The mitochondrial metabolic reprogramming agent trimetazidine as an ‘exercise mimetic’ in cachectic C26-bearing mice

Animals and experimental design

Balb-c male mice (5 weeks old) were used. They were obtained from Charles River and were maintained on a regular dark–light cycle (light from 08:00 to 20:00), with free access to food (Piccioni) and water during the whole experimental period, including the night before sacrifice. Experimental animals were cared for in compliance with the Italian Ministry of Health Guidelines (no 86609 EEC, permit number 106/2007-B) and the Policy on Humane Care and Use of Laboratory Animals (NIH 1996). The animals were randomized according to their body weight on the day before the treatments and divided into four groups, namely, controls and tumour bearers, treated or not with TMZ (Sigma-Aldrich). Because we evidenced the instability of TMZ powder, we always used fresh TMZ, not older than 1 month after package opening. Sample size has been defined on the basis of previous studies. In particular, 6–7 mice/experimental group (depending if healthy or tumour-bearing) have been estimated sufficient to detect a difference of 10 in the means of each parameter with a SD of 8 and with 90% power, a significance level of 95% at Student's t-test, and a dropout rate of 10%. In more detail, the first group (Ctrl; n = 6) served as controls and included healthy mice inoculated with vehicle (saline); the second group (TMZ; n = 6) included healthy mice receiving intraperitoneal injection of 5 mg/kg TMZ once a day for 12 consecutive days; the third group (C26; n = 7) included tumour-bearing mice inoculated subcutaneously dorsally with 5 * 10⁵ C26 carcinoma cells; the fourth group (C26-TMZ; n = 7) was inoculated with C26 cells and the same day started receiving TMZ intraperitoneal injection of 5 mg/kg TMZ once a day for 12 consecutive days. Based on the most recent indications suggesting that, in order to possibly treat cachexia, it is necessary to act on of pre-cachexia conditions, in the present work, we aimed to investigate the possibility that TMZ administration could interfere with the onset and progression of cachexia; therefore, to start the treatment concomitant with tumour inoculation was the best option. Future studies could be developed in order to understand if TMZ might have a curative action in milder models of cancer cachexia, starting to treat the animals when the tumour becomes palpable. Animal weight and food intake were recorded daily. The day of sacrifice (12 days after tumour transplantation), the animals were anesthetized by isoflurane inhalation. The blood was collected by cardiac puncture and used to monitor glycaemia (Glucocard G-sensor; Menarini Diagnostics). The mice were then sacrificed by cervical dislocation and tibialis anterior (TA) and gastrocnemius (GSN) muscles were rapidly excised, weighed consecutively numbered and frozen in liquid N₂-cooled isopentane, and finally stored at −80°C. When subsequent analyses are performed, the operator just knows the numbers but not the corresponding experimental groups.

Skeletal muscle function analysis

A forelimb grip strength blinded test was performed at Days 6 and 12 after tumour implantation by using a commercial digital grip strength metre (Columbus Instruments). Mice held by the tail were gently allowed to grasp a wire grid with the fore paws. The mice were then gently pulled by the tail until they released their grip. The force achieved by the mouse was recorded during three trials and averaged.

Immunofluorescence and cross-sectional area evaluation

Myofibre CSA measurement was performed on TA. Serial muscle sections (9 μm) were obtained from the midbelly region of the TA muscles that had been embedded in OCT. A CM1900 cryostat (Leica) at −20°C was used. Sections were fixed in 4% paraformaldehyde and stained with an antilaminin (L9393) antibody and an antislow MyHC (M8421) from Sigma-Aldrich. The Alexa Fluor 488 anti-rabbit IgG (A11008) from Life Technologies was used as secondary antibodies. Nuclei were visualized with the DNA dye 40,6-diamidino-2-phenylindole, and the samples were mounted in SlowFade Gold mounting media (Life Technologies). The images were acquired with a Leica TCS SP5 confocal microscope. In the stained muscle sections, automated CSA determination along the laminin-stained border of each fibre was evaluated by using Image J.46 Because errors in fibre border recognition might occur (i.e. either the fibres might not be recognized or several fibres/non-fibre regions might be interpreted as a single fibre), a manual correction of myofibre border misinterpretation was performed.

Histochemistry

Enzymatic staining for the succinate dehydrogenase (SDH) activity was performed on TA cross sections. SDH incubation media were prepared by dissolving 1 mg/ml of nitrotetrazolium blue chloride and 27 mg/ml of sodium succinate in PBS.

Quantitative real-time polymerase chain reaction

Dissociation of samples was performed with Qiagen TissueRuptor, and RNA isolation was performed by using TRIreagent (Sigma-Aldrich) following the manufacturer's instructions. For RT-PCR, cDNA was synthesized with oligo-dT by adding 1 μg of RNA with GoScript Reverse Transcription System (Promega). Comparative real-time PCR was performed with the SYBR-green master mix (Promega) by using the Stratagene MX3000 (Thermo Fisher Scientific). Data were normalized to 18S, and a calibrator was used as internal control. Resulting data were analysed by the MX3PRO (v4.10), and fold change was determined by using the 2-ΔΔCT method.47, 48 All reactions were performed in triplicate. The following primers were used:

Protein isolation and western blotting

Tissue samples from GSN and TA muscles were homogenized by using a GentleMACS Dissociator (Miltenyi Biotec) and lysed in ice cold Cosper and Leinwand Myosin Extraction Buffer (300 mM NaCl, 0.1 M NaH₂PO₄, 0.05 M Na₂HPO₄, 0.01 M Na₄P2O7, 1 mM MgCl₂, 10 mM EDTA, and 1 mM DTT pH 6.5) or in lysis buffer (10 mM Tris/HCl, pH 7.4, 5 mM EDTA, 5 mM EGTA, 1% Triton X-100, 130 mM NaCl, 0.1% SDS, and 0.1% sodium deoxycholate) supplemented with a protease inhibitor cocktail (Roche) and a phosphatase inhibitor cocktail (Sigma-Aldrich). A clear supernatant was obtained by centrifugation of lysates at 13 000 g for 20 min at 4°C. Protein concentration in the supernatant was determined by Bradford protein assay (Bio-Rad). Aliquots of total cell lysates were then separated by SDS-PAGE by using Miniprotean precast gels (BioRad), and proteins were transferred to nitrocellulose membranes (BioRad) or to polyvinylidene difluoride membranes (BioRad). Membranes were blocked 1 h at RT at 4°C with 5% non-fat milk in Tris-buffered saline with 0.05% Tween 20 and then probed by using the following antibodies directed against: Atrogin-1 (AP2041; ECMBiosciences), Laminin (L9393; Sigma-Aldrich), MyHC (MF20 recognizing all isoforms of skeletal muscle myosin heavy chain; Developmental Studies Hybridoma Bank at the University of Iowa), slow MyHC (M8421; Sigma-Aldrich), fast MyHC (M4276; Sigma-Aldrich), Desmin (D1033; Sigma-Aldrich), α-tubulin (T5168; Sigma-Aldrich), pAkt (9271; Cell Signalling), Tom20 (sc-11415; Santa Cruz Biotechnology), TFAM (sc-23588; Santa Cruz Biotechnology) pFoxo 3A (9466S; Cell Signalling), PGC1α (AB3242; Millipore), βchain-ATP Synthase (MAB3494; Millipore), NDGR1 (3217; Cell Signalling), p-GSK3β (ser9) (ab75814; Abcam), p-RPS6 (2215S; Cell Signalling), p-p38MAPK (9216; Cell Signalling), VE cadherin (ab91064, Abcam), p-NFATc2 (ab200819, Abcam), and pACC (05-673 Millipore). The appropriate secondary horseradish peroxidase-conjugated antibodies from Jackson Immunoresearch were used in blocking solution for 1 h at room temperature. Immunoreactive bands were visualized by SuperSignal West Pico Chemioluminescent substrate kit (Pierce). Equal loading of samples was confirmed by α-tubulin (T5168; Sigma-Aldrich) or actin (A3853; Sigma-Aldrich) normalized and quantified by densitometry by using the ImageQuant TL software from GE Healthcare Life Sciences. The first WB was performed by an operator who was not aware of the group assignment, then in order to properly present data, samples were divided in the experimental groups and re-probed.

Cell culture and RNA extraction

Murine C2C12 skeletal myoblasts were grown at 37°C in 5% CO2 in an air-humidified chamber in high-glucose DMEM (Gibco) with glutamax, supplemented with 20% foetal bovine serum and 1% penicillin/streptomycin (Euro-Clone).49 As the cells approached confluency, the growth medium was replaced with a differentiation medium: DMEM supplemented with 2% horse serum and 1% penicillin/streptomycin. The medium was changed every second day. On Day 4 of differentiation, myotubes were treated with TMZ (10 μM) and RNA was isolated with an RNeasy Micro Kit (Qiagen Instrumentation Laboratory, Milan, Italy), according to the manufacturer's instructions.

Assessment of ΔΨm

ΔΨm was measured by using tetramethylrhodamine ethyl ester (TMRE, Molecular Probes) as previously reported.50 Murine C2C12 skeletal myoblasts treated or not with TMZ (10 μM) were incubated at 37°C for 15 min in media containing TMRE (50 nM). As a control for ΔΨm dissipation, cells were treated with 10 μM carbonyl cyanide p-(trifluoromethoxy) phenylhydrazone. Cells were then rinsed in fresh medium and detached from the dish. TMRE fluorescence was detected by flow cytometry on a FACScalibur flow cytometer (Becton-Dickinson).

Ex vivo measurement of skeletal muscle contractile properties

The contractile properties of both EDL and Soleus muscles excised from 3-month-old C57/BL6 mice were measured ex vivo as previously described.51 The muscle to be tested was vertically mounted in an oxygenated and temperature controlled chamber containing Krebs-Ringer bicarbonate buffer. One end of the muscle was linked to a fixed clamp while the other was connected to the lever arm of an actuator/transducer (Aurora Scientific Inc. 300B). The isolated muscle was electrically stimulated, by means of two platinum electrodes, with 3 single 0.1 ms pulses and with 2 pulse trains at tetanic frequency (180 Hz for the EDL and 80 Hz for the Soleus). From the single pulse stimulations, we measured the twitch force and the kinetics parameters, namely, the time to peak, the half relaxation time, and the force derivative during both the contractile and the relaxation phases, while through the stimulation at the tetanic frequency, we measured the muscle maximum force and the resistance to isotonic fatigue. All the forces were then normalized with reference to the muscle cross sectional area52 to obtain the specific force values. Within this technique, we tested both the systemic and the direct effects of TMZ on skeletal muscles. In the first case, the animals were treated with 5 mg/kg TMZ once a day for 12 days. In the second case, each muscle was incubated in Krebs-Ringer bicarbonate buffer containing TMZ for 15 min before being stimulated. Two different TMZ concentrations were tested: 10 and 100 μM.

Statistical analysis

All experiments were performed at least three times, unless otherwise indicated. Data are presented as mean ± standard error of the mean. Statistical differences between groups were verified by Student's t-test (2-tailed). *p < 0.05 was considered significant. For contractile properties measurements, differences in the mean of muscles incubated ex vivo with TMZ were assessed with one-way ANOVA followed by Tukey's multiple comparisons test.

The metabolic reprogramming triggered by trimetazidine leads to grip strength increase in mice

Mice bearing the C26 colon carcinoma and healthy animals (controls; Ctrl) were treated with the metabolic modulator TMZ for 12 days. Grip strength was measured following 6 and 12 days of TMZ administration (Figure 1). In both Ctrl and C26-bearing mice, a significant increase in grasping strength was recorded after 6 days of TMZ treatment (Figure 1A); at day 12, this increase was persistent in Ctrl, but no more detectable in C26 hosts (Figure 1B). In line with our previous results obtained on C2C12 myotubes,43 blood glucose was reduced and the expression of the glucose receptor Glut 4 was increased in TMZ-treated Ctrl mice but not in the C26 hosts (Figures 1C and S1). The expression of Glut 4 increased also in C2C12 myotubes treated with TMZ (Figure S1).

Trimetazidine partially restores myofibre cross-sectional area in C26-bearing mice

Based on our previous results showing that TMZ reduces myotube atrophy induced by TNFα and serum deprivation,43 we investigated the effect of this metabolic modulator on cancer-induced muscle wasting. The C26 tumour becomes palpable 5–6 days after implantation; the mice start losing weight at day 9 and are sacrificed at day 12, when the final body weight reaches about 75–80% of Ctrl animals (Figure 2A). TMZ did not prevent the decrease in body weight occurring in the C26-bearing mice (Figure 2A), nor did TMZ treatment reverse the decrease in TA, GSN weight, and adipose tissue mass recorded at 12 days after tumour implantation (Figure 2B and C). Moreover, tumour growth was not affected by TMZ (Figure 2D).

Conversely, TMZ partially protects myofibre CSA (Figure 3A and C; C26 vs. C26-TMZ) from the reduction associated to cancer (Figure 3A and B; Ctrl vs. C26). This is also reflected by the median value of CSA in the C26 hosts treated with TMZ, which was significantly higher compared with that of the untreated C26 bearers (Figure 3C; red vs. black bar). Of note, TMZ did not influence CSA in healthy mice (Figures 3A and S2; Ctrl vs. TMZ).

Trimetazidine triggers the activation of pathways involved in the maintenance of muscle mass

In order to evaluate the effect of TMZ treatment in the skeletal muscle also at the molecular level, we analysed the expression of genes associated to muscle mass maintenance such as those coding for myofibrillar proteins, namely, Desmin and MyHC. In the present study, we found that Desmin and MyHC mRNA levels increase in both the TA and the GSN of TMZ-treated C26 hosts compared with the untreated mice (Figure 4A). Because the two muscles were comparable, we performed further analysis, throughout this study, on one muscle type only.

Trimetazidine effects on the proteasomal degradative pathway were evaluated by measuring the transcript levels of the atrogenes Muscle Ring Finger protein 1 and atrogin-1. While the former tends to decrease (although not significantly) upon TMZ treatment in muscles of C26-bearing mice (Figure 4A), the latter does not significantly change (data not shown).

Finally, we also measured the levels of muscle Carnitine palmitoyl transferase 1 transcript, a transporter of fatty acid into mitochondria, which we used as a marker of TMZ effectiveness. We observed that the muscle Carnitine palmitoyl transferase 1 mRNA resulted up-regulated in TMZ-treated animals, possibly as a compensative response to the reduction of free fatty acid β-oxidation (Figure 4A).

Consistent with the data obtained for mRNAs, we found that, also at the protein level, the decrease of Desmin in muscle of C26-bearing mice is partially counteracted by TMZ treatment (Figure 4B). Accordingly, the protein levels of atrogin-1 increase in C26 hosts vs. Ctrl (likely contributing to the loss of muscle mass) while they are significantly reduced upon TMZ treatment (Figure 4B). In the attempt to shed some light on the signalling leading to myofibrillar protein up-regulation triggered by TMZ, we analysed some members of the PI3K/AKT-mediated molecular pathways, known to act on both protein degradation and synthesis. Along with atrogin-1 decrease, we found an increase in ribosomal protein S6 (RPS6)-phosphorylation (Figure 4B), this indicating activation of the mammalian target of rapamycin complex 1 target S6 kinase (S6 kinase 1) and suggesting that protein synthesis is enhanced. We have found that TMZ treatment also enhances GSK3-β phosphorylation in Ctrl mice (Figure 4B). Such an enhancement cannot be observed in the C26 hosts; however, in these animals, pGSK3-β levels are higher than in Ctrl, so the possibility that a further increase is not achievable should be considered (Figure 4B and Penna et al.19). All these results suggest that the PI3K/AKT pathway is involved in the effects evoked by TMZ and indicate the activation of mechanisms counteracting muscle mass loss.

A shift towards a slow-twitch phenotype is triggered by trimetazidine

Total MyHC protein levels in TMZ-treated C26-bearing mice (Figure 5A) did not reflect the strong increase in MyHC gene expression (Figure 4A). Therefore, we decided to investigate if TMZ treatment was able to influence specific MyHC isoforms, which are differentially associated with myofibres having specific metabolic assessments. TMZ being a metabolic modulator, we set out to study the influence that this drug exerts on the metabolic features of the skeletal muscle, which is known to adaptively change in response to environmental alterations. Our analysis revealed that, in both GSN and TA muscles, TMZ resulted in increased slow MyHC protein levels, whereas fast MyHC and total MyHC did not appear to vary significantly (Figure 5A); no effect was found in healthy animals (data not shown). Similarly, TMZ triggers a robust increase of the slow MyHC transcript levels (Figure 5B). Interestingly, we have found that NFAT2c, whose transcriptional activity includes slow gene expression,53, 54 becomes dephosphorylated—this indicating activation and nuclear translocation—following TMZ treatment (Figure 5C).

The up-regulation of the slow MyHC isoform triggered by TMZ suggests a metabolic shift, from glycolytic to oxidative, within the myofibre, but also that the number of type I fibres could be increased. To assess this last hypothesis, we performed an immunofluorescence analysis on TA muscle cryosections, which indicated that the number of type I myofibres expressing the slow MyHC isoform (red in Figure 5D) tends to increase, though not significantly, in C26-TMZ animals.

Three isoforms of fast MyHC have been described in rodents; type IIa, type IIx/d, and type IIb. We analysed these fast transcripts in TA and we observed that, in line with the increased expression of slow MyHC, their levels decrease in C26 hosts upon TMZ treatment (Figure 6A). A fibre-type conversion towards a more slow-twitch phenotype might also occur within type II fibres (e.g. from type IIb or type IIx/d towards type IIa) similarly to what happens upon exercise (reviewed by Gundersen54 and Lira55). For this reason, a more specific histological analysis will be part of further investigations. Lastly, we checked the expression levels of the slow isoforms of Troponin T, I, and C, some of which tend to increase, though not significantly, in TMZ-treated C26 animals (Figure 6B). In sum, these data support our hypothesis of a fast-to-slow twitch phenotype shift induced by TMZ in skeletal muscle.

Trimetazidine promotes mitochondrial biogenesis, oxidative metabolism, and angiogenesis

Slow fibres are characterized by a high reliance on oxidative metabolism; therefore, an increase in slow fibre proportion following TMZ administration should imply an enhanced oxidative metabolism. To test this hypothesis, we assessed if TMZ might impact on mitochondrial biogenesis by measuring the expression of the master regulator of mitochondrial biogenesis PGC1α.56, 57 We observed that PGC1α mRNA was up-regulated upon TMZ treatment in both GSN and TA muscles of the C26 hosts (Figure 7A). Accordingly, PGC1α expression is higher in TMZ-treated C26 hosts vs. the untreated ones, also at the protein level (Figure 7B). PGC1α activation might be achieved by the phosphorylation and activation of the metabolic sensor AMPK.58 Due to technical problems in detecting pAMPK, we evaluated the phosphorylation of the acetyl CoA carboxylase, one of the main targets of AMPK, and we found that pACC increases upon TMZ treatment in C26-bearing mice (Figure 7C), indicating an increased activation of AMPK, which is consistent with PGC1α up-regulation. The increase of mitochondrial biogenesis was further investigated by evaluating the expression of the mitochondrial protein Tom20, the mitochondrial transcription factor A (TFAM), and the β-subunit of the catalytic portion of mitochondrial ATP synthase, which, as expected, resulted definitively increased in TMZ-treated animals compared with the untreated ones (Figure 7B and C). Interestingly, PGC1α, Tom20, and TFAM protein levels decrease because of the disease (PGC1α; Ctrl vs. C26 p ≤ 0.01, Tom20; Ctrl vs. C26 p ≤ 0.01 and TFAM; Ctrl vs. C26 p ≤ 0.05) and such a reduction is partially rescued by TMZ (Figure 7B). Of note, mitochondrial biogenesis is controlled by several mechanisms, some of which may be independent on PGC1α; this explains the up-regulation of Tom20 and TFAM not associated to PGC1α increase in TMZ-treated Ctrl mice. Because PGC1α is a transcriptional co-activator, we also measured its protein levels in the nuclear fractions of GSN muscles, confirming the expression pattern observed in total lysates (data not shown).

As discussed in the succeeding texts, all our data suggest that the effects of TMZ administration to C26-bearing parallel those triggered by endurance exercise, as reported in the literature.59, 60 It has also been reported that exercise-induced mitochondrial biogenesis is largely controlled via PGC1α.61, 62, 59 To gain further insights into this similarity, we evaluated if the phosphorylation of p38MAPK, characterizing exercise and ex vivo muscle contraction,63, 64 is also activated by TMZ in C26 animals, and we found that the levels of p-p38MAPK are enhanced by TMZ in the C26 hosts (Figure 7C).

It is also known that exercise, as well as PGC1α overexpression, is able to trigger angiogenesis. As a further similarity among the effect of TMZ and exercise, we interestingly found that TMZ induces an increase of the vascular endothelial cadherin (Figure 7C) and, in line with this, we also revealed increased levels of the vascular endothelial growth factor—which is indeed a transcriptional target of PGC1α65—in skeletal muscle of TMZ-treated C26 mice (Figure 7E).

The experiments reported in the preceding texts show that TMZ triggers an increased mitochondrial biogenesis, likely reflecting an improved energy metabolism. To investigate this point, we evaluated the activity of the oxidative mitochondrial enzyme SDH in muscle cryosections. The percentage of oxidative fibres stained for SDH activity significantly increased in the C26 hosts upon TMZ treatment, while no difference was detected in TMZ-treated Ctrl mice (Figure 8; histogram on the left). The specific analysis of regions characterized by mixed glycolytic and oxidative fibres showed a trend towards an increased percentage of SDH-positive fibres also in TMZ-treated Ctrl mice (Figure 8; histogram on the right).

Finally, we measured the mitochondrial membrane potential (ΔΨm) in C2C12 myoblasts treated or not with TMZ; TMRE staining revealed an increase of ΔΨm triggered by TMZ, suggesting an increased mitochondria metabolism (Fig S3).

Autophagy and apoptosis are not altered by trimetazidine in C26-bearing mice

Because we have found an effect of TMZ on autophagy in vitro,43 we investigated the occurrence of autophagy and apoptosis in muscles upon metabolic modulation by TMZ. We evaluated both LC3 and p62 protein levels (Figure 9A), and we found that LC3-I decreases and LC3-II increases in C26 hosts (Ctrl vs. C26 p ≤ 0.01 for both LC3-I and LC3-II), indicating an increased autophagy in cancer cachexia, as previously reported.66 However, we found no significant variation of LC3-I and LC3-II levels associated with TMZ treatment; this demonstrates that, in contrast to our previous in vitro data, TMZ does not interfere with autophagy in vivo. This hypothesis is further supported by the observation that TMZ did not influence p62 expression levels in C26-bearing mice, despite p62 up-regulation found in Ctrl mice. To monitor apoptosis, we analysed both Caspase-3 and PARP cleavage67 in GSN lysates but observed no influence of TMZ in C26-bearing mice (Figure 9B).

Skeletal muscle is a direct in vivo target of trimetazidine

Finally, in order to study the specific effect of TMZ on skeletal muscle and to exclude the occurrence of an indirect role mediated by the known beneficial action of TMZ on cardiac function, we performed an ex vivo analysis on isolated hind limb muscles exposed to TMZ (see 2 section). Despite no significant effect being detected in the Soleus (data not shown), we observed a direct dose-dependent effect of TMZ on the fast EDL muscle (Figure 10). TMZ triggered a change of the EDL contraction kinetics inducing a shift towards a slow-twitch contractile phenotype, with the kinetics being significantly slower in the EDL treated with 100 μM TMZ compared with untreated EDL muscles (Figure 10A and B; df/df and −df/dt). The contraction rate of treated muscles was about 32% lower in comparison with the untreated ones (Figure 10A; df/df), and the relaxation rate was almost halved (Figure 10A; −df/df). This change towards a slow contractile phenotype was also found in the EDL excised from mice systemically treated with TMZ as shown in Figure 10B (df/df and −df/dt). This kinetic slowdown is partly explained by the decreased twitch force generated following both TMZ treatment of excised muscles and systemic TMZ treatment (Table 1; F_{tw_sp}), although the specific tetanic force and the resistance to fatigue did not significantly change upon TMZ treatment (Table 1; F_{0_sp} and T_fat). However, also considering the reduced twitch force, TMZ has a neat effect on the decrease of relaxation time (Figure 10A; −df/df) and on the kinetic slowdown. These experiments confirm that TMZ has the ability to directly act on skeletal muscle and to modify its contractile kinetics.

Ethical standard statement

Experimental animals were cared for in compliance with the Italian Ministry of Health Guidelines (no 86609 EEC, permit number 106/2007-B) and the Policy on Human Care and Use of Laboratory Animals (NIH 1996). The experimental protocol was approved by the Bioethical Committee of the University of Turin and by the Italian Ministry of Health, Italy, and has therefore been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments.