Potential adverse effects to the retina of cancer therapy targeting pyruvate kinase M2

The Warburg effect was fi rst observed in 1920s when Otto Warburg and his team demonstrated that cancer cells make lactate from glycolysis despite the presence of oxygen [1]. His team also noted that normal mammalian retinal explants displayed aerobic glycolysis, but the fi nding was dismissed as experimental artefact because it did not fi t neatly with Warburg ’ s beliefs about cancer pathogenesis [1]. However, subsequent studies have confi rmed that the mammalian retina does, in fact, display a strong Warburg effect [2]. The presence of the Warburg effect in cancer and normal embryonic tissue can be explained by the cellular biosynthetic demands [3]. When energy supply (in the form of glucose) is suffi cient, prolif-erating cells direct the metabolic pathways away from oxidative phosphorylation towards biomass synthesis via aerobic glycolysis [3]. Cancer cells have the ability to oscillate between biosynthesis and energy production to meet the metabolic requirements. In adult mammalian retina, such biosynthetic demands exist due to the constant prodigious photoreceptor turnover in the rod outer segments [4]. Pyruvate kinase is a glycolytic enzyme which catalyses the conversion of phosphoenolpyruvate into pyruvate in the fi nal step of glycolysis [3]. How-ever, hypoxia-inducible factor 1 alpha (HIF-1 α) is a master regulator of transcription processes in response to hypoxia [3]. Both pyruvate kinase M2 isoform (PKM2) and HIF-1 α are established key mediators of the Warburg effect in cancer, with PKM2 reportedly coactivates HIF-1 α to promote the reprogramming of glucose metabolism [3]. A greater understanding of the molecular underpin-nings of the Warburg effect in recent years has moti-vated attempts to modulate PKM2 as a novel cancer therapy [5]. Goldberg et al. identifi ed several small interfer-ing RNAs (siRNAs) namely si25, si155 and si156 which specifi cally silenced PKM2 mRNA to induce apoptosis in cancer cells in vitro [6]. Interestingly, intratumoural delivery of si156 into xenograft tumour in mice demonstrated tumour regression [6]. In a high-throughput screen, Vander Heiden et al. identifi ed Compound 3 as an inhibitor of PKM2, postulated to target the allosteric regulatory site of PKM2 [6]. More potent shikonin and alkan-nin were subsequently shown to inhibit PKM2 selectively and downregulate glycolysis [7]. Con-versely, small-molecule activators of PKM2 such as DASA-58 and TEPP-46 also illustrated inhibitory effect on tumour xenograft growth in mice, possibly by interfering with the anabolic biosynthesis [8]. All the aforementioned agents are currently still in the preclinical development phase with in vitro and in vivo animal studies. Although the mechanism of the Warburg effect in the retina has not been elucidated, it is likely that key aspects of the molecular mechanism are conserved between the retina and cancer. Recent evidence indicates that both PKM2 and HIF-1 α are present in normal physiological mammalian