Conformational changes associated with activation of bee venom phospholipase A2

Ultrastructural variability of mitochondrial cristae induced in vitro by bee (Apis mellifera) venom and its derivatives, melittin and phospholipase A2, in isolated rat adrenocortical mitochondria

We tested the ability of bee venom (BV), melittin (Mlt), and phospholipase A2 (PLA) - used in 5 concentrations each (5, 10, 15, 20 and 40 μg/100 μl) - to promote ultrastructural changes and reorganization of cristae in vitro in mitochondria isolated from rat adrenal cortex after a protocol optimized by us. Thus, apart from two control grups (CI and CS), in which the mitochondria were suspended into saline buffer and isolation medium respectively, 15 more groups of mitochondria were constituted, corresponding to the five different doses of the three substance tested (BV5 to M40; M5 to M40 and P5 to P40). The ultrastructural effects were quantified on transmission electron micrographs using a morphometry software. Values of 84.49 nm and 95.45 nm were calculated for median diameters of mitochondrial cristae in two control groups. Large and very large vesicular cristae, many with 2 or 3 membranes, were generated depending on dose among normal cristae in all treated groups. In the BV and Mlt treated groups, after an initial increase (up to 127.27 nm in V15 group and 151.2 nm in M10 group) due to stimulation of cristae fusion, the cristae diameter diminished as the doses increased, mainly by the collapse of the cristae. In the PLA treated groups, the cristae diameter increased continuously from 83.84 nm to 136.01 nm, by stimulated fusion of cristae, only the two largest doses promoting the collapse of cristae in some mitochondria. The highest percentage of abnormal cristae was found in the Mlt treated groups and next in BV treated groups. All substances tested produced pronounced ultrastructural variability of mitochondrial cristae in vitro: they also changed (depending on dose) mitochondrial shapes, generated matrix debris and the highest concentrations of BV and Mlt were responsible for mitochondrial breakdown. These ultrastructural alterations of mitochondrial criste in the presence of the BV molecules suggest a reduced capacity of adrenocortical mitochondria to synthetize steroid hormones consequently to BV envenomations and partially explain the toxic effects of the BV.

Nonspecific Binding Domains in Lipid Membranes Induced by Phospholipase A2

Phospholipase A2 (PLA2) is a peripheral membrane protein that can hydrolyze phospholipids to produce lysolipids and fatty acids. It has been found to play crucial roles in various cellular processes and is thought as a potential candidate for triggering drug release from liposomes for medical treatment. Here, we directly observed that PLA2 hydrolysis reaction can induce the formation of PLA2-binding domains at lipid bilayer interface and found that the formation was significantly influenced by the fluidity of the lipid bilayer. We prepared supported lipid bilayers (SLBs) with various molar ratios of 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) and 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) to adjust the reactivity and fluidity of the lipid bilayers. A significant amount of the PLA2-induced domains was observed in mixtures of DPPC and DOPC (1,2-dioleoyl-sn-glycero-3-phosphocholine) but not in either pure DPPC or pure DOPC bilayer, which might be the reason that previous studies rarely observed these domains in lipid bilayer systems. The fluorescently labeled PLA2 experiment showed that newly formed domains acted as binding templates for PLA2. The AFM result showed that the induced domain has stepwise plateau structure, suggesting that PLA2 hydrolysis products may align as bilayers and accumulate layer by layer on the support, and the hydrophobic acyl chains at the side of the layer structure may be exposed to the outside aqueous environment. The introduced hydrophobic region could have hydrophobic interactions with proteins and therefore can attract the binding of not only PLA2 but also other types of proteins such as proteoglycans and streptavidin. The results suggest that the formation of PLA2-induced domains may convert part of a zwitterionic nonsticky lipid membrane to a site where biomolecules can nonspecifically bind.

Interfacial binding dynamics of bee venom phospholipase A2 investigated by dynamic light scattering and quartz crystal microbalance

Bee venom phospholipase A(2) (bvPLA(2)) is part of the secretory phospholipase A(2) (sPLA(2)) family whose members are active in biological processes such as signal transduction and lipid metabolism. While controlling sPLA(2) activity is of pharmaceutical interest, the relationship between their mechanistic actions and physiological functions is not well understood. Therefore, we investigated the interfacial binding process of bvPLA(2) to characterize its biophysical properties and gain insight into how membrane binding affects interfacial activation. Attention was focused on the role of membrane electrostatics in the binding process. Although dynamic light scattering experiments indicated that bvPLA(2) does not lyse lipid vesicles, a novel, nonhydrolytic activity was discovered. We employed a supported lipid bilayer platform on the quartz crystal microbalance with dissipation sensor to characterize this bilayer-disrupting behavior and determined that membrane electrostatics influence this activity. The data suggest that (1) adsorption of bvPLA(2) to model membranes is not primarily driven by electrostatic interactions; (2) lipid desorption can follow bvPLA(2) adsorption, resulting in nonhydrolytic bilayer-disruption; and (3) this desorption is driven by electrostatic interactions. Taken together, these findings provide evidence that interfacial binding of bvPLA(2) is a dynamic process, shedding light on how membrane electrostatics can modulate interfacial activation.

Activation of phospholipase A2 by temporin B: formation of antimicrobial peptide-enzyme amyloid-type cofibrils

Phospholipases A2 have been shown to be activated in a concentration dependent manner by a number of antimicrobial peptides, including melittin, magainin 2, indolicidin, and temporins B and L. Here we used fluorescently labelled bee venom PLA2 (PLA2D) and the saturated phospholipid substrate 1,2-dipalmitoyl-glycero-sn-3-phosphocholine (L-DPPC), exhibiting a lag-burst behaviour upon the initiation of the hydrolytic reaction by PLA2. Increasing concentrations of Cys-temporin B and its fluorescent Texas red derivative (TRC-temB) caused progressive shortening of the lag period. TRC-temB/PLA2D interaction was observed by Förster resonance energy transfer (FRET), with maximum efficiency coinciding with the burst in hydrolysis. Subsequently, supramolecular structures became visible by microscopy, revealing amyloid-like fibrils composed of both the activating peptide and PLA2. Reaction products, palmitic acid and 1-palmitoyl-2-lyso-glycero-sn-3-phosphocholine (lysoPC, both at >8 mol%) were required for FRET when using the non-hydrolysable substrate enantiomer 2,3-dipalmitoyl-glycero-sn-1-phosphocholine (D-DPPC). A novel mechanism of PLA2 activation by co-fibril formation and associated conformational changes is suggested.

Production of synthetically created phospholipase A(2) variants with industrial impact

Phospholipases A(2) (PLA(2)) play an important role for the production of lysophospholipids. Presently they are mainly obtained from porcine or bovine pancreas but these mammalian sources are not accepted in several fields of application. To make accessible a non-mammalian PLA(2) to industrial application, synthetic genes encoding PLA(2) from honey bee (Apis mellifera) with modified N-termini were constructed and expressed in Escherichia coli. While expression of the gene with an N-terminal leader sequence to direct the protein into the periplasm failed, four variants with slightly modified N-termini (I1A-PLA(2), I1V-PLA(2), His(6)-tagged PLA(2) and PLA(2) still containing the start methionine) were successfully expressed. In all cases, the PLA(2) variants were produced as inclusion bodies. Their protein content amounted to 26-35% of total cell protein. The optimized renaturation procedure and subsequent purification by cation-exchange chromatography yielded pure active enzymes in yields of 4-11 mg L(-1). The recombinant PLA(2) variants showed activities, far-UV CD and fluorescence spectra similar to the glycosylated PLA(2) isolated from the venom glands of honey bee (bv-PLA(2)). The thermodynamic stabilities of the recombinant enzymes calculated from the transition curves of guanidine hydrochloride induced unfolding were also nearly identical to the stability of bv-PLA(2). For the variant I1A-PLA(2) high-cell density fermentation in 10 L-scale using mineral salt medium was shown to increase the volumetric enzyme yield considerably.