Dichloroacetate as a novel pharmaceutical treatment for cancer-related fatigue in melanoma

Abstract

Cancer-related fatigue (CRF) is one of the most common complications in patients with multiple cancer types and severely affects patients’ quality of life. However, there have only been single symptom-relieving adjuvant therapies but no effective pharmaceutical treatment for the CRF syndrome. Dichloroacetate (DCA), a small molecule inhibitor of pyruvate dehydrogenase kinase, has been tested as a potential therapy to slow tumor growth, based largely on its effects in vitro to halt cell division. We found that although DCA did not affect rates of tumor growth or the efficacy of standard cancer treatment (immunotherapy and chemotherapy) in two murine cancer models, DCA preserved physical function in mice with late-stage tumors by reducing circulating lactate concentrations. In vivo liquid chromatography-mass spectrometry/mass spectrometry studies suggest that DCA treatment may preserve membrane potential, postpone proteolysis, and relieve oxidative stress in muscles of tumor-bearing mice. In all, this study provides evidence for DCA as a novel pharmaceutical treatment to maintain physical function and motivation in murine models of CRF.

NEW & NOTEWORTHY We identify a new metabolic target for cancer-related fatigue, dichloroacetate (DCA). They demonstrate that in mice, DCA preserves physical function and protects against the detrimental effects of cancer treatment by reducing cancer-induced increases in circulating lactate. As DCA is already FDA approved for another indication, these results could be rapidly translated to clinical trials for this condition for which no pharmaceutical therapies exist beyond symptom management.

INTRODUCTION

Cancer-related fatigue (CRF) is a common and long-lasting complication in patients with cancer. Cancer treatments, such as chemotherapy, radiotherapy, and surgery, often worsen CRF (1–3). More than 50% of patients—and, in some studies of patients with late-stage tumors under treatment, almost 100% of patients—have reported experiencing CRF (4–6). The exhaustion patients with CRF experience differs dramatically from exercise-induced fatigue (7). CRF cannot be reduced with rest and drastically worsens quality of life. CRF is also correlated with a high prevalence of mental health issues, including anxiety, depression, and sleeping disorders, in patients with advanced cancer (8–10). Moreover, according to patients’ self-reports, CRF hinders many patients from fully adhering to potentially curative therapies (11).

CRF is diagnosed based on questionnaire evaluation or self-report, so it is understudied; among thousands of patients with cancer in the Yale-New Haven Hospital system, for instance, only two have been diagnosed with CRF according to a search of deidentified patient data in the Epic Slicer Dicer database. Underdiagnosis of CRF renders targeted therapeutic strategies challenging to identify (12). Current CRF treatments include nonpharmacological interventions, such as yoga (13, 14), acupuncture (15), and dietary supplements (16), that exclusively target symptoms of fatigue. Evidence indicating significant beneficial effects of nonpharmacological treatments, however, is limited (17). Pharmacological interventions, such as antidepressants and hematopoietic growth factors, have been tested in clinical trials in patients with CRF, yet evidence of clinically meaningful efficacy is generally lacking, and increased adverse effects have been documented (18, 19). Therefore, there is an urgent need to develop new pharmacological treatments for CRF. Such treatments would likely increase the quality of life and treatment adherence in patients with late-stage cancer.

Some propose that in patients with late-stage cancer, elevated peripheral inflammatory cytokines affect the central nervous system and induce fatigue (20, 21). However, one recent study suggests that CRF (as measured by decreased wheel-running in mice) precedes an increase of IL1ββ, an inflammatory cytokine, in the brain. Moreover, anti-inflammatory drugs do not eliminate CRF (22). Others suggest that CRF may be caused by blood-loss-related anemia, chemotherapy-induced neutropenia, hypothalamic-pituitary axis-mediated neuroendocrine perturbation, and disrupted circadian rhythms (3). Despite the seeming plausibility of these proposed mechanisms, pharmacological treatments developed based on these possible mechanisms have failed to show significant efficacy in treating CRF.

Metabolic changes during tumor progression could contribute to CRF but have yet to be explored in depth (23). Most attention to cancer-related metabolic changes has been focused on tumor tissue and to a lesser extent immune cells (24–26), neglecting potential metabolic changes within tissues not directly related to the tumor, such as skeletal muscle, as well as systemic metabolic changes. Lactate is closely associated with muscle fatigue (27, 28). Although whole body lactate turnover, likely produced by the tumor, is elevated in patients with cancer (29, 30), its impact on CRF remains unknown. In this study, we present a new therapeutic option for CRF using dichloroacetate (DCA), a small molecule inhibitor of pyruvate dehydrogenase kinase, and consequently, an activator of entry of carbons derived from glucose (pyruvate per se) into the tricarboxylic acid (TCA) cycle (31). DCA has been used in multiple clinical trials to reduce tumor progression, as many tumors are thought to switch their main energy resource from oxidative phosphorylation to glycolysis (i.e., the Warburg effect) (32). Although clinical trials using DCA have generally failed to show a survival benefit in patients with cancer (33–35), studies have not focused on the impact of the lactate-lowering effect of DCA on CRF in patients with cancer (36–38), and thus may miss a potentially beneficial on-target effect not directly related to tumor progression.

In this study, we propose a new therapeutic use for DCA in treating cancer-related fatigue. We used two tumor-bearing mouse models to demonstrate that DCA improves muscle performance and motivation during tumor progression and mitigates the decline in physical function following treatment with low-dose chemotherapy. This study identifies a potential practice-changing approach, harnessing metabolic adjuvant therapy to treat cancer-related fatigue.

METHODS

Cells

YUMMER1.7 cells were a generous gift from the Bosenberg Lab and were maintained in Dulbecco’s modified Eagle medium (MEM)/Nutrient Mixture F-12 (DMEM/F-12) containing 10% fetal bovine serum, 2.5 mM glutamine, 0.5 mM sodium pyruvate, 1,200 mg/L sodium bicarbonate, 0.1 mM nonessential amino acids, and penicillin/streptomycin. MC38 cells were purchased from Kerafast (Cat. No. ENH204-FP) and maintained in Dulbecco’s modified MEM with 10% fetal bovine serum, 2.5 mM glutamine, 0.1 mM nonessential amino acids, 1 mM sodium pyruvate, 10 mM HEPES, and penicillin/streptomycin. Cells were injected into mice when passage number was <20.

Mice

The Yale Institutional Animal Care and Use Committee approved all animal studies. We purchased male C57bl/6J mice from Jackson Laboratories at 8 wk of age. After 1 wk of acclimation to our animal facility, we injected mice subcutaneously in the left chest (or right flank for the forced swim test) with 5 × 10⁵ YUMMER1.7 or 1 × 10⁶ MC38 tumor cells. A blinded investigator measured tumor size three times a week starting from 1 wk after tumor implantation using calipers. Mice were randomly assigned to control or dichloroacetate (DCA) groups, and in some cases, to immunotherapy or chemotherapy, before studies. DCA was administered through drinking water (300 mg/L) immediately following tumor injection. Anti-PD1 immunotherapy [InVivoMAb anti-mouse PD-1 (CD279), BioXCell Cat. No. BE0273] was intraperitoneally injected three times a week (10 mg/kg) starting from day 13 after YUMMER1.7 tumor cell injection. Chemotherapy (5-fluorouracil, Sigma Cat. No. 343922-1GM) was intraperitoneally injected three times a week (100 mg/kg) starting on day 12 after tumor injection in MC38 tumor-bearing mice. For infusion studies, surgery was performed under isoflurane anesthesia to implant catheters in the jugular vein 15 days after tumor implantation. In accordance with regulatory rules from Yale Institutional Animal Care and Use Committee, mice were euthanized when tumor size reached 2 cm³ or mice had another severe health issue such as tumor ulceration.

Grip Strength Test

Using a grip strength meter (Columbus Instruments), grip strength tests were performed 1 wk and 3 wk after tumor injection. Each mouse was gently pulled by its tail while both forelimbs solidly gripped the sensor and the torso remained horizontal. The maximal grip strength shown on the screen was recorded and the trial was repeated three times for each mouse.

Maximal Speed Test

Maximal speed tests were performed on the Exer 3/6 Treadmill (Columbus Instruments). Mice were acclimated to the treadmill using a 3-day protocol beginning 3 days before the first test day and reacclimated before the subsequent test: on the first day, they explored a stationary treadmill for 5 min. On the two subsequent days, the treadmill speed was set to 10 m/min for a 10 min run. For the test, mice started running at 10 m/min, with an increase in speed by 1 m/min each minute. The mouse’s running capacity was recorded as the maximum speed at which the mouse could remain on the treadmill for 5 consecutive seconds without touching the shock pad. This test was performed both 1 and 3 wk after tumor injection.

V̇o₂ Peak Test

V̇o₂ peak tests were performed using metabolic treadmills 1 wk before, 1 wk after, and 3 wk after tumor injection. Using a similar method as described for the maximal speed test, after 3 days of acclimation, mice started running at 6 m/min followed by an increase of 1 m/min per minute until exhaustion (defined as unable to stay on the treadmill for 5 consecutive seconds). V̇o₂ and V̇co₂ were measured by the ZIRCONIA oxygen sensor (Columbus Instruments) and carbon dioxide sensor (Columbus Instruments). The respiratory exchange ratio (RER) was calculated as:

$$RER=\hspace{0.17em}\frac{\dot{\mathrm{V}}{\mathrm{C}\mathrm{o}}_2}{\dot{\mathrm{V}}{\mathrm{o}}_2}$$

Forced Swim Test

The motivation for movement was evaluated by forced swim test as described by Can et al. (39). In short, mice were placed in room temperature water sufficiently deep to prevent contact of their tails or limbs with the container base. Their behavior was recorded for 6 min after placed in water. After 2 min of acclimation, mobility time was manually analyzed by two blinded investigators with stopwatches from the last 4 min of recording. Original videos are accessible from the corresponding author upon request.

Lactate Injection and Lactate Measurement

Sodium lactate (1 g/kg) was intraperitoneally injected into nonfasting healthy control mice. Plasma samples (10 μL) were obtained from mouse tail blood at 0, 5, 15, and 30 min after injection, followed by centrifugation at 12,000 rpm for 30 s. Plasma lactate concentration was measured by 2500 Biochemistry Analyzer (YSI). Measurements of grip strength, maximum running speed, and forced swim test following protocols mentioned above were started 5 min and finished within 30 min after sodium lactate injection, during which plasma lactate concentration remained high. Similarly, sodium lactate (1 g/kg) was injected into DCA-treated late-stage tumor-bearing mice and physical performance tests were conducted as previously described, with the control group injected with the same volume of PBS. In the studies to measure the impact of lactate injection on muscle amino acids, mice were injected with the same dose of sodium lactate and euthanized, with quadriceps freeze-clamped 20 min after injection.

Skeletal Muscle Isolation, Mitochondrial Membrane Potential, and Fiber Type Measurement

Skeletal muscle fibers were obtained from mouse quadriceps femoris muscle 3 wk after tumor injection. Muscle tissues were minced on ice and digested in DMEM with 2 g/L collagenase type II (MP Biomedicals, Cat. No. 100502) at 37°C for 1 h. Isolated cells were collected through 70-μm cell strainer by centrifuge. Mitochondrial membrane potential was measured by staining with a TMRE-Mitochondrial Membrane Potential Assay Kit (abcam, Cat. No. ab113852) according to the manufacturer’s protocol and detected by BD LSRII flow cytometry in the Yale Flow Cytometry Facility. Skeletal muscle TNNI1 and TNNI2 content was measured by ELISA (Aviva Systems Biology OKEH05236 and OKCD08424, respectively).

Ad Libitum Wheel Running

Mice were singly housed for 1 wk for acclimation with access to running wheel (Columbus Instruments), chow diet, and control or DCA-containing drinking water before being injected with YUMMER1.7 tumors. Mice were maintained in running wheel cages for 3 wk before being euthanized for inferior vena cava blood to be used in complete blood count experiments.

Metabolic Cages

Mice were placed into the CLAMS apparatus from Columbus Instruments on day 5 (early-stage) or day 19 (late-stage) after tumor injection. Mice were provided with chow diet and control or DCA-containing drinking water. The first 48-h data were not analyzed to avoid any confounding effects of acclimation to the metabolic cages.

Flux Studies

Mice underwent a primed (3X)-continuous infusion of [U-¹³C₆] glucose (1 mg/kg body wt/min) for 5 min, followed by 115 min of continuous (1X) infusion. Stable isotopes were obtained from Cambridge Isotopes (Tewksbury, MA). Within 30 s of euthanizing animals with IV Euthasol (Covetrus; Portland, ME), we isolated and freeze-clamped tumors and muscle tissue in liquid nitrogen.

Tracer Bolus Study

Stable isotopes were obtained from Cambridge Isotopes (Tewksbury, MA). Mice were gavaged with [U-¹³C₆] glucose (2 g/kg body wt) after 6-h fasting. After 30 min, mice were euthanized by isoflurane. Within 1 min, we obtained IVC blood and isolated and freeze-clamped tumors and muscle tissue in liquid nitrogen.

Metabolite extraction.

To extract metabolites, tissue samples were transferred to 2-mL round-bottom Eppendorf Safe-Lock tubes on dry ice. Samples were then ground into powder with a cryomill machine (Retsch) for 30 s at 25 Hz and maintained at a cold temperature using liquid nitrogen. For every 20-mg tissue, 800 μL of −20°C 40:40:20 methanol:acetonitrile:water (extraction solvent) was added to the tube, vortexed for 10 s, and then centrifuged at 21,000 g for 20 min at 4°C. The supernatants were then transferred to plastic vials for LC-MS analysis. A procedure blank was generated identically without tissue, which was used later to remove the background ions.

Metabolite measurement by LC-MS/MS.

A quadrupole-orbitrap mass spectrometer (Q Exactive Plus, Thermo Fisher Scientific, San Jose, CA) operating in negative mode was coupled to hydrophobic interaction chromatography (HILIC) via electrospray ionization. Scans were performed from 70 to 1,000 m/z at 1 Hz and 140,000 resolution. LC separation was on a XBridge BEH Amide column (2.1 mm × 150 mm × 2.5 mm particle size, 130 Å pore size; Water, Milford, MA) using a gradient of solvent A (20 mM ammonium acetate, 20 mM ammonium hydroxide in 95:5 water:acetonitrile, pH 9.45) and solvent B (acetonitrile). Flow rate was 150 mL/min. The LC gradient was: 0 min, 85% B; 2 min, 85% B; 3 min, 80% B; 5 min, 80% B; 6 min, 75% B; 7 min, 75% B; 8 min, 70% B; 9 min, 70% B; 10 min, 50% B; 12 min, 50% B; 13 min, 25% B; 16 min, 25% B; 18 min, 0% B; 23 min, 0% B; 24 min, 85% B. Autosampler temperature was 5°C, and injection volume was 15 μL.

For improved detection of fructose-1,6-bisphosphate and 3-phosphoglycerate, selected ion monitoring (SIM) scans were added. For 3-phosphoglycerate and fructose-1,6-bisphosphate, scans were performed from 180 to 190 m/z and 336 to 350 m/z, respectively, from 13–15 min of the 25-min gradient run with 70,000 resolution and maximum IT of 500 ms and AGC target of 3e6. Data were analyzed using the El-MAVEN (Elucidata) software. For tracer experiments, isotope labeling was corrected for ¹³C natural abundances using the AccuCor package (41).

Cytokine Analysis

Blood from the inferior vena cava was obtained on day 21 after YUMMER1.7 tumor injection and centrifuged for 30 s at 13,000 rpm, then immediately frozen down in −80°C freezer. Samples were sent to Eve Technologies and tested with the Mouse Cytokine/Chemokine 31-Plex Discovery Assay Array (MD31).

Complete Blood Count

To assess the impact of DCA on the complete blood count in mice, 3 wk after tumor cell injection, 100-µL blood from the inferior vena cava was obtained, immediately transferred to EDTA-coated Eppendorf tubes, and mixed with 10-µL 0.5 M EDTA under room temperature. Complete blood count was conducted on a veterinary hematology analyzer (Drew Scientific, HEMAVET 950FS), provided by the Cooperative Center of Excellence in Hematology at the Yale School of Medicine.

Statistical Analysis

We used the two-tailed unpaired Student’s t test to compare two groups and ANOVA with Tukey’s multiple comparisons test to compare three or more groups in GraphPad Prism Version 9. A P value <0.05 was deemed significant. In all cases, we verified that the data met the assumptions of the statistical test used. All measurements shown are from distinct samples (i.e., animals), rather than repeated measurements from the same sample.

RESULTS

DCA Preserves Muscle Strength in Mice with Late-Stage Melanoma

In this study, we used mice with xenograft YUMMER1.7 murine melanoma (42). Mice were treated with 300 mg/L DCA via drinking water (∼30 mg/kg DCA per day) starting from day 0 of tumor injection. DCA did not decelerate YUMMER1.7 tumor growth (Fig. 1A), indicating that any effects of DCA on behavior are not attributable to a smaller tumor. However, in replicable experiments, we consistently observed that DCA-treated mice were more active than untreated mice, with activity rates almost equivalent to those of healthy mice. This observation continued until mouse tumors reached a humane endpoint at which point euthanasia was required (Supplemental Videos S1, S2, and S3). We, therefore, hypothesized that DCA might be used to treat CRF and preserve physical performance during late-stage cancer.

Figure 1.

DCA has no effect on YUMMER1.7 tumor growth but preserves physical performance in mice with advanced-stage tumors. A: YUMMER1.7 tumor volume. DCA treatment started on day 1 after tumor injection (n = 8). B: grip strength (n = 8). C: maximal running speed (n = 10). D: peak oxygen uptake (V̇o_2peak) before tumor injection, 1 wk after tumor injection and 3 wk after tumor injection (n = 5). E: daily voluntary running distance in home cages containing ad libitum running wheels or immovable wheels (n = 8). Tumor cells were injected after 6-day acclimation to the cages. F: forced swimming time at the early and late stages of tumor progression (n = 10). In all panels, SE is shown; n represents number of animals; *P < 0.05, ***P < 0.001, ****P < 0.0001). DCA, dichloroacetate.

We first verified that these tumor-bearing mice did not have cachexia, another late-stage cancer complication that might cause fatigue (43). In contrast to cachexia, we observed a body mass increase in both DCA-treated and untreated mice (Supplemental Fig. S1A). Body composition measurements showed no difference in lean mass at both early and late stages of tumor development, and the increases in lean mass in both groups corresponded to estimated tumor mass (Supplemental Fig. S1B). Fat mass decreased in untreated mice, but not in the DCA-treated group (Supplemental Fig. S1C). Moreover, by directly measuring muscle mass, we found that the quadriceps muscle weight was identical among healthy control, treated tumor-bearing, and untreated tumor-bearing mice (Supplemental Fig. S1D). Together, these data demonstrate that cachexia cannot explain the CRF phenotype observed in our late-stage tumor-bearing mice.

We then conducted multiple noninvasive behavioral tests to study muscle function in these mice. We observed significantly higher grip strength in DCA-treated mice than untreated mice at late stages of tumor progression (Fig. 1B). The different grip strength between treated and untreated groups on day 9 in Fig. 1B was not observed in subsequent experiments (Fig. 4B) and may be a result of random grouping. We then conducted a maximal running speed test and found that DCA-treated mice have an average decrease in their maximal speed of just 13%, as compared with their early-stage measurements, whereas untreated mice have an average decrease of 31% (Fig. 1C). We also measured peak oxygen consumption (V̇o_2peak), which is the most conclusive and translational readout of overall fitness available in mice. We found that V̇o_2peak did not differ at an early stage compared with the pretumor stage in either the treated or untreated groups. V̇o_2peak did, however, significantly decrease at a late stage only in the control group but not in the DCA-treated group (Fig. 1D). The respiratory exchange ratio when V̇o_2peak was achieved was not altered at any stage in either group (Supplemental Fig. S2A), demonstrating that both groups of mice were exercising to a similarly metabolically challenging extent. These behavioral studies indicate that late-stage YUMMER1.7 tumor-bearing mice experience CRF and exhibit decreased muscle function, but DCA preserves their physical performance.

DCA Preserves Motivation for Movement in Mice with Late-Stage Melanoma

Both the literature and our previous study suggest that animals with late-stage tumors significantly reduce voluntary movement (22). Therefore, we examined whether DCA would preserve motivation for movement. We first used a wheelcage to study daily voluntary running. We observed higher daily running distance in DCA-treated mice than untreated mice with late-stage tumors only, but no differences were observed before tumor injection or before tumors were palpable (Fig. 1E and Supplemental Fig. S3, A and B). We then conducted the gold standard test of motivation in animal models of depression-like behavior: the forced swimming test. We observed fully conserved motivation (as measured by time spent actively swimming) in DCA-treated mice with late-stage tumors, but a 36% decrease in swimming time in untreated mice (Fig. 1F). These results demonstrate that DCA can preserve motivation for movement despite unchanged tumor progression.

DCA Does Not Affect Food/Water Intake in Tumor-Bearing Mice

We verified that DCA does not alter physical performance in healthy control mice (Supplemental Fig. S4, A–C), indicating that the effect of DCA is exclusive in tumor-bearing mice. We then conducted studies to identify the underlying mechanism of DCA’s effect on CRF. As nutrient deficiency is considered one possible cause of CRF (16), we used metabolic cages to study the effect of DCA on food and water intake, as well as other metabolic parameters, at the early and late stages of tumor progression. As expected, based on its mechanism of action, we observed a significant increase in the respiratory exchange ratio in DCA-treated mice (Fig. 2A) as compared with untreated mice throughout the day, suggesting that DCA is successfully delivered and enhances glucose oxidation systemically. Food intake, however, was not different between treated and untreated groups at all stages of tumor progression (Fig. 2B). Water intake increased at the late tumor stage but remained similar between untreated and DCA-treated groups (Fig. 2C). These results indicate that DCA does not preserve muscle strength by promoting food and/or water intake. Similarly, and as predicted by the lack of difference in food intake or fat mass between late-stage DCA-treated and control mice, plasma insulin concentrations were not different between DCA-treated and control animals at a late stage of tumor progression (Supplemental Fig. S4D).

Figure 2.

DCA neither alters food/water intake nor improves systemic inflammation or anemia. Basal respiratory exchange ratio (A), ad libitum food (B), and water intake (C) of control and DCA-treated mice with early- and late-stage tumors, measured using metabolic cages (n = 8). D: plasma concentrations of inflammatory cytokines (n = 6). E: white blood cell counts (WBC, white blood cell; NE, neutrophil; LY, lymphocyte; MO, monocyte; EO, eosinophil). F: red blood cell number in untreated and DCA-treated tumor-bearing mice (n = 8). SE is shown; n represents number of animals; **P < 0.01, ****P < 0.0001. DCA, dichloroacetate.

DCA Does Not Affect Systemic Inflammation or Anemia

One common explanation of CRF is that cancer causes systemic inflammation (20). Our measurements of peripheral inflammatory cytokines from tumor-bearing mice blood plasma, however, revealed little difference in inflammatory cytokine concentrations among DCA-treated and untreated groups (Fig. 2D). This suggests DCA does not treat CRF by reducing inflammation. To be noted, we observed a twofold increase in concentrations of granulocyte colony-stimulating factor (G-CSF), which stimulates granulocyte production and release into the bloodstream (44, 45) and reduces inflammation (46), in DCA-treated mice compared with untreated mice. However, the results from complete blood count additionally indicated identical leukocyte numbers in DCA-treated and untreated groups (Fig. 2E), further suggesting a mechanism other than alterations in inflammation or blood cell counts. Anemia has been proposed as another possible cause of fatigue (7). However, we observed no difference in erythrocyte numbers between DCA-treated and untreated mice in complete blood count (Fig. 2F), ruling out the possibility that DCA treatment improves CRF by mitigating anemia in tumor-bearing mice.

DCA Alters Metabolic Patterns in Muscle Tissue

Tumor progression may also induce metabolic changes that affect physical performance. We conducted stable isotope tracer studies using [U-¹³C₆]-glucose in tumor-bearing mice and measured the impact of DCA on the fractional contributions of various precursors to products in muscle tissue at steady state. We found that, regardless of DCA treatment or tumor progression, the fractional contribution of circulating glucose to muscle glycolytic intermediates (3-phosphoglycerate and pyruvate) remained identical among different groups (Fig. 3, A–C). As expected, the contribution of glucose to TCA cycle intermediates was increased by DCA treatment in the early-stage group (Fig. 3, D and E). However, the contribution of pyruvate to TCA cycle intermediates was enhanced in both late-stage groups independent of DCA treatment (Fig. 3F), implying metabolic alterations in nontumor tissue during cancer progression but arguing against metabolic alterations in muscle as a primary explanation for the improvements in muscle function observed in DCA-treated mice. Consistent with this hypothesis, we observed no difference in muscle fiber type, as indicated by equal troponin type I (TNNI1) and type II (TNNI2) concentrations in skeletal muscle (Supplemental Fig. S6, A and B). Moreover, given that results from infusion studies represent muscular metabolism in the resting state only, we also used a [U-¹³C₆]-glucose tracer bolus to investigate glucose utilization in muscle tissue. We found no significant differences between enrichments of labeled glycolytic intermediates except a statistically but likely not physiologically significant decrease in m + 3 pyruvate in late-stage DCA-treated mice compared with early-stage DCA-treated mice (Fig. 3, G and H). The enrichment of TCA cycle intermediates remained similar among different groups as well (Fig. 3I). Notably, we observed a significant increase in ¹³C enrichment of m + 2 glutamate as compared with other TCA cycle intermediates. This may result from tissue contamination from blood or interstitial fluid, as it has been previously shown that blood has higher m + 2 glutamate compared with muscle (47).

Figure 3.

DCA preserves muscle mitochondrial membrane potential and enhances metabolic flux through TCA cycle. A: tracer labeling in glycolysis with [U-¹³C₆]-glucose as tracer. Adapted from “glycolysis,” by BioRender.com (2022), retrieved from https://app.biorender.com/biorender-templates. ¹³C enrichment of glycolytic intermediates (B) and fractional contribution of glucose to pyruvate (a readout of glycolysis) in muscle tissue (C). D: tracer labeling scheme reflecting the first turn of the TCA cycle infused with [U-¹³C₆] glucose (n = 5). Adapted from “Kreb’s [sic] Cycle,” by BioRender.com (2022). Retrieved from https://app.biorender.com/biorender-templates. Enrichment of TCA cycle intermediates (E) and fractional contribution of m + 3 pyruvate to m + 2 citrate/isocitrate in muscle tissue from [U-¹³C₆] glucose-infused tumor-bearing mice (F). Enrichment of glycolytic intermediates (G), fractional contribution of pyruvate to the TCA cycle (H), and TCA cycle intermediates in skeletal muscle (I) obtained from tumor-bearing mice gavaged with [U-¹³C₆] glucose (n = 5). The mean is presented with SE; n represents number of animals; *P < 0.05, DCA, dichloroacetate.

We did, however, observe differences in the absolute concentrations of free amino acids in muscle tissue. We identified increases in concentrations of arginine, asparagine, leucine, lysine, methionine, and threonine when comparing late-stage to early-stage in untreated groups (Supplemental Fig. S5, A–F), consistent with increased proteolysis and/or less amino acid utilization in muscle tissue during advanced-stage tumor progression. However, such an increase was not observed in DCA-treated mice. We also discovered decreased alanine concentration in the late-stage DCA-treated group compared with the untreated group (Supplemental Fig. S5G), which may result from shunting pyruvate (and, therefore, the metabolites with which it equilibrates: alanine and lactate) into the TCA cycle. Such changes were not observed in DCA-treated tumor-bearing mice injected with sodium lactate to induce transient increases in lactate concentrations (Supplemental Fig. S6C).

DCA Treats CRF by Reducing the Circulating Lactate Concentration

In addition to measuring substrate fractional contributions to glycolysis and to the TCA cycle in muscle tissue, we also measured systemic metabolic changes in these tumor-bearing mice. In particular, we found that untreated tumor-bearing mice had elevated circulating lactate concentration in plasma at advanced tumor stage in the resting state, whereas such changes were not found in DCA-treated mice (Fig. 4A). Isotopic tracer studies using [U-¹³C₆]-glucose also showed increased m + 3 plasma lactate enrichment in late-stage untreated mice compared with early-stage, but this change was not observed in DCA-treated mice (Fig. 4B). These data are consistent with the effect of DCA to shunt pyruvate and lactate, with which pyruvate rapidly equilibrates, into the TCA cycle. Next, we conducted an intervention to investigate if preventing reductions in lactate concentrations reversed the effect of DCA on physical function. We found that, in DCA-treated late-stage tumor-bearing mice, the maximal running speed (Fig. 4C) and motivation for movement (Fig. 4D) were both decreased following an acute injection of lactate (Supplemental Fig. S7). However, there were no differences in grip strength between two groups (Fig. 4E). Similar results were observed in healthy control mice injected with sodium lactate (Fig. 4, F–H), which further validated that increased plasma lactate level altered physical performance, likely dependent on an effect of lactate to impair motivation considering that lactate injection did not alter grip strength, while markedly improving both swimming time and maximal running speed.

Figure 4.

DCA preserves physical function by reducing circulating lactate concentration in tumor-bearing mice. A: plasma lactate concentration in tumor-bearing mice (early control n = 10; early DCA n = 11; late control n = 12; late DCA n = 12). B: ¹³C lactate enrichment (m + 3/m + 0) from plasma obtained at 120 min of infusion study using [U-¹³C₆]-glucose as tracer n = 5, with the exception of early DCA for which n = 7. Maximum running speed (C), forced swim time (D), and grip strength (E) measured within 30 min after intraperitoneal sodium lactate injection in DCA-treated tumor-bearing mice with late-stage tumor (for C, D, E, G, and H, n = 5 per group in both groups; for F, n = 4 control and n = 5 lactate). Maximum running speed (F), forced swim time (G), and grip strength (H) measured within 30 min after intraperitoneal sodium lactate injection in healthy control mice. The mean is presented with SE; n represents number of animals *P < 0.05, **P < 0.01. DCA, dichloroacetate.

DCA Relieves Oxidative Stress in Muscle Tissue

We observed a decrease in NADH/NAD+ ratio in muscle tissue from late-stage DCA-treated mice compared with early-stage (Supplemental Fig. S8A), indicating less oxidative stress in muscle cells. This change was not found in untreated mice. Metabolomics data also suggested that DCA could increase the lactate/pyruvate ratio in muscle tissue (Supplemental Fig. S8B), though not significantly, which may suggest a less oxidative microenvironment. Meanwhile, DCA also preserves mitochondrial function; using TMRE (tetramethylrhodamine, ethyl ester), we found a significant increase in muscle cell mitochondrial potential in untreated mice, which may induce excessive reactive oxygen species (ROS) production. DCA treatment reversed this increase (Supplemental Fig. S8C). In addition, concentrations of serine and glycine, precursors for antioxidants, were decreased in muscle of late-stage DCA-treated mice compared with early-stage, but not different in untreated mice (Supplemental Fig. S5, H and I). These findings suggest that in addition to metabolic changes, DCA may also preserve physical function by reducing oxidative stress in muscle tissue.

DCA Relieves Treatment-Induced CRF in Murine Models of Both Melanoma and Colon Cancer

Clinical investigations have suggested that traditional cancer treatments (e.g., chemotherapy, immunotherapy, and surgery) may worsen CRF in patients with multiple types of cancers (48, 49). Therefore, we combined DCA with other treatments to examine whether DCA could reduce the decrement in physical performance induced by cancer treatment. Immunotherapy is a standard first-line treatment for advanced melanoma (50). In this study, we treated YUMMER1.7 melanoma tumor-bearing mice with anti-PD1 treatment and/or DCA starting at day 13, to mimic the typical treatment start time in human patients with advanced tumors. Initiation of treatment, once tumors are established, reduces the efficacy of anti-PD-1 in inhibiting tumor growth, so any differences in behavior or physical function could not be attributed to differences in tumor size. We found that without altering tumor volumes (Fig. 5A), mice treated with only immunotherapy exhibited reduced grip strength (Fig. 5, B and C) and maximal speed (Fig. 5, D and E) compared with untreated mice. In addition, immunotherapy tended to compromise motivation for movement: immunotherapy-treated mice showed a 39% reduction in forced swim time compared with a 34% reduction in the control group (Fig. 5, F and G). However, adding DCA treatment could at least partially restore the immunotherapy-induced muscle function loss and preserved motivation for movement (Fig. 5, B–G).

Figure 5.

DCA relieves immunotherapy-induced CRF in YUMMER1.7 tumor-bearing mice. A: YUMMER1.7 tumor volume on day 18 in mice treated with immunotherapy (anti-PD1 monoclonal antibody) and/or DCA treatment. DCA treatment started on day 1 and immunotherapy started on day 12 after tumor injection. Grip strength (B), maximum running speed (D), and forced swimming time (F) measured at the early and late stage of tumor growth, and the percentage of decreases (C, E, and G), respectively. SE is shown; *P < 0.05, **P < 0.01, ***P < 0.001. n = 5 animals. CRF, cancer-related fatigue; DCA, dichloroacetate.

Chemotherapy is another common treatment for multiple cancer types that can induce CRF. Therefore, we studied the combination of DCA and chemotherapy in MC38, a murine colon cancer model. We started DCA treatment, together with a low dose of 5-fluorouracil, a common chemotherapy for colon cancer (51), on day 13, at which point tumor volumes were ∼1,000 mm³. Although low-dose chemotherapy did not significantly reduce MC38 tumor volume (Fig. 6A), chemotherapy alone reduced muscle function in late-stage tumor-bearing mice, as shown by a 20% decrease in grip strength (Fig. 6, B and C) and a 29% decrease in maximal speed (Fig. 6, D and E). This reduction could be rescued with DCA treatment without changing tumor volumes. Finally, we observed an increase in motivation for movement in DCA-treated mice treated with chemotherapy, in contrast to a significant decrease in untreated or chemotherapy-treated mice (Fig. 6, F and G). Taken together, these results suggest that without impairing efficacy of treatments or accelerating tumor growth, DCA holds promise to improve the adverse effects of CRF induced by immunotherapy or chemotherapy in murine models of melanoma and colon cancer, respectively.

Figure 6.

DCA relieves chemotherapy-induced CRF in MC38 tumor-bearing mice. A: MC38 tumor volume in chemotherapy (5-fluorouracil) and/or DCA-treated on day 18 after tumor injection. DCA treatment started on day 1 and chemotherapy started on day 13 after tumor injection. Grip strength (B), maximal running speed (D), and forced swimming time (F) measured at the early and late stage, and the percentage of decreases (C, E, and G), respectively. SE is shown; *P < 0.05, **P < 0.01, ****P < 0.0001. n = 5 animals. CRF, cancer-related fatigue; DCA, dichloroacetate.

DISCUSSION

CRF is one of the most debilitating effects of cancer. CRF compromises patient’s quality of life and reduces adherence to treatment (7, 11), but has received little attention in basic research. In preclinical research, choosing a multimodal strategy to assess CRF is essential, as asking animals how they feel is impossible. Therefore, we conducted multiple noninvasive behavioral tests to assess CRF during tumor progression. We acknowledge that our results do not specifically reflect muscle contractility as most are related to motivation for movement, with the possible exception of the grip strength test. However, as CRF is defined as patients lacking energy (7) and to our knowledge no evidence in humans has shown decreased muscle contractility in CRF, we believe that the current murine models are the most appropriate to reflect the patients’ condition.

Here, we reveal DCA as the first metabolism-targeting treatment for CRF. DCA was previously shown to hinder tumor growth (31); however, its impact on nontumor tissues in patients with cancer has been overlooked. We were gratified to discover that DCA preserved muscle strength and motivation for movement in mice with late-stage cancer and prevented CRF exacerbated by treatment with both chemotherapy and immunotherapy. This suggests the potential for DCA to prevent CRF alongside existing cancer treatments.

We next investigated the mechanism underlying DCA’s ability to preserve performance in tumor-bearing mice and identified elevated plasma lactate as a key factor driving CRF. Plasma lactate concentration and lactate production significantly increased in late-stage untreated mice compared with the early stage of tumor progression. DCA, previously recognized as a lactate-lowering drug during exercise in healthy individuals (36), eliminated the elevation in both plasma lactate concentration and production in late-stage treated mice, and consequently preserved physical performance.

Because exogenous lactate had limited impact on grip strength, we used stable isotope tracing techniques to study other alterations in in vivo metabolic patterns and measured fractional contributions of precursors to products in glycolysis and the TCA cycle. Not surprisingly because DCA does not regulate glycolysis, we did not observe any changes in glucose contribution to glycolytic metabolites in mice treated with DCA. As expected, the contribution of glucose to TCA cycle intermediates was increased in early-stage DCA-treated mice. However, both DCA-treated and untreated groups exhibited a similarly increased contribution of pyruvate to TCA cycle intermediates in muscle from late-stage tumor-bearing mice. This result emphasizes that in addition to metabolic alterations in tumor cells, cancer induces systemic metabolic stress and affects nontumor tissues.

In addition to the key role of muscle metabolism to provide energy, muscle also serves as a nutrient reservoir including amino acids generated from proteolysis (52). Therefore, we also measured the relative concentrations of amino acids in muscle with LC-MS/MS. As compared with early-stage tumor-bearing mice, animals with late-stage tumors exhibited increased muscle concentrations of six amino acids in the untreated group, whereas DCA-treated mice exhibited no significant differences. The increased concentrations of many amino acids observed in late-stage untreated tumor-bearing mice imply greater proteolysis in muscle from untreated tumor-bearing mice, while DCA may postpone cancer-induced muscle wasting. Moreover, we identified an increase in mitochondrial potential and hallmarks of decreased oxidative stress in muscle from tumor-bearing mice that were restored to baseline by DCA. Taken together, these results suggest that muscles in late-stage tumor-bearing mice may face greater oxidative stress, which can be reversed by DCA.

We also discovered decreased alanine concentration in late-stage DCA-treated muscle as compared with untreated mice. Alanine released from muscle serves as a substrate for gluconeogenesis. Therefore, decreased alanine in muscle may suggest increased demand for gluconeogenesis in DCA-treated tumor-bearing mice. Moreover, changes in amino acid concentrations were not observed in DCA-treated tumor-bearing mice injected with sodium lactate, which resulted in a transient increase in lactate concentration. This implies that changes in amino acid metabolism are more likely to be a long-term effect during tumor progression. Further investigation should focus on metabolic cross talk between muscle and liver in cancer. Adipose tissue may be another player in metabolic cross talk in cancer. WAT mass decreased in control but not in DCA-treated mice as cancer progressed (Supplemental Fig. S1C). Although not tested here, it is possible that these data may reflect increased WAT lipogenesis resulting from diversion of pyruvate carbons from oxidation to fat synthesis.

Taken together, this study tackles the long-understudied conundrum of CRF treatment from multiple dimensions of muscle performance and metabolism. Our data reveal the critical role of nontumor tissue metabolism in CRF induction and positions DCA as a potential therapeutic target for CRF in patients.

DATA AVAILABILITY

Supplemental material (8 Extended Data figures and 3 videos), as well as the main figures, are publicly available in FigShare: https://doi.org/10.6084/m9.figshare.22530181.

SUPPLEMENTAL DATA

Supplemental Figs. S1–S8 and Supplemental Videos S1–S3: https://doi.org/10.6084/m9.figshare.22530181.

GRANTS

This study was funded by a Young Investigator Award from the Melanoma Research Alliance (to R.J.P.). The National Institutes of Health Medical Scientist Training Program Training Grant T32GM007205 supported B.P.L. The Core is supported in part by an NCI Cancer Center Support Grant NIH P30 CA016359. The BD Symphony was funded by shared instrument grant NIH S10 OD026996.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

X.Z. and R.J.P. conceived and designed research; X.Z., W.D.L., W.Z., A.F., Z.L., R.C.G., A.A.H., and B.R. performed experiments; X.Z., W.D.L., and B.P.L. analyzed data; X.Z. and R.J.P. interpreted results of experiments; X.Z. prepared figures; X.Z. drafted manuscript; X.Z. and W.D.L. edited and revised manuscript; X.Z., W.D.L., B.P.L., W.Z., A.F., Z.L., R.C.G., A.A.H., B.R., J.D.R., and R.J.P. approved final version of manuscript.