The simulated dynamics of the insulin monomer and their relationship to the molecule's structure

Progress in Simulation Studies of Insulin Structure and Function

Insulin is a peptide hormone known for chiefly regulating glucose level in blood among several other metabolic processes. Insulin remains the most effective drug for treating diabetes mellitus. Insulin is synthesized in the pancreatic β-cells where it exists in a compact hexameric architecture although its biologically active form is monomeric. Insulin exhibits a sequence of conformational variations during the transition from the hexamer state to its biologically-active monomer state. The structural transitions and the mechanism of action of insulin have been investigated using several experimental and computational methods. This review primarily highlights the contributions of molecular dynamics (MD) simulations in elucidating the atomic-level details of conformational dynamics in insulin, where the structure of the hormone has been probed as a monomer, dimer, and hexamer. The effect of solvent, pH, temperature, and pressure have been probed at the microscopic scale. Given the focus of this review on the structure of the hormone, simulation studies involving interactions between the hormone and its receptor are only briefly highlighted, and studies on other related peptides (e.g., insulin-like growth factors) are not discussed. However, the review highlights conformational dynamics underlying the activities of reported insulin analogs and mimetics. The future prospects for computational methods in developing promising synthetic insulin analogs are also briefly highlighted.

Evidence of oligomerization of bovine insulin in solution given by NMR

The protein hormone insulin exists in several forms in nature, and a large number of modified sequences are used in pharmacy. They differ by physicochemical properties and efficiency of biological action. Pancreatic bovine insulin was studied in an acidic solution by nuclear magnetic resonance spectroscopy. [Formula: see text]H and [Formula: see text]C NMR signal assignment of backbone and side chains was made by analysis of a set of 2D spectra obtained on a sample with natural isotope abundance. The presence of certain secondary structure elements was revealed on a qualitative level based on nuclear Overhauser effect spectroscopy, which are similar to those observed in the crystal structure. The C-terminus of the B-chain possessed a remarkable flexibility. The molecule was shown to exist in exchange with oligomers based on its self-diffusion coefficient and correlation time measurements performed at different concentrations. Certain signals in the NOESY and HSQC spectra are consistent with the presence of minor conformers; this is an obstacle in simulating the molecular structure under the conditions used in the experiment.

Theoretical and computational studies of peptides and receptors of the insulin family

Synergistic interactions among peptides and receptors of the insulin family are required for glucose homeostasis, normal cellular growth and development, proliferation, differentiation and other metabolic processes. The peptides of the insulin family are disulfide-linked single or dual-chain proteins, while receptors are ligand-activated transmembrane glycoproteins of the receptor tyrosine kinase (RTK) superfamily. Binding of ligands to the extracellular domains of receptors is known to initiate signaling via activation of intracellular kinase domains. While the structure of insulin has been known since 1969, recent decades have seen remarkable progress on the structural biology of apo and liganded receptor fragments. Here, we review how this useful structural information (on ligands and receptors) has enabled large-scale atomically-resolved simulations to elucidate the conformational dynamics of these biomolecules. Particularly, applications of molecular dynamics (MD) and Monte Carlo (MC) simulation methods are discussed in various contexts, including studies of isolated ligands, apo-receptors, ligand/receptor complexes and intracellular kinase domains. The review concludes with a brief overview and future outlook for modeling and computational studies in this family of proteins.

Protein flexibility: multiple molecular dynamics simulations of insulin chain B

Multiple molecular dynamics simulations totaling more than 100 ns were performed on chain B of insulin in explicit solvent at 300 K and 400 K. Despite some individual variations, a comparison of the protein dynamics of each simulation showed similar trends and most structures were consistent with NMR experimental values, even at the elevated temperature. The importance of packing interactions in determining the conformational transitions of the protein was observed, sometimes resulting in conformations induced by localized hydrophobic interactions. The high temperature simulation generated a more diverse range of structures with similar elements of secondary structure and populated conformations to the simulations at room temperature. A broad sampling of the conformational space of insulin chain B illustrated a wide range of conformational states with many transitions at room temperature in addition to the conformational states observed experimentally. The T-state conformation associated with insulin activity was consistently present and a possible mechanism of behavior was suggested.

Protein structure and the spandrels of San Marco: insulin's receptor-binding surface is buttressed by an invariant leucine essential for its stability

Insulin provides a model of induced fit in macromolecular recognition: the hormone's conserved core is proposed to contribute to a novel receptor-binding surface. The core's evolutionary invariance, unusual among globular proteins, presumably reflects intertwined constraints of structure and function. To probe the architectural basis of such invariance, we have investigated hydrophobic substitutions of a key internal side chain (Leu(A16)). Although the variants exhibit perturbed structure and stability, moderate receptor-binding activities are retained. These observations suggest that the A16 side chain provides an essential structural buttress but unlike neighboring core side chains, does not itself contact the receptor. Among invertebrate insulin-like proteins, Leu(A16) and other putative core residues are not conserved, suggesting that the vertebrate packing scheme is not a general requirement of an insulin-like fold. We propose that conservation of Leu(A16) among vertebrate insulins and insulin-like growth factors is a side consequence of induced fit: alternative packing schemes are disallowed by lack of surrounding covariation within the hormone's hidden receptor-binding surface. An analogy is suggested between Leu(A16) and the spandrels of San Marco, tapering triangular spaces at the intersection of the dome's arches. This celebrated metaphor of Gould and Lewontin emphasizes the role of interlocking constraints in the evolution of biological structures.

Structure and stability of the insulin dimer investigated by molecular dynamics simulation

Molecular dynamics simulation indicates that the dynamical behaviour of the insulin dimer is asymmetric. Atomic level knowledge of the interaction modes and protein conformation in the solvation state identifies dynamical structures, held by hydrogen bonds that stabilize, mainly in one monomer, the interaction between the chains. Dynamic cross-correlation analysis shows that the two insulin monomers behave asymmetrically and are almost independent. Solvation energy, calculated to evaluate the contribute of each interface residue to the dimer association pattern, well compares with the experimental association state found in protein mutants indicating that this parameter is an important factor to explain the association properties of mutated insulin dimers.

Solution structures of the R6 human insulin hexamer,

The three-dimensional solution structure of the phenol-stabilized 36 kDa R6 insulin hexamer was determined by NMR spectroscopy and restrained molecular dynamics. The hexamer structures were derived using a stepwise procedure. Initially, 60 monomers were obtained by distance geometry from 665 NOE-derived distance restraints and three disulfide bridges. Subsequently, the hexamer structures were calculated by simulated annealing, using 30 hexamers constructed from the best 36 monomer structures as the starting models. The NMR data show that the aromatic ring of residue Phe(B25) can take two different orientations in the solution hexamer: one in which it points inward (molecule 1, about 90%) and one in which it points outward from the surface of the monomer (molecule 2, about 10%). Therefore, two hexamer structures were calculated: a symmetric hexamer consisting of six molecule 1 monomers and a nonsymmetric hexamer consisting of five molecule 1 monomers and one molecule 2 monomer. For each of the six monomers, the restraints used in the calculations of the hexamer structures include, in addition to the intramonomeric restraints, 25 NOEs between insulin and phenol, 23 NOEs and two hydrogen bonds across the dimer interface, nine NOEs across the trimer interface, and five intramonomeric or two intermonomeric NOEs, respectively, specifying the different orientations of the Phe(B25) ring. The coordination of the two Zn atoms was defined by eight distance restraints. Thus, a total of 4394 and 4391 distance restraints, respectively, were used in the two hexamer calculations. The NOE restraints were classified in an iterative process as intra- or intermonomeric on the basis of their consistency or inconsistency with the structure of the monomer. The assignment of the dimer- and trimer-specific NOEs was made using the crystal structure of the R6 hexamer as the starting model. For both solution hexamers, the average backbone rms deviation is 0.81 A, if the less well-defined N- and C-terminal residues are excluded. The corresponding rms deviations for all heavy atoms are 1.17 and 1.19 A for the nonsymmetric and symmetric hexamer, respectively. The overall solution structure of the R6 insulin hexamer is compact, rigid, and symmetric and resembles the corresponding crystal structure. However, the extension of the B-chain alpha-helix, which characterizes the R state, is shorter in the solution structure than in the crystal structure. Also, the study shows that the orientation of the Phe(B25) ring has no effect on the structure of the rest of the molecule, within the uncertainty of the structure determination. The importance of these findings for the current model for the insulin-receptor interaction is discussed.

Comparative studies on the dynamics of crosslinked insulin

Molecular dynamics simulations were carried out on an insulin crosslinked between the N-terminal A chain and the C-terminal B chain to form a so-called mini-proinsulin: N alpha-A1-N epsilon-B29-diaminosuberoyl insulin (DASI). To investigate the influence of crosslinking on the dynamics of the insulin moiety, the bridge was removed from a transient DASI structure and simulation was carried on independently with the then unlinked (ULKI) as well as with the crosslinked species. The effects of crystal packing and quaternary interactions were checked by simulating both types of monomers and dimers known from the hexamer structure. All simulations were compared to previous ones of native insulin. DASI shows general similarity to the native simulations in most parts of the structure. Deviations are visible in the segments to which the bridge is directly connected, i.e. their flexibility is reduced. Upon removal of the bridge the ULKI simulations reapproach those of native insulin. The influence of the bridge spreads over the whole molecule, but all of its main structural features remain intact. The simulations suggest that the displacement of the C-terminal B chain of native insulin, considered important for receptor interaction, is prevented by the bridge, which also partially shields some binding residues. This is in accordance with the poor biological potency of A1-B29-crosslinked insulins.

Acid stabilization of insulin

The effect of pH on the conformational stability of insulin was studied. Surprisingly, the Gibbs free energy of unfolding increased approximately 30% by acidification. pH titration of insulin's conformational stability is described by a transition involving a single proton with an apparent pK(a) of 7.0. The acid stabilization of insulin's conformation was attributed to the protonation of histidine at position 5 on the B-chain (HB5) as determined by 1H-NMR of the histidines, selective amino acid alteration, and enthalpies of ionization. Further acidification (at least to pH 2) does not decrease the free energy of unfolding. A conformational change in the tertiary structure, as indicated by the near-UV circular dichroism spectrum, accompanies this change in stability. We propose that this acid stabilization of insulin is physiologically important in maintaining insulin stability in the acid environment of the secretory/storage granules of the beta-cell of the pancreatic islets of Langerhans.

Three-dimensional solution structure of an insulin dimer. A study of the B9(Asp) mutant of human insulin using nuclear magnetic resonance, distance geometry and restrained molecular dynamics

The solution structure of the B9(Asp) mutant of human insulin has been determined by two-dimensional 1H nuclear magnetic resonance spectroscopy. Thirty structures were calculated by distance geometry from 451 interproton distance restraints based on intra-residue, sequential and long-range nuclear Overhauser enhancement data, 17 restraints on phi torsional angles obtained from 3JH alpha HN coupling constants, and the restraints from 17 hydrogen bonds, and the three disulphide bridges. The distance geometry structures were optimized using restrained molecular dynamics (RMD) and energy minimization. The average root-mean-square deviation for the best 20 RMD refined structures is 2.26 A for the backbone and 3.14 A for all atoms if the less well-defined N and C-terminal residues are excluded. The helical regions are better defined, with root-mean-square deviation values of 1.11 A for the backbone and 2.03 A for all atoms. The data analysis and the calculations show that B9(Asp) insulin, in water solution at the applied pH (1.8 to 1.9), is a well-defined dimer with no detectable difference between the two monomers. The association of the two monomers in the solution dimer is relatively loose as compared with the crystal dimer. The overall secondary and tertiary structures of the monomers in the 2Zn crystal hexamer is found to be preserved. The conformation-averaged NMR structures obtained for the monomer is close to the structure of molecule 1 in the hexamer of the 2Zn insulin crystal. However, minor, but significant deviations from this structure, as well as from the structure of monomeric insulin in solution, exist and are ascribed to the absence of the hexamer and crystal packing forces, and to the presence of monomer-monomer interactions, respectively. Thus, the monomer in the solution dimer shows a conformation similar to that of the crystal monomer in molecular regions close to the monomer-monomer interface, whereas it assumes a conformation similar to that of the solution structure of monomeric insulin in other regions, suggesting that B9(Asp) insulin adopts a monomer-like conformation when this is not inconsistent with the monomer-monomer arrangement in the dimer.

Detection of an equilibrium intermediate in the folding of a monomeric insulin analog

To determine the conformational properties of the C-terminal region of the insulin B-chain relative to the helical core of the molecule, we have investigated the fluorescence properties of an insulin analog in which amino acids B28 and B29 have been substituted with a tryptophan and proline residue respectively, ([WB28,PB29]insulin). The biological properties and far-UV circular dichroism (CD) spectrum of the molecule indicate that the conformation is similar to that of native human insulin. Guanidine hydrochloride (GdnHCl)-induced equilibrium denaturation of the analog as monitored by CD intensity at 224 nm indicates a single cooperative transition with a midpoint of 4.9 M GdnHCl. In contrast, when the equilibrium denaturation is observed by steady-state fluorescence emission intensity at 350 nm, two distinct transitions are observed. The first transition accounts for 60% of the observed signal and has a midpoint of 1.5 M GdnHCl. The second transition roughly parallels that observed by CD measurements with an approximate midpoint of 4.5 M GdnHCl. The near-UV CD spectrum, size-exclusion, and ultracentrifugation properties of [WB28,PB29]insulin indicate that this analog does not self-associate in a concentration-dependent manner as does human insulin. Thus, the observed fluorescence changes must be due to specific conformational transitions which occur upon unfolding of the insulin monomer with the product of the first transition representing a stable folding intermediate of this molecule.

Substrate mobility in a deeply buried active site: analysis of norcamphor bound to cytochrome P-450cam as determined by a 201-psec molecular dynamics simulation

While cytochrome P-450cam catalyzes the hydroxylation of camphor to 5-exo-hydroxycamphor with 100% stereospecificity, norcamphor is hydroxylated by this enzyme yielding 45% 5-exo-, 47% 6-exo-, and 8% 3-exo-hydroxynorcamphor (Atkins, W.M., Sligar, S.G., J. Am. Chem. Soc. 109:3754-3760, 1987). The present study describes a 201-psec molecular dynamics (MD) stimulation of norcamphorbound cytochrome P-450cam to elucidate the relationship between substrate conformational mobility and formation of alternative products. First, these data suggest that the product specificity is, at least partially, due to the mobility of the substrate within the active site. Second, the high mobility of norcamphor in the active site leads to an average increase in separation between the heme iron and the substrate of about 1.0 A; this increase in separation may be the cause of the uncoupling of electron transfer when norcamphor is the substrate. Third, the active site water located in the norcamphorbound crystal structure possesses mobility that correlates well with the spin-state equilibrium of this enzyme-substrate complex.

Structure and dynamics of des-pentapeptide-insulin in solution: the molten-globule hypothesis

Structures of insulin in different crystal forms exhibit significant local and nonlocal differences, including correlated displacement of elements of secondary structure. Here we describe the solution structure and dynamics of a monomeric insulin analogue, des-pentapeptide-(B26-B30)-insulin (DPI), as determined by two-dimensional NMR spectroscopy and distance geometry/restrained molecular dynamics (DG/RMD). Although the solution structure of DPI exhibits a general similarity to its crystal structure, individual DG/RMD structures in the NMR ensemble differ by rigid-body displacements of alpha-helices that span the range of different crystal forms. These results suggest that DPI exists as a partially folded state formed by coalescence of distinct alpha-helix-associated microdomains. The physical reality of this model is investigated by comparison of the observed two-dimensional nuclear Overhauser enhancement (NOE) spectroscopy (NOESY) spectrum with that predicted from crystal and DG/RMD structures. The observed NOESY spectrum contains fewer tertiary contacts than predicted by any single simulation, but it matches their shared features; such "ensemble correspondence" is likely to reflect the effect of protein dynamics on observed NOE intensities. We propose (i) that the folded state of DPI is analogous to that of a compact protein-folding intermediate rather than a conventional native state and (ii) that the molten state is the biologically active species. This proposal (the molten-globule hypothesis) leads to testable thermodynamic predictions and has general implications for protein design.

The solution structure of a monomeric insulin. A two-dimensional 1H-NMR study of des-(B26-B30)-insulin in combination with distance geometry and restrained molecular dynamics

The solution conformation of des-(B26-B30)-insulin (DPI) has been investigated by 1H-NMR spectroscopy. A set of 250 approximate interproton distance restraints, derived from two-dimensional nuclear Overhauser enhancement spectra, were used as the basis of a structure determination using distance geometry (DG) and distance-bound driven dynamics (DDD). Sixteen DG structures were optimized using energy minimization (EM) and submitted to short 5-ps restrained molecular dynamics (RMD) simulations. A further refinement of the DDD structure with the lowest distance errors was done by energy minimization, a prolonged RMD simulation in vacuo and a time-averaged RMD simulation. An average structure was obtained from a trajectory generated during 20-ps RMD. The final structure was compared with the des-(B26-B30)-insulin crystal structure refined by molecular dynamics and the 2-Zn crystal structure of porcine insulin. This comparison shows that the overall structure of des-(B26-B30)-insulin is retained in solution with respect to the crystal structures with a high flexibility at the N-terminal part of the A chain and at the N-terminal and C-terminal parts of the B chain. In the RMD run a high mobility of Gly A1, Asn A21 and of the side chain of Phe B25 is noticed. One of the conformations adopted by des-(B26-B30)-insulin in solution is similar to that of molecule 1 (Chinese nomenclature) in the crystal structure of porcine insulin.

Two-dimensional NMR studies on des-pentapeptide-insulin. Proton resonance assignments and secondary structure analysis

The shortened analogue of insulin, des-(B26-B30)-pentapeptide insulin, has been characterized by two-dimensional 1H NMR. The 1H resonance assignments and the secondary structure in water solution are discussed The results indicate that the secondary structure in solution is very similar to that reported for the crystalline state. A high flexibility of both A and B chains is observed. Of the two conformations seen in the 2-Zn insulin crystals and indicated as molecules 1 and 2 (Chinese nomenclature), the structure of the analogue is more similar to that of molecule 1.

Using molecular dynamics simulations on crambin to evaluate the suitability of different continuum dielectric and hydrogen atom models for protein simulations

Molecular dynamics simulations of enzymes with enough explicit waters of solvation to realistically account for solute-solvent interactions can burden the computational resources required to perform the simulation by more than two orders of magnitude. Since enzyme simulations even with an implicit solvation model can be imposing for a supercomputer, it is important to assess the suitability of different continuum dielectric models for protein simulations. A series of 100-picosecond molecular dynamics simulations were performed on the X-ray crystal structure of the protein crambin to examine how well computed structures, obtained using seven continuum dielectric and two hydrogen atom models, agreed with the X-ray structure. The best level of agreement between computed and experimental structures was obtained using a constant dielectric of 2 and the all-hydrogen model. Continuum dielectric models of 1, 1r, and 2r also led to computed structures in reasonably good agreement with the X-ray structure. In all cases, the all-hydrogen model gave better agreement than the united atom model, although, in one case, the difference was not significant. Dielectric models of 4, 80, and 4r with either hydrogen model yielded significantly poorer fits. It is especially noteworthy that the observed trends did not semiquantitatively converge until about 50 picoseconds into the simulations, suggesting that validation studies for protein calculations based on energy minimizations or short simulations should be viewed with caution.

Changes in secondary structure follow the dissociation of human insulin hexamers: a circular dichroism study

Vacuum UV circular dichroism spectra measured down to 178 nm for hexameric 2-zinc human insulin, zinc-free human insulin, and the two engineered and biologically active monomeric mutants, [B/S9D] and [B/S9D,T27E] human insulin, show significant differences. The secondary structure analysis of the 2-zinc human insulin (T6) in neutral solution was determined: 57% helix, 1% beta-strand, 18% turn, and 24% random coil. This is very close to the corresponding crystal structure showing that the solution and solid structures are similar. The secondary structure of the monomer shows a 10-15% increase in antiparallel beta-structure and a corresponding reduction in random coil structure. These structural changes are consistent with an independent analysis of the corresponding difference spectra. The advantage of secondary structure analyses of difference spectra is that the contribution of odd spectral features stemming mainly from side chain chromophores is minimized and the sensitivity of the analyses improved. Analysis of the CD spectra of T6 2-zinc, zinc-free human insulin and monomeric mutant insulin by singular value decomposition indicates that the secondary structure changes following the dissociation of hexamers into dimers and monomers are two-state processes.

Underlying assumptions in the estimation of secondary structure content in proteins by circular dichroism spectroscopy--a critical review

Recombinant DNA technology has made possible the large-scale production of proteins for pharmaceutical applications. As a result, there has been a renaissance in methodology which can provide information on the structural stability and character of these materials. Circular dichroism (CD) spectroscopy, with its sensitivity to the secondary structure adopted by the polypeptide chain, is a powerful tool in this regard. Quantitative analysis of the CD spectra of proteins is now wide-spread, aided by the availability of such algorithms on commercial instrumentation. However, there are basic assumptions made when conducting these calculations, many of which have not been addressed or summarized. Some of these assumptions are independent of the selection of basis spectra and the algorithm employed. These assumptions are listed and the available data concerning their validity is presented and discussed.

Synthesis and properties of [A19-(p-fluorophenylalanine)] insulin

The synthesis of [Phe(F)A19]insulin (porcine) is described. First the protected [Phe(F)19]A-chain was assembled by segment condensation of [1-12] and [13-21] using the dicyclohexyldiimide/1-hydroxybenzotriazole procedure. [Phe(F)19]A-chain was purified by ion exchange chromatography after removal of all the protecting groups (Boc, But, OBut and S-Trt) and its conversion into the tetra-S-sulfonated derivative. [Phe(F)A19]insulin was prepared by combination with porcine B-chain and purified by gel filtration and ion-exchange chromatography. The in vitro biological activity of this analogue was 60%. CD spectra in the near and far UV are qualitatively very similar to those of insulin.

Molecular dynamics simulation of despentapeptide insulin in a crystalline environment

Despentapeptide insulin (DPI), an analogue of insulin, is the only analogue of insulin existing as a monomer in crystalline form so far discovered. A 31 picosecond molecular dynamics simulation of DPI was carried out in the unit cell that includes four DPI molecules, 398 water molecules, four Cd2+, eight Na+ and four Cl-. The molecular dynamics results were compared with the available X-ray structure in terms of atomic positions, backbone dihedral angles, hydrogen bonds and positional fluctuations. Good agreement was found between the molecular dynamics results and the X-ray data. The largest root-mean-square fluctuations were observed at the N-terminal part of the B chain. This fits with the experimental observation.