Cyclin D1 Inhibits Mitochondrial Activity in B Cells

Cyclin D1 is a critical mitogenic sensor linking growth factor signals to the cell cycle machinery and promoting cell proliferation. It is the regulatory subunit of complexes with the cyclin-dependent kinases CDK4/6. Cyclin D1/CDK complexes phosphorylate retinoblastoma protein (RB), inactivating this protein and thus abolishing its transcriptional repressive activity and promoting cell cycle progression (1). Cyclin D1 also associates physically with transcription factors, transcriptional coregulators and chromatin remodelling proteins, positively or negatively regulating their activities thereby controlling cell growth, cell migration and differentiation (2–4).

The human CCND1 gene encoding cyclin D1 was initially cloned as a breakpoint locus in parathyroid adenoma (5). Subsequent studies showed that cyclin D1 was overproduced in many solid cancers and malignant hemopathies (6). Uncontrolled cell proliferation is a hallmark of cancer, and the subversion of cell cycle control due to cyclin D1 overproduction drives tumorigenesis. Other mechanisms controlled by cyclin D1, such as cell metabolism, growth, invasion, and metastasis also contribute to cell transformation (7–9).

Mantle cell lymphoma (MCL) is a very aggressive lymphoid neoplasm characterized by the proliferation of mature B cells that disseminate throughout the body. Multiple myeloma (MM) is also an aggressive hemopathy in which clonal malignant plasma cells accumulate within the bone marrow. The hallmark of MCL is the reciprocal translocation t (11; 14) (q13; q32) that juxtaposes the CCND1 gene and a strong enhancer of the IGH locus. This translocation is also present in 25% of MM cases. It leads to the production of cyclin D1, which is normally absent from cells of the B lineage. CCND1 activation is considered to be an early oncogenic event in both these diseases (10, 11). It has been suggested that cyclin D1 production leads to the bypassing of the restriction point at the transition between the G1 and S phases, thus favoring uncontrolled proliferation. Consistent with this hypothesis, we and others have recently shown that cyclin D1 regulates the G1-S transition in MCL and MM cells (12, 13). The contribution of other cyclin D1 functions to the oncogenic action of this protein in B cells remains to be determined.

In this study, we investigated the effects of cyclin D1 expression in mature B cells. We overcame the need to use transfected cells by transducing B-cell lines with TAT-cyclin D1 fusion proteins. The “protein transduction domain” (PTD) of the TAT protein encoded by human immunodeficiency virus mediates the transfer of proteins across cell membranes without any major change to cellular homeostasis (14, 15). We thought that the introduction of TAT-cyclin D1 into B cells could be used to mimic the behavior of mature malignancies overexpressing cyclin D1, such as MCL and MM cells. Using this strategy, we provide evidence that cyclin D1 inhibits the mitochondrial activity of mature malignant and immortalized B cells. Our findings suggest that cyclin D1 contributes to mature B-cell malignancies through concomitant interference with several key processes, such as proliferation and metabolism.

Antibodies (Abs) against cyclin D1 (sc-718), BECN1 (sc-11427), β-tubulin (sc-9104), CDK6 (sc-177), VDAC1/2/3 (sc-98708), BCL2 (sc-492) were obtained from Santa Cruz Technologies; against COXII (#A6404) from Molecular Probes (Invitrogen); against HA (12CA5) from Roche Diagnostics; against hexokinase 2 (HK2, #2106) from Cell Signaling Tech. and against GAPDH (AM4300) from Applied Biosystems. ImmunoPure goat anti-rabbit and anti-mouse IgG (H+L) peroxidase-conjugated Abs and DyLight 488-conjugated affinity-purified donkey anti-rabbit IgG were purchased from Thermo Scientific Pierce Protein Research Products; Alexa Fluor 546-conjugated goat anti-rabbit IgG was obtained from Molecular Probes.

The complete cyclin D1 cDNA was inserted between the KpnI and EcoRI restriction sites of pTAT-HA plasmid, to generate the pTAT-D1wt plasmid (16). The pTAT-D1K112E, pTAT-D1T156A, and pTAT-D1T286A plasmids with the corresponding point mutations were constructed from pTAT-D1wt according to standard methods. The pTAT-D1ΔCD plasmid, containing a cyclin D1 gene encoding a protein lacking amino acids 142 to 253, was obtained by inserting KpnI and EcoRI restriction sites by PCR. The pTAT-D1CDonly was obtained by inserting the fragment excised by digestion with KpnI and EcoRI into pTAT-HA. All plasmids were sequenced to check the integrity of the coding sequences. The BL21(DE3)pLysS thermocompetent E. coli (Promega) cells were transformed with the various plasmids and selected on Luria Broth (LB) medium supplemented with 50 μg/mL ampicillin and 50 μg/mL chloramphenicol (Invitrogen). A single colony was cultured overnight in 10 mL of LB supplemented with antibiotics. This culture was then diluted 1/10, the bacteria were grown for 3 hours and protein production was stimulated by incubation for 3 hours with 100 μmol/L isopropyl-β-D-1-thiogalactopyranoside (IPTG; Sigma-Aldrich). Bacterial cultures were centrifuged for 15 minutes at 15,000 g, and the pellet was washed in PBS. The pellet was then suspended in 1 mL of buffer A (8 mol/L urea, 500 mmol/L NaCl, 20 mmol/L HEPES), sonicated 10 times, for 10 seconds each, at 3 W, with an ultrasonic processor (Vibra Cell 75022, Fischer Bioblock Scientific) and mixed with an equal volume of buffer B (500 mmol/L NaCl, 20 mmol/L HEPES, 60 mmol/L imidazole). Bacterial lysates were applied to Ni-NTA agarose (Qiagen) columns (Poly-Prep Chromatography Column, BioRad), which were washed 4 times with buffer C (equal volumes of buffer A and buffer B). Three elution fractions were obtained after three 10-minute incubations with 500 μL of buffer E (4 mol/L urea, 500 mmol/L NaCl, 20 mmol/L HEPES, 250 mmol/L imidazole). Purified proteins were dialyzed overnight against 0.9% NaCl. We then subjected 10 μL of the protein solution and BSA standards (0.2 to 1.6 μg) to SDS-PAGE. The gel was stained with Coomassie blue (Sigma-Aldrich) and the purified proteins were quantified by densitometry (FluorSImager and QuantityOne software, Bio-Rad).

The Ramos human Burkitt lymphoma cell line (ATCC: CRL-1596), the human MM U266 (DSMZ: ACC9), Karpas 620 (ACC 163), NCI-H929 (ACC 402), RPMI 8226 (ATCC: CCL-155) cell lines, the human MCL JeKo-1 (ACC 553), HBL-2, NCEB-1 (CRL-3005), and Granta-519 (ACC342) cell lines and CM immortalized mature B cells were cultured in RPMI 1640 medium (Lonza) supplemented with 10% fetal calf serum (FCS, Lonza) and 1% penicillin/streptomycin (Lonza), at 37°C, in a humidified atmosphere. For TAT-fusion protein transduction, Ramos cells used to seed plates at a density of 2 × 10⁵ cells/mL and cells were counted directly by trypan blue exclusion. For proteasome inhibition, Ramos cells were treated with 10 nmol/L MG132 (Calbiochem) for 2 hours. For G1/S arrest, they were treated with 1 mmol/L hydroxyurea (HU, Sigma-Aldrich) for 18 hours.

Whole-cell lysates, nuclear and cytoplasmic fractions were prepared as previously described (17). For immunoprecipitation (IP), protein extracts were obtained from 3 × 10⁷ cells resuspended in IP buffer (10 mmol/L Tris-HCl, 150 mmol/L NaCl, 5 mmol/L EDTA, 20% glycerol, and a protease inhibitor cocktail) and incubated for 15 minutes on ice. Cells were centrifuged for 15 minutes at 16,000 g and the supernatant was recovered for protein determinations with the Bradford assay. Protein extracts (500 μg) were incubated overnight at 4°C with 2 μg of relevant Ab. Protein A-agarose beads (50 μL, Roche Diagnostics) were then added and the mixture was incubated for 1 hour at 4°C. Beads were washed 3 times in IP buffer, resuspended in 20 μL of Laemmli buffer, boiled for 3 minutes and then removed by centrifugation. The methods used for immunoblotting (IB) have been described in detail elsewhere (17).

Cells (10⁵ cells per spot) were cytospun on Superfrost glass slides, at 500 g for 3 minutes, then fixed in 4% paraformaldehyde (PFA) and permeabilized by incubation with 0.5% Triton-X100 (v/v) for 5 minutes. In accordance with standard methods, slides were incubated with primary Abs and fluorescent secondary Abs: AlexaFluor 546-labeled donkey anti-mouse or anti-rabbit IgG (Molecular Probes, Invitrogen). Nuclei were counterstained with DAPI. Slides were mounted, and analyzed with a Fluoview FV 1000 confocal microscope and Fluoview Viewer software (Olympus).

Cells (5 × 10⁵ cells/condition) were fixed in 80% ethanol in PBS and treated with 100 μg/mL RNase A and 20 μg/mL propidium iodide (PI) for 30 minutes at 37°C. They were then analyzed in an Epics XL flow cytometer with the Expo32 software (Beckman Coulter).

Cells (5 × 10⁵ cells/condition) were stained for 30 minutes with 100 nmol/L of MitoTracker633 Deep Red (MTK, Invitrogen; 18). They were then analyzed by FACS or examined under a confocal fluorescence microscope. MTK stains functional mitochondria in living cells. Exponentially growing cells were harvested and suspended in a buffer containing 120 mmol/L KCl, 5 mmol/L KH₂PO₄, 10 mmol/L HEPES, 1 mmol/L MgSO₄, 2 mmol/L EGTA, 0.4 mmol/L ADP, pH7.4 at 37°C. Oxygen consumption was measured by polarography (OXYTHERM System, Hansatech Instruments). Cells were permeabilized with 0.1 mg/mL digitonin. Mitochondrial respiration was recorded for 5 minutes after the addition of 5 mmol/L glutamate and 5 mmol/L malate as substrates for complex I, or 10 mmol/L succinate for complex II. The specificity of the reaction was shown with 1 mmol/L KCl. Total ATP levels were determined in cultured cells with the ATP Determination Kit (Molecular Probes, A22066), a bioluminescence assay, according to the manufacturer's instructions.

Cells (5 × 10⁵ cells/condition) were stained for 10 minutes with 10 μg/mL 5′,5′,6,6-tetrachloro-1,1′,3,3′-tetraethylbenzimidazolcarbocyanine iodide cationic dye (JC-1, Molecular Probes) before harvesting. Depending on the mitochondrial membrane potential, JC-1 monomers can form aggregates, resulting in a shift in fluorescence emission from green to red. The intensity of red/green fluorescence decreases as the membrane becomes increasingly depolarized (19). Mitochondrial transmembrane potential (ΔΨm) was quantified by flow cytometry (Epics XL flow cytometer), on the basis of the aggregates/monomers fluorescence intensity ratio.

Cells were fixed with 2.5% glutaraldehyde in 0.1 mol/L Sorensen phosphate buffer pH 7.4 and postfixed with 1% OsO₄ in the same buffer. Cells were then dehydrated through a graded series of alcohol solutions, embedded in Epon 812 resin, and cut into thin sections (80 nm). Sections were stained with uranyl acetate and lead citrate, and observed with a JEOL 1011 transmission electron microscope equipped with a Megaview III camera.

Student's t test was used to determine the significance of differences between 2 experimental groups. Data were analyzed with 2-tailed tests and P < 0.05 was considered significant.

The direct intracellular delivery of proteins was made possible by the use of TAT fusion proteins (14–16). Purified TAT fusion proteins were added directly to the culture medium of Ramos cells, which were obtained from a Burkitt lymphoma that did not express cyclin D1 or cyclin D2 (20). Cyclin D2 is recognized as the only other G1 cyclin capable of compensating in vivo for cyclin D1 function in B cells (21, 22). Following the addition of 250 nmol/L TAT-cyclin D1wt, the fusion protein entered the cells within an hour. The protein was still detected at 24 hours, but not at 48 hours (Fig. 1A). Six hours after the addition of 250 nmol/L TAT-cyclin D1wt protein, cyclin D1 levels were similar to those observed in MCL and cyclin D1-expressing MM cell lines (Fig. 1A). In TAT-cyclin D1wt-transduced Ramos cells treated with MG132, cyclin D1 detection continued until 48 hours (Fig. 1A), consistent with the degradation of TAT-cyclin D1wt by the ubiquitin/proteasome pathway known to be the main effector of cyclin D1 degradation (23). TAT-cyclin D1wt and CDK6, the kinase partner of cyclin D1, were coimmunoprecipitated from transduced cell extracts, showing their ability to associate (Fig. 1A). Confocal microscopy examination of transduced Ramos cells confirmed that the TAT-cyclin D1wt protein was present in both the nuclear and cytoplasmic compartments (Fig. 1B), like the endogenous protein in cyclin D1-expressing cells (24). Flow cytometry analyses showed that transduction with TAT-cyclin D1wt in exponentially growing asynchronous cells had no effects on the cell cycle; whereas it induced cell cycle reentry in G1/S-arrested Ramos cells (Fig. 1C). Indeed, TAT-cyclin D1wt decreased the percentage of cells in the G0/G1 fraction. The TAT-fusion cyclin D1 protein may therefore be considered functional.

It is widely accepted that cyclin D1 overproduction results in deregulation of the G1/S checkpoint, thus contributing to cell transformation (1). Exponentially growing Ramos cells transduced with TAT-cyclin D1wt displayed a dose- and time-dependent decrease in total cell number in culture (Fig. 2A). Direct cell counting after trypan blue staining revealed no necrosis (data not shown), and PI-staining (Fig. 1C) and annexin V/PI-double staining (data not shown) of transduced cells showed that there was no apoptosis. Moreover, the subcellular distribution of the microtubule-associated protein (MAP) LC-3 and beclin-1 proteins, which are known to relocate to the autophagosomes membrane during autophagy (25), was not affected by transduction with the TAT fusion protein (data not shown). These results were confirmed by western blotting (not shown). Based on these data, we concluded that the decrease in the number of Ramos cells was not due to cell death of any type. A slowing down of cell proliferation could account for these differences. Consistent with this hypothesis, the doubling time of Ramos cells increased from 22 hours to 30 hours after a treatment with 250 nmol/L TAT-cyclin D1wt (Supplemental Table 1).

Mutant proteins were investigated to evaluate the role of the functional domains of cyclin D1 in the slowing down of cell proliferation (Supplementary Fig. 1A). The lysine 112 (K112) residue of cyclin D1 is required for the catalytic activity of cyclin D1/CDK4/6 complexes because the cyclin D1K112E mutant cannot activate CDKs (26). The cyclin D1T156A mutant prevents the nuclear import of CDK4 and its phosphorylation by CDK-activating kinase, and cannot phosphorylate RB (27). The threonine 286 (T286) residue is required for the phosphorylation of cyclin D1 by GSK-3β and the subsequent nuclear export and proteasomal degradation of the protein (23). The T286A mutant form of cyclin D1 is constitutively nuclear and displays oncogenic potential in B cells (28). The cyclin D1 central domain (CD, 142–253) is involved in interactions with transcriptional factors and nuclear receptors (2). Thus, the D1ΔCD mutant cyclin lacking the CD is likely to display impaired transcriptional regulation. We also constructed a mutant containing only the CD, and the resulting fusion protein is referred to as TAT-cyclin D1CDonly. The corresponding proteins were produced in bacteria, purified, added to Ramos cell cultures, and their subcellular distributions were determined. The TAT-cyclin D1wt, TAT-cyclin D1K112E, and TAT-cyclin D1ΔCD fusion proteins seemed to be present in both the cytoplasm and the nucleus. The TAT-cyclin D1T286A protein was exclusively nuclear, whereas TAT-cyclin D1T156A and TAT-cyclin D1CDonly proteins were strictly cytoplasmic (Supplementary Fig. 1A). The molecular weights of all fusion proteins were consistent with expected values (data not shown). TAT-cyclin D1K112E, TAT-cyclin D1T156A, and TAT-cyclin D1CDonly protein treatments led to a significant decrease in the total number of cells (Supplementary Fig. 1B). By contrast, nuclear TAT-cyclin D1T286A and TAT-cyclin D1ΔCD proteins had no effect. The effects of TAT-cyclin D1 on Ramos cells were: a) independent of the ability of cyclin D1 to bind CDK6 and to phosphorylate RB; b) dependent on the central domain of the protein (142–243 aa) and c) dependent on cyclin D1 being present in the cytoplasm.

Electronic microscopy observations of Ramos cells transduced with TAT-cyclin D1wt revealed no gross change in cell morphology or size (Fig. 2B). However, the mitochondria appeared clearer and less well defined, with the disappearance of cristae. Mitochondrial activity was assessed by MTK staining and flow cytometry sorting. Transduction with TAT-cyclin D1wt protein was associated with a dose-dependent decrease in mitochondrial activity (Fig. 2C). The number of stained cells decreased from 99.6% to 79.1% 6 hours after TAT-cyclin D1wt transduction. This decrease in MTK mitochondria staining was confirmed by confocal microscopy (Fig. 2C). Consistent with our previous results, TAT-cyclin D1T156A, TAT-D1K112E, TAT-D1DConly proteins had similar effects on mitochondrial activity whereas TAT-cyclin D1T286A and -cyclin D1ΔCD had no effect (Supplementary Fig. 2A). Transduction with the TAT-cyclin D1wt protein decreased mitochondrial activity in the CM immortalized and non transformed mature B-cell line (Supplementary Fig. 2B). We conclude that the inhibition of mitochondrial activity by cyclin D1 occurs in various B cells and is independent of transformation status.

Confocal microscopy studies of both MTK and TAT-cyclin D1wt staining indicated that the fusion protein colocalized with mitochondria (Fig. 3A). We carried out the same analysis for the endogenous cyclin D1 present in the U266 MM cell line to exclude nonrelevant binding due to the PTD part of the fusion protein and to confirm its colocalization with mitochondria (Fig. 3A). TAT-cyclin D1CDonly (Fig. 3B) and TAT–cyclin D1K112E (Supplementary Fig. 2C) that also inhibited mitochondria activity had similar distributions within the cell. We further analyzed fluorescence signals for BCL2 (an outer mitochondrial membrane protein), COXII (cytochrome c oxidase subunit 2, an inner mitochondrial membrane protein), and TAT-cyclin D1wt, by 3-dimensional (3D) confocal microscopy. BCL2 and TAT-cyclin D1wt were found at the outer mitochondrial membrane (OMM) whereas COXII and TAT-cyclin D1 were not (Fig. 3C).

The cyclin D1 protein interacts with various partners in the accomplishment of its physiological functions. As the changes in mitochondrial activity occurred in the absence of apoptosis, we searched for a putative partner involved in the regulation of mitochondrial metabolism, located at the OMM. The voltage-dependent anion channel (VDAC) is the most abundant OMM protein; it mediates the trafficking of metabolites and ions across the OMM resulting in energy production (29). We confirmed, by IP, that TAT-cyclin D1wt bound VDAC (Fig. 4A). Consistent with our previous results, TAT-cyclin D1CDonly in Ramos cells and endogenous cyclin D1 in U266 cells also bound VDAC (Fig. 4A). Closure of the VDAC is known to result in the inhibition of total mitochondrial activity (30). We then assessed mitochondrial respiratory activity by polarography. Both complex I (electron donors: glutamate and malate) and complex II (electron donor: succinate, complex I inhibitor: rotenone) displayed a 20% to 25% decrease in activity in Ramos cells transduced with TAT-cyclin D1wt (Fig. 4B, top panel). In addition to the inhibition of mitochondrial respiration, we also observed changes in proton/electron flux. The JC-1 aggregate/monomer ratio decreased in a dose-dependent manner following transduction with TAT-cyclin D1wt, showing a decrease in ΔΨm (Fig. 4B, bottom panel). Studies with TAT-cyclin mutants confirmed those results: monomer fluorescence increased for TAT-cyclin D1CDonly but not for the TAT-cyclin T286A and -cyclin D1ΔCD proteins (Supplementary Fig. 3B). Thus, our data suggest that cyclin D1 binds VDAC, thereby impairing mitochondrial activity through the closure of this channel. Total ATP production was measured in Ramos cells transduced with TAT-cyclin D1wt fusion, 24 hours after treatment. The total amount of ATP decreased as a function of fusion protein concentration (Fig. 4C), showing that the impairment of mitochondrial activity resulted in a global decrease in energy production. This result was confirmed by studies of TAT-cyclin D1 mutant proteins. The total amount of ATP decreased in Ramos cells treated with TAT-cyclin D1CD only, but not in those treated with TAT–cyclin D1T286A or TAT–cyclin D1ΔCD (Supplementary Fig. 3A).

The first reaction in glucose metabolism (conversion of glucose to glucose-6-phosphate) is catalyzed by HK2, a protein kinase bound to VDAC (31). Our results may also reflect the disruption of VDAC/HK2 interaction by competition between cyclin D1 and HK2 for VDAC binding, decreasing the supply of pyruvate to the mitochondria. We investigated HK2/VDAC/cyclin D1 interactions in Ramos cells transduced with 50 or 250 nmol/L TAT-cyclin D1wt protein for 6 hours. Whole-cell extracts were analyzed by western blotting with anti-HK2 antibody (input) or IP with either a nonimmune serum (ns) or an anti-VDAC antibody (Fig. 5A). In experiments with nonimmune serum, HK2 proteins were present in the unbound fraction (u), whereas in transduced cells, HK2 proteins were coimmunoprecipitated with VDAC (bound fraction, b). Moreover, the amount of HK2 bound to VDAC decreased with increasing levels of cyclin D1 (50 vs. 250 nmol/L fusion protein). The same IP experiment was carried out with 2 MM cell lines: NCI-H929 and U266 having different endogenous cyclin D1 levels (Fig. 1A). In NCI-H929 cells, the fraction of HK2 bound to VDAC was much higher than the fraction bound to cyclin D1. By contrast, in U266 cells, the fraction of HK2 bound to VDAC was lower than the fraction bound to cyclin D1 (Fig. 5A). These data indicate that cyclin D1 competes with HK2 for binding to VDAC.

The presence of cyclin D1 in cycling Ramos cells results in a paradoxical inhibition of cell proliferation. By contrast, the presence of cyclin D1 in G1-arrested cells leads to their entry into S phase. The fine regulation of cyclin D1 levels at the G1/S boundary is required for the correct completion of the cell cycle (32). This regulation is lost in cyclin D1-expressing MCL and MM cells, in which the CCND1 gene is controlled by strong enhancers at the IgH locus rather than by extracellular signals (10, 20). The introduction of cyclin D1 in exponentially growing cells may therefore mimic events in MM and MCL. This model revealed an unexpected function of cyclin D1 in the B-cell mitochondrial metabolism.

Two reports from the same group reported such a function for cyclin D1 in epithelial cells and fibroblasts (9, 33). Cyclin D1 represses mitochondrial biogenesis and activity by phosphorylating and inactivating nuclear respiratory factor 1 (NRF1) and, thus, mitochondrial transcription factor A (mtTFA), which is required for respiration (34). This is a nonnuclear function of cyclin D1 requiring the binding of this molecule to CDK. In the mammary epithelium, endogenous cyclin D1 represses mitochondrial activity by inhibiting the expression of nuclear genes encoding mitochondrial proteins (9). This function is conserved in immortalized and transformed epithelial cells. We have confirmed the inhibitory effect of cyclin D1 on mitochondrial activity in immortalized and malignant B cells, by a different mechanism. We show here that the mitochondrial uncoupling mediated by cyclin D1 involves binding to VDAC.

The form of cyclin D1 active against mitochondria is cytoplasmic and is not bound to CDK4 (data not shown). However, the central domain (142–253 aa), which is necessary for interactions with transcription factors or nuclear receptors (34–36), is required for this activity. Cyclin D1 is located at the OMM, where it binds VDAC through its central domain. The VDAC is the most abundant protein of the OMM and the mean regulator of the mitochondrial metabolism (29). Cell proliferation, oxygen consumption, and mitochondrial transmembrane potential (ΔΨm) were all inhibited in the presence of cyclin D1, suggesting that this protein acts through a global mechanism. The VDAC interacts with F0/F1ATPase and promotes ADP/ATP flux through the OMM, establishing a link between oxidative phosphorylation and the tri-carboxylic acid cycle. We suggest that the interaction between cyclin D1 and VDAC inhibits the transport of ATP, ADP, and other metabolites to the mitochondria thus resulting in a decrease in mitochondrial metabolism. The downregulation of VDAC1 expression by specific shRNA has been shown to decrease the rate of cell proliferation, thereby disrupting energy production (37). The mechanical obstruction of VDAC by the C-terminal tail of β-tubulin also inhibits mitochondrial respiration (38).

We also show here the existence of competition between cyclin D1 and HK2 for VDAC binding. Unlike mitochondria from most normal tissues, the mitochondria of cancer cells display binding of HK2 to VDAC (39). Once bound to VDAC, HK2 couples oxidative phosphorylation and glycolysis by capturing the ATP produced by mitochondria and exported through the pore. Cyclin D1 disrupts VDAC-HK2 interactions, thereby reducing ATP production as schematized in Figure 5B.

We have previously shown that cyclin D1 regulates the mitochondrial apoptotic pathway through the heat-shock protein (HSP)70 chaperone effects on pre and postmitochondrial proapoptotic effectors (40). The VDAC seems to serve as an anchoring point for proteins of the BCL2 family and forms part of the cytochrome c release channel (41). The release of proapoptotic molecules from the mitochondria is a key event in cell death signaling. The closure of VDAC by cyclin D1 may have prevented the onset of apoptosis and may be considered an alternative mechanism of cyclin D1-mediated resistance to anticancer drugs. In good agreement with that statement, we have preliminary data indicating that TAT-cyclin D1-transduced Ramos cells display resistance to dexamethasone-induced apoptosis. Interestingly, MM and MCL pathologies are particularly resistant to anticancer treatments.

In mammary tumors induced by cyclin D1 and on cyclin D1 antisense arrays, the genes involved in mitochondrial metabolism are reciprocally expressed (9). Cyclin D1 inhibits mitochondrial metabolism in both normal and transformed epithelial mammary cells, but these data suggest that this newly identified function of cyclin D1 may contribute to mammary tumorigenesis. We also found that cyclin D1 inhibited mitochondrial metabolism in both nonmalignant and malignant B cells. As cyclin D1 expression is restricted to malignant B cells, we can also speculate that this new function of cyclin D1 may be involved in the transformation process.

No potential conflicts of interest were disclosed.

We thank A Barbaras for technical assistance, Dr G Roué (Hospital Clinic, Barcelona, Spain) for providing the EBV+ immortalized CM B-cell line, Dr SF Dowdy (Howard Hughes Medical Institute, Chevy Chase, MD, USA) for the pTAT-HA plasmid, Dr G Peters (London Research Institute, London, UK) for the complete cyclin D1 cDNA, Didier Goux (CMaBio, IFR 146, Université de Caen, France) for assistance with electron microscopy and Dr S Allouche (Laboratoire de Biochimie, CHU de Caen, Caen, France) for assistance with polarography assays.

This work was financially supported by the Ligue contre le Cancer–Comité du Calvados and Comité de la Manche (B. Sola).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.