No abstract

β-neurotoxins are enzymes, secreted phospholipases A2, that inhibit neurotransmission in neuromuscular synapses by poisoning the motoneuron. They were reviewed extensively several years ago (Pungerčar and Križaj, 2007). Here we present and critically discuss the most important experimental facts reported since then. Evidence has been presented for specific internalization of β-neurotoxins into the nerve endings of motoneurons, their in vivo binding to some cytosolic proteins, direct action on mitochondria, disruption of Ca(2+) homoeostasis and inhibition of amphiphysin function. New insights have led to a more confident interpretation of the action of these toxins at the molecular level. The most important questions that remain to be answered are listed.

The β-neurotoxic secreted phospholipases A2 (sPLA2s) block neuro-muscular transmission by poisoning nerve terminals. Damage inflicted by such sPLA2s (β-ntx) on neuronal mitochondria is characteristic, very similar to that induced by structurally homologous endogenous group IIA sPLA2 when its activity is elevated, as, for example, in the early phase of Alzheimer’s disease. Using ammodytoxin (Atx), the β-ntx from the venom of the nose-horned viper (Vipera a. ammodytes), the sPLA2 receptor R25 has been detected in neuronal mitochondria. This receptor has been purified from porcine cerebral cortex mitochondria by a new Atx-affinity-based chromatographic procedure. Mass spectrometry analysis revealed R25 to be the subunit II of cytochrome c oxidase (CCOX), an essential constituent of the respiratory chain complex. CCOX was confirmed as being the first intracellular membrane receptor for sPLA2 by alternative Atx-affinity-labellings of purified CCOX, supported also by the encounter of Atx and CCOX in PC12 cells. This discovery suggests the explanation of the mechanism by which β-ntx hinders production of ATP in poisoned nerve endings. It also provides a new insight into the potential function and dysfunction of endogenous GIIA sPLA2 in mitochondria.

Following the finding that ammodytoxin (Atx), a neurotoxic secreted phospholipase A2 (sPLA2) in snake venom, binds specifically to protein disulfide isomerase (PDI) in vitro we show that these proteins also interact in living rat PC12 cells that are able to internalize this group IIA (GIIA) sPLA2. Atx and PDI co-localize in both differentiated and non-differentiated PC12 cells, as shown by fluorescence microscopy. Based on a model of the complex between Atx and yeast PDI (yPDI), a three-dimensional model of the complex between Atx and human PDI (hPDI) was constructed. The Atx binding site on hPDI is situated between domains b and b’. Atx interacts hPDI with an extensive area on its interfacial binding surface. The mammalian GIB, GIIA, GV and GX sPLA2s have the same fold as Atx. The first three sPLA2s have been detected intracellularly but not the last one. The models of their complexes with hPDI were constructed by replacement of Atx with the respective mammalian sPLA2 in the Atx—hPDI complex and molecular docking of the structures. According to the generated models, mammalian GIB, GIIA and GV sPLA2s form complexes with hPDI very similar to that with Atx. The contact area between GX sPLA2 and hPDI is however different from that of the other sPLA2s. Heterologous competition of Atx binding to hPDI with GV and GX sPLA2s confirmed the model-based expectation that GV sPLA2 was a more effective inhibitor than GX sPLA2, thus validating our model. The results suggest a role of hPDI in the (patho)physiology of some snake venom and mammalian sPLA2s by assisting the retrograde transport of these molecules from the cell surface. The sPLA2–hPDI model constitutes a valuable tool to facilitate further insights into this process and into the (patho)physiology of sPLA2s in relation to their action intracellularly.

No abstract