The moonlighting concept in molecular biology
The term “moonlighting” in protein biology refers to proteins that carry out two or more biochemically distinct functions — functions that are not simply the same enzymatic activity applied in a different location, but genuinely different activities arising from the same polypeptide chain. The concept was formally articulated by Jeffery in a series of papers beginning in 1999, though individual examples had been described earlier. Moonlighting appears to be far more common than initially appreciated, partly because most biochemical studies characterise proteins in isolation in defined assays, making it easy to miss secondary activities that require specific post-translational modifications or interaction partners.
Glycolytic enzymes are, collectively, among the most intensively studied moonlighting proteins. They are abundant, highly conserved across evolution, and present in essentially every cell type. Their canonical glycolytic functions are well-characterised. But beginning in the 1990s and accelerating through the 2000s and 2010s, studies began revealing that glycolytic enzymes — particularly GAPDH, alpha-enolase (ENO1), and the M2 isoform of pyruvate kinase (PKM2) — have secondary activities in the nucleus, at the cell surface, and in signalling pathways that are entirely distinct from glycolysis. In cancer biology especially, these moonlighting functions have taken on considerable importance because cancer cells dramatically upregulate glycolytic enzyme expression (as part of the Warburg effect), and higher protein abundance means more substrate available for nuclear and non-metabolic functions.
GAPDH: from glycolysis to the nucleus via S-nitrosylation
Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) catalyses the sixth step of glycolysis — the oxidative phosphorylation of glyceraldehyde-3-phosphate to 1,3-bisphosphoglycerate, using NAD+ as electron acceptor. It is a tetrameric enzyme, among the most abundant proteins in mammalian cells, and its mRNA and protein are so constitutively expressed that GAPDH became the default “housekeeping” normalisation standard for Western blotting — a practice that subsequent moonlighting research has complicated, since GAPDH expression is not in fact invariant across conditions.
The nuclear translocation of GAPDH is triggered by S-nitrosylation — the covalent addition of a nitric oxide (NO) group to the active-site cysteine (Cys152 in human GAPDH). S-nitrosylation is catalysed by nitric oxide synthase (NOS) isoforms, particularly inducible NOS (iNOS) during inflammatory stimulation. When Cys152 is S-nitrosylated, GAPDH loses its glycolytic activity but gains affinity for the nuclear import protein importin-α (also known as KPNA3/4). The GAPDH-importin-α complex is then transported into the nucleus.
Once nuclear, S-nitrosylated GAPDH has several documented activities. It S-nitrosylates SIRT1 (in a transnitrosylation reaction), which inhibits SIRT1 deacetylase activity on p53. Because SIRT1 normally deacetylates p53 and limits its transcriptional activity, GAPDH-mediated SIRT1 inhibition indirectly activates p53 — contributing to the apoptotic response to nitrosative stress. GAPDH also participates in nuclear DNA repair through its interaction with the base excision repair enzyme APE1 (apurinic/apyrimidinic endonuclease 1). In addition, nuclear GAPDH has been implicated in mRNA stability and in autophagy induction through a distinct mechanism involving its interaction with mTOR: GAPDH can directly bind to mTORC1 and inhibit it under energy stress conditions, providing an additional node of mTOR regulation that is independent of the canonical AMPK-TSC2 axis.
The significance for cancer biology is considerable. Tumours with high iNOS activity (common in many inflammatory cancers) generate nitric oxide that S-nitrosylates GAPDH and promotes nuclear accumulation. But the consequences of nuclear GAPDH are not simply pro-apoptotic — the outcome depends on the cellular context and the downstream binding partners available. In some settings, nuclear GAPDH appears to promote survival rather than apoptosis, possibly through its DNA repair functions. The directionality of the effect is therefore context-dependent.
Alpha-enolase (ENO1): glycolytic enzyme, plasminogen receptor, and MBP-1
Alpha-enolase (ENO1) catalyses the ninth step of glycolysis — the dehydration of 2-phosphoglycerate to phosphoenolpyruvate (PEP). Like GAPDH, it is a homodimeric enzyme expressed in essentially all tissues. The moonlighting story for ENO1 unfolds at two distinct locations: the cell surface and the nucleus.
At the cell surface, ENO1 functions as a receptor and activator for plasminogen, the precursor to plasmin — the primary fibrinolytic enzyme. ENO1 at the cell surface binds plasminogen with high affinity, and the membrane-bound complex recruits tissue plasminogen activator (tPA) to enhance plasmin generation locally. This cell surface protease activity enhances cell invasion and extracellular matrix degradation. In cancer biology, surface ENO1 expression correlates with invasive phenotype, and several cancer types overexpress ENO1 on their surface in ways that appear to promote metastasis. ENO1 antibodies are being investigated as diagnostic biomarkers and potential therapeutic targets in pancreatic cancer and small cell lung cancer.
In the nucleus, ENO1 undergoes an entirely different transformation. A shorter isoform of ENO1, produced from an internal ribosome entry site within the ENO1 mRNA, lacks the N-terminal sequence of the full-length protein and adopts a different localisation — nuclear rather than cytosolic. This shorter isoform is known as MBP-1 (c-myc promoter binding protein-1). MBP-1 was originally identified as a transcriptional repressor that binds to the TATA box of the c-myc promoter and inhibits c-myc transcription. Because c-myc is a major oncogene that drives proliferation, ribosome biogenesis, and metabolic reprogramming, MBP-1-mediated c-myc repression is an anti-proliferative function. Reduced MBP-1 expression has been documented in breast cancer, and restoration of MBP-1 in cancer cell lines reduces c-myc expression and suppresses growth.
The same gene, ENO1, therefore contributes both a pro-invasive surface protease activity and an anti-proliferative transcriptional repressor activity, depending on which translation initiation site is used and which cellular compartment the resulting protein traffics to. The balance between these functions is regulated at multiple levels and is sensitive to the metabolic and stress state of the cell.
PKM2: protein kinase, STAT3 phosphorylator, and HIF-1α coactivator
Pyruvate kinase M2 (PKM2) is perhaps the most intensively studied glycolytic moonlighting protein in contemporary cancer biology. PKM2 catalyses the terminal glycolytic step — transfer of the high-energy phosphate from PEP to ADP to generate pyruvate and ATP. This reaction is thermodynamically irreversible and represents a major ATP-generating step of glycolysis.
PKM2 exists in two quaternary states with profoundly different properties. The tetrameric form is the catalytically active pyruvate kinase. The dimeric form has low pyruvate kinase activity. When PKM2 is in the dimeric state, glycolytic flux to pyruvate decreases — but rather than being a simple loss of function, the dimer has entirely new activities. The PKM2 dimer translocates to the nucleus, where it functions as both a protein kinase and a transcriptional coactivator.
As a nuclear protein kinase, PKM2 phosphorylates STAT3 at Tyr705 — the same activating phosphorylation normally attributed to JAK2 and other tyrosine kinases. This was described by Guo and colleagues in a 2012 Molecular Cell paper that demonstrated PKM2-driven STAT3 activation promoting expression of MEK5 and driving cell proliferation in glioblastoma. STAT3 is itself a major oncogenic transcription factor — its activation drives expression of BCL-2, cyclin D1, and c-myc, among others.
PKM2 also functions as a transcriptional coactivator for HIF-1α (hypoxia-inducible factor 1-alpha). Under hypoxic conditions or conditions of EGF receptor signalling, nuclear PKM2 binds directly to HIF-1α and promotes its transactivation of target genes including LDHA, GLUT1, and PDK1 — glycolytic genes that further entrench the Warburg phenotype. This creates a positive feedback loop: PKM2 dimers in the nucleus coactivate HIF-1α, which transcribes more glycolytic enzymes including PKM2 itself, maintaining high levels of the dimeric nuclear form.
The acetylation of PKM2 at Lys433 by the acetyltransferase p300 is a key post-translational modification that promotes nuclear translocation. Acetyl-CoA availability — which varies with metabolic state — therefore influences how much PKM2 enters the nucleus. This connects PKM2 moonlighting to the broader metabolic epigenomics theme in which metabolite concentrations directly regulate chromatin-associated activities.
The tetramer-to-dimer switch is regulated by several allosteric mechanisms. Fructose-1,6-bisphosphate (FBP) stabilises the tetramer. Phosphotyrosine peptides (from activated growth factor receptors) bind PKM2 and displace FBP, destabilising the tetramer and promoting dimer formation. AMPK-mediated phosphorylation of PKM2 at Ser37 promotes nuclear translocation of the dimeric form. Conversely, conditions that elevate FBP (high glycolytic flux) or conditions that activate AMPK in ways that favour tetrameric assembly will oppose nuclear moonlighting.
How spirulina intersects with glycolytic moonlighting
The connections between spirulina and glycolytic moonlighting follow from three well-documented effects of spirulina’s components on cell metabolism: reduction of LDHA activity and glycolytic flux, AMPK activation, and modulation of nitric oxide signalling.
Phycocyanin has been shown in multiple studies to reduce LDHA expression and activity in cancer cell lines. This reduction in LDHA reduces the conversion of pyruvate to lactate — one metric of glycolytic flux through the terminal steps. Reduced glycolytic flux is expected to lower fructose-1,6-bisphosphate concentrations (an upstream glycolytic intermediate whose levels track with glycolytic rate), and since FBP stabilises PKM2 tetramer, reduced FBP would shift the equilibrium toward PKM2 dimer formation. This sounds counterproductive — if spirulina reduces glycolysis, why would it promote the pro-cancer dimeric form of PKM2?
The answer lies in the concurrent AMPK activation. Spirulina components, particularly phycocyanobilin, activate AMPK. The effects of AMPK on PKM2 are not straightforwardly through FBP but through direct phosphorylation that has complex effects on localisation. More importantly, the net effect of AMPK activation on cancer cell metabolism is broadly anti-proliferative: AMPK inhibits mTORC1, reduces protein synthesis, shifts metabolism toward oxidative phosphorylation, and activates FOXO transcription factors that promote cell cycle arrest. In this context, even if some PKM2 dimer formation occurs, the broader AMPK-driven reprogramming limits the pro-proliferative consequences.
For GAPDH moonlighting, the relevant spirulina connection is nitric oxide signalling. Phycocyanobilin inhibits NADPH oxidase (NOX) enzymes, which reduces superoxide production. Superoxide reacts with nitric oxide to form peroxynitrite, consuming NO and reducing its effective concentration. By reducing superoxide through NOX inhibition, phycocyanobilin paradoxically preserves NO availability — the so-called NOX-eNOS uncoupling argument. But the specific consequence for GAPDH S-nitrosylation depends on whether iNOS-derived NO (the pathological, high-flux NO of inflammation) is reduced or whether eNOS-derived NO (the physiological, low-flux NO of vascular tone) is preserved. Phycocyanin and phycocyanobilin appear to reduce iNOS expression (consistent with their NF-κB inhibitory activity), which would reduce the nitrosative stress responsible for pathological GAPDH nuclear translocation in inflammatory contexts.
For ENO1/MBP-1, the connection is more speculative. The ratio of full-length ENO1 to the MBP-1 isoform depends on internal ribosome entry site usage, which is regulated by the translational stress response and metabolic state. A detailed study of how spirulina treatment affects ENO1 isoform ratios has not been published. This remains a hypothesis worth investigating.
Frontier science: what direct studies would be needed
The glycolytic moonlighting literature is genuinely on the frontier of cancer cell biology. Many of the key findings — PKM2 as a nuclear kinase, GAPDH transnitrosylation, ENO1/MBP-1 switching — date from the 2010s and are still being characterised mechanistically. The application of these findings to spirulina research requires studies that measure not just total enzyme expression but subcellular localisation (cytosolic versus nuclear), post-translational modification state (S-nitrosylation of GAPDH, acetylation of PKM2), and isoform balance (ENO1 full-length versus MBP-1).
Standard transcriptomic or proteomic studies of spirulina-treated cells would not capture these dynamics unless specifically designed to do so. What would be required are: nuclear fractionation followed by specific immunoblotting for nuclear GAPDH, PKM2, and ENO1/MBP-1 in spirulina-treated versus untreated cancer cells; ChIP experiments asking whether spirulina treatment affects PKM2 occupancy at HIF-1α target gene promoters; and iNOS knockdown experiments asking whether the effects of phycocyanin on GAPDH nuclear localisation are dependent on iNOS activity.
None of these studies appear to have been published at the time of writing. The moonlighting connections described in this post are therefore mechanistic hypotheses — well-grounded in the independent literatures of glycolytic moonlighting and spirulina biochemistry, but not yet directly tested at their intersection. That is what makes this genuinely frontier science, and it is worth saying clearly.