Perturbation of transforming growth factor (TGF)-beta1 association with latent TGF-beta binding protein yields inflammation and tumors

Transforming growth factor-beta (TGF-beta) activity is controlled at many levels including the conversion of the latent secreted form to its active state. TGF-beta is often released as part of an inactive tripartite complex consisting of TGF-beta, the TGF-beta propeptide, and a molecule of latent TGF-beta binding protein (LTBP). The interaction of TGF-beta and its cleaved propeptide renders the growth factor latent, and the liberation of TGF-beta from this state is crucial for signaling. To examine the contribution of LTBP to TGF-beta function, we generated mice in which the cysteines that link the propeptide to LTBP were mutated to serines, thereby blocking covalent association. Tgfb1(C33S/C33S) mice had multiorgan inflammation, lack of skin Langerhans cells (LC), and a shortened lifespan, consistent with decreased TGF-beta1 levels. However, the inflammatory response and decreased lifespan were not as severe as observed with Tgfb1(-/-) animals. Tgfb1(C33S/C33S) mice exhibited decreased levels of active TGF-beta1, decreased TGF-beta signaling, and tumors of the stomach, rectum, and anus. These data suggest that the association of LTBP with the latent TGF-beta complex is important for proper TGF-beta1 function and that Tgfb1(C33S/C33S) mice are hypomorphs for active TGF-beta1. Moreover, although mechanisms exist to activate latent TGF-beta1 in the absence of LTBP, these mechanisms are not as efficient as those that use the latent complex containing LTBP.

The role of antibodies specific for toxic sPLA2s and haemorrhagins in neutralizing potential of antisera raised against Vipera ammodytes ammodytes venom

The contribution of antibodies directed against the two main toxic groups of proteins in the Vipera ammodytes ammodytes venom, haemorrhagic metalloproteinases (H) and neurotoxic sPLA2s (Atxs), to the overall protective efficacy of the whole venom antisera was investigated. Using ELISA assays we established a high correlation between the protective efficacy of the whole venom antisera in mice and their anti-Atxs antibody content. As the haemorrhage is the prevailing toxic effect of the venom in human, the lack of correlation also with anti-H IgG content exposed that the mouse model might not be optimal to evaluate the neutralizing potential of the venom-specific antisera for human therapy. We further revealed that Atxs and structurally very similar but non-toxic AtnI2 from the venom are not immuno cross-reactive.

Binding to the high-affinity M-type receptor for secreted phospholipases A2 is not obligatory for the presynaptic neurotoxicity of ammodytoxin A

R180, isolated from porcine brain cortex, is a high-affinity membrane receptor for ammodytoxin A (AtxA), a secreted phospholipase A(2) (sPLA(2)) and presynaptically active neurotoxin from venom of the long-nosed viper (Vipera ammodytes ammodytes). As a member of the M-type sPLA(2) receptors, present on the mammalian plasma membrane, R180 has been proposed to be responsible for one of the first events in the process of presynaptic neurotoxicity, the binding of the toxin to the nerve cell. To test this hypothesis, we prepared and analyzed three N-terminal fusion proteins of AtxA possessing a 12 or 5 amino acid residue peptide. The presence of such an additional "propeptide" prevented interaction of the toxin with the M-type receptor but not its lethality in mouse and neurotoxic effects on a mouse phrenic nerve-hemidiaphragm preparation. In addition, antibodies raised against the sPLA(2)-binding C-type lectin-like domain 5 of the M-type sPLA(2) receptor were unable to abolish the neurotoxic action of AtxA on the neuromuscular preparation. The specific enymatic activities of the fusion AtxAs were two to three orders of magnitude lower from that of the wild type, yet resulting in a similar but less pronounced neurotoxic profile on the neuromuscular junction. This is in accordance with other data showing that a minimal enzymatic activity suffices for presynaptic toxicity of sPLA(2)s to occur. Our results indicate that the interaction of AtxA with the M-type sPLA(2) receptor at the plasma membrane is not essential for presynaptic activity of the toxin. Interaction of AtxA with two intracellular proteins, calmodulin and the R25 receptor, was affected but not prevented by the presence of the N-terminal fusion peptides, implying that these proteins may play a role in the sPLA(2) neurotoxicity.

Fibronectin is required for integrin alphavbeta6-mediated activation of latent TGF-beta complexes containing LTBP-1

Transforming growth factor-betas (TGF-beta) are secreted as latent complexes consisting of the TGF-beta dimer, the TGF-beta propeptide dimer, and the latent TGF-beta binding protein (LTBP). Although the bonds between TGF-beta and its propeptide are cleaved intracellulary, the propeptide associates with TGF-beta by electrostatic interactions, thereby conferring latency to the complex. We reported that a specific sequence of LTBP-1 is required for latent TGF-beta activation by the integrin alphavbeta6. Here we describe a 24 amino acid sequence from the hinge domain required for activation. The LTBP-1 polypeptide rL1N, which includes the hinge, associates with fibronectin in binding assays. We present evidence that fibronectin null cells minimally activate latent TGF-beta and poorly incorporate the active hinge sequence into their matrix. In addition, cells missing the fibronectin receptor alpha5beta1 exhibit defective activation of latent TGF-beta by alphavbeta6 and decreased matrix incorporation. The results indicate specificity for integrin-mediated latent TGF-beta activation that include unique sequences in LTBP-1 and an appropriate matrix molecule.

The C-terminal and β-wing regions of ammodytoxin A, a neurotoxic phospholipase A2 from Vipera ammodytes ammodytes, are critical for binding to factor Xa and for anticoagulant effect

Ammodytoxin A (AtxA) from the venom of Vipera ammodytes ammodytes belongs to group IIA secreted phospholipase A2 (sPLA2), for which the major pathologic activity is presynaptic neurotoxicity. We show here that this toxin also affects hemostasis because it exhibits strong anticoagulant activity. AtxA binds directly to human coagulation factor Xa (FXa) with Kdapp of 32 nM, thus inhibiting the activity of the prothrombinase complex with an IC50 of 20 nM. To map the FXa-interaction site on AtxA, various mutants of AtxA produced by site-directed mutagenesis and expressed in Escherichia coli were tested in the study. In surface plasmon resonance (SPR) measurements, with FXa covalently attached to the sensor chip, we show that the FXa-binding site on AtxA includes several basic amino acid residues at the C-terminal and beta-wing regions of the molecule. Applying an in vitro biological test for inhibition of prothrombinase activity, we further demonstrate that the same residues are also very important for the anticoagulant activity of AtxA. We conclude that the anticoagulant site of AtxA is located in the C-terminal and beta-wing regions of this phospholipase A2. Synthetic peptides comprising residues of the deduced anticoagulant site of AtxA provide a basis to synthesize novel anticoagulant drugs.

Identification of a novel binding site for calmodulin in ammodytoxin A, a neurotoxic group IIA phospholipase A2

The molecular mechanism of the presynaptic neurotoxicity of snake venom phospholipases A2 (PLA2s) is not yet fully elucidated. Recently, new high-affinity binding proteins for PLA2 toxins have been discovered, including the important intracellular Ca2+ sensor, calmodulin (CaM). In the present study, the mode of interaction of group IIA PLA2s with the Ca2+-bound form of CaM was investigated by mutational analysis of ammodytoxin A (AtxA) from the long-nosed viper (Vipera ammodytes ammodytes). Several residues in the C-terminal part of AtxA were found to be important in this interaction, particularly those in the region 115-119. In support of this finding, introduction of Y115, I116, R118 and N119, present in AtxA, into a weakly neurotoxic PLA2 from Russell's viper (Daboia russellii russellii) increased by sevenfold its binding affinity for CaM. Furthermore, two out of four peptides deduced from different regions of AtxA were able to compete with the toxin in binding to CaM. The nonapeptide showing the strongest inhibition was that comprising the AtxA region 115-119. This stretch contributes to a distinct hydrophobic patch within the region 107-125 in the C-terminal part of the molecule. This lacks any substantial helical structure and is surrounded by several basic residues, which may form a novel binding motif for CaM on the molecular surface of the PLA2 toxin.

The neurotoxic phospholipase A2 associates, through a non-phosphorylated binding motif, with 14-3-3 protein gamma and epsilon isoforms

Two novel acceptors for ammodytoxin C, a presynaptically neurotoxic phospholipase A(2) from snake venom, have been purified from porcine cerebral cortex by a toxin-affinity-based procedure. Using tandem mass spectrometry, the isolated acceptors were identified as 14-3-3 gamma and epsilon isoforms, highly conserved cytoplasmic proteins involved in the regulation of numerous physiological processes. The interaction between ammodytoxin C and 14-3-3 proteins is direct and not mediated by calmodulin, a high-affinity acceptor for both ammodytoxin C and 14-3-3 proteins, as demonstrated in pull-down experiments and by surface plasmon resonance. The latter technique gave an apparent dissociation constant of 1.0+/-0.2 microM for the interaction between chip-immobilized 14-3-3 and ammodytoxin C. 14-3-3 usually interacts with proteins through specific phospho-Ser/Thr motifs. Ammodytoxin C is not a phospho-protein, therefore the interaction must occur through a non-phosphorylated binding site, most probably the KEESEK sequence at its C-terminal end. The interaction we describe suggests an explanation for the pathophysiological effects evoked by some secreted phospholipases A(2), such as the inhibition of protein phosphorylation, of terminal ion currents, and of neurotransmission, as well as the initiation of neuronal cell death, all processes regulated by 14-3-3 proteins.

The C-terminal region of ammodytoxins is important but not sufficient for neurotoxicity

Ammodytoxins (Atxs) are presynaptically acting snake venom phospholipase A2 (PLA2) toxins the molecular mechanism of whose neurotoxicity is not completely understood. Two chimeric PLA2s were prepared by replacing the C-terminal part of a nontoxic venom PLA2, ammodytin I2, with that of AtxA(K108N). The chimeras were not toxic, but were able to bind strongly to an Atxs-specific neuronal receptor, R25. They also showed an increased affinity for calmodulin, a recently identified high-affinity binding protein for Atxs, whereas affinity for a neuronal M-type PLA2 receptor remained largely unchanged. The results show that the C-terminal region of Atxs, which is known to be involved in neurotoxicity, is critical for their interaction with specific binding proteins, but that some other part of the molecule also contributes to toxicity.

Charge reversal of ammodytoxin A, a phospholipase A2-toxin, does not abolish its neurotoxicity

The positive charge concentrated at the C-terminal region of ammodytoxin (Atx) A, which is involved in presynaptic toxicity, has been reversed. A six-site mutant of AtxA (K108N/K111N/K127T/K128E/E129T/K132E , where K108N=Lys(108)-->Asn etc. ) was prepared, in which five out of seven C-terminal basic amino acid residues were substituted with neutral or acidic ones. The mutant was approximately 30-fold less lethal, but still neurotoxic. Consistent with this, its binding affinity for the neuronal receptors decreased by only a factor of five. Additionally, a single-site mutant of AtxA was prepared, with substitution at only one position (K127T) out of six mutated in the six-site mutant. Its toxicity indicated that most, if not all, of the six mutated residues partially contribute to the decreased lethality of the multiple-site mutant.