhibition of translation [41,42]. Recently, the AU-rich-element-binding protein HuR has been shown to interact with the human 3′UTR of β-mRNA [30]. However, HuR seems to play an ancillary role in β-F1-ATPase expression in human cells [30], consistent with the finding that it provides a relevant independent marker of breast cancer prognosis [30]. Since the control of β-F1-ATPase expression is mostly exerted at the level of translation [4,5,13], and we have observed no specific interactions of the human 3′UTR with protein extracts derived from breast and lung cancer biopsies, the possibility exists that miRNAs [41,42] could play a role in post-transcriptional regulation of β-mRNA expression in cancer. Taken together, we have unveiled the existence of a specific translation repression mechanism for human β-mRNA that might explain the diminished bioenergetic signature of the tumours. Deciphering the complete mechanistic picture that accompanies translational repression of β-mRNA will certainly contribute to our understanding of what transforms cancer into a chronic disease.

Abbreviations

     
  • ARF

    ADP-ribosylatian factor

  •  
  • DTT

    dithiothreitol

  •  
  • β-F1-ATPase

    the catalytic subunit of the mitochondrial H+-ATP synthase

  •  
  • GAPDH

    glyceraldehyde-3-phosphate dehydrogenase

  •  
  • GFP

    green fluorescent protein

  •  
  • miRNA

    microRNA

  •  
  • β-mRNA

    β-F1-ATPase mRNA

  •  
  • pβF1

    precursor of β-subunit

  •  
  • qPCR

    quantitative PCR

  •  
  • RT

    reverse transcription

  •  
  • 3′UTR

    3′ untranslated region

AUTHOR CONTRIBUTION

Imke Willers and Antonio Isidoro designed and performed the experiments, and interpreted the data. Álvaro Ortega provided conceptual input and advice. The pathologist Pedro Fernández analysed and contributed patient samples for the protein studies. José Cuezva conceived the project, wrote the manuscript and helped with the analysis of results. All of the authors contributed to a critical review of the paper and approved the final version.

We are grateful to Dr M. López de Heredia and Mrs M. Chamorro for preliminary experiments with the human 3′UTR of β-mRNA and expert technical assistance respectively. The Xarxa de Bancs de Tumors de Catalunya is acknowledged. I.M.W./A.I. is/was the recipient of pre-doctoral fellowships from Plan de Formación de Profesorado Universitario, Ministerio de Educación y Ciencia and Fundación Ramón Areces respectively.

FUNDING

This work was supported by the Ministerio de Educación y Ciencia [grant number BFU2007-65253]; the Centro de Investigación Biomédica en Red de Enfermedades Raras (CIBERER) and Fondo de Investigación Sanitaria [grant number PI080274]; and the ISCIII, Madrid and Comunidad de Madrid [grant number S-GEN-0269]. The CBMSO (Centro Bidogía Molecular Severo Ochoa) receives an institutional grant from Fundación Ramón Areces.

References

References
1
Scarpulla
R. C.
Transcriptional paradigms in mammalian mitochondrial biogenesis and function
Physiol. Rev.
2008
, vol. 
88
 (pg. 
611
-
638
)
2
Cuezva
J. M.
Sanchez-Arago
M.
Sala
S.
Blanco-Rivero
A.
Ortega
A. D.
A message emerging from development: the repression of mitochondrial β-F1-ATPase expression in cancer
J. Bioenerg. Biomembr.
2007
, vol. 
39
 (pg. 
259
-
265
)
3
Lopez-Rios
F.
Sanchez-Arago
M.
Garcia-Garcia
E.
Ortega
A. D.
Berrendero
J. R.
Pozo-Rodriguez
F.
Lopez-Encuentra
A.
Ballestin
C.
Cuezva
J. M.
Loss of the mitochondrial bioenergetic capacity underlies the glucose avidity of carcinomas
Cancer Res.
2007
, vol. 
67
 (pg. 
9013
-
9017
)
4
Izquierdo
J. M.
Cuezva
J. M.
Control of the translational efficiency of β-F1-ATPase mRNA depends on the regulation of a protein that binds the 3′ untranslated region of the mRNA
Mol. Cell. Biol.
1997
, vol. 
17
 (pg. 
5255
-
5268
)
5
Izquierdo
J. M.
Ricart
J.
Ostronoff
L. K.
Egea
G.
Cuezva
J. M.
Changing patterns of transcriptional and post-transcriptional control of β-F1-ATPase gene expression during mitochondrial biogenesis in liver
J. Biol. Chem.
1995
, vol. 
270
 (pg. 
10342
-
10350
)
6
Egea
G.
Izquierdo
J. M.
Ricart
J.
San Martín
C.
Cuezva
J. M.
mRNA encoding the β-subunit of the mitochondrial F1-ATPase complex is a localized mRNA in rat hepatocytes
Biochem. J.
1997
, vol. 
322
 (pg. 
557
-
565
)
7
Tvrdik
P.
Kuzela
S.
Houstek
J.
Low translational efficiency of the F1-ATPase β-subunit mRNA largely accounts for the decreased ATPase content in brown adipose tissue mitochondria
FEBS Lett.
1992
, vol. 
313
 (pg. 
23
-
26
)
8
Martinez-Diez
M.
Santamaria
G.
Ortega
A. D.
Cuezva
J. M.
Biogenesis and dynamics of mitochondria during the cell cycle: significance of 3′UTRs
PLoS ONE
2006
, vol. 
1
 pg. 
e107
 
9
Margeot
A.
Blugeon
C.
Sylvestre
J.
Vialette
S.
Jacq
C.
Corral-Debrinski
M.
In Saccharomyces cerevisiae, ATP2 mRNA sorting to the vicinity of mitochondria is essential for respiratory function
EMBO J.
2002
, vol. 
21
 (pg. 
6893
-
6904
)
10
Margeot
A.
Garcia
M.
Wang
W.
Tetaud
E.
di Rago
J. P.
Jacq
C.
Why are many mRNAs translated to the vicinity of mitochondria: a role in protein complex assembly?
Gene
2005
, vol. 
354
 (pg. 
64
-
71
)
11
Di Liegro
C. M.
Bellafiore
M.
Izquierdo
J. M.
Rantanen
A.
Cuezva
J. M.
3′-Untranslated regions of oxidative phosphorylation mRNAs function in vivo as enhancers of translation
Biochem. J.
2000
, vol. 
352
 (pg. 
109
-
115
)
12
Izquierdo
J. M.
Cuezva
J. M.
Internal-ribosome-entry-site functional activity of the 3′-untranslated region of the mRNA for the β subunit of mitochondrial H+-ATP synthase
Biochem. J.
2000
, vol. 
346
 (pg. 
849
-
855
)
13
de Heredia
M. L.
Izquierdo
J. M.
Cuezva
J. M.
A conserved mechanism for controlling the translation of β-F1-ATPase mRNA between the fetal liver and cancer cells
J. Biol. Chem.
2000
, vol. 
275
 (pg. 
7430
-
7437
)
14
Cuezva
J. M.
Krajewska
M.
de Heredia
M. L.
Krajewski
S.
Santamaria
G.
Kim
H.
Zapata
J. M.
Marusawa
H.
Chamorro
M.
Reed
J. C.
The bioenergetic signature of cancer: a marker of tumor progression
Cancer Res.
2002
, vol. 
62
 (pg. 
6674
-
6681
)
15
Wallace
D. C.
A mitochondrial paradigm of metabolic and degenerative diseases, aging, and cancer: a dawn for evolutionary medicine
Annu. Rev. Genet.
2005
, vol. 
39
 (pg. 
359
-
407
)
16
Cuezva
J. M.
Ortega
A. D.
Willers
I.
Sánchez-Cenizo
L.
Aldea
M.
Sánchez-Aragó
M.
The tumor suppresor function of mitochondria: translation into the clinics
Biochim. Biophys. Acta
2009
, vol. 
1792
 (pg. 
1145
-
1158
)
17
DeBerardinis
R. J.
Lum
J. J.
Hatzivassiliou
G.
Thompson
C. B.
The biology of cancer: metabolic reprogramming fuels cell growth and proliferation
Cell Metab.
2008
, vol. 
7
 (pg. 
11
-
20
)
18
Isidoro
A.
Martinez
M.
Fernandez
P. L.
Ortega
A. D.
Santamaria
G.
Chamorro
M.
Reed
J. C.
Cuezva
J. M.
Alteration of the bioenergetic phenotype of mitochondria is a hallmark of breast, gastric, lung and oesophageal cancer
Biochem. J.
2004
, vol. 
378
 (pg. 
17
-
20
)
19
Cuezva
J. M.
Chen
G.
Alonso
A. M.
Isidoro
A.
Misek
D. E.
Hanash
S. M.
Beer
D. G.
The bioenergetic signature of lung adenocarcinomas is a molecular marker of cancer diagnosis and prognosis
Carcinogenesis
2004
, vol. 
25
 (pg. 
1157
-
1163
)
20
Isidoro
A.
Casado
E.
Redondo
A.
Acebo
P.
Espinosa
E.
Alonso
A. M.
Cejas
P.
Hardisson
D.
Fresno Vara
J. A.
Belda-Iniesta
C.
, et al. 
Breast carcinomas fulfill the Warburg hypothesis and provide metabolic markers of cancer prognosis
Carcinogenesis
2005
, vol. 
26
 (pg. 
2095
-
2104
)
21
Unwin
R. D.
Craven
R. A.
Harnden
P.
Hanrahan
S.
Totty
N.
Knowles
M.
Eardley
I.
Selby
P. J.
Banks
R. E.
Proteomic changes in renal cancer and co-ordinate demonstration of both the glycolytic and mitochondrial aspects of the Warburg effect
Proteomics
2003
, vol. 
3
 (pg. 
1620
-
1632
)
22
Chen
J.
Kahne
T.
Rocken
C.
Gotze
T.
Yu
J.
Sung
J. J.
Chen
M.
Hu
P.
Malfertheiner
P.
Ebert
M. P.
Proteome analysis of gastric cancer metastasis by two-dimensional gel electrophoresis and matrix assisted laser desorption/ionization mass spectrometry for identification of metastasis-related proteins
J. Proteome Res.
2004
, vol. 
3
 (pg. 
1009
-
1016
)
23
He
Q. Y.
Chen
J.
Kung
H. F.
Yuen
A. P.
Chiu
J. F.
Identification of tumor-associated proteins in oral tongue squamous cell carcinoma by proteomics
Proteomics
2004
, vol. 
4
 (pg. 
271
-
278
)
24
Bi
X.
Lin
Q.
Foo
T. W.
Joshi
S.
You
T.
Shen
H. M.
Ong
C. N.
Cheah
P. Y.
Eu
K. W.
Hew
C. L.
Proteomic analysis of colorectal cancer reveals alterations in metabolic pathways: mechanism of tumorigenesis
Mol. Cell. Proteomics
2006
, vol. 
5
 (pg. 
1119
-
1130
)
25
Lin
P. C.
Lin
J. K.
Yang
S. H.
Wang
H. S.
Li
A. F.
Chang
S. C.
Expression of β-F1-ATPase and mitochondrial transcription factor A and the change in mitochondrial DNA content in colorectal cancer: clinical data analysis and evidence from an in vitro study
Int. J. Colorectal Dis.
2008
, vol. 
23
 (pg. 
1223
-
1232
)
26
Hernlund
E.
Hjerpe
E.
Avall-Lundqvist
E.
Shoshan
M.
Ovarian carcinoma cells with low levels of β-F1-ATPase are sensitive to combined platinum and 2-deoxy-D-glucose treatment
Mol. Cancer Ther.
2009
, vol. 
8
 (pg. 
1916
-
1923
)
27
Santamaria
G.
Martinez-Diez
M.
Fabregat
I.
Cuezva
J. M.
Efficient execution of cell death in non-glycolytic cells requires the generation of ROS controlled by the activity of mitochondrial H+-ATP synthase
Carcinogenesis
2006
, vol. 
27
 (pg. 
925
-
935
)
28
Shin
Y. K.
Yoo
B. C.
Chang
H. J.
Jeon
E.
Hong
S. H.
Jung
M. S.
Lim
S. J.
Park
J. G.
Down-regulation of mitochondrial F1F0-ATP synthase in human colon cancer cells with induced 5-fluorouracil resistance
Cancer Res.
2005
, vol. 
65
 (pg. 
3162
-
3170
)
29
Livak
K. J.
Schmittgen
T. D.
Analysis of relative gene expression data using real-time quantitative PCR and the 2−ΔΔCT method
Methods
2001
, vol. 
25
 (pg. 
402
-
408
)
30
Ortega
A. D.
Sala
S.
Espinosa
E.
Gonzalez-Baron
M.
Cuezva
J. M.
HuR and the bioenergetic signature of breast cancer: a low tumor expression of the RNA-binding protein predicts a higher risk of disease recurrence
Carcinogenesis
2008
, vol. 
29
 (pg. 
2053
-
2061
)
31
Wang
T.
Marquardt
C.
Foker
J.
Aerobic glycolysis during lymphocyte proliferation
Nature
1976
, vol. 
261
 (pg. 
702
-
705
)
32
Klevecz
R. R.
Bolen
J.
Forrest
G.
Murray
D. B.
A genomewide oscillation in transcription gates DNA replication and cell cycle
Proc. Natl. Acad. Sci. U.S.A.
2004
, vol. 
101
 (pg. 
1200
-
1205
)
33
Frezza
C.
Gottlieb
E.
Mitochondria in cancer: not just innocent bystanders
Semin. Cancer Biol.
2009
, vol. 
19
 (pg. 
4
-
11
)
34
Mamane
Y.
Petroulakis
E.
LeBacquer
O.
Sonenberg
N.
mTOR, translation initiation and cancer
Oncogene
2006
, vol. 
25
 (pg. 
6416
-
6422
)
35
Hanahan
D.
Weinberg
R. A.
The hallmarks of cancer
Cell
2000
, vol. 
100
 (pg. 
57
-
70
)
36
Pagliarini
D. J.
Calvo
S. E.
Chang
B.
Sheth
S. A.
Vafai
S. B.
Ong
S. E.
Walford
G. A.
Sugiana
C.
Boneh
A.
Chen
W. K.
, et al. 
A mitochondrial protein compendium elucidates complex I disease biology
Cell
2008
, vol. 
134
 (pg. 
112
-
123
)
37
Gallie
D. R.
The cap and poly(A) tail function synergistically to regulate mRNA translational efficiency
Genes Dev.
1991
, vol. 
5
 (pg. 
2108
-
2116
)
38
Luciakova
K.
Kuzela
S.
Increased steady-state levels of several mitochondrial and nuclear gene transcripts in rat hepatoma with a low content of mitochondria
Eur. J. Biochem.
1992
, vol. 
205
 (pg. 
1187
-
1193
)
39
Abaza
I.
Gebauer
F.
Trading translation with RNA-binding proteins
RNA
2008
, vol. 
14
 (pg. 
404
-
409
)
40
Shyu
A. B.
Wilkinson
M. F.
van Hoof
A.
Messenger RNA regulation: to translate or to degrade
EMBO J.
2008
, vol. 
27
 (pg. 
471
-
481
)
41
Jackson
R. J.
Standart
N.
How do microRNAs regulate gene expression?
Sci. STKE 2007
2007
pg. 
re1
 
42
Pillai
R. S.
Bhattacharyya
S. N.
Filipowicz
W.
Repression of protein synthesis by miRNAs: how many mechanisms?
Trends Cell Biol.
2007
, vol. 
17
 (pg. 
118
-
126
)

Author notes

1

These authors equally contributed to the present study.

Supplementary data