The influence of PFK-II overexpression on neuroblastoma patients’ survival may be dependent on the particular isoenzyme expressed, PFKFB3 or PFKFB4

Nowadays, the metabolic reprogramming of cancer cells is considered a very promising field for development of novel therapeutic strategies. Apparently, the most evident alterations in cancer metabolism are related to glucose breakdown. As initially observed by Otto Warburg, in comparison to normal cells, the cancer cells ATP is generated mainly by glycolysis, even in normoxic conditions [1, 2]. Although glycolysis is less efficient than oxidative phosphorylation, it allows the generation of ATP at a much faster pace, which is highly beneficial for proliferating cancer cells. Additionally, elevation of glycolysis enables sufficient production levels of glycolytic intermediates required for de novo nucleotide, lipid, amino acid, and NADPH synthesis. The role of enhanced glucose breakdown in carcinogenesis is well confirmed by studies, showing that attenuation of glycolysis inhibits growth and decreases survival of cancer cells [3, 4]. Without doubt, an attractive anti-cancer approach would be to target glycolysis by inhibiting the aforementioned enzymes crucial for this process. Nevertheless, to do so, there must be a significant difference in the particular enzyme expression and/or activity between cancer cells and normal proliferating cells. In fact, studies in recent years revealed that PFKFB3/PFKFB4 isoenzymes are excellent candidates for glycolysis targeting, especially in cancer cells [5, 6]. PFKFB3 and PFKFB4 are cancer specific isoenzymes of the bifunctional 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase (PFKFB) enzyme group, responsible for controlling the steady-state cytoplasmic levels of fructose-2,6-bisphosphate, which allosterically activates phosphofructokinase-I (PFK-I), the key enzyme that catalyzes the rate-limiting step of glycolysis [5, 7]. Several in vitro studies revealed that targeting PFKFB3 and PFKFB4 in cancer cells results in glycolysis inhibition and in consequence, attenuation of tumor growth [8‐12]. Without doubt, it is worth noting that PFKFB4 and PFKFB3 isoenzymes are not only involved in glucose metabolism but may also regulate other processes crucial for carcinogenesis, having a multi-level anti-cancer effect [13‐16]. Taking into consideration all of the aforementioned, and the fact that there is limited clinical data related to this topic, in our current paper, using publicly available data sets deposited in the R2 data base, we analyzed the influence of PFKFB3 and PFKFB4 expression on overall survival of 49 independent, available cancer data sets. The results of the analysis indicates the importance of PFKFB3/4 co-expression for the clinical outcome of cancer patients.

Currently, several literature data indicates PFK-II as a potential marker of cancer prognosis [19‐25]. Nevertheless, up to this day, most research on PFK-II refers to one of the two cancer specific isoenzymes PFKFB3 or PFKFB4 [8‐12, 19, 20, 22‐25], undoubtedly neglecting the importance of co-expression of individual isoenzymes. Our analysis of publicly available data sets unquestionably indicates that in several tumors, high PFKFB3 and/or PFKFB4 expression correlates with poor survivability. Importantly, our analysis confirmed the previously published negative impact of PFKFB4 on the prognosis of glioma patients [23] and negative impact of PFKFB3 on the prognosis of hepatocellular cancer [24]. Interestingly, recently published data also revealed the negative impact of PFKFB4 on breast and bladder cancer patients but it was not noticable in the data sets we have analyzed. Nevertheless, Li et al. [22] and Ling et al. [20] analyzed the expression of PFKFB3 and PFKFB4 at the protein level in lung and breast cancer, which definitely may explain the discrepancy between the outcomes. Moreover, it may suggest that the same analysis, as we have done so far, could be performed at the protein level revealing more tumors for which expression of PFKFB3/PFKFB4 constitutes a potent prognostic marker. Undoubtedly, attention should be paid to the fact that in the case of neuroblastoma solely high expression of isoenzyme PFKFB3 correlates with poor prognosis whereas solely high expression of PFKFB4 is a positive prognostic factor for these patients. According to our current knowledge, there is no single study analyzing the influence of PFK-II on neuroblastoma patients’ prognosis, whereas our analysis clearly indicates that cancer specific isoenzymes may have opposite effects on prognosis in this group of specific cancer patients.

Although both genes code for phosphofructokinase II, according to the literature, kinase activity dominates in the PFKFB3 isoenzyme [26, 27], enhancing the glycolytic breakdown of glucose, whereas PFKFB4 seems to have higher FBPase-2 activity [5, 28], stimulating the flow of glucose toward the pentose phosphate pathway, providing the source of NADPH, crucial as a reducing factor for lipid biosynthesis and ROS-detoxifying enzymes. Altogether, we speculate, that the group with high PFKFB3 expression and low PFKFB4 may be characterized by higher glycolysis rates in relation to other groups whereas the PFKFB4 solely high group will have the lower glycolytic activity in comparison to other patients. If PFKFB3 dependent glycolytic cell activity in neuroblastoma indeed has significant impact on neuroblastoma cell proliferation [29], as determined by Almeida et al., one would expect to have opposite effects of PFKFB3 and PFKFB4 expression on patient survival as observed in our study. In addition, the analysis of the impact of isoenzymes on patients’ survival rates should also take into account that both, PFKFB3 and PFKFB4 are vastly involved in other important biological processes in a non-glycolysis-dependent manner [13‐16, 30‐32], which could also explain their opposite effect on neuroblastoma patients’ prognosing. For instance, it has been shown that PFKFB3 can be localized in the nucleus, which results in proliferation enhancement without increasing of the glycolytic rate. As reported by Yalcin et al. both, kinase activity of PFKFB3 and nuclear localization are needed for its effect on cancer cell proliferation [32]. As for PFKFB4, it has been reported that this enzyme is somewhat involved, be it directly or indirectly, in phosphorylation of the CBP-interacting domain of oncogenic steroid receptor coactivator-3 (SRC-3), enhancing its transcriptional activity, which results in higher expression of its target genes; transketolase, adenosine monophosphate deaminase-1 (AMPD1) and xanthine dehydrogenase (XDH), involved in the metabolism of nucleotides [13]. Regardless of the molecular mechanism of action of PFKFB3/4 isoenzymes our data clearly indicates that in analyzing the effect of PFK-II on the prognosis of cancer patients, the expression of both PFKFB3/4 isozymes should be considered. This is particularly important in the case of potential use of PFK-II inhibitors, because, as our analysis has undoubtedly shown, the effectiveness of inhibition may be dependent on the suitable inhibitor application for a specific isoenzyme overexpressed in specific cancer cells. What is more, our analysis indicates worryingly, that application of PFKFB4 inhibitors for neuroblastoma patients characterized by overexpression of both isoenzymes could even have a negative impact on the patient prognosis. Consequently, the results of our analysis provide important insight for future clinical oncology by presenting the importance of metabolic enzymes as a likely destination for modern targeted anti-tumor therapy.