Uncovering metabolism in rhabdomyosarcoma

Abstract

Rhabdomyosarcoma (RMS) is a myogenic tumor classified as the most frequent soft tissue sarcoma affecting children and adolescents. The histopathological classification includes 5 different histotypes, with 2 most predominant referred as to embryonal and alveolar, the latter being characterized by adverse outcome. The current molecular classification identifies 2 major subsets, those harboring the fused Pax3-Foxo1 transcription factor generating from a recurrent specific translocation (fusion-positive RMS), and those lacking this signature but harboring mutations in the RAS/PI3K/AKT signaling axis (fusion-negative RMS). Since little attention has been devoted to RMS metabolism until now, in this review we summarize the “state of art” of metabolism and discuss how some of the molecular signatures found in this cancer, as observed in other more common tumors, can predict important metabolic challenges underlying continuous cell growth, oxidative stress resistance and metastasis, which could be the subject of future targeted therapies.

Cancer metabolism

The multiple genetic lesions accumulating in cancer cells drive the molecular adaptations to escape cell death and sustain continuous cell division. These alterations are preferentially selected by the tumor microenvinronment and have a great impact on the metabolic status, which undergoes consistent adaptations that change dinamically depending on the stage of tumor progression. Metabolism of cancer cells, especially during tumor growth expansion, is characterized by predominance of anabolic reactions at the expense of catabolic reactions, in order to provide daughter cells with newly synthesized nucleotides, lipids and amino acids.^1-4 A central aspect of cancer metabolism is that tumor cells often metabolize glucose anaerobically also under aerobic conditions. This adaptative process, known as Warburg effect, allows that glucose-derived carbon skeletons can be used for macromolecule biosynthesis rather than for complete oxidation through the tricarboxylic acid (TCA) cycle in the mitochondria.^5-9 As a result, cancer cells will alternatively use the glucose backbone in the pentose phosphate pathway to produce ribose for nucleotide synthesis and NADPH for lipid biosynthesis and maintainance of cell's redox status. Pushing the Warburg effect progressively triggers depletion of α-ketoacid intermediates into TCA cycle, so that cancer cells have high metabolic demands of glutamine, a non-essential amino acid precursor for a myriad of components that also serves as substrate for gluconeogenesis and as alternative energy source in rapidly dividing cells. Glutamine utilization via glutaminase (glutaminolysis) provides cancer cells with glutamate that, becoming in turn oxaloacetate, may enter the TCA cycle and partially substitutes for glucose (glutamine anapleurosis).^10-13 In addition to dependency on glutaminolysis, the majority of human cancers, including breast, colon, ovary, lung, and prostate tumors express high levels of fatty acid synthase (FAS), a key metabolic enzyme that is functional to catalyze the synthesis of long chain saturated fatty acids for supporting the increased demand for membrane biogenesis.^14,15 The therapies currently employed to limit tumor expansion mostly utilize cocktails of antineoplastic drugs that interfere with the cell cycle progression; these agents include cell cycle specific drugs like plant alkaloids (etoposide, topotecan) or DNA synthesis inhibitors (5-fluorouracil, methotrexate) and cell cycle non-specific drugs like crosslinking agents (cyclophosphamide, ifosfamide, cisplatin) or intercalating anthracycline antibiotics (doxorubicin, daunorubicin). Some of them, such as doxorubicin, cisplatin and the anti-tumor peptide actinomycin, are also known to trigger cell death by increasing the reactive oxygen species (ROS) levels through various mechanisms.¹⁶ Although increasing oxidative stress is considered as a valuable strategy for overcoming primary tumors and metastasis,¹⁷ the Darwinian selection of oxidatively stressed cancer cells can allow the survival of clonal cells that expressing supra-physiological levels of detoxifying enzymes, such as the mitochondria-located manganese superoxide dismutase (MnSOD or SOD2)¹⁸ and gluthatione synthetase,¹⁹ can be responsible of collapse and chemoresistance.

How do the genetic lesions underlying neoplastic transformation shape the metabolism in cancer cells? An oversimplified but commonly accepted view is consistent with the notion that aberrant activation of tyrosin-kinase receptors (RTK) concurrently with sustained Myc activity and loss of p53 function drive important metabolic challenges, ultimately promoting the Warburg effect, synthesis of nucleotides, proteins and lipids, glutamine anauplerosis, and resistance to oxidative stress^1-3 (Fig. 1).

Figure 1.

Multiple genetic lesions impact the metabolism in cancer cells. Sustained activation of RTK pathways, when occurring concurrently with gain of Myc and loss of p53 functions, is responsible of metabolic challenges that allow cancer cells to proliferate under normoxia or hypoxia via increased glucose consumption, nucleotide-protein-lipid synthesis and glutamine anapleurosis. In addition, these pathways may confer resistance to oxidative stress.

Rhabdomyosarcoma

Soft tissue sarcomas (STS) are tumors arising from mesenchymal cells that harbor distinctive chromosomal aberrations and de-regulation of pathways governing morphogenesis of tissues like fat, muscles, nerves, fibrous tissues, blood vessels, or subcutaneous tissues.²⁰ Rhabdomyosarcoma (RMS) is a tumor showing distinctive traits of skeletal muscle lineage that, with about 350 new diagnoses in the United States per year, is classified as the most common STS in children and adolescents.^21,22 Patients who have localized RMS have a 5-year survival greater than 70% following a multimodal approach that includes chemotherapy, radiation therapy, and surgery; yet, overall survival of patients with metastasis remains poor.^23,24

Cells mainly deriving from myogenic lineages have been shown to contribute to RMS development in mouse^25-28 and zebrafish models,²⁹ yet even nonmyogenic lineages could participate in the formation of RMS.³⁰ As a result, these tumors can arise in or near to skeletal muscle districts as well as in sites that lack skeletal muscle, such as the biliary and genitourinary tract.³¹ The current histopathological classification comprises 5 subtypes, with 2 most predominant referred as to alveolar (ARMS) and embryonal (ERMS), accounting for about 25% and 60% of all RMS, respectively, while the remaining are classified as botryoid RMS, spindle cell RMS, and undifferentiated sarcomas.³² ARMS commonly develops in the musculature of trunk and limb extremities in children and adolescents and is characterized by poor prognosis, while ERMS commonly presents in disparate anatomical sites of patients younger than 10 y of age. The current molecular classification of RMS includes fusion-positive and fusion-negative subsets³³ depending on the presence or not of the fused Pax3-Foxo1 transcription factor (Fig. 2), a distinctive molecular signature generating from a chromosomal translocation t(2;13)(q35;q14) that juxtaposes the DNA binding domains of the PAX3 gene in frame with the DNA activation domain of the FOXO1 gene.^34,35 Pax3-Foxo1 is found in 70% ARMS cases and is considered a strong predictor of poor prognosis, while fusion-negative ARMS have better resolution and are clinically and molecularly indistinguishable from the larger group of ERMS in the majority of patients.³⁶ Because of the ability of Pax3-Foxo1 to drive the transcription of genes like fibroblast growth factor receptor 4 (FGFR4) and insulin-like growth factor 2 (IGF-2), fusion-positive RMS are characterized by sustained activation of the RTK/RAS/phosphatidylinositol-3-kinase (PI3K) axis;³³ in addition, these tumors often carry high copy number of N-Myc gene^37-39 and exhibit IGF-2 overexpression due to loss of imprinting (LOI) at 11p15.5 locus⁴⁰ (Fig. 2). To date, turning off the expression and/or activity of Pax3-Foxo1 represents one of the major milestones for researchers, and considerable results have been already obtained using targeted inhibitors which, destabilizing Pax3-Foxo1 oncoprotein, elicit tumor regression in xenograft and transgenic mouse models.^41-43 In fusion-negative RMS the genomic landscape is wider, and includes IGF-2 overexpression due to loss of heterozygosity (LOH) at 11p15.5,^40,44 mutations in genes that deliberately activate the RTK/RAS/PI3K signaling axis,^33,45–47 such as platelet-derived growth factor receptor A (PDGFRA), erb-b2 receptor tyrosine kinase 2 (ERBB2), FGFR4, and transducers like NRAS, KRAS and HRAS, phosphatidylinositol-4,5-bisphosphate 3-kinase, catalytic subunit α (PIK3CA); in addition, somatic mutations in cell cycle genes such as CTNNB1, FBXW7, BCOR³³ and p53,⁴⁸ and gain of chromosomes 2, 8 and 13^33,36 are frequently observed (Fig. 2).

Figure 2.

The genomic landscape in fusion-positive and -negative RMS. Fusion-positive RMS have a relatively low burden of somatic mutations; they are dominated by the fused Pax3-Foxo1 signature and high copy number of N-Myc gene. Instead, fusion-negative RMS are characterized by a number of different alterations, including mutations in the p53 pathway or cell cycle genes and gain of chromosomes. Both the fusion-positive and -negative subsets share the aberrant activation of RTK/RAS/PI3K axis as well as the IGF-2 overexpression due to LOI and LOH at 11p15.5 locus, respectively.

In the following paragraphs we describe the molecular hallmarks having impact on RMS metabolism.

Molecular hallmarks driving metabolic changes in RMS

Pax3-Foxo1 oncoprotein

Pax3-Foxo1 orchestrates the expression of numerous Pax3 downstream genes with increased amplitude and without feedback control, impairing the apoptosis and differentiation processes^49-53 and conferring a G2 checkpoint-dependent resistance to irradiation in vitro and in vivo.⁵⁴ For example, the Pax3-Foxo1-dependent transcription of genes like hepatocyte growth factor receptor (HGFR or c-MET), FGFR4, IGF-2 and C-X-C chemokine receptor type 4 (CXCR4) contributes to increase tumor aggressiveness and metastasis recurrence.⁵³ In addition, the Pax3-Foxo1 fusion protein drives the oncogenic transcription of N-MYC.³⁷ In the metabolic context, Pax3-Foxo1 has been shown to drive transcription of GLUT4 gene, thereby eliciting increased glucose uptake in ARMS cells⁵⁵ (Fig. 3). Consistent with this, the administration of 2-deoxyglucose (2-DG), a glycolytic inhibitor commonly utilized as a cancer therapeutic,⁵⁶ has been reported to sensitize different fusion-positive RMS lines to cell death,⁵⁷ suggesting that the Pax3-Foxo1-dependent increased glucose consumption could be rate-limiting for tumor growth. Pax3-Foxo1 also has the ability to elicit downregulation of phosphate and tensin homolog deleted on chromosome 10 (PTEN) gene,⁵⁸ a phosphatase which functionally antagonizes PI3K and downstream protein kinases such as AKT,^59,60 therefore increasing the activity of PI3K/AKT/mammalian target of rapamycin (mTOR) pathway⁶¹ (Fig. 3). In contrast to what commonly observed in cancer cells, where a long-lasting AKT activity correlates with increased lipid biogenesis,⁶² the activity of AKT in ARMS cells concurrently with a Pax3-Foxo1-dependent transcription of the carnitine palmitoyltransferase gene (CPT1A), the outer mitochondrial enzyme responsible for the formation of acyl carnitines, was reported to facilitate lipid degradation⁶³ (Fig. 3). This mechanism has been supposed to provide ARMS cells with the burst of energy required to migrate and metastasize.⁶³ This characteristic appears to be rather unique and is clearly indicative of how the Pax3-Foxo1 signature may confer unique metabolic settings to fusion-positive RMS. In this regard, the identification of new putative Pax3-Foxo1 target genes involved in metabolic processes could be useful to better understand the molecular adaptations underlying the aggressiveness of fusion-positive RMS.

Figure 3.

Molecular signatures influencing the metabolism in cancer and RMS cells. The cartoon depicts the molecular effectors that in cancer cells increase the glucose uptake and glycolytic consumption, nucleotide synthesis, lipogenesis, protein synthesis and glutamine anapleurosis. As observed in different tumors, RMS subsets share many of these molecular hallmarks; moreover, the more aggressive fusion-positive RMS express the Pax3-Foxo1 oncoprotein, which is actively involved in the transcription of different targets, including metabolic genes such as GLUT4 and CPT1A. In addition, Pax3-Foxo1 increases the expression of N-Myc and decreases that of PTEN. Thin arrows indicate the interconversion occurring between organic substrates, whereas thick arrows indicate the activating (in green) or inhibitory (in red) actions of the protein effectors on different processes through enzymes (blue box outlined) or transporters (red box outlined). Abbreviations: ACL, ATP citrate lyase; CPT1A, carnitine palmitoyl transferase 1A; FAS, fatty acid synthase; GLUTs, GLUT transporters; G6PD, glucose-6-phosphate dehydrogenase; HIF, hypoxia inducible factor; HK, glycolytic hexokinase; LDH, lactate dehydrogenase; MAPK, mitogen activated protein kinase; PDC, pyruvate dehydrogenase complex; PDK1, pyruvate dehydrogenase kinase isozyme 1; PFK, phosphofructokinase; PI3K, phosphatidylinositol-3-kinases; mTOR, mammalian target of rapamycin; RTK, tyrosine kinase receptors; PTEN, phosphate and tensin homolog deleted on chromosome 10; SREBPs, Sterol Regulatory Element-Binding Proteins.

Deliberate activation of RTK pathways

As previously outlined, both fusion-positive and -negative RMS subsets share a deliberate activation of RTK-dependent pathways.³³ These include 2 highly conserved signaling systems widely used by cancer cells to respond to growth factors, i.e., the PI3K/AKT and RAS pathways, which are responsible of important challenges in cancer metabolism. The PI3K/AKT axis stimulates the uptake of glucose, amino acids, and other nutrients,⁶⁴ in turn promoting glycolysis and lactate production,⁶⁵ lipogenesis,⁶⁶ and protein synthesis via activation of mTOR⁶⁷ (Fig. 3). In addition, the AKT/mTOR pathway is responsible of the activation of Hypoxia inducible factors (HIF), which allow survival of solid tumors under low oxygen tension⁶⁸ (Fig. 3). The RAS pathway usually activates in response to mitogens and integrates the outcome of the PI3K/AKT pathway by activation of Myc transcription factors^69,70 (Fig. 3). These latter, in turn, drive the expression of at least 1000 genes and microRNAs involved in disparate processes, including cell division and tumor formation, and promote metabolic changes such as the Warburg effect, lipogenesis and glutamine anapleurosis^71-73 (Fig. 3).

The activation of the PI3K/AKT and extracellular regulated kinases 1/2 (ERK1/2) pathways observed in different ERMS and ARMS cell lines under hypoxia was shown to increase cell survival through an increased production and secretion of the proangiogenic interleukin-8 (IL-8).⁷⁴ This molecule is a member of the α (C-X-C) subfamily of chemokines, is secreted by several haematopoietic cells, fibroblasts, hepatocytes and various cell lines, and has been found to act as a mitogenic autocrine growth factor for some human cancers⁷⁵ by interacting with 2 7-transmembrane G-protein-coupled receptors, namely CXCR1 and CXCR2.⁷⁶ Although human RMS cells were found to be negative for CXCR1 and CXCR2 expression,⁷⁴ the release of IL-8 by RMS cells was hypothesized to be a paracrine mechanism for stimulating stroma cells to increase de novo vascularization and counteract hypoxia. Consistent with this, shRNA-mediated downregulation of IL-8 expression in the alveolar RH30 cells was sufficient to limit the formation of xenotransplanted tumors in vivo.⁷⁴ In response to hypoxia, an increased glucose uptake was also observed in the A204 cell line to be protective against cell death⁷⁷; however, it should be noted that this cell line is no longer retained as truly representative of RMS.⁷⁸ Finallly, the cardiac-specific deletion of the von Hippel-Lindau protein (VHL), an E3 ubiquitin ligase responsible of suppression of HIF levels during normoxia, has been described to favor the formation of highly metastatic cardiac RMS in a mouse model.⁷⁹ Taken together, different lines of evidence indicate that the PI3K/AKT and ERK1/2 pathways, often concurrently with the HIF pathway, play a central role for RMS progression; their targeted inhibition was indeed shown to enhance cell death in RMS lines under both normoxic⁸⁰ and hypoxic conditions.^81,82

Loss of p53 activity

Different regulators in healthy tissues control the glucose metabolism, including p53, AKT, Myc and HIF proteins. P53, in particular, inhibits the expression of glucose transporters (GLUTs), such as GLUT1 and GLUT4⁸³ or GLUT3,⁸⁴ and reduces glucose utilization in the glycolytic pathway^85-87 (Fig. 3). P53 can also sustain the expression of PTEN, which is a negative regulator of the glycolysis and of PI3K/AKT/mTOR and HIF pathways⁸⁸ (Fig. 3). As a result, p53-deficient cells are characterized by high sugar transport across the cell membrane and high glucose consumption, 2 hallmarks underlying neoplastic transformation. A gene signature consistent with the “p53 off” state has been significantly found in fusion-negative RMS with an incidence from 5% to 19%,^33,89 whereas to a lesser extent in fusion-positive tumors.^42,89 P53 gene inactivation is indeed an experimentally validated cooperating lesion in zebrafish and mouse models of ERMS²⁹ and ARMS,²⁵ respectively. In the human RD cell line, one of the most widely used human ERMS cell lines,⁷⁸ loss of p53 activity was reported to rise the glucose uptake by increased expression of the GLUT1 and GLUT4 transporters⁸³ (Fig. 3). Moreover, GLUT1 overexpression in RD cells, increasing glucose uptake and glycolytic use, was a condition permissive to enhance cell proliferation and invasiveness via a matrix metalloprotease type 2 (MMP2)-dependent mechanism.⁹⁰ It is worthy mentioning that loss of p53 in cancer has not only been implicated in increasing sugar metabolism and Warburg effect, but also lipogenesis and nucleotide biosynthesis.^91-94 For example, p53 can favor the production of acetyl-CoA by activation of the pyruvate dehydrogenase complex (PDC)⁹⁵ and push the oxidative phosphorylation through upregulation of cytochrome c oxidase 2 gene⁹⁶ (Fig. 3). In addition, p53 can support glutamine anapleurosis by driving the expression of phosphate-activated mitochondrial glutaminase (GLS2)⁹⁷ (Fig. 3), which converting glutamine to glutamate represents a key reaction for the anaplerotic reactions of TCA cycle and for the synthesis of glutathione (GSH).

Overall, besides glucose consumption, further experimental evidence is required to address the question of whether p53 loss may impact other key metabolic pathways in RMS.

Oxidative stress pathways

Killing cancer cells with oxidative stress represents a promising therapy.¹⁷ Whole-genome and RNA sequencing technologies applied to human RMS samples have recently allowed to identify a number of genetic anomalies, especially in ERMS, which would predict susceptibility to oxidative stress⁸⁹ (see comments in).⁹⁸ In particular, the incidence of RAS pathway mutations was found to increase with ERMS tumor risk group classification, being found in 75% of high-risk ERMS, 45% of intermediate-risk ERMS, and 0% of low-risk ERMS, indicating that deliberate activation of this pathway correlates with ERMS aggressiveness.⁸⁹ Interestingly, prior observations in lung tumors had indicated that activating RAS pathway mutations are responsible of the increased production of mitochondrial ROS levels, which are in turn required to promote the anchorage-independent cell growth of cancer cells via MAPK activation.⁹⁹ In this regard, RAS-positive ERMS cells seemingly have a number of characteristics ascribable to increased oxidative stress; for instance, in comparison to other childhood tumor types like T-cell acute lymphoblastic leukemia and medulloblastoma, ERMS tumors are characterized by higher rate of G→T transversions, which are mutations typically caused by oxidative damage⁸⁹; moreover, ERMS displayed methylation of several genes implicated in regulation of metabolism, mitochondrial function, and oxidative stress, including PTK2, COX7A1, NOSIP, NOS1, ATP2A3, DDAH1, GLRX, and TXNDC12.⁸⁹ Hence, if on the one side ERMS tumors may benefit on increased ROS levels for cell growth, on the other side they can progressively accumulate ROS-induced gene mutations making them more susceptible to cell death. Thus, oxidative stress inducers such as actinomycin-D, one of the most active agents used in the treatment of RMS, may help to overcome tumor growth.

Another valuable strategy to rise oxidative stress in RMS could be targeting the detoxifying enzymes. In this regard, inhibition of the thioredoxin (TXN) reductase via auranofin administration has been reported to limit consistently the growth of ERMS xenografts, especially in combination with HDACs inhibitors such as panobinostat.⁸⁹ TXN reductase and GSH reductase are the main enzymes responsible of the ROS neutralization in the cytosol. To do this, they need to be constantly regenerated by NADPH. Thus, depletion of NADPH levels, besides affecting lipid synthesis, can perturb the cellular redox status via impairment of the activity of TXN and GSH reductases. Major cellular sources of NADPH are the pentose phosphate pathway, the cytosolic isocitrate dehydrogenase and NADP-dependent malic enzymes, and recent findings have shown that silencing the glucose-6-phosphate dehydrogenase (G6PD), the enzyme catalyzing the first rate-limiting step of the pentose phosphate shunt, impairs proliferation of ERMS cells and soft agar growth.¹⁰⁰

Also reducing the endogenous TXN and GSH levels may reflect on increased oxidative stress. GSH availability depends primarily on its synthesis obtained by condensation of cysteine, glutamate and glycine. Cysteine transport is under the control of the glutamate-cystine xCT transporter,^101,102 whereas the availability of glutamate depends on activity of glutaminase, target of p53 and Myc factors. Glycine is not only a component of GSH, but also represents a major source for biosynthesis of purines and heme group,¹⁰³ and of methyl groups for the so-called one-carbon metabolism, a complex network that based on folate compounds, well-known targets of antifolate chemotherapy. The GSH-dependent antioxidant response of cancer cells may benefit depending on the availability of serine, which can be transformed into glycine through the serine hydroxymethyltransferase (SHMT).^104-106 Interestingly, it has been demonstrated that cancer cells lacking p53 are protected from oxidative stress-induced by serine depletion in the presence of p73, a p53 family member that has the ability to drive serine biosynthesis de novo^107,108 and to increase the pentose phosphate shunt and nucleotide biosynthesis via G6PD.^109,110 Thus, the p73 overexpression detected in various primary RMS samples¹¹¹ may have an important role for oxidative stress resistance.

Finally, it should be mentioned that the expression levels of GSH reductase, TXN reductase and catalase are all under direct transcriptional control of the nuclear factor erythroid 2-related factor (NRF2),¹¹² which is expressed in a broad spectrum of sarcomas, including RMS.¹¹³ In addition, NRF2 is a downstream target of p53, which also drives the expression of other several genes with antioxidant functions like sestrin, GSH peroxidase, aldehyde dehydrogenase, GLS2, TIGAR, and tumor protein p53-inducible nuclear protein 1 (TP53INP1).^93,114-119 Considering the central role of both NRF2 and p53 in modulating oxidative stress resistance in cancer,¹²⁰ further studies are required to establish their contribute in RMS.

Targeting intratumor cell heterogeneity: A difficult but promising task

Multiple factors contribute to cancer progression. Tumor cells constantly undergo clonal selection as a result of genetic and epigenetic changes, a process culminating with the presence of distinct subpopulations within the bulk of tumor.^121,122 The clonal composition of a tumor will change depending on the various stages of tumor propagation, hence those mutations that enhance cell division will be selected during the early tumor stage for rapid mass growth, whereas others permitting increased invasiveness and chemoresistance during the later stages.¹²³ Recent discoveries have shown that RMS tumors seem to be composed of a dominant clone, although each tumor contains subclonal populations with a unique mutational profile, which may provide selective advantage to a relapse or metastatic tumor.¹²⁴ Through whole-genome sequencing (WGS) carried out on 44 primary RMS tumors it has been further shown that loss of heterozygosity of chromosome 11p15.5 and point mutations in members of the RAS pathway are early events occurring in majority of cases in fusion-negative RMS tumors, while PAX3-FOXO1 fusion event occurs prior to a whole genome duplication event which results in tetraploidy of fusion-positive RMS.¹²⁴ A distinctive characteristic of RMS is the coexistence of cell subpopulations displaying a variable degree of myogenic differentiation, especially in ERMS, which may play distinct roles on tumor evolution. In this regard, in vivo real-time confocal imaging of primary heterogeneous ERMS arising in Zebrafish model has provided evidence that more differentiated ERMS tumor cells are highly migratory, able to invade neighboring normal tissue and to intravasate. Once they have established a new tumor site, the less differentiated myf5-positive cells migrate into the newly forming tumor, ultimately driving tumor expansion and progression.^125,126 This implicates that the less differentiated RMS-propagating cells may take advantage of specific nutrients produced by the neighboring more differentiated cells during tumor dissemination, a process that might be even influenced by the proportion of fast or slow twitch muscle fibers potentially releasing specific metabolites around the tumor niche (at least when the tumor arises in muscle beds). This scenario seemingly resembles the so-called reverse Warburg effect, in which epithelial cancer cells were found to corrupt the stroma associated fibroblasts undergoing myo-fibroblastic differentiation to release micronutrients in the milieu, such as lactate and pyruvate, which are then taken up by cancer cells to produce energy in the mitochondria.¹²⁷ RMS cells might also exploit nutrients deriving from the fusion with other cells, since the fusion process has been observed occurring between ARMS cells and muscle satellite cells through a IL-4R-dependent mechanism.¹²⁸ In light of these important findings, the future lines of research should be devoted to a better understanding of the metabolism in RMS cells in both the undifferentiated and differentiated status.

Could the design of targeted metabolic inhibitors improve the current standard therapy of RMS?

Therapy for RMS requires a multimodality approach, including surgery, radiotherapy and chemotherapy, which results in acute toxicities and long-term side effects.²¹ The staging and risk stratification for RMS based on the results of successive Intergroup Rhabdomyosarcoma Study Group (IRSG) clinical trials. The Stage classification recommended by the Soft Tissue Sarcoma Committee of the Children's Oncology Group (COG) utilizes the so-called TNM classification (ie, tumor, nodes, metastases), a pretreatment clinical step that provides a system based on the site, size and invasiveness of the primary tumor, lymph nodes status, and the presence or absence of metastases. In addition to TNM staging, patients following surgery are assigned to a clinical group based on the completeness of tumor excision and the evidence of tumor metastasis. Treatment is tailored based on the patient's status as having low, intermediate or high risk for treatment failure. For patients wth low-risk disease, chemotherapy regimens based on vincristine, D-actinomycin and cyclophosphamide or ifosfamide (the so-called VAC/VAI) result in significant improvement, while survival for high-risk patients, typically having ARMS at unfavourable sites and/or metastasis and recurrent disease, remains poor, not exceeding 50%.¹²⁹ Currently, the main molecular targets under investigation to improve therapies for advanced, recurrent and metastatic RMS include receptors (IGF-1R, c-Met, PDGFR,c-Kit), intracellular signaling molecules (mTOR, MEK/ERK), cell cycle (CDK4/CDK6) and apoptotic proteins (p53, Bcl-2, TRAIL), the proteasome machinery, HSP90, histone deacetylases, angiogenetic molecules (VEGFR, VEGF), and the Pax3-Foxo1 oncoprotein (all these targets have already been extensively discussed in ref.¹³⁰). While preclinical in vitro and in vivo studies have suggested the effectiveness of inhibitors targeting RTK like IGF-1R and PDGF-R,¹³¹ and pathways like the AKTmTOR, results from recently published clinical trials have indicated that these therapies are not sufficient to improve the outcome in patients with recurrent disease (Table 1). This evidence has suggested to implement the therapies in early-phase clinical trials and preclinical models using multityrosine kinase inhibitors like pazopanib (VEGFR-1, -2, and -3, PDGFR-α and -β and KIT), crizotinib (dual ALK/c-MET inhibitor), vismodegib (an hedgehog signaling inhibitor), alisertib (an Aurora Kinase A inhibitor) and TH-302 (nitroimidazole, a mustard prodrug that releases the alkylating agent bromo-isophosphoramide when exposed to hypoxic conditions, all reviewed in).¹³² Currently, there are no pending clinical trials exploiting the use of targeted metabolic inhibitors against RMS. In this regard, only one phase II study has shown that pemetrexed, a multitargeted antifolate inhibiting the nucleotide biosynthesis, has no anti-tumor activity on refractory and recurrent RMS (Table 1). Yet, there is a number of metabolic inhibitors targeting the glycolysis, TCA and fatty acid synthesis, including inhibitors of GLUT1 and GLUT4 (phloretin, WZB117, ritonavir), hexokinase (2-deoxyglucose), pyruvate kinase M2 (ionidamine), lactate dehydrogenase A (FX11, oxamate), pyruvate dehydrogenase kinase (dichloroacetate) and FAS (cerulenin, C75, orlistat), which are currently under investigation in pre-clinical and clinical studies on different types of cancer (colon, lung, leukemia, fibrosarcoma, prostate, pancreatic and breast cancer)⁵⁶ and might therefore be helpful to improve the RMS therapy.

Table 1.

List of the more recently completed phase I/II clinical trial studies on drugs against RMS.

Drugs	Molecular target	Eligibility	Patient age	Phase	Clinical efficiency
Cixutumumab (IMC-A12)	IGF1-R monoclonal antibody	Recurrent or refractory solid tumors	≥2 and ≤30 years	II	Cixutumumab is well tolerated in children with refractory solid tumors. Limited objective single-agent activity of cixutumumab was observed; however, prolonged stable disease was observed in 15% of patients. Ongoing studies are evaluating the toxicity and benefit of cixutumumab in combination with other agents that inhibit the IGF pathway 138
Cixutumumab (IMC-A12)	IGF1-R monoclonal antibody	Advanced or metastatic rhabdomyosarcoma, leiomyosarcoma, adipocytic sarcoma, synovial sarcoma or Ewing family of tumors	≥12 years	II	Patients with adipocytic sarcoma may benefit from treatment with cixutumumab. Cixutumumab treatment was well tolerated, with limited gastrointestinal AEs, fatigue and hyperglycaemia 139
R1507	IGF1-R monoclonal antibody	Recurrent or refractory RMS, osteosarcoma, synovial sarcoma	≥2 years	II	R1507 is safe and well tolerated but has limited activity in patients with recurrent or refractory bone and soft tissue sarcomas 140
Temsirolimus (TEM) with irinotecan (IRN) and temozolomide (TMZ)	mTOR	Recurrent/refractory solid tumors, including RMS and central nervous system tumors.	≥1 and ≤22 years	I	The combination of is well tolerated in children. Phase 2 trials of this combination are ongoing 141
Cixutumumab in combination with temsirolimus	IGF1-R and mTOR	Relapsed or refractory osteosarcoma; Ewing sarcoma; RMS; and non-RMS soft tissue sarcoma (NRSTS)	≥1 and ≤30 years	II	Despite encouraging preclinical data, the combination of cixutumumab and temsirolimus did not result in objective responses 142
Imatinib	PDGFR	Advanced sarcoma	≥10 years	I/II	Imatinib is not an active agent 143
Trabectedin	DNA intercalant	Recurrent RMS, Ewing sarcoma and non-RMS soft tissue sarcomas	≥1 and ≤21 years	II	Trabectedin did not demonstrate sufficient activity as a single agent for children with relapsed pediatric sarcomas 144
Pemetrexed	multi-targeted antifolate that inhibits key enzymes involved in nucleotide biosynthesis	Refractory or recurrent solid tumors, including RMS.	≥6 months and ≤21 years	II	Pemetrexed, was tolerable in children and adolescents with refractory solid tumors, but did not show evidence of objective anti-tumor activity in the childhood tumors studied 145

Obtained from clinicaltrials.gov website.

Conclusions

Uncovering the molecular drivers that shape metabolism in cancer cells represents a pivotal step to elaborate targeted therapeutic strategies. Little attention has been devoted to metabolism in RMS so far, though there are several elements indicating that these tumors share several molecular signatures with common malignancies that would predict important metabolic adaptations. Taking advantage of the advent of the whole-genome sequencing technology, current studies are greatly contributing to dissect the intratumoral heterogeneity in RMS and delineate how specific lesions drive tumor evolution from the onset at the primary site until releapse and metastasis. Over the past years the Computational Molecular Phenotyping is emerging as a potential cancer biomarker strategy. This novel metabolomics technique quantitatively profiles metabolites (all the natural amino acids, glutathione, and many others) with cellular resolution by ultrathin tissue section arrays using hapten-specific antibodies.^133-137 Potential application of this tecnique is the ability to interrogate, cell-by-cell, metabolite profiles associated with biological processes. Hopefully, these different strategies will contribute to implement the current therapies through the characterization of the dominant pathways that shape metabolism within cell classes that make up RMS tumors.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Funding

This work was supported by the University of Brescia research fund (ex 60%) to AF and EM.