Predicted calmodulin-binding sequence in the gamma subunit of phosphorylase b kinase

Pathogen-driven coevolution across the CBP60 plant immune regulator subfamilies confers resilience on the regulator module

Components of the plant immune signaling network need mechanisms that confer resilience against fast-evolving pathogen effectors that target them. Among eight Arabidopsis CaM-Binding Protein (CBP) 60 family members, AtCBP60g and AtSARD1 are partially functionally redundant, major positive immune regulators, and AtCBP60a is a negative immune regulator. We investigated possible resilience-conferring evolutionary mechanisms among the CBP60a, CBP60g and SARD1 immune regulatory subfamilies. Phylogenetic analysis was used to investigate the times of CBP60 subfamily neofunctionalization. Then, using the pairwise distance rank based on the newly developed analytical platform Protein Evolution Analysis in a Euclidean Space (PEAES), hypotheses of specific coevolutionary mechanisms that could confer resilience on the regulator module were tested. The immune regulator subfamilies diversified around the time of angiosperm divergence and have been evolving very quickly. We detected significant coevolutionary interactions across the immune regulator subfamilies in all of 12 diverse core eudicot species lineages tested. The coevolutionary interactions were consistent with the hypothesized coevolution mechanisms. Despite their unusually fast evolution, members across the CBP60 immune regulator subfamilies have influenced the evolution of each other long after their diversification in a way that could confer resilience on the immune regulator module against fast-evolving pathogen effectors.

Structural diversity of calmodulin binding to its target sites

Calmodulin (CaM) is a ubiquitous, highly conserved, eukaryotic protein that binds to and regulates a number of diverse target proteins involved in different functions such as metabolism, muscle contraction, apoptosis, memory, inflammation and the immune response. In this minireview, we analyze the large number of CaM-complex structures deposited in the Protein Data Bank (i.e. crystal and nuclear magnetic resonance structures) to gain insight into the structural diversity of CaM-binding sites and mechanisms, such as those for CaM-activated protein kinases and phosphatases, voltage-gated Ca(2+)-channels and the plasma membrane Ca(2+)-ATPase.

Calmodulin signaling: analysis and prediction of a disorder-dependent molecular recognition

Calmodulin (CaM) signaling involves important, wide spread eukaryotic protein-protein interactions. The solved structures of CaM associated with several of its binding targets, the distinctive binding mechanism of CaM, and the significant trypsin sensitivity of the binding targets combine to indicate that the process of association likely involves coupled binding and folding for both CaM and its binding targets. Here, we use bioinformatics approaches to test the hypothesis that CaM-binding targets are intrinsically disordered. We developed a predictor of CaM-binding regions and estimated its performance. Per residue accuracy of this predictor reached 81%, which, in combination with a high recall/precision balance at the binding region level, suggests high predictability of CaM-binding partners. An analysis of putative CaM-binding proteins in yeast and human strongly indicates that their molecular functions are related to those of intrinsically disordered proteins. These findings add to the growing list of examples in which intrinsically disordered protein regions are indicated to provide the basis for cell signaling and regulation.

Interactions between the juxtamembrane domain of the EGFR and calmodulin measured by surface plasmon resonance

One early response to epidermal growth factor receptor (EGFR) activation is an increase in intracellular calcium. We have used surface plasmon resonance (SPR) to study real-time interactions between the intracellular juxtamembrane (JM) region of EGFR and calmodulin. The EGFR-JM (Met(644)-Phe(688)) was expressed as a GST fusion protein and immobilised on a sensor chip surface. Calmodulin specifically interacts with EGFR-JM in a calcium-dependent manner with a high on and high off rate. Chemical modification of EGFR-JM by using arginine-selective phenylglyoxal or deletion of the basic segment Arg(645)-Arg(657) inhibits the interaction. Phosphorylation of EGFR-JM by protein kinase C (PKC) or glutamate substitution of Thr(654) inhibits the interaction, suggesting that PKC phosphorylation electrostatically interferes with calmodulin binding to basic arginine residues. Calmodulin binding was also inhibited by suramin. Our results suggest that EGFR-JM is essential for epidermal growth factor (EGF)-mediated calcium-calmodulin signalling and for signal integration between other signalling pathways.

The Ca2+-dependent binding of calmodulin to an N-terminal motif of the heterotrimeric G protein beta subunit

Ca2+ ion concentration changes are critical events in signal transduction. The Ca2+-dependent interactions of calmodulin (CaM) with its target proteins play an essential role in a variety of cellular functions. In this study, we investigated the interactions of G protein betagamma subunits with CaM. We found that CaM binds to known betagamma subunits and these interactions are Ca2+-dependent. The CaM-binding domain in Gbetagamma subunits is identified as Gbeta residues 40-63. Peptides derived from the Gbeta protein not only produce a Ca2+-dependent gel mobility shifting of CaM but also inhibit the CaM-mediated activation of CaM kinase II. Specific amino acid residues critical for the binding of Gbetagamma to CaM were also identified. We then investigated the potential function of these interactions and showed that binding of CaM to Gbetagamma inhibits the pertussis toxin-catalyzed ADP-ribosylation of Galphao subunits, presumably by inhibiting heterotrimer formation. Furthermore, we demonstrated that interaction with CaM has little effect on the activation of phospholipase C-beta2 by Gbetagamma subunits, supporting the notion that different domains of Gbetagamma are responsible for the interactions of different effectors. These findings shed light on the molecular basis for the interactions of Gbetagamma with Ca2+-CaM and point to the potential physiological significance of these interactions in cellular functions.

Characterization of the interaction between beta2-glycoprotein I and calmodulin, and identification of a binding sequence in beta2-glycoprotein I

beta2-Glycoprotein I was shown to bind reversibly to calmodulin in a Ca2+-dependent manner with a 1:1 stoichiometry, a Kd of 3 x 10(-9) M and a Hill coefficient of 1.4. A sequence in beta2-glycoprotein I (Lys-Pro-Gly-Tyr-Val-Ser-Arg-Gly-Gly-Met-Arg-Lys-Phe-Ile-) limited by Cys-32 and Cys-47 is suggested to be the calmodulin-binding region. This sequence was the only one in beta2-glycoprotein I theoretically having the ability to form a basic amphiphilic alpha-helix typical of a calmodulin binding sequence. The peptide corresponding to this sequence was synthesized and found to inhibit the interaction between beta2-glycoprotein I and calmodulin with an IC50 value of 0.38 x [beta2-glycoprotein I] and to displace the beta2-glycoprotein I from the beta2-glycoprotein I/calmodulin complex with an IC50 value of 0.90 x [beta2-glycoprotein I].

Identification, characterization, and comparison of the calmodulin-binding domains of the endothelial and inducible nitric oxide synthases

The calmodulin (CaM)-binding regions in bovine endothelial nitric oxide synthase (eNOS) and murine inducible nitric oxide synthase (iNOS) are identified in this study as eNOS residues 493-512 and iNOS residues 501-532. Peptides corresponding to eNOS 493-512 and NOS 501-532 produce a (Ca2+)-dependent, electrophoretic mobility shift of CaM on 4 M urea gels. The two peptides are also potent inhibitors of the CaM-mediated activation of neuronal nitric oxide synthase and have dissociation constants for CaM binding of 4.0 and 1.5 nM respectively. Substitution of eNOS and iNOS CaM-binding domains in eNOS/iNOS chimeric proteins produces major alterations in the Ca2+ and CaM dependence of the intact enzymes expressed and purified from a baculovirus/Sf9 insect cell system. Replacement of aligned NOS sequence with eNOS 493-512 creates a functional, chimeric iNOS that is both (Ca2+)- and CaM-dependent. Replacement of aligned eNOS sequence with NOS 501-532 creates a functional, chimeric eNOS that is CaM-independent but that remains (Ca2+)-dependent. Specific amino acid residues critical for CaM binding by eNOS are also identified in this study as Phe-498, Lys-499, and Leu-511 in the bovine eNOS sequence.

The anticalmodulin effect of aflatoxin B1 on purified erythrocyte Ca(2+)-ATPase

The genotoxic carcinogen aflatoxin B1 (AFB1) inhibited the calmodulin-stimulated membrane-bound (Ca2+Mg2+)-ATPase. Using the purified enzyme, 12 nmoles per ml of AFB1 caused maximum inhibition of 28% and 50%, of the acidic phospholipid-stimulated and calmodulin-activated Ca(2+)-ATPase activity respectively. Treatment of red cell ghosts with increasing concentrations of Triton X-100, a non-ionic detergent caused a progressive loss of both the basal and calmodulin-stimulated Ca(2+)-ATPase activity. The activity of the phospholipid-free, detergent-solubilized enzyme was almost fully restored by phosphatidyl serine (PS) and its sensitivity to calmodulin was restored in the presence of phosphatidyl choline (PC). Analysis of the results obtained using varying concentrations of ATP shows that AFB1 did not affect the Km and Vmax of the unstimulated enzyme whereas these parameters were reduced by about 75% and 50%, respectively, in the presence of calmodulin. Using the product of limited proteolysis by trypsin i.e. the 90 kDa fragment which still retains its calmodulin binding-domain and the 76 kDa fragment which has lost this domain, kinetic studies on the enzyme activity revealed that AFB1 inhibited the calmodulin-activated 90 kDa fragment by about 50% while the 76 kDa was not affected at all by the toxin and calmodulin. The toxin had no significant affect on the basal activity of the 90 kDa limited proteolysis fragment of the enzyme. These observations suggest that AFB1 inhibits the activated Ca(2+)-ATPase by binding to an important site in the calmodulin-binding domain of the enzyme. It seems likely that the toxin binds to tryptophan in the calmodulin-binding domain, thus causing a reduction in the rate at which this domain can interact with Ca(2+)-calmodulin or acidic phospholipids. The implication of these observations is that Ca(2+)-extrusion and other calmodulin-activated enzymes and processes may be slowed down during prolonged exposure to AFB1 because of its anticalmodulin effect.

Molecular cloning of the human type VIII adenylyl cyclase

A cDNA coding for a human type VIII adenylyl cyclase has been isolated from human newborn brain-stem tissue. This cDNA is 6,005 bp long and encodes for a protein of 1251 amino acids, exhibiting the two sets of six transmembrane spanning regions and the hydrophobicity profile typical of other mammalian adenylyl cyclases. Comparison with the rat form shows that they share 97% identity in amino acids. Type VIII adenylyl cyclase is unique in that it has both a long carboxy terminal and a long amino terminal tail. This is the first report on a complete cDNA clone coding for a human adenylyl cyclase. The distribution and regulation of this particular adenylyl cyclase suggest that it may be involved in learning, in memory and in drug dependence.

A model for the calmodulin-peptide complex based on the troponin C crystal packing and its similarity to the NMR structure of the calmodulin-myosin light chain kinase peptide complex

In the crystal structure of troponin C, the holo C-domain is bound in a head-to-tail fashion to the A-helix of the apo N-domain of a symmetry-related molecule. Using this interaction, we have proposed a model for the calmodulin-peptide complex. We find that the interaction of the C-domain with the A-helix is similar to that observed in the NMR structure of the calmodulin-myosin light chain kinase (MLCK) peptide complex. This similarity in binding has enabled us to make a precise sequence alignment of the target peptides in the calmodulin-binding cleft and to rationalize the amino acid sequence-dependent binding strengths of various peptides. Our model differs from that proposed by Strynadka and James (Proteins Struct. Funct. Genet. 7, 234-248, 1990) in that the peptides are rotated by 100 degrees in the calmodulin binding cleft.

Calmodulin binding to alpha 1-purothionin: solution binding and modeling of the complex

CD and fluorescence spectroscopic measurements show that calmodulin (CaM) binds to purothionins (alpha 1-purothionin: alpha 1-PT; beta-purothionin: beta-PT) in 1:1 stoichiometry with an affinity similar to that exhibited with the tightest binding CaM-binding peptides. Using the available crystal structures of CaM and alpha 1-PT, a model has been built for the interaction of CaM and alpha 1-PT and subjected to potential energy minimization. In the model, there is a bend in the central helix of CaM similar to that suggested by Persechini and Kretsinger (J. Card. Pharm. 12:501-512, 1988). alpha 1-PT fits snugly into the cavity formed by the bent CaM molecule with each of its two helices making apolar interactions with each of the two hydrophobic clefts situated at the terminal domains of CaM. The complex is further stabilized by numerous polar and electrostatic interactions on the rims of the clefts. Our model is compared with two other similar models previously reported for the CaM complexes with other helical peptides and generalizations about the mode of CaM binding to target proteins are made, which have wide relevance to the function of CaM. By analogy, a similar model is predicted for a CaM-beta-PT complex.

Calmodulin interacts with a C-terminus peptide from the lens membrane protein MIP26

Lens fiber cells are coupled by communicating junctions that comprise over 50% of their appositional surfaces. The main intrinsic protein (MIP26) of lens fibers is a 28.2 kDa protein that forms large gap junction-like channels in reconstituted systems. Previously, we have shown that Ca(++)-activated calmodulin (CaM) regulates the permeability of reconstituted MIP26 channels and changes the conformation of MIP26, as measured by intrinsic fluorescence and circular dichroism spectroscopy. Examination of the MIP26 amino acid sequence has revealed a basic amphiphilic alpha-helical segment (Pep C) on the C-terminus with residue distribution similar to that found in other CaM binding proteins. To test the interaction between the amphiphilic segment and CaM, both a 20-mer peptide and trp-substituted fluorescent analog have been synthesized and purified by HPLC. Evidence from spectrofluorometric titration shows that the Pep C binds with CaM in 1:1 stoichiometry and with a kd of approximately 10 nM. Neither Ca++ nor H+ alone affects the conformation of the Pep C. However, when mixed with CaM the Pep C undergoes both a dramatic blue-shift in tryptophan fluorescence emission, indicative of strong hydrophobic interaction, and an increase in circular dichroism absorption in the alpha-helical region. Additional fluorescence blue-shift and alpha-helical content occur when Ca++ is added to the CaM:Pep C complex.

Evidence for domain organization within the 61-kDa calmodulin-dependent cyclic nucleotide phosphodiesterase from bovine brain

The complete amino acid sequence of the 61-kDa calmodulin-dependent, cyclic nucleotide phosphodiesterase (CaM-PDE) from bovine brain has been determined. The native protein is a homodimer of N alpha-acetylated, 529-residue polypeptide chains, each of which has a calculated molecular weight of 60,755. The structural organization of this CaM-PDE has been investigated with use of limited proteolysis and synthetic peptide analogues. A site capable of interacting with CaM has been identified, and the position of the catalytic domain has been mapped. A fully active, CaM-independent fragment (Mr = 36,000), produced by limited tryptic cleavage in the absence of CaM, represents a functional catalytic domain. N-Terminal sequence and size indicate that this 36-kDa fragment is comprised of residues 136 to approximately 450 of the CaM-PDE. This catalytic domain encompasses a approximately 250 residue sequence that is conserved among PDE isozymes of diverse size, phylogeny, and function. CaM-PDE and its PDE homologues comprise a unique family of proteins, each having a catalytic domain that evolved from a common progenitor. A search of the sequence for potential CaM-binding sites revealed only one 15-residue segment with both a net positive charge and the ability to form an amphiphilic alpha-helix. Peptide analogues that include this amphiphilic segment were synthesized. Each was found to inhibit the CaM-dependent activation of the enzyme and to bind directly to CaM with high affinity in a calcium-dependent manner. This site is among the sequences cleaved from a 45-kDa chymotryptic fragment that has the complete catalytic domain but no longer binds CaM. These results indicate that residues located between position 23 and 41 of the native enzyme contribute significantly to the binding of CaM although the involvement of residues from additional sites is not excluded.

Sequence comparison of the 63-, 61-, and 59-kDa calmodulin-dependent cyclic nucleotide phosphodiesterases

Partial protein sequences from the 59-kDa bovine heart and the 63-kDa bovine brain calmodulin-dependent phosphodiesterases (CaM-PDEs) were determined and compared to the sequence of the 61-kDa isozyme reported by Charbonneau et al. [Charbonneau, H., Kumar, S., Novack, J. P., Blumenthal, D. K., Griffin, P. R., Shabanowitz, J., Hunt, D. F., Beavo, J. A. & Walsh, K. A. (1991) Biochemistry (preceding paper in this issue)]. Only a single segment (34 residues) at the N-terminus of the 59-kDa isozyme lacks identity with the 61-kDa isozyme; all other assigned sequence is identical in the two isozymes. Peptides from the 59-kDa isozyme that correspond to residues 23-41 of the 61-kDa protein bind calmodulin with high affinity. The C-terminal halves of these calmodulin-binding peptides are identical to the corresponding 59-kDa sequence; the N-terminal halves differ. The localization of sequence differences within this single segment suggests that the 61- and 59-kDa isozymes are generated from a single gene by tissue-specific alternative RNA splicing. In contrast, partial sequence from the 63-kDa bovine brain CaM-PDE isozyme displays only 67% identity with the 61-kDa isozyme. The differences are dispersed throughout the sequence, suggesting that the 63- and 61-kDa isozymes are encoded by separate but homologous genes.

Solution conformations of the N-terminal CNBr fragment of glycogen phosphorylase and its interaction with calmodulin

The CNBr peptides, CBPa and CBPb, corresponding to the N-terminal 1-91 amino acid residues of glycogen-phosphorylase a and b, respectively, were purified and characterized. CD, 31P-NMR and fluorescence spectroscopy were used to assess the structural organization of the cyanogen bromide peptides in solution. The cyanogen bromide peptides yielded 21% of alpha-helical structures by CD compared to a calculated value of 36.3%. These peptides interact with calmodulin which induces measurable alpha-helices in the cyanogen bromide peptides. The helix stabilizing reagent, trifluoroethanol, induces high numbers of alpha-helices in CBP, thereby demonstrating the conformation fluidity of this peptide. The dissociation constants for calmodulin and CBP estimated by fluorescence titrations were 36.0 and 29.9 nM for CBPb in the presence of Ca2+ and EGTA, respectively. The phosphorylated residue in CBPa causes a decrease in binding interactions with calmodulin and corresponding values obtained for CBPa by fluorescence titration are 56.0 and 141.0 nM, respectively. The Ser-P-14 of CBPa is titratable, yielding a pKa = 5.45 and a Hill coefficient of 1.5. A helical wheel analysis using a computer program in PC/GENE of the CBP shows that peptide stretches in the alpha-1 and alpha-2 helices are most basic and fairly amphiphilic and therefore represent the most probable segment for CaM binding. It is this structural character of these segments which presumably confer the ability to bind CaM and facilitate some of the allosteric transitions of glycogen phosphorylase.

Regulation of smooth muscle myosin light chain kinase by calmodulin

The mutagenesis work described in this paper has been instrumental in furthering our understanding of how CaM binds to and activates MLCK. Figure 2 schematically represents this interaction. The inactive MLCK appears to have a catalytic domain that is repressed by a substrate inhibitory domain that overlaps with the CaM binding domain, a basic amphipathic helix. In the presence of Ca2+, CaM undergoes a conformational change that exposes two hydrophobic pockets, one in each globular lobe, that are important for binding to MLCK. Upon binding CaM, MLCK undergoes a conformational change that derepresses the catalytic site, allows substrate access and light chain phosphorylation. Calmodulin antagonist drugs intercalate within these hydrophobic pockets to interfere with target enzyme binding. The total loss of activity if W800 is altered to A illustrates the importance of these hydrophobic interactions within the enzyme. The basic residues are also important; most of the basic residues in the binding domain of MLCK appear to aid in CaM binding but are not in themselves crucial, this includes the RRK triad. However, a specific electrostatic interaction between R812 of MLCK and CaM is suggested by the complete failure in MLCK activation if this residue is changed to an A. Electrostatic interactions between MLCK and CaM are also indicated by the TaM-BM1 mutant. This mutant can bind to but not activate MLCK. It is hypothesized that TaM-BM1 will bind to the basic amphipathic helix of MLCK but that the alterations in the surface charges (especially E14 and T34) and/or hydrophobicity (S38) prevent the proper conformational change in MLCK necessary for light chain phosphorylation. The resulting MLCK-CaM complex is therefore, inactive but can bind TaM-BM1. The exact interaction of these amino acids in CaM with MLCK will have to await the elucidation of a CaM-MLCK co-crystal.

Charge distributions and amphipathicity of receptor-binding alpha-helices

Contacts between ligands and cell-surface receptors result in cellular activation. Defining principles which govern these important interactions are of interest and might facilitate pharmacologic intervention. We examined receptor-binding alpha-helical segments of polypeptide hormones and globular proteins for distinguishing amino acid content and distributions. There was a slight excess of basic residues in both sets of alpha-helices compared with a panel of control helices. Helical concentrations of charged residues were quantitated using the hydrophobic moment algorithm, adapted to obtain the vector sum of side chain charges. By this analysis we detected increased concentrations of the set of basic residues (arginine, lysine and histidine) on one side of the receptor-binding alpha-helices of the polypeptide hormones, and to a lesser extent the protein ligands. Comparable data were obtained for "lytic" venom peptides and calmodulin-regulated kinase segments. There was an even greater correlation between receptor-associating alpha-helical segments and large hydrophobic moments. Receptor-binding helical segments of polypeptide hormones, and to a lesser extent those of protein ligands, often are basic and amphipathic.

Functional and structural similarities between the inhibitory region of troponin I coded by exon VII and the calmodulin-binding regulatory region of the catalytic subunit of phosphorylase kinase

A sequence homology has been noted between the carboxyl quarter of the catalytic gamma subunit of phosphorylase kinase and the region of troponin I coded by exon VII. Because this portion of troponin I contains the inhibitory region that interacts with actin and troponin C, we have examined whether the gamma subunit of phosphorylase kinase can functionally mimic troponin I by also interacting with actin and troponin C. We have found that troponin C not only activates the isolated gamma subunit of phosphorylase kinase but also binds with approximately the same affinity as calmodulin. Although actin had no effect on the activity of the gamma subunit alone, it did inhibit the activity of gamma-calmodulin and gamma-troponin C complexes. Conversely, the gamma subunit was able to inhibit actomyosin ATPase in a process that could be overcome by calmodulin. These results suggest that actin and calmodulin (or troponin C) compete for binding to the gamma subunit. Moreover, the structural and functional similarities between the gamma subunit and troponin I suggest that the gamma subunit of phosphorylase kinase may have evolved from the fusion of a protein kinase protogene with a progenitor of exon VII of troponin I.

How calmodulin binds its targets: sequence independent recognition of amphiphilic alpha-helices

Calmodulin (CaM) is a protein capable of recognizing positively charged, amphiphilic alpha-helical peptides independent of their precise amino acid sequences; this structural feature has also been found in many CaM-binding proteins. Recent work involving crystallography and site-directed mutagenesis of CaM along with studies of photoreactive and fluorescent CaM-binding peptides have helped define how calmodulin interacts with amphiphilic helices.

Model for the interaction of amphiphilic helices with troponin C and calmodulin

Crystals of troponin C are stabilized by an intermolecular interaction that involves the packing of helix A from the N-terminal domain of one molecule onto the exposed hydrophobic cleft of the C-terminal domain of a symmetry related molecule. Analysis of this molecular recognition interaction in troponin C suggests a possible mode for the binding of amphiphilic helical molecules to troponin C and to calmodulin. From the template provided by this troponin C packing, it has been possible to build a model of the contact region of mastoporan as it might be bound to the two Ca2+ binding proteins. A possible binding mode of melittin to calmodulin is also proposed. Although some of the characteristics of binding are similar for the two amphiphilic peptides, the increased length of melittin requires a significant bend in the calmodulin central helix similar to that suggested recently for the myosin light chain kinase calmodulin binding peptide (Persechini and Kretsinger: Journal of Cardiovascular Pharmacology 12:501-512, 1988). Not only are the hydrophobic interactions important in this model, but there are several favorable electrostatic interactions that are predicted as a result of the molecular modeling. The regions of troponin-C and calmodulin to which amphiphilic helices bind are similar to the regions to which the neuroleptic drugs such as trifluoperazine have been predicted to bind (Strynadka and James: Proteins 3:1-17, 1988).

A design approach to the structural analysis of interleukin-2

A semi-synthetic protein design approach has been employed for the structural investigation of a putative helical region at the C-terminus of Interleukin-2. With crystallographic or NMR derived conformational data as yet unavailable, we have relied only on primary sequence information and computer-assisted modelling to direct the analysis. By employing both chemical peptide synthesis and recombinant DNA methods, the C-terminus of IL-2 was modified according to a strategy designed to stabilize helical secondary structure. A semi-synthetic protein incorporating 12 simultaneous amino acid replacements was constructed, which possessed potentiated biological activity and displayed a far UV circular dichroism spectrum comparable to a hybrid protein with the authentic sequence. By comparison, another hybrid protein containing a C-terminal region designed to contain helix breaking residues was totally devoid of bioactivity. These findings provide evidence that the modelling method correctly identified a helix necessary for the formation of a bioactive tertiary fold. Moreover, by employing semi-synthesis it was possible to circumvent the difficulties associated with the preparation, purification and analysis of multiple recombinant proteins, and also to avoid the unreliability of total chemical synthesis for proteins greater than 100 residues.