Prediction of the packing arrangement of strands in β-sheets of globular proteins

A method is proposed for predicting the adjacency order in which strands pack in a -sheet in a protein, on the basis of its amino acid sequence alone. The method is based on the construction of a predicted contact map for the protein, in which the probability that various residue pairs are close to each other is computed from statistically determined average distances of residue pairs in globular proteins of known structure. Compact regions, i.e., portions of the sequence with many interresidue contacts, are determined on the map by using an objective search procedure. The proximity of strands in a -sheet is predicted from the density of contacts in compact regions associated with each pair of strands. The most probable -sheet structures are those with the highest density of contacts. The method has been tested by computing the probable strand arrangements in a five-strand -sheet in five proteins or protein domains, containing 62–138 residues. Of the theoretically possible 60 strand arrangements, the method selects two to eight arrangements as most probable; i.e., it leads to a large reduction in the number of possibilities. The native strand arrangement is among those predicted for three of the five proteins. For the other two, it would be included in the prediction by a slight relaxation of the cutoff criteria used to analyze the density of contacts.     

Specific recognition in the tertiary structure of β-sheets of proteins

The frequency of occurrence of nearest neighbour residue pairs on adjacent antiparallel (β_A) and parallel (β_P) strands is obtained from 30 known protein structures. The specificity of interstrand recognition due to such pairing as a factor in the folding of β-sheets is studied by statistical methods. Residues of sufficiently high count for statistical analysis are treated individually while the rest are combined into small groups of similar size, polarity, and/or genetic exchangeability. The hypothesis of specific recognition between individuals and small groups is contrasted with the alternative hypothesis of non-specific recognition between broad classes (hydrophobia, neutral, polar) of residues. A χ² test of pair correlations favours specific recognition against non-specific recognition with a high level of confidence. The largest and most significant correlations are: Ser/Thr (1.9 ± 0.3), Ile/Val (1.7 ± 0.3) and Lys-Arg/Asp-Gln (1.8 ± 0.3) in β_A, and Ile/Leu (1.9 ± 0.4) in β_P. The pair Gly/Gly never occurs in any β-sheet. The specific residue-pair correlations derived here may be useful in statistical prediction methods of protein tertiary structure.     

A contact scoring matrix for qualitative prediction of change in folding of α-helices in globular proteins caused by a mutation

The atomic pairs in contact for atoms from pairs of amino-acid residues on pairs of helices in a protein database consisting of 48 proteins of known tertiary structure from the Brookhaven Protein Data Bank are searched and counted to construct a primary scoring system. Each score in the primary scoring system is weighted further with the possibility of occurrence of each residue pair in the protein database to give a final scoring matrix. Scores for predicting change in folding of α-helices in a mutant protein are calculated by assuming that every pair of helices in the protein can closely interact with each other. It is shown that the change in folding of α-helices in several mutant proteins are reflected in both the change of the contact scores and the helix geometry calculated.     

The structure of the ends of α-helices in globular proteins: effect of additional hydrogen bonds and implications for helix formation

We prepared a set of about 2000 α-helices from a relational database of high-resolution three-dimensional structures of globular proteins, and identified additional main chain i ← i+3 hydrogen bonds at the ends of the helices (i.e., where the hydrogen bonding potential is not fulfilled by canonical i ← i+4 hydrogen bonds). About one-third of α-helices have such additional hydrogen bonds at the N-terminus, and more than half do so at the C-terminus. Although many of these additional hydrogen bonds at the C-terminus are associated with Schellman loops, the majority are not. We compared the dihedral angles at the termini of α-helices having or lacking the additional hydrogen bonds. Significant differences were found, especially at the C-terminus, where the dihedral angles at positions C2 and C1 in the absence of additional hydrogen bonds deviate substantially from those occurring within the α-helix. Using a novel approach we show how the structure of the C-terminus of the α-helix can emerge from that of constituent overlapping α-turns and β-turns, which individually show a variation in dihedral angles at different positions. We have also considered the direction of propagation of the α-helix using this approach. If one assumes that helices start as a single α-turn and grow by successive addition of further α-turns, the paths for growth in the N → C and C → N directions differ in a way that suggests that extension in the C → N direction is favored.     

On the computation of the tertiary structure of globular proteins—IV. Use of secondary structure information

The distance geometry approach for computing the tertiary structure of globular proteins emphasized in this series of papers (Goelet al., J. theor. Biol. 99, 705–757, 1982) is developed further. This development includes incorporation of some secondary structure information—the location of alpha helices in the primary sequence—in the algorithm to compute the tertiary structure of alpha helical globular proteins. An algorithm is developed which estimates the interresidue distances between chain-proximate helices. These distances, in conjunction with the global statistical average distances obtainable from a database of real proteins and determined by the primary sequence of the protein under study, are used to determine the tertiary structure. Five proteins, parvalbumin, hemerythrin, human hemoglobin, lamprey hemoglobin, and sperm whale myoglobin, are investigated. The root mean square (RMS) errors between the calculated structures and those determined by X-ray diffraction range from 4.78 to 7.56 Å. These RMSs are 0.21–2.76 Å lower than those estimated without the secondary structure information. Contact maps and three-dimensional backbone representations also show considerable improvements with the introduction of secondary structure information.     

Studies on the rules of β-strand alignment in a protein β-sheet structure

To further disclose the underlying mechanisms of protein β-sheet formation, studies were made on the rules of β-strands alignment forming β-sheet structure using statistical and machine learning approaches. Firstly, statistical analysis was performed on the sum of β-strands between each β-strand pairs in protein sequences. The results showed a propensity of near-neighbor pairing (or called “first come first pair”) in the β-strand pairs. Secondly, based on the same dataset, the pairwise cross-combinations of real β-strand pairs and four pseudo-β-strand contained pairs were classified by support vector machine (SVM). A novel feature extracting approach was designed for classification using the average amino acid pairing encoding matrix (APEM). Analytical results of the classification indicated that a segment of β-strand had the ability to distinguish β-strands from segments of α-helix and coil. However, the result also showed that a β-strand was not strongly conserved to choose its real partner from all the alternative β-strand partners, which was corresponding with the ordination results of the statistical analysis each other. Thus, the rules of “first come first pair” propensity and the non-conservative ability to choose real partner, were possible important factors affecting the β-strands alignment forming β-sheet structures.     

Identification and analysis of extended strands and β-sheets in globular proteins

A method to identify β-sheets in globular proteins from extended strands, using only α-carbon positions, has been developed. The strands that form β-sheets are picked up by means of simple distance criteria. The method has been tested by applying it to three proteins with accurately known secondary structures. It has also been applied to ten other proteins wherein only α-carbon coordinates are available, and the list of β-sheets obtained. The following points are worth noting: (i) The sheets identified by the algorithm are found to agree satisfactorily with the reported ones based on backbone hydrogen bonding, wherever this information is available. (ii) β-Strands that do not form parts of any sheet are a common feature of protein structures. (iii) Such isolated β-strands tend to be short. (iv) The conformation corresponding to the preferred right-handed twist of the sheet is overwhelmingly observed in both the sheet-forming and isolated β-strands.     

Minimal model systems for β-sheet secondary structure in proteins

Use of model systems to explore the forces that control β sheet formation was stymied for many years by the perception that small increments of β sheet necessarily aggregate. Recently, however, a number of short peptides (9–16 residues in length) that fold into two-stranded antiparallel β sheets (‘β hairpins’) have been reported; several short peptides (20–24 residues in length) that fold into three-stranded antiparallel β sheets have also been described. These model systems are beginning to provide fundamental insights into the origins of β sheet conformational stability.     

Relationship between structure and amino acid sequence of strongly twisted and coiled β-hairpins in globular proteins

β-Hairpins are widespread in proteins, and it is possible to find them both within β-sheets and separately. In this work, a comparative analysis of amino acid sequences of β-strands within strongly twisted β-hairpins from different structural protein subclasses has been conducted. Strongly twisted and coiled β-hairpin generates in the space a right double helix out of β-strands that are connected by a loop region (connections). The frequencies of amino acid residues on the internal (concave) and external (convex) surfaces of strongly twisted β-hairpins have been determined (220 β-hairpins from nonhomologous proteins were studied). The concave surface of these β-hairpins is mainly generated by hydrophobic residues, while the convex surface by hydrophilic residues; accordingly, the alternation of hydrophobic internal and hydrophilic external residues is observed in their amino acid sequences. Amino acid residues of glycine and alanine (especially in places of the largest twisting of the strands) were anomalously frequently found in internal positions of strongly twisted and coiled β-hairpins. It was established that internal positions never contain the proline residues, while external positions in the twisting region contain them in a relatively large amount. It was demonstrated that at least one amino acid residue in α_L- or ε-conformation is required for generation of relatively short (up to 7 amino acid residues) connection. As a rule, these positions are occupied by glycines. Thus, not only the alternation of hydrophobic and hydrophilic amino acid residues, but also the presence of one or two glycine residues in the connection region and the excess of glycines and alanines in the places of the largest strand twisting on the concave surface, as well as the presence of prolines on the convex surface, are required to generate a strongly twisted and coiled β-hairpin.     

Quasirandom distribution of α and β regions in globular proteins

Distributions of α and β regions in globular proteins among clusters containing different numbers of adjacent α-helices, adjacent β regions, and “overlapping” βαβ units are considered. It is shown that these distributions do not differ greatly from what can be expected for random distributions of α and β regions along protein chains. In particular, it is shown that the amounts of relatively long α, β and βαβ clusters (which provide the basis for the conventional classification of globular proteins or domains into α, β, or α/β types) in random sequences also do not differ very much from those in real globular proteins. It follows that the possibility of structural classification of globular proteins (domains) does not imply the existence of a correlation in protein primary structure. This possibility exists even in random sequences of amino acid residues and therefore may not be the result of biological evolution.     

BamE structure: the assembly of β-barrel proteins in the outer membranes of bacteria and mitochondria

A study recently published in EMBO reports solves the solution structure of E. coli BamE, thus providing the basis for a better understanding of the mechanism of β-barrel assembly in bacterial and mitochondrial outer membranes.EMBO Rep (2011) advance online publication. doi: 10.1038/embor.2010.202β-barrel membrane proteins are found exclusively in the outer membrane of Gram-negative bacteria and the outer membranes of eukaryotic organelles of prokaryotic origin, mitochondria and chloroplasts. In contrast to the inner membrane, the bacterial outer membrane is an asymmetrical bilayer that consists mainly of lipopolysaccharides in the outer leaflet and phospholipids in the inner leaflet. Bacterial β-barrel outer membrane proteins (OMPs) mediate many cellular functions, for example, passive or selective diffusion of small molecules through the β-barrel pores across the outer membrane. By contrast, only a few mitochondrial β-barrel outer membrane proteins (MBOMPs) have been identified so far. The central machineries that mediate insertion and assembly of OMPs/MBOMPs are the β-barrel assembly machine (BAM) complex in the bacterial outer membrane and the topogenesis of outer-membrane β-barrel proteins (TOB)/sorting and assembly machinery (SAM) complex in the mitochondrial outer membrane (Knowles et al, 2009; Endo & Yamano, 2010; Stroud et al, 2010; Fig 1). However, the molecular mechanisms of β-barrel protein topogenesis in bacterial and mitochondrial outer membranes remain poorly understood.Open in a separate windowFigure 1β-barrel protein assembly in bacterial and mitochondrial outer membranes. (A) Bacteria. Ribbon models of the structures of the Sec complex, SurA, BamA (Clantin et al, 2007; Kim et al, 2007), BamE and OMP. The upper and lower inserts show the surface of BamE (residues 20–108; viewed after approximately 90° rotation of the ribbon model around the horizontal axis toward the reader). Residues important for BamD binding are shown in red and residues with NMR signals that were perturbed by BamD binding are shown in yellow. The residue (Phe 74) important for PG binding is shown in red and the residues with NMR signals that were perturbed by PG binding are shown in yellow. (B) Mitochondria. Ribbon models were drawn for the structures of small Tim and MBOMP. IM, inner membrane; IMS, intermembrane space; MBOMP, mitochondrial β-barrel outer membrane protein; OM, outer membrane; OMP, outer membrane protein; PG, phosphatidylglycerol; POTRA, polypeptide transport-associated domain.Bacterial OMPs are synthesized in the cytosol as precursor proteins with an amino-terminal signal sequence that guides the proteins to the Sec machinery for crossing the inner membrane and is cleaved off in the periplasm. Periplasmic chaperones then escort OMPs through the aqueous periplasmic space in a partly unfolded state. On reaching the outer membrane, OMPs assemble into a β-barrel structure and insert into the outer membrane with the help of the BAM complex. The bacterial OMP insertion pathway can be compared to the assembly pathway of MBOMPs from the mitochondrial intermembrane space into the outer membrane. MBOMPs are synthesized in the cytosol and imported into the intermembrane space by the outer membrane translocator TOM40. The subsequent chaperone-mediated escort across the intermembrane space and insertion into the outer membrane by the TOB complex is similar to the OMP assembly process. Notably, the BAM and TOB complexes share the homologous β-barrel proteins BamA and Tob55/Sam50, respectively, as the central components of their insertion machineries. The BAM complex in Escherichia coli consists of BamA (YaeT/Omp85) and four accessory lipoproteins: BamB (YfgL), BamC (NlpB), BamD (YfiO) and BamE (SmpA). BamA and BamD are essential for cell growth, yet deletion of dispensable BamB, BamC or BamE leads to outer membrane defects manifested in hypersensitivity to antibiotics. Although BamAB and BamCDE can form distinct subcomplexes, they become functional only after formation of the entire BAM complex with all five subunits (Hagan et al, 2010).In this issue of EMBO reports, Knowles et al (2011) solve the nuclear magnetic resonance (NMR) solution structure of E. coli BamE, which sheds light on the roles of one of the Bam subunits in β-barrel protein assembly. The structure of BamE consists of a three-stranded antiparallel β-sheet packed against a pair of α-helices (Fig 1).As the ΔbamE mutant cannot grow in the presence of vancomycin, the authors identify functionally important residues of BamE by testing the effects of amino-acid substitutions in BamE on its inability to complement the growth defects of ΔbamE, without destabilizing BamE itself. Many of the identified residues are conserved among BamE proteins from different organisms and map to a single surface area on BamE. Interestingly, NMR signals of the residues around this region are sensitive to the addition of micelles containing the lipid phosphatidylglycerol, but not phosphatidylethanolamine or cardiolipin. In parallel, the authors analyse perturbation of the NMR spectra of BamE after the addition of purified BamB, C and D proteins. Only BamD affects the NMR spectra of BamE, and the BamD interacting region of BamE is found to overlap partly with the residues involved in phosphatidylglycerol binding. As the addition of BamD and phosphatidylglycerol have different effects on the NMR spectra of BamE, the binding of BamD and phosphatidylglycerol to BamE seem to take place simultaneously. What is the biological relevance of the observed interactions of BamE with both BamD and phosphatidylglycerol? As phosphatidylglycerol was found to help the insertion of OMPs into lipid liposomes (Patel et al, 2009), BamE might recruit the BAM complex through BamD to phosphatidylglycerol-rich regions in the outer membrane, or might directly recruit phosphatidylglycerol to form assembly points for OMP insertion and folding.What are the roles of other subunits of the BAM complex in β-barrel protein assembly? The essential subunit of the E. coli BAM complex BamA consists of two domains: the N-terminal polypeptide transport-associated (POTRA) domain repeat in the periplasm and the carboxy-terminal β-barrel domain, embedded in the outer membrane. The number of POTRA domains ranges from one to five in BamA homologues from different organisms. Of these POTRA domains, the one nearest to the C-terminal that is most connected to the β-barrel domain is essential for cell viability and its deletion leads to disassembly of the BAM complex (Kim et al, 2007). Structural studies of the E. coli BamA POTRA domains suggest that each POTRA domain has a common fold, whereas conformational rigidity might differ between inter-domain linkers (Gatzeva-Topalova et al, 2010; Fig 1). As individual POTRA domains have some affinity for unfolded substrate proteins, the periplasmic tandem POTRA repeat probably provides several substrate binding sites that slide the substrate progressively towards the BamA β-barrel domain. The β-barrel domain of BamA probably functions as a scaffold to facilitate the formation of β-strands, possibly through β-augmentation and subsequent spontaneous membrane insertion of the β-barrel. Yet, it is not clear whether this cradle for β-strand formation is provided by the pore formed within the monomer or oligomeric forms of the BamA β-barrel domain. Alternatively, membrane insertion and folding of OMPs might take place at the interface between BamA and the outer membrane lipid bilayer.How much of the β-barrel assembly process is conserved during the evolution of mitochondria from Gram-negative bacteria? Although the central subunits BamA and Tob55 of the BAM and TOB complexes are conserved, other subunits of these complexes are unrelated to each other. The POTRA domains of BamA are essential for recognition and assembly of bacterial OMPs, whereas that of Tob55 is dispensable for MBOMP assembly in the mitochondrial outer membrane. Nevertheless, the mitochondrial TOB complex facilitates assembly of bacterial OMPs at low efficiency (Walther et al, 2009) and, in turn, the bacterial BAM complex can mediate assembly of mitochondrial porin. Therefore, the basic mechanism of β-barrel assembly in the outer membranes of bacteria and mitochondria seems to be conserved. High-resolution structures of each component of the BAM and TOB complexes—including that of BamE in this study—will thus provide the basis for a better understanding of the mechanism of β-barrel assembly in evolutionarily related bacterial and mitochondrial outer membranes.     

Reversible transition between α-helix and β-sheet conformation of a transmembrane domain

Despite the important functions of protein transmembrane domains, their structure and dynamics are often scarcely known. The SNARE proteins VAMP/synaptobrevin and syntaxin 1 are implicated in membrane fusion. Using different spectroscopic approaches we observed a marked sensitivity of their transmembrane domain structure in regard to the lipid/peptide ratio. In the dilute condition, peptides corresponding to the complete transmembrane domain fold into an α-helix inserted at ∼ 35° to the normal of the membranes, an observation in line with molecular simulations. Upon an increase in the peptide/lipid ratio, the peptides readily exhibited transition to β-sheet structure. Moreover, the insertion angle of these β-sheets increased to 54° and was accompanied by a derangement of lipid acyl chains. For both proteins the transition from α-helix to β-sheet was reversible under certain conditions by increasing the peptide/lipid ratio. This phenomenon was observed in different model systems including multibilayers and small unilamellar vesicles. In addition, differences in peptide structure and transitions were observed when using distinct lipids (DMPC, DPPC or DOPC) thus indicating parameters influencing transmembrane domain structure and conversion from helices to sheets. The putative functional consequences of this unprecedented dynamic behavior of a transmembrane domain are discussed.     

Toxic prefibrillar α-synuclein amyloid oligomers adopt a distinctive antiparallel β-sheet structure

Parkinson's disease is an age-related movement disorder characterized by the presence in the mid-brain of amyloid deposits of the 140-amino-acid protein AS (α-synuclein). AS fibrillation follows a nucleation polymerization pathway involving diverse transient prefibrillar species varying in size and morphology. Similar to other neurodegenerative diseases, cytotoxicity is currently attributed to these prefibrillar species rather than to the insoluble aggregates. Nevertheless, the underlying molecular mechanisms responsible for cytotoxicity remain elusive and structural studies may contribute to the understanding of both the amyloid aggregation mechanism and oligomer-induced toxicity. It is already recognized that soluble oligomeric AS species adopt β-sheet structures that differ from those characterizing the fibrillar structure. In the present study we used ATR (attenuated total reflection)-FTIR (Fourier-transform infrared) spectroscopy, a technique especially sensitive to β-sheet structure, to get a deeper insight into the β-sheet organization within oligomers and fibrils. Careful spectral analysis revealed that AS oligomers adopt an antiparallel β-sheet structure, whereas fibrils adopt a parallel arrangement. The results are discussed in terms of regions of the protein involved in the early β-sheet interactions and the implications of such conformational arrangement for the pathogenicity associated with AS oligomers.     

Statistical mechanical treatment of α-helices and extended structures in proteins with inclusion of short- and medium-range interactions

A statistical mechanical model of protein conformation with medium-range interactions between theith and (i+k)^th residues (k<-4) is presented. Two two-state models, an α-helix-coil and an extended-structure-coil model, are formulated using the same form of the partition function, but the two models are applied independently to predict the locations of α-helical, extended, and coil segments; in the relatively few cases (<2%) where the predictions from the two models are in conflict, the prediction is scored as an incorrect one. Two independent sets of statistical weights (one set for each model) are derived to describe the interactions between the 20 amino acid residues for each range of interactionk; they are evaluated by minimizing an objective function so that the probability profiles for the α-helix or extended structure, respectively, in proteins computed from these statistical weights correlate optimally with the experimentally observed native conformations of these proteins. Examination of the resulting statistical weights shows that those for the interactions between hydrophobic residues and between a hydrophobic and a hydrophilic residue have reasonable magnitudes compared to what would be expected from the spatial arrangements of the side chains in the α-helix and the extended structure, and that those for the α-helix-coil model correlate well with experimentally determined values of the Zimm-Bragg parameterss and σ of the helix-coil transition theory. From the point of view of a method to predict the conformational states (i.e., α-helix, extended structure, and coil) of each residue, the statistical weights (as inall empirical prediction schemes) depend very much on the proteins used for the data base, since the presently available set of proteins of known structure is still too small for very high predictability; as a result, the correctness of the prediction is not very good for proteins not included in the data base. However, the correctness of the prediction, at least for the 37 proteins utilized as the data base in this study, is 91% and 87% for the α-helix-coil and the extended-structure-coil models, respectively; further, 79% of all the residues are predicted correctly when both the α-helix-coil and extended-structure-coil models are applied independently.     

Evaluation of methods for predicting the topology of β-barrel outer membrane proteins and a consensus prediction method

Background  

Prediction of the transmembrane strands and topology of β-barrel outer membrane proteins is of interest in current bioinformatics research. Several methods have been applied so far for this task, utilizing different algorithmic techniques and a number of freely available predictors exist. The methods can be grossly divided to those based on Hidden Markov Models (HMMs), on Neural Networks (NNs) and on Support Vector Machines (SVMs). In this work, we compare the different available methods for topology prediction of β-barrel outer membrane proteins. We evaluate their performance on a non-redundant dataset of 20 β-barrel outer membrane proteins of gram-negative bacteria, with structures known at atomic resolution. Also, we describe, for the first time, an effective way to combine the individual predictors, at will, to a single consensus prediction method.     

β-sheet topology prediction with high precision and recall for β and mixed α/β proteins

The prediction of the correct β-sheet topology for pure β and mixed α/β proteins is a critical intermediate step toward the three dimensional protein structure prediction. The predicted beta sheet topology provides distance constraints between sequentially separated residues, which reduces the three dimensional search space for a protein structure prediction algorithm. Here, we present a novel mixed integer linear optimization based framework for the prediction of β-sheet topology in β and mixed α/β proteins. The objective is to maximize the total strand-to-strand contact potential of the protein. A large number of physical constraints are applied to provide biologically meaningful topology results. The formulation permits the creation of a rank-ordered list of preferred β-sheet arrangements. Finally, the generated topologies are re-ranked using a fully atomistic approach involving torsion angle dynamics and clustering. For a large, non-redundant data set of 2102 β and mixed α/β proteins with at least 3 strands taken from the PDB, the proposed approach provides the top 5 solutions with average precision and recall greater than 78%. Consistent results are obtained in the β-sheet topology prediction for blind targets provided during the CASP8 and CASP9 experiments, as well as for actual and predicted secondary structures. The β-sheet topology prediction algorithm, BeST, is available to the scientific community at http://selene.princeton.edu/BeST/.     

Application of protein knockdown strategy targeting β-sheet structure to multiple disease-associated polyglutamine proteins

Polyglutamine diseases are a class of neurodegenerative diseases associated with the accumulation of aggregated mutant proteins. We previously developed a class of degradation-inducing agents targeting the β-sheet-rich structure typical of such aggregates, and we showed that these agents dose-, time-, and proteasome-dependently decrease the intracellular level of mutant huntingtin with an extended polyglutamine tract, which correlates well with the severity of Huntington’s disease. Here, we demonstrate that the same agents also deplete other polyglutamine disease-related proteins: mutant ataxin-3 and ataxin-7 in cells from spino-cerebellar ataxia patients, and mutant atrophin-1 in cells from dentatorubral-pallidoluysian atrophy patients. Targeting cross-β-sheet structure could be an effective design strategy to develop therapeutic agents for multiple neurodegenerative diseases.     

Single amino acid substitutions producing instability of globular proteins. Calculation of their frequencies in the entire mutational spectra of the α- and β-subunits of human hemoglobin

Summary The frequencies of substitutions resulting in protein instability were calculated by a method estimating changes in stability produced by amino acid substitutions. The method takes into account the accessibility of an amino acid position to a solvent and changes in the specificity of amino acid interactions. When tested on human mutant hemoglobins, the method yielded predictions with a preciseness of 80%. The consideration of the evolutionary homologous proteins in the analysis allowed us to estimate the evolutionary constraints imposed on stability of their spatial structure. With these limitations, approximately 50% of amino acid substitutions in the entire mutational spectra of the - and -subunits of human hemoglobin were found to damage the spatial structure of the globular proteins.