Association of cancer metabolism-related proteins with oral carcinogenesis – indications for chemoprevention and metabolic sensitizing of oral squamous cell carcinoma?

Cancer is regarded as an acquired genetic disease. The genetic model of multistep carcinogenesis describes the rise of malignant tumors from a single transformed cell (monoclonal theory of carcinogenesis) and subsequent development through morphologically and clinically detectable precancerous stages [1]. The carcinogenesis of oral squamous cell carcinoma (OSCC) is a highly complex multifocal process that occurs when squamous epithelium is affected by several genetic alterations [2]. Understanding the mechanistic basis await the availability of molecular tools to experimentally and selectively manipulate this multistep process with subsequent clinical implications for therapy of precursor lesions and OSCC.

OSCC is an aggressive tumor with low response to chemotherapy and basic resistance to most standard of care anticancer drugs [3, 4]. Tumor metabolism [5] with a special focus on increased hypoxia/glycolytic activity is regarded as a crucial factor for the carcinogenesis of OSCC and is associated with radio- and, chemotherapy resistance, as well as tumor recurrence [6‐9].

Cancer can be considered as integrated metabolic ecosystem and includes several pathways of carcinogenesis associated with metabolic phases of transformation [10]. Glycolysis [11], mitochondrial oxidative phosphorylation (OXPHOS) [12], and glutaminolysis have been shown to play key roles in tumor metabolism. Mitochondria have an important role in carcinogenesis due to their roles in mediating apoptosis [13]. They act as a major source of endogenous reactive oxygen species (ROS) that escape from the electron transport chain (ETC.) during OXPHOS [14]. Although glycolysis is a major characteristic of tumor cell metabolism this pathway alone cannot account for energy usage in all types of cancer cells. Finally, the dominant metabolic process can be either glycolysis or mitochondrial oxidative metabolism based on the tumor type [15]. Both metabolic phenotypes have been associated with subsequent nutritional consequences [16‐19].

The generation of adenosine triphosphate (ATP) in glycolysis has a lower efficiency, but a faster rate than OXPHOS [11, 20]. This enhanced rate of ATP generation has been postulated to be beneficial for rapidly proliferating cells. However, several studies have suggested that OXPHOS is the major source of cellular ATP in proliferating and non-proliferating [21] cancer cells [11, 21‐23].

A recent study by Vander Heiden [24] indicated that the induction of the Warburg effect in cancer cells is more the consequence of the activation of protooncogenes (e.g., Myc), transcription factors (e.g., hypoxia-inducible factor-1, HIF-1), and signaling pathways (e.g., PI3K), as well as the inactivation of tumor suppressors (e.g. p53) rather than the primary generation of much needed energy [11]. Moreover, it has been stated that tumor cells profit from the enhanced glycolytic activity in glycolytic intermediates, which are shunted into subsidiary pathways (e.g. by the pentose phosphate pathway [PPP]) to fuel metabolic pathways that generate de novo nucleotides, lipids, amino acids, and nicotinamide adenine dinucleotide phosphate (NADPH) [11, 25, 26]. Frezza et al.[12] showed that defects in mitochondrial enzymes or complexes within the electron transport chain are not frequently observed in cancer. Therefore, investigation of OXPHOS provides a clear rational for future anti-cancer therapy strategies in OSCC [27].

Today, it is estimated that more than 30% of all tumor entities may be due to dietary factors [17]. Studies have clearly linked diabetes and obesity to cancer [28]. Hyperinsulinemia leads to increased production of insulin-like growth factor-1 (IGF-1) [29], which activates insulin-like growth factor-1 receptor (IGF-1R). IGF-1R is a receptor tyrosine kinase (RTK) that stimulates protein synthesis by activating the mammalian target of rapamycin (mTOR), and in turn mTOR mediated upregulation of glycolytic enzymes may promote tumor development [30, 31]. Therefore, the IGF-1R pathway is an emerging therapeutic target in oncology [32‐34] but has not yet been described for the carcinogenesis of OSCC.

Hexokinase 2 (HK 2) is expressed in insulin-sensitive tissues such as muscle and adipose [11], is one of the rate-limiting enzymes of glucose catabolism in tumor cells, is upregulated in many cancers [35, 36], and was recently described for OSCC [37]. Phosphofructokinase-1 (PFK-1) [38] is a key enzyme in glycolysis where it forms a rate-limiting step, but its expression has not been described for OSCC. Among glycolytic enzymes PFK-1 has been more extensively studied than other enzymes, which is likely to be due to its various regulatory mechanisms.

Recently, we have demonstrated glucose transporter 1 (GLUT-1) (solute carrier family 2 [facilitated glucose transporter], member 1 [SLC2A1]) [9], transketolase-like-1 (TKTL1) [7], and lactate dehydrogenase A (LDHA/LDH5) [39] as adverse prognostic factors for the survival of patients with OSCC. However, the expression of GLUT-1, HK 2, PFK-1, LDHA, and TKTL1 during a multi-step carcinogenesis has not been described yet.

More recently, characterization of OXPHOS in cancer was performed by describing succinate dehydrogenase SDHA, SDHB (respiratory complex II in mitochondria), and ATP synthase (respiratory complex V in mitochondria) [40, 41]. None of these enzymes have yet been described for OSCC.

The purpose of this study was to examine the relationship between metabolism-related proteins [8] with a multistep carcinogenesis. This is the first study describing glycolysis-related PFK-1, OXPHOS-related SDHA, SDHB, and ATP synthase in OSCC.

In our study, we investigated cancer metabolism-related proteins in the carcinogenesis of OSCC. For the first time, we found increased expression of mitochondrial enzymes (SDHA, SDHB, ATP synthase) in OSCC compared with normal oral mucosa. However, very few data is available describing a mitochondrial oxidative metabolism [27] in OSCC. Authors assume that OXPHOS is an important pathway for the generation of ATP [11, 22, 23] and ROS [18, 55‐58] during the carcinogenesis of OSCC. The TUNEL assay demonstrated that tumor cells do not undergo apoptosis and therefore, increased ROS generation by OXPHOS does not reach toxic levels. Based on our results and as currently stated by Whitaker-Menezes et al.[57] in the context of breast cancer we assume that mitochondria are the ‘Achilles heel’ and ‘powerhouse’ in the carcinogenesis of OSCC [23, 56‐59]. Increased levels of ROS in tumor cells are generated by altered metabolic activity, oncogene activation, and deregulated proliferation [60]. Oncogenic transformation promotes the production of excessive ROS, which would become toxic if not counteracted, while low levels of ROS can help to promote cell proliferation. This is the reason why many cancer cells may show an increased expression of antioxidant proteins [26] such as LDHA [39] and TKTL1 [7] as indicated by our observation, which contribute to the survival and success of the tumor. Indeed, this dependence on antioxidants can make cancer cells more vulnerable to the inhibition of these detoxifying systems than normal cells, which do not harbor such a high burden of oxidative stress [61‐63]. On the other hand, an increase in ATP production by OXPHOS has been shown in response to hypoxic stress and protects cells from a critical energy crisis [64]. However, we do not know which metabolic pathway (glycolysis vs. OXPHOS) has been upregulated in carcinogenesis of OSCC as first.

In the literature, bioactive food components [5, 17, 65, 66] have been demonstrated to mediate the reversal of a glycolytic phenotype in cancer cells, thus leading to growth inhibition and induction of apoptosis (Table 2). The reprogramming of energy metabolism [67‐70] has been suggested for targeting of mitochondria [18, 19, 21, 23, 55, 58, 61‐63, 71‐74] and subsequent induction of apoptosis [71] as a valid anti-cancer strategy [18] for which bioactive food components [19] have been suggested. Rapidly proliferating cells are more sensitive to radio-, and chemotherapy, which have been shown to be less effective in non-dividing cancer cells [75]. Activation of mitochondrial OXPHOS [58, 61] and other mechanisms in cancer cells by natural compounds may induce apoptosis even in therapy resistant cancer cells [55]. Because OXPHOS is the predominant supplier of ATP in (proliferating and) non-proliferating cancer cells [21] targeted anti-mitochondrial therapies could be of interest for apoptosis induction in quiescent (non-proliferating) but metabolically active cancer cells, which rely on mitochondrial lipid β-oxidation [76]. Therefore, bioactive food components inducing apoptosis by ROS generation (Table 2) and other mechanisms play an emerging role in cancer therapy. According to other tumor entities several other natural compounds have been shown to activate ROS [58, 61] in OSCC [77‐81] and subsequent apoptosis in cancer cells and may therefore provide a clear rational to study them in further pre-clinical and clinical trials (Table 2). Moreover, phytochemicals [82] and vitamins have different hypoxia-inducible factor-1 (HIF-1) binding capacities (inhibitory activity: lycopene > curcumin > tocopherol > ascorbic acid) suggestive for their impact on the decrease in tumor hypoxia and antioxidative properties in normal tissue [83].

Lactate, pyruvate, gluthathione, and NADPH generated in glycolysis and/or the PPP effectively scavenge free radicals and ROS, thereby protecting the tumor cell from free radical-mediated DNA damage [26] (e.g. radiation therapy) or other ROS-inducing therapies by natural compounds leading to apoptosis. Most likely, modulation of one pathway will be not effective in most cases [17]. Therefore, synchronous [59] targeting of glycolysis (e.g. carbohydrate-restricted diets [16, 204‐217] or natural compounds, Table 2) with anti-mitochondrial therapies [18, 19, 21, 23, 55, 58, 61‐63, 71‐74] increasing ROS (Table 2) may act as sensitizer for adjuvant therapies in OSCC or could be useful for chemoprevention. Based on the literature a synergistic effect of a carbohydrate-restricted diet with an anti-mitochondrial therapy can be concluded, since carbohydrate-restricted diets may induce enhanced OXPHOS and lead to inhibition of mTOR [218], which is responsible for synthesis of glycolytic enzymes [30, 31]. Specifically observed in patients with head and neck cancer a ketogenic diet decreased the in vivo production of lactate in tumor cells [213].

However, it must be stated that natural compounds like phytochemicals [65, 219‐225] and vitamins may also prevent ROS-mediated carcinogenicity in cancer chemoprevention. During carcinogenesis ROS may act as a double-edged sword [226]. ROS are important intermediates of cellular signaling that suppress and promote tumorigenesis at once. They make both mitochondrial DNA and nuclear DNA susceptible to damage, and mutations in these two DNA pools are reported to lead to carcinogenesis [227]. However, targeted anti-mitochondrial therapies inducing apoptosis probably require functional active mitochondria without mutations that may respond to radiotherapy/chemo-radiotherapy in OSCC [228].

With specific regard to SDHA and SDHB, vitamin E (α-tocopheryl succinat, target: respiratory complex II in mitochondria) [229] and resveratrol (target: respiratory complex V in mitochondria, ATP synthase) [230] were shown to induce apoptosis in cancer cells. Metformin has been demonstrated to block respiratory complex I in mitochondria [231] as an effective anti-cancer agent [232] and prevented the development of OSCC from carcinogen-induced premalignant lesions [233]. More recently, a synthetic modified thiamine analog oxybenfotiamine [195] specifically inhibits TKTL1 in the PPP [87], of which elevated levels have been detected in the carcinogenesis of OSCC [7]. Targeting the PPP [87] as a detoxifying system [26] may revise tumor hypoxia and resistance to radio- and chemotherapy [7, 9]. Therefore, small molecules like oxybenfotiamine [195] provide new opportunities for targeted therapies in cancer and specifically OSCC. Nevertheless, the cytoprotective function of the PPP is not limited to defending against ROS but also expands to helping DNA damage repair [70].

However, it remains unclear whether phytochemicals are standardized effective for chemoprevention [2, 17, 65, 66, 84, 88, 115, 120, 219, 221, 223, 234, 235] in the treatment of precursor lesions or OSCC development as suggested for multistep carcinogenesis [2] but they provide a clear rational for further in-vitro, in-vivo, and clinical studies in the carcinogenesis of OSCC (Table 2) [2, 84, 88, 115, 120, 219, 234‐236]. Polyphenols like flavonoids and anthocyanidins have been well investigated in pre-clinical and clinical trials for the treatment of oral precursor lesions and OSCC [84, 115, 234]. For example, in 1999 Li et al. have already been reported of the chemopreventive impact of green tea on oral leukoplakia with increased rate of partial regression (systemically, oral capsules with 1.2 g polyphenols, and topical tea extract in glycerine over a period of 6 months) [236].

Proliferating cells have intrinsic increased metabolic activities compared to non-proliferating cells [21, 69]. This is supported by our data showing a significantly correlation of proliferating cancer cells with both glycolysis-related proteins (GLUT-1, TKTL1), and OXPHOS-related enzymes (SDHA, SDHB, ATP synthase). In this context glycolysis-related proteins may act as detoxifying system [26] (LDHA, TKTL1) of increased ATP producing (and ROS generating) OXPHOS-related proliferating cancer cells. These findings can be clinically addressed by differentiating cancer patients into metabolic responders and non-responders for malignancies such as SCC of the esophagus or head and neck squamous cell carcinoma [237‐239].

As for OSCC, there are several reports for glycolysis [9] as the predominant energy metabolism pathway. Glycolysis is involved in aggressive tumor behavior because it causes radio-, and chemotherapy resistance, creates a tumor microenvironment favorable for tumor cell migration, induces angiogenesis, and contributes to the immunologic escape of tumors [26]. However, a previous study by Yi et al. demonstrated that inhibition of the glycolysis-related PFK-1 activity redirects the glucose flux through the PPP [240], thereby conferring a selective growth advantage on cancer cells. Our results are well in line with this hypothesis showing increased TKTL1 expression and decreased PFK-1 expression in OSCC (significant inverse correlation). Zhang et al.[8] presented a similar mechanism describing a metabolic shift from glycolysis into the PPP [67] in OSCC. The authors conclude that the highly robust nature of OSCC metabolism implies that a systematic medical approach targeting multiple metabolic pathways is needed to improve cancer treatment. Downregulation of PFK-1 as observed in our study can be explained by an increase of natural inhibitors such as ATP, which is generated by OXPHOS, and citrate (from the citric acid cycle) that inhibits PFK-1 expression [241]. Therefore, we assume a metabolic shift [8, 67, 240, 241] of glucose from glycolysis towards the PPP mediated by the increased presence of PFK-1 inhibitors like ATP/citrate generated in OXPHOS (indicated by SDHA, SDHB, ATP synthase expression) during the carcinogenesis of OSCC.

If not provided by glycolysis, metabolites (pyruvate) for lactate production are available from amino acids [242]. Amino acid catabolism from the citric acid cycle (e.g. glutaminolysis) supports pyruvate anabolism leading to lactate and NADPH production [69, 242]. NADPH, pyruvate, and lactate itself have been proven to scavenge free radicals, thus protecting cancer cells from apoptosis [26]. However, this hypothesis of lactate anabolism through amino acids catabolism requires further studies in OSCC. Glutamine metabolism is also a cancer cell metabolic pathway important for both ATP production and providing intermediates for macromolecular synthesis. However, Glucose, not glutamine, was described as the dominant energy source required for proliferation and survival of head and neck squamous carcinoma cells [243]. This result does not automatically exclude lactate generation by amino acid catabolism, as the glutamine pathway has not been described for OSCC and has yet to be revealed. Finally, focusing on combination strategies [116, 158, 186, 244] (Table 2) with different signaling pathways (e.g. mTOR) [245] that have the potential to eradicate malignant and premalignant clones are warranted [245, 246].

For the first time, our study provides evidence of increased IGF-1R in OSCC. The expression of IGF-1R has been described for in-vitro analysis of an OSCC cell line [247] but not for carcinogenesis of OSCC yet. The authors state that IGF-1R activation is associated with resistance of EGFR-tyrosine-kinase inhibitor (TKI) treatment. Therefore, targeting IGF-1R pathway, reversal of hyperinsulinemia and IGF by dietry recommendations [16, 34, 199, 204‐217, 248] or metformin [232] may decrease resistance of EGFR-TKI as well as reduce the risk of cancer recurrence in tumor patients [34].

This study provides the first evidence of the expression of glycolysis-related proteins GLUT-1, HK 2, PFK-1, LDHA, TKTL1 and mitochondrial enzymes SDHA, SDHB, ATP synthase in the multi-step carcinogenesis of OSCC. It seems that both, hypoxia-related glucose metabolism and mitochondrial oxidative phosphorylation characteristics are associated with the carcinogenesis of OSCC. Acidosis and OXPHOS may drive a metabolic shift towards the PPP [67]. Therefore, inhibition of the PPP and glycolysis, as well as targeted anti-mitochondrial therapies (ROS generation) by natural compounds (polyphenol mix, selective vitamins) or synthetic vitamin derivatives (e.g. oxybenfotiamine) may act as sensitizer for apoptosis in cancer cells mediated by adjuvant therapies in OSCC. Summarizing in other words, targeting detoxifying systems (e.g. TKTL1, LDHA) make cancer cells or (oral) precursor lesions more vulnerable to apoptosis.