Chemical synthesis of mouse cripto CFC variants.

We report for the first time the chemical synthesis of refolded CFC domain of mouse Cripto (mCFC) and of two variants bearing mutations on residues W107 and H104 involved in Alk4 binding. The domains undergo spontaneous and quantitative refolding in about 4 h, yet with very different kinetics. Disulfide linkages have been assessed by enzyme digestion and mass spectrometry analysis of resulting fragments, and the first experimental studies on structural organization have been conducted by circular dichroism spectroscopy under different pH conditions. Upon refolding, the domains considerably change their conformations, although they do not assume canonical structures, and become highly resistant to enzyme degradation. A comparative study of receptor binding shows that the CFC domain can bind Alk4 and confirms the importance of W107 and H104 for receptor recognition.

Solution structure of mouse Cripto CFC domain and its inactive variant Trp107Ala.

We report here for the first time the solution structures at pH 3 and pH 6 of the synthetic CFC domain of mouse Cripto and of the point mutated variant W107A that is unable to bind to the Alk4 Cripto receptor. NMR data confirm that the CFC domain has a C1-C4, C2-C6, C3-C5 disulfide pattern and show that structures are rather flexible and globally extended, with three noncanonical antiparallel strands. His104 and Trp107 side chains protrude from a protein edge and are strongly exposed to solvent, supporting previous evidence of direct involvement in receptor binding. On the opposite molecule side, several nonpolar residues are gathered, forming a large hydrophobic patch that supposedly acts as interface with the cell membrane or the adjacent EGF-like domain. A second hydrophilic patch surrounding His104 and Trp107 is present only in the wild type variant, suggesting a possible involvement in modulating Alk4 recognition.

Structural investigations on the Nodal-Cripto binding: a theoretical and experimental approach.

Nodal, a member of the transforming growth factor-ß superfamily, is a potent embryonic morphogen also implicated in tumor progression. Up to date structural information on the interaction of Nodal with its molecular partners are unknown. To deepen our understanding about mechanisms underlying both embryonic development and Nodal/Cripto-dependent tumor progression, we present here a molecular model of activin receptor-like kinase 4/Cripto/Nodal complex built by homology modeling as well as docking tests aimed at identifying potential binding epitopes. Starting from this model, we have predicted a large interaction surface on Nodal, which encompasses residues 43-69 and includes the prehelix loop and the H3 helix. This hypothesis has been subsequently assessed by surface plasmon resonance binding assays between the full-length Cripto and synthetic peptides reproducing the selected Nodal regions. In addition, the binding affinity between the full-length Nodal and Cripto proteins has been evaluated for the first time.

Structural insights into the interaction between the Cripto CFC domain and the ALK4 receptor.

The protein Cripto is the founding member of the extra-cellular EGF-CFC growth factors, which are composed of two adjacent cysteine-rich domains: the EGF-like and the CFC. Members of the EGF-CFC family play key roles in embryonic development and are also implicated in tumourigenesis. Cripto is highly over-expressed in many tumours, while it is poorly detectable in normal tissues. Although both Cripto domains are involved in its tumourigenic activity, the CFC domain appears to play a crucial role. Indeed, through this domain, Cripto interferes with the onco-suppressive activity of Activins, either by blocking the Activin receptor ALK4 or by antagonising proteins of the TGF-beta family. We have undertaken the chemical synthesis and the structural characterisation of human CFC Cripto domain. Using a combined NMR and computational approach, supported by binding studies by SPR, we have investigated the molecular basis of the interaction between h-CFC and ALK4. Binding studies indicate that the synthetic h-CFC interacts with the ALK4 receptor with a K(D) in micro M range, whereas it does not recognise the ActRIIB receptor. The NMR study shows that the h-CFC overall topology is determined by the presence of three disulfide bridges and that residues H120 and W124 are located between the first strand and the first loop with the side chains externally exposed. A model of the CFC-ALK4 complex has also been obtained by molecular docking and shows that all residues indicated by prior mutagenesis studies can contribute to the ALK4-CFC interaction at the protein-protein interface.

Abeta(25-35) and its C- and/or N-blocked derivatives: copper driven structural features and neurotoxicity.

The toxic properties of beta-amyloid protein, Abeta(1-42), the major component of senile plaques in Alzheimer's disease, depend on nucleation-dependent oligomerization and aggregation. In addition, Abeta(1-42) toxicity is favored by the presence of trace metals, which affect the secondary structure of the peptide. A peptide comprising 11 residues within Abeta(1-42) [Abeta(25-35)] aggregates and retains the neurotoxic activity of Abeta(1-42). We have used both Abeta(25-35) and its C-amidated or N-acetylated/C-amidated derivatives to investigate the role of copper(II) in modulating the conformation and aggregation state as well as the neurotoxic properties of amyloid peptides. Electrospray ionization mass spectrometry (ESI-MS) and electron paramagnetic resonance (EPR) measurements were performed to verify the formation of copper(II)/Abeta(25-35) complexes and to determine the coordination mode, respectively. Abeta(25-35) and its derivatives were analyzed by circular dichroism spectroscopy to assess their secondary structure, subjected to thioflavine-T (Th-T) binding assay to reveal beta-sheet structured aggregates formation, and imaged by scanning force microscopy. Toxicity was assessed on mature cultures of rat cortical neurons. We found that beta-sheet-structured species of Abeta(25-35) were neurotoxic, whereas the random-coil-structured derivatives were devoid of effect. Interestingly, copper promoted the random-coil/beta-sheet transition of Abeta(25-35), with ensuing peptide toxicity, but it induced the toxicity of the N-acetylated/C-amidated derivative without affecting peptide folding. Moreover, copper did not influence either the folding or the activity of the C-amidated Abeta(25-35), suggesting that blockade of the C-terminus of Abeta peptides might be sufficient to prevent Abeta toxicity.

The chemical synthesis of the GstI protein by NCL on a X-Met site.

The small GstI protein (63 amino acids) of Rhizobium leguminosarum is the endogenous inhibitor of the glnII (glutamine synthetase II) gene expression. It has been suggested that GstI has a predominantly beta-structure and mediates the block of translation and stabilization of glnII mRNA through direct binding to its 5' untranslated region. Because of the unavailability of adequate amounts of purified recombinant protein, the mechanism as well as the protein tridimensional structure remain very poorly understood. To obtain the full-length protein, we have undertaken the chemical synthesis of the protein by different approaches. In a first attempt, the stepwise synthesis was unsuccessful, with strong aggregation experienced on the N-terminal side, after residue 44 from the C-terminus. In a second approach, we set up the conditions to carry out a native chemical ligation (NCL). Albeit the protein contains two Cysteine residues, located at positions 40 and 47, to minimize the size of the N-terminal segment to be synthesized, we have devised an alternative strategy of ligation on Met32, utilizing homoCys as the ligating moiety and then alkylating the resulting polypeptide with methyl iodide. New conditions to quantitatively methylate thiol groups in complex polypeptides have been conceived, obtaining the protein in very good yields and purity. A CD spectroscopy investigation has revealed that the protein does not adopt canonical secondary structures but is very rich in beta-structure (approximately 60%), in agreement with a previous study carried out on samples obtained by recombinant methods.

A new ligand for immunoglobulin g subdomains by screening of a synthetic peptide library.

By screening a synthetic peptide library of general formula (NH(2)-Cys1-X2-X3-X4)(2)-Lys-Gly-OH, a disulfide-bridged cyclic peptide, where X2-X3-X4 is the tripeptide Phe-His-His, has been selected as a ligand for immunoglobulin G (IgG). The peptide, after a preliminary chromatographic characterization, has proved useful as a new affinity ligand for the purification of polyclonal as well as monoclonal antibodies from biological fluids, with recovery yields of up to 90% (90% purity). The ligand is able to bind antibody fragments containing both Fab and Fc from different antibody isotypes, a fact suggesting the presence of at least two different antibody-binding sites. While the recognition site on Fab is unknown, comparative binding studies with Fc, in association with the striking similarities of the peptide (named Fc-receptor mimetic, FcRM) with a region of the human FcgammaRIII receptor, strongly indicate that the peptide could recognize a short amino acid stretch of the lower hinge region, which has a key role in autoimmune disease triggering. The unique properties make the ligand attractive for both the purification of antibody fragments and as a lead for the generation of Fc-receptor antagonists.