Introduction

In healthy humans, skeletal muscle makes up to 40% of the total body mass (Janssen et al, 2000), of which 20% are constituted by proteins. Massive skeletal muscle atrophy is the hallmark of a multi-organ wasting disorder known as cachexia, which causes severe asthenia and intolerance to therapies in patients with chronic diseases such as cardiac failure, COPD, and notably cancer, leading to poor clinical outcomes (Fearon et al, 2011; Porporato, 2016). Indeed, the prevalence of cachexia reaches 80% in advanced-stage cancer patients and has been estimated to directly cause at least 20% of all cancer-related deaths (Tisdale, 2002).

In cancer cachexia, systemic alterations contribute to the uncontrollable decrease in quality of life, insulin resistance, liver dysfunction, chronic inflammation, and even altered gut microbiota and nutrient absorption (Porporato, 2016). Remarkably, iron deficiency is diagnosed in more than half of patients afflicted with colorectal, lung, and pancreatic cancers, which are also associated with high prevalence of cachexia (Ludwig et al, 2013). Chronic inflammation hampers iron absorption from the diet and causes iron retention in reticuloendothelial cells, which results in insufficient iron availability to meet the body’s needs. Iron is indeed a versatile cofactor essential to a multitude of vital metabolic processes including oxygen supply, DNA synthesis, redox homeostasis, or energy metabolism. Energy production directly depends on the availability of iron. It is indispensable for the activity of several mitochondrial enzymes involved in the TCA cycle and the electron transport chain, where iron is found under the form of heme or iron–sulfur cluster (ISC). Moreover, iron has been shown to directly regulate mitochondrial biogenesis, highlighting the sensitivity of these organelles to iron availability (Rensvold et al, 2016). Notably, both iron accumulation and iron deficiency are detrimental to mitochondrial function. Cellular iron homeostasis is thus a tightly regulated process involving a broad variety of proteins responsible for its transport (transferrin), uptake (transferrin receptor/TFR1), storage (ferritin/FT), and export (ferroportin/FPN1). The fine tuning of intracellular iron metabolism is made possible by the Iron Responsive Element/Iron Responsive Protein (IRE/IRP) system exerting a major control on the translation of several key iron-related proteins.

In the skeletal muscle, iron is particularly needed to support the high metabolic activity required for ATP generation, a requisite for contraction and movement. While mitochondrial dysfunction (in particular, decreased oxidative capacity, inefficient energy production, and altered mitochondrial dynamics) has been proven to promote skeletal muscle wasting in cachexia (Boengler et al, 2017; Abrigo et al, 2019), little is known about the consequence of altered iron levels on skeletal muscle function and mass. Importantly, iron deficiency is present in the vast majority of cancer patients and has been associated to advanced stage and poor prognosis (Ludwig et al, 2013). Hence, we decided to investigate the role of iron metabolism in cancer cachexia-related muscle wasting.

Results Iron deficiency induces skeletal muscle atrophy

Iron deficiency is highly prevalent in cancer patients and has been associated to advanced stage and poor prognosis (Ludwig et al, 2013). To assess the effects of cancer-induced iron deficiency on skeletal muscle metabolism, we analyzed the transcript levels of the main cellular iron importer, transferrin receptor 1 (TFR1) in a cohort of cancer patients presenting body weight loss and anemia (Fig 1A and B). During iron deficiency, cells normally increase iron import through TFR1 to maintain homeostasis (Camaschella et al, 2020). Surprisingly, the patients displayed decreased TFR1 (Fig 1C). To simulate the condition of iron-deficient anemia typical of cancer patients, we induced severe anemia in mice by combined phlebotomy and iron-free diet (Fig EV1A). This treatment resulted in muscle atrophy (Figs 1D and EV1B), supporting the hypothesis of an involvement of iron homeostasis in the onset of cancer-associated muscle wasting. As expected, iron deficiency promoted TFR1 upregulation in liver (Fig 1E), presumably to ensure the necessary supply of iron to the organ (Camaschella et al, 2020). However, TFR1 was downregulated in skeletal muscle of iron-deficient mice (Figs 1F and EV1C), suggesting a different response of this tissue to iron deprivation. To study the role of TFR1 expression on muscle mass, we transfected TFR1-silencing or TFR1-overexpressing plasmid by electroporating the tibialis anterior of healthy mice. TFR1 silencing was sufficient to induce fiber atrophy (Figs 1G and EV1D), while TFR1 overexpression triggered hypertrophy in the positive fibers (Fig 1H). Coherently, inhibition of iron import by silencing TFR1 induced significant myotube atrophy in vitro and decrease of the labile iron pool (Figs 1I and EV1E and F). Similarly, blocking intracellular iron mobilization by silencing NCOA4 (a cytoplasmic protein that mediates autophagic degradation of ferritin (Bellelli et al, 2016)), also caused significant myotube atrophy (Figs 1J and EV1G). Furthermore, to assess the direct impact of iron deficiency on muscle, we evaluated the effect of iron chelation on murine and human myotubes. Treatment with different iron chelators, namely deferoxamine (DFO), bathophenanthroline disulfonic acid (BPS), and apotransferrin (Apo-Tf), which is known to decrease transferrin saturation, led to a reduction in C2C12 myotube diameter and labile iron pool (Figs 1K and EV1H). Consistently, iron chelation by DFO exerted the same atrophic effect on human myotubes (Figs 1L). In summary, we found that cachectic cancer patients have decreased muscular TFR1 expression, and decreased iron availability is sufficient to induce skeletal muscle atrophy in vivo and myotube diameter reduction in vitro.

Figure 1. Iron deficiency induces skeletal muscle atrophy

A, B. Hemoglobin (A) and hematocrit (B) levels of healthy subjects and cachectic cancer patients presenting a body weight loss superior to 10% of initial body weight (19 healthy subjects, 17 cachectic patients). C. TFR1 mRNA levels in muscle biopsies from cancer patients of late stage cachexia with at least 10% total body weight loss (19 healthy subjects, 17 cachectic patients). D. Gastrocnemius weight in mice subjected to iron deprivation by feeding with an iron-deficient diet (IDD) combined to a phlebotomy (PHL) (n = 5–6). E. TFR1 mRNA levels in the liver of mice subjected to iron deprivation by feeding with an iron-deficient diet (IDD) combined to a phlebotomy (PHL) (n = 5–6). F. TFR1 mRNA levels in the gastrocnemius of mice subjected to iron deprivation by feeding with an iron- deficient diet (IDD) combined to a phlebotomy (PHL) (n = 5–6). G. Cross-sectional area of skeletal muscle fibers transfected with shSCR (scramble) and shTFR1 (n = 3–4) and representative picture of shTFR1 transfected fibers. Scale bar = 50 µm. H. Cross-sectional area of skeletal muscle fibers transfected with TFR-pHuji (n = 4) and representative picture of TFR-pHuji transfected fibers. Scale bar = 50 µm. I, J. Diameter of TFR1 (I) or NCOA4 (J) knocked down C2C12 myotubes at day 3 post-transfection (n = 7 and n = 3 respectively). K. Diameter of C2C12 myotubes after 48 h treatment with Deferoxamine (DFO), bathophenanthroline disulfonate (BPS), or apo-transferrin (Apo-Tf). L. Representative pictures and diameter measurements of human myoblast-derived myotubes after 48 h treatment with DFO (n = 3). Scale bar = 50 µm.

Data information: For all data, n represents the number of biological replicates. Statistical significance was calculated by unpaired, two-tailed Student’s t-test. Data are mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.

Source data are available online for this figure.

Click here to expand this figure.

Figure EV1. Iron deficiency induces skeletal muscle atrophy

Hematocrit in mice subjected to iron deprivation by a combination of iron-deficient diet (IDD) and phlebotomy (PHL) (n = 5–6). Gastrocnemius weight in mice subjected to iron deprivation by iron-deficient diet (IDD), phlebotomy (PHL), or a combination of both (n = 4–8). Gastrocnemius TFR1 mRNA levels in mice after iron deprivation by iron-deficient diet (IDD), phlebotomy (PHL) or a combination of both (n = 3–6). Representative Western blot of TFR1 and ferritin after transfection with shTFR1-pGFP in 3T3 cells (n = 3). Representative Western blot of TFR1 after knockdown in C2C12 myotubes (n = 3). Labile iron pool in C2C12 myotubes after TFR1 knockdown (n = 3). Representative Western blot of NCOA4 after knockdown in C2C12 myotubes (n = 3). Labile iron pool in C2C12 myotubes treated with iron chelators DFO, BPS, or apo-transferrin (n = 3).

Data information: For all data, n represents the number of biological replicates. Statistical significance was calculated by unpaired, two-tailed Student’s t-test (A, F) or one-way Anova with Bonferroni’s correction (B, C, and H). Data are mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.

Altered iron metabolism in the skeletal muscle is a feature of cancer-induced cachexia

To confirm the link between cancer cachexia and iron metabolism in the skeletal muscle, we recreated cancer cachexia in mice using the C26-colon cancer model in Balb/C mice, which led to significant hematocrit reduction, total body weight loss, and muscle mass reduction (Figs 2A and EV2A–C). In line with the human data, cachectic mice showed a drastic reduction of TFR1 in the skeletal muscle (Fig 2B and C) despite no change in liver TFR1 or hepatic iron content (Fig EV2D and E), suggesting that the regulation of iron metabolism is organ-specific. Muscle TFR1 downregulation was further confirmed in two different murine cachexia models, namely, LLC (Lewis Lung Carcinoma) and BaF3 (murine interleukin 3-dependent pro-B cell line) (Fig EV2F–K). Moreover, we assessed iron-sensing RNA-binding proteins mediating post-transcriptional regulation of iron metabolism in mammalian cells (Meyron-Holtz et al, 2004; Wang et al, 2020). We observed a decrease in cytosolic aconitase activity (hence a switch to iron-regulatory protein/IRP1) (Fig 2D), and an upregulation of iron-regulatory protein 2 (IRP2) (Fig 2E), indicating an iron-deficient state in the skeletal muscle of tumor-bearing mice. As IRP activity should drive TFR1 expression via iron-responsive element (IRE), we measured the activity of the IRE-IRP system in cachectic muscles by RNA electrophoretic mobility shift assay (REMSA). Despite the decreased aconitase function of IRP1 and the overexpression of IRP2, we observed a lower RNA-binding activity to the IRE site of ferritin (FT) in cachectic samples compared to control in native conditions (Figs 2F and EV2L). The phenotype appeared to be linked to protein oxidation. Indeed, by performing the assay in reducing condition, we evidenced the opposite pattern, with cachectic samples presenting a higher IRE-binding, suggesting an oxidative damage, which is known to negatively regulate IRP2 activity (Gehring et al, 1999). The oxidative stress in skeletal muscle of C26 tumor-bearing mice was further confirmed by upregulated protein carbonylation (Fig 2G). In addition, we observed an overexpression of FT (Fig 2H), which is in line with impaired IRP activity (Cairo et al, 1996). Coherently, cachectic muscles showed significantly increased protein-chelated iron despite no change in total iron content (Fig 2I and J).

Figure 2. Altered iron metabolism in the skeletal muscle is a feature of cancer-induced cachexia

A. Gastrocnemius weight normalized to tibial length in C26 tumor-bearing mice on day 12 post C26-injection (n = 6–7). B. TFR1 mRNA levels normalized to 18s (n = 6–7). C. TFR1 protein expression and densitometric quantification in mouse gastrocnemius (n = 5). D. Cytosolic aconitase activity in mouse quadriceps (n = 4) measured following subcellular fractionation (n = 4). E. Representative Western blot of IRP2 in mouse gastrocnemius and densitometric quantification (n = 5). F. Binding of IRE-BPs to the ferritin IRE. The biotin-labeled IRE probe was incubated with cytosolic gastrocnemius extracts from mice, in native or reducing conditions (with EDTA and DTT) (n = 3). G. Representative protein carbonylation blot and densitometric quantification in mouse quadriceps (n = 6–7). H. Representative Western blot of ferritin in mouse gastrocnemius and densitometric quantification (n = 4). I, J. ICP-MS quantification of total (I) and protein-bound (J) iron in mouse quadriceps (n = 4–5).

Data information: For all data, n represents the number of biological replicates. Statistical significance was calculated by unpaired, two-tailed Student’s t-test. Data are mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.

Source data are available online for this figure.

Click here to expand this figure.

Figure EV2. Altered iron metabolism in the skeletal muscle is a feature of cancer-induced cachexia

A. Hematocrit levels of C26 tumor-bearing mice at day 12 post C26 injection (n = 3–6). B. Body weight evolution of mice after C26 injection (n = 6). C. Quadriceps weight of C26 tumor-bearing mice normalized to tibial lenght (n = 6). D. Liver TFR1 mRNA levels normalized to 18s in C26 tumor-bearing mice (n = 5). E. Total liver iron content in C26 tumor-bearing mice (n = 11). F, G. Total body weight (F) and gastrocnemius weight (G) in LLC tumor-bearing mice (n = 5–6). H. TFR1 mRNA levels normalized to 18s in the gastrocnemius of LLC tumor-bearing mice (n = 3). I, J. Total body weight gain (I) and gastrocnemius weight (J) of BaF-transplanted mice (n = 6–8). K. TFR1 mRNA levels normalized to 18s in the gastrocnemius of BaF-transplanted mice (n = 6–8). L. Raw blots of Fig 2F RNA electrophoretic mobility shift assay (REMSA). The biotin-labeled IRE probe was incubated without (lane 1) or with cytosolic gastrocnemius extracts from CTR and C26 tumor-bearing mice, in native (lane 2) or reducing conditions (lane 4). Where indicated, unlabeled IRE probe was added in 200-fold molar excess (lanes 3 and 5) (n = 3).

Data information: For all data, n represents the number of biological replicates. Statistical significance was calculated by unpaired, two-tailed Student’s t-test (A, C-K) or ordinary two-way Anova (B). Data are mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.

Cachectic muscles are characterized by mitochondrial iron deficiency and impaired oxidative metabolism

In most cells, a major amount of iron is taken up by mitochondria for the production of ISCs and heme. In the skeletal muscle of C26 tumor-bearing mice, we found a significant reduction of mitochondrial iron and total heme content (Fig 3A and B), as well as upregulated levels of mitochondrial iron importer mitoferrin 2 (MFRN2) and of the rate-limiting enzyme of heme synthesis aminolevulinic acid synthase 2 (ALAS2) (Fig 3C and D) (Barman-Aksozen et al, 2019). Given that iron is essential for several enzymes involved in the TCA cycle and mitochondrial oxidative metabolism (OXPHOS) (Xu et al, 2013), we assessed the enzymatic activity of two iron–sulfur proteins, aconitase (ACO) and succinate dehydrogenase (SDH), and found a 50% reduction in the activity of both enzymes in cachectic muscles (Fig 3E and F). Along with these alterations, we observed a drop in mitochondrial ATP (Fig 3G) and increased AMPK phosphorylation, denoting mitochondrial dysfunction in cachectic muscles (Zhao et al, 2016) (Fig 3H). In summary, tumor-bearing mice display remarkable alterations in muscle iron metabolism coupled with mitochondrial dysfunction, which has been linked to muscle atrophy.

Figure 3. Cachectic muscles are characterized by mitochondrial iron deficiency and impaired oxidative metabolism

A. ICP-MS quantification of mitochondrial iron in mouse quadriceps (n = 6–8). B. Gastrocnemius heme content quantified by fluorescent heme assay (n = 4–6). C, D. mRNA levels of mitochondrial iron importer MFRN2 (C) and the rate limiting enzyme of heme synthesis ALAS2 (D) normalized to 18s in mouse gastrocnemius (n = 4–5). E. Aconitase activity in mouse quadriceps lysates normalized to protein content (n = 6–8). F. Succinate Dehydrogenase activity staining in transversal sections of mouse gastrocnemius and corresponding intensity quantification (n = 3–4). Scale bar = 50 µm. G. Mitochondrial ATP content in mouse quadriceps (n = 5–9). H. Representative Western blot and densitometric quantification of phospho-AMPK and total AMPK in mouse gastrocnemius (n = 6).

Data information: For all data, n represents the number of biological replicates. Statistical significance was calculated by unpaired, two-tailed Student’s t-test. Data are mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.

Source data are available online for this figure.

Iron supplementation prevents mitochondrial dysfunction and atrophy in vitro

In line with the in vivo data, we found considerably decreased mitochondrial DNA and proteins in myotubes treated with C26 conditioned medium (CM) (Fig 4A and B). To verify the hypothesis that cancer-associated iron shortage could cause mitochondrial dysfunction, a known feature of muscle atrophy (Liu et al, 2016), we supplemented atrophic C2C12-myotubes with iron. Iron supplementation fully rescued the C26 CM-induced reduction of the oxygen consumption rate (OCR) (Fig 4C–F), while showing no effect on control myotubes (Fig EV3A). Moreover, microscopic analysis confirmed that iron supplementation prevents C26-induced diameter decreased in both murine and human myotubes in vitro (Fig 4G and H). Noteworthy, although iron substantially increases the diameter of C26-treated myotubes over time, we observed no change in the fusion index, excluding a potential effect on myogenesis (Fig EV3B and C). Consistently with the C26 model, iron supplementation prevented other cancer CM- and activin A (ActA)-induced atrophy (Zhou et al, 2020) in murine myotubes (Fig EV3D and E). Similarly, normalization of iron levels using the iron–ionophore hinokitiol (Grillo et al, 2017) also rescued myotube atrophy induced by TFR1-silencing or C26 CM (Fig EV3F and G). Importantly, low-dose rotenone (complex 1 inhibitor) caused a mild atrophy that cannot be rescued by iron, indicating that the protective effects of iron are mediated by mitochondrial function (Fig 4I). Altogether, these data demonstrate that C26 CM treatment recapitulates the mitochondrial dysfunction observed in vivo (Fig 3), which can be fully rescued in vitro by iron supplementation.

Figure 4. Iron enhances mitochondrial function and prevents cancer-induced myotube atrophy

A. qPCR analysis of mitochondrial DNA (mtDNA) on nuclear DNA (nDNA) in C2C12 myotubes treated for 48 h with C26 CM and ferric citrate (n = 6). B. Western blot of mitochondrial OXPHOS respiratory complexes in C2C12 myotubes treated for 48 h with C26 CM and ferric citrate (n = 3). C–F. Profile of oxygen consumption rate OCR (C), basal OCR (D), maximal OCR (E), and OCR used for mitochondrial ATP production (F) in C2C12 myotubes after 48 h treatment with C26 CM and ferric citrate supplementation. Data normalized to protein content (n = 9–12). G, H. Representative microscopic pictures and diameter of C2C12 myotubes stained for myosin heavy chain (G) or human myoblast-derived myotubes (H) after 48 h treatment with C26 CM and ferric citrate (n = 3 per condition). Scale bars = 50 and 100 µm, respectively. I. Diameter of C2C12 myotubes treated with rotenone and ferric citrate (n = 3).

Data information: For all data, n represents the number of biological replicates. Statistical significance was calculated by unpaired, two-tailed Student’s t-test (D-E), or one-way Anova with Bonferroni’s correction (F-G). Data are mean ± SEM. **P < 0.01, ***P < 0.001 compared to control and ##P < 0.01, ###P < 0.001 compared to C26 CM-treated group.

Source data are available online for this figure.

Click here to expand this figure.

Figure EV3. Iron enhances mitochondrial function and prevents cancer-induced myotube atrophy

Profile of oxygen consumption rate OCR in C2C12 myotubes after 48 h treatment with ferric citrate (n = 6). Myotube diameter normalized to Day 0 values (n = 3). Fusion index of C2C12 myotubes treated with C26 CM and ferric citrate for 48 h (n = 7–8). Diameter of C2C12 myotubes treated with LLC CM and iron citrate for 48 h (n = 3). Diameter of C2C12 myotubes treated with Activin A (ActA) and ferric citrate for 48 h (n = 3). Diameter of TFR1-silenced C2C12 myotubes after 24 h treatment with iron ionophore hinokitiol (HNK) (n = 3). Diameter of C2C12 myotubes treated with C26 CM and HNK for 48 h (n = 3).

Data information: For all data, n represents the number of biological replicates. Statistical significance was calculated by one-way Anova with Bonferroni’s correction (A-E). Data are mean ± SEM. ***P < 0.001 compared to control and ###P < 0.001 compared to conditioned medium, ActA, or esiTFR-treated group.

Iron supplementation rescues skeletal muscle mass and mitochondrial function

To assess if iron supplementation could prevent cancer-induced muscle atrophy in vivo, C26 tumor-bearing mice were i.v. treated with ferric carboxymaltose (FeCM) every 5 days post C26 injection. Remarkably, intravenous injections of iron resulted in healthier (smooth fur, no orbital discharge, no humpback) and more physically active mice that survived far beyond the usually fatal 2 weeks (Fig 5A and B). Of note, iron improved notably the grip strength within 24 h (Fig EV4A) and the protection was preserved until the end-point of the experiment (Fig 5C), without affecting hematocrit (Fig EV4B). In addition, the loss of body weight and muscle mass occurring at day 12 after C26 injection was prevented in FeCM-treated mice (Figs 5D–F and EV4C). Noteworthy, iron supplementation downregulated TFR1 in the tumor without affecting tumor growth (Figs 5E and EV4F). Consistently, immunostaining for myosin heavy chain revealed larger muscle fibers in the gastrocnemius of FeCM-treated mice, especially in the fast-twitch fibers, the most susceptible to atrophy (Fig 5G, in green). The protection from atrophy was further confirmed by the cross-sectional area (CSA) distribution, showing a shift toward bigger areas in FeCM-treated mice compared to C26 tumor-bearing, untreated animals, and the average CSA (Figs 5G and H, and EV4D and E). These findings were further reinforced by a significant drop of FBXO32 (ATRO1), TRIM63 (MURF1), and DDIT4 (REDD1) mRNA levels, which are indicators of skeletal muscle atrophy (Fig 5I–K).

Figure 5. Iron supplementation prevents cancer-induced cachexia

A. Representative images of C26 tumor-bearing mice receiving saline solution (left) or FeCM 15 mg/kg I.V. injection (right) taken at day 12 after C26 injection. B. Kaplan–Meier depicting the survival of C26 tumor-bearing mice after I.V. injection of saline or iron every 5 days post C26-injection (3-month-old Balb/C, n = 8–11). C. Grip strength of mice measured at day 12 post C26 injection, normalized to average strength of the control group (n = 5–9). D. Final body weight of C26 tumor-bearing mice after iron supplementation at day 12 post C26 injection (n = 5–12). E. Final weight of total tumor mass extracted from mice after sacrifice (n = 17). F. Gastrocnemius weight normalized to tibial length of C26 tumor-bearing mice after iron supplementation at day 12 post C26 injection (n = 5–12). G, H. Immunofluorescent staining of myosin heavy chain fast (green) and slow (red) of transversal sections of gastrocnemius (midbelly) with corresponding frequency distribution (G) and average cross-sectional areas (H) (n = 3–5). Scale bar = 100 µm. I–K. mRNA levels of Murf 1 (I), Atrogin 1 (J), and REDD1 (K) normalized to GAPDH in gastrocnemius (n = 5–14).

Data information: For all data, n represents the number of biological replicates. Statistical significance was calculated by one-way Anova with Bonferroni’s correction, or chi-square test for the survival curves (B). Data are mean ± SEM. *P < 0.05 and ***P < 0.001 compared to control and #P < 0.05, ##P < 0.01, ###P < 0.001 compared to C26-untreated group.

Source data are available online for this figure.

Click here to expand this figure.

Figure EV4. Iron supplementation prevents cancer-induced cachexia

A. Grip strength of C26 tumor-bearing mice measured at day 11 and normalized to day 10 (n = 3–4). B. Hematocrit in tumor-bearing mice measured on day 12 post C26 injection (n = 3) C. Quadriceps weight normalized to tibial length of C26 tumor-bearing mice at day 12 post C26 injection (n = 5–12). D, E. Average cross-sectional area of fast (D) and slow (E)–twitch muscle fibers (n = 2–5). F. Representative Western blot of iron metabolism proteins in tumor extracts from C26 tumor-bearing mice showing a physiological response to iron loading (n = 3). G. Mitochondrial aconitase activity in quadriceps of C26 tumor-bearing mice supplemented with FeCM (n = 4–6).

Data information: For all data, n represents the number of biological replicates. Statistical significance was calculated by unpaired, two-tailed Student’s t-test (A) or one-way Anova with Bonferroni’s correction (B-D). Data are mean ± SEM. *P < 0.05, ***P < 0.001 compared to control and #P < 0.05, ##P < 0.01compared to C26-untreated group.

Iron supplementation refuels mitochondrial oxidative metabolism and energy generation

Since our findings in vitro indicate that iron can refuel mitochondrial metabolism and maintain myotube mass, we sought to validate our hypothesis in vivo. After verifying the replenishment of mitochondrial iron in skeletal muscle (Fig 6A), we measured the activity of aconitase and succinate dehydrogenase and observed a significant recovery of enzymatic functionality following iron injection (Figs 6B and C, and EV4G). In agreement with these findings in mice showing restored mitochondrial metabolism upon iron treatment, we also found a significant increase in mitochondrial ATP content (Fig 6D) and coherently a drop in AMPK phosphorylation (Fig 6E). As AMPK drives fatty acid oxidation (FAO), which has been functionally linked to the cachectic process (Fukawa et al, 2016), we next evaluated the effect of iron injection on FAO. Consistently, iron supplementation reduced the C26-induced upregulation of FAO genes (Fig 6F–H). Altogether, these data indicate that iron supplementation of tumor-bearing mice mitigates energetic stress and catabolic pathways, which mediate the increase in muscle functionality and mass.

Figure 6. Iron supplementation rescues mitochondrial function

A. Mitochondrial iron quantified by ICP-MS in quadriceps of C26 tumor-bearing mice after FeCM supplementation (n = 7–8). B. Aconitase activity of quadriceps lysates normalized to protein content (n = 3–7). C. Succinate Dehydrogenase activity staining of gastrocnemius transversal sections and resulting intensity quantification (n = 3–7). D. Mitochondrial ATP content determined by luminescence assay in quadriceps (n = 3–9). E. Representative Western blot and densitometric quantification of phospho-AMPK and total AMPK (stripped and re-blotted) in the gastrocnemius of C26 tumor-bearing mice after iron carboxymaltose supplementation (n = 6). F–H. mRNA levels of ACOT 1 (F), MCD (G), and CPT1B (H) normalized to GAPDH in gastrocnemius (n = 5–10).

Data information: For all data, n represents the number of biological replicates. Statistical significance was calculated by unpaired, two-tailed Student’s t-test. Data are mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 compared to control and #P < 0.05, ##P < 0.01, ###P < 0.001 compared to C26-untreated group.

Source data are available online for this figure.

Iron supplementation improves muscle strength in cancer patients

Based on the results obtained in the pre-clinical model of cancer-associated muscle wasting, we measured the handgrip force in cancer patients with severe anemia, who reported muscle weakness, before and after FeCM injection (Table EV1). Improved strength was observed in the dominant hand of all patients, while more than half showed also increased force in the non-dominant hand (Fig 7A and B) as short as 3 days after the injection. Together with our data reporting TFR1 downregulation in the skeletal muscle of cachectic patients (Fig 1C), these findings indicate that altered iron metabolism may contribute to muscle weakness in cachectic patients. Consequently, these results highlight the contribution of iron on both muscle mass and functionality and suggest a new promising therapeutic strategy to counteract cancer-induced skeletal muscle wasting. Altogether, our findings show that iron supplementation prevents cancer-induced cachexia through a recovery of mitochondrial function (Fig 7C).

Figure 7. Iron supplementation improves muscle function in iron deficient cancer patients

A, B. Grip force of dominant or non-dominant arm in iron-deficient cancer patients, expressed in percentage of baseline force (A) and absolute values normalized to height (B) before/after single dose of iron carboxymaltose (15 mg/kg, 7 subjects).