Polymorphic assemblies of double strands of sickle cell hemoglobin: Manifold pathways of deoxyhemoglobin S crystallization

Crystallization of sickle cell hemoglobin proceeds by distinctive pathways which depend upon the pH and the ionic composition of the crystallizing milieu. The pathways differ in that after fibers form they associate into different intermediates which then crystallize. We term the pathways “high pH” and “low pH”. The value of the transition pH between the high pH and low pH pathways depends upon the specific ionic species present in the hemoglobin solution. Over the pH range studied the mechanism of crystallization is pH-dependent but the structure of the crystals ultimately formed is not.In this paper we describe two newly discovered intermediates involved in the crystallization of deoxyhemoglobin S via the low pH pathway. The first of these consists of a class of particles we call macrofibers. Optical diffraction patterns of fibers and macrofibers have similar intensity distributions and layer-line spacings suggesting that macrofibers and fibers are assembled from a common structural unit which we take to be the Wishner-Love double strand.The second new structure is a paracrystalline form of deoxyhemoglobin S. The paracrystal is built from layers of double strands of molecules in an arrangement similar to that within the crystals. Optical diffraction of electron micrographs of paracrystals reveals that longitudinal disorder is present between double strands. Projections of the electron density down the c axis of the crystal provide images very similar to those in electron micrographs of negatively stained paracrystals. The patterns appearing in the paracrystal due to the disorder can be fully simulated by shifts between the layers of double strands.     

Macrofiber structure and the dynamics of sickle cell hemoglobin crystallization

Fibers of deoxyhemoglobin S undergo spontaneous crystallization by a mechanism involving a variety of intermediate structures. These intermediate structures, in common with the fiber and crystal, consist of Wishner-Love double strands of hemoglobin S molecules arranged in different configurations. The structure of one of the key intermediates linking the fiber and crystal, called a macrofiber, has been studied by a variety of analytical procedures. The results of the analysis indicate that the intermediates involved in the fiber to crystal transition have many common structural features. Fourier analysis of electron micrographs of macrofibers confirms that they are composed of Wishner-Love double strands of hemoglobin molecules. Electron micrographs of macrofiber cross-sections reveal that the arrangement of the double strands in macrofibers resembles that seen in micrographs of the a axis projection of the crystal. This orientation provides an end-on view of the double strands which appear as paired dumb-bell-like masses. The structural detail becomes progressively less distinct towards the edge of the particle due to twisting of the double strands about the particle axis. Serial sections of macrofibers confirm that these particles do indeed rotate about their axes. The twist of the particle is right handed and its average pitch is 10,000 Å. The effect of rotation on the appearance of macrofiber cross-sections 300 to 400 Å thick can be simulated by a 15 ° rotation of an a axis crystal projection. The relative polarity of the double strands in macrofibers and crystals can be determined easily by direct inspection of the micrographs. In both macrofibers and crystals they are in an anti-parallel array.On the basis of these observations we conclude that crystallization of macrofibers involves untwisting and alignment of the double strands.     

Studies of the fiber to crystal transition of sickle cell hemoglobin in acidic polyethylene glycol

The crystallization of deoxygenated sickle cell hemoglobin in acidic (pH 5.2) polyethylene glycol (10%) has been studied in order to determine if the mechanism of crystal formation under such conditions has features in common with the mechanism of crystal formation at higher pH values in the absence of polyethylene glycol. The existence of a common mechanism of crystallization under different conditions is relevant in validating the use of the known high resolution crystal structure to interpret the fiber structure. Our findings indicate that deoxygenated sickle cell hemoglobin crystallization in acidic polyethylene glycol is initiated by fiber formation. Fibers, in turn, convert to larger structures called macrofibers within several hours (Wellems et al., 1981). Fibers and macrofibers (and their respective optical transforms) formed in acidic polyethylene glycol appear to have the same structure as their counterparts formed at higher pH values in the absence of polyethylene glycol. Early in the transition one can observe macrofibers in the process of alignment and fusion. The structural characterization of the intermediates leaves little doubt that crystallization in acidic polyethylene glycol is mediated by the same mechanism as that occurring under more physiological conditions, and that fibers are a metastable intermediate whose ultimate fate is to crystallize.     

Analysis of the stability of hemoglobin S double strands.

The deoxyhemoglobin S (deoxy-HbS) double strand is the fundamental building block of both the crystals of deoxy-HbS and the physiologically relevant fibers present within sickle cells. To use the atomic-resolution detail of the hemoglobin-hemoglobin interaction known from the crystallography of HbS as a basis for understanding the interactions in the fibers, it is necessary to define precisely the relationship between the straight double strands in the crystal and the twisted, helical double strands in the fibers. The intermolecular contact conferring the stability of the double strand in both crystal and fiber is between the beta6 valine on one HbS molecule and residues near the EF corner of an adjacent molecule. Models for the helical double strands were constructed by a geometric transformation from crystal to fiber that preserves this critical interaction, minimizes distortion, and makes the transformation as smooth as possible. From these models, the energy of association was calculated over the range of all possible helical twists of the double strands and all possible distances of the double strands from the fiber axis. The calculated association energies reflect the fact that the axial interactions decrease as the distance between the double strand and the fiber axis increases, because of the increased length of the helical path taken by the double strand. The lateral interactions between HbS molecules in a double strand change relatively little between the crystal and possible helical double strands. If the twist of the fiber or the distance between the double strand and the fiber axis is too great, the lateral interaction is broken by intermolecular contacts in the region around the beta6 valine. Consequently, the geometry of the beta6 valine interaction and the residues surrounding it severely restricts the possible helical twist, radius, and handedness of helical aggregates constructed from the double strands. The limitations defined by this analysis establish the structural basis for the right-handed twist observed in HbS fibers and demonstrates that for a subunit twist of 8 degrees, the fiber diameter cannot be more than approximately 300 A, consistent with electron microscope observations. The energy of interaction among HbS molecules in a double strand is very slowly varying with helical pitch, explaining the variable pitch observed in HbS fibers. The analysis results in a model for the HbS double strand, for use in the analysis of interactions between double strands and for refinement of models of the HbS fibers against x-ray diffraction data.     

Structural basis and dynamics of the fiber-to-crystal transition of sickle cell hemoglobin

The kinetics of the assembly of structurally distinct, polymeric aggregates constituting the fiber-to-crystal transition of sickle cell hemoglobin in slowly stirred, deoxygenated solutions has been studied with the use of electron microscopy as a function of pH, as a function of the crystal structures of mutant forms of human deoxyhemoglobins employed as nucleating seeds, and as a function of hemoglobin S chemically modified at the Cys F9 (beta 93) position. The temporal order of appearance of fibers of approximately 210 A diameter, bundles of aligned fibers, macrofibers of greater than or equal to 650 A diameter, and microcrystals is observed. Microscopic fragments of end-stage crystals formed under slowly stirred conditions and introduced as nucleating seeds enhance the rate of crystallization only when added prior to the formation of large bundles of aligned fibers, while microscopic seed crystals added after the formation of bundles of aligned fibers do not alter the rate of crystallization. Over the pH range 6.3 to 7.1, the presence of macrofibers does not influence modulation of the kinetics of the transition with seed crystal fragments. Microscopic seed crystals of deoxyhemoglobin S and deoxyhemoglobin C formed under acidic conditions (pH less than 6.5) have a comparable influence on the kinetics of the fiber-to-crystal transition to that of end-stage crystals. Microscopic seed crystals of deoxyhemoglobin C formed under alkaline conditions (pH greater than 6.5) enhance the formation of macrofibers but do not alter the rate of crystallization. Under conditions associated with enhanced formation of macrofibers, metastable microscopic crystals having axial periodicities of approximately 64 A and approximately 210 A are observed in the intermediate phase of the transition, while end-stage crystals have axial unit cell dimensions identical to those of deoxyhemoglobin S crystallized from polyethylene glycol solutions of pH less than 6.5. Although the metastable crystals may arise from fragments of macrofibers, it is shown that they cannot be transformed directly into end-stage crystals under slowly stirred conditions without undergoing dissolution. These results stipulate that the pathway of the fiber-to-crystal transition proceeds according to the reaction: (Formula: see text) wherein the rate-limiting step is the alignment of fibers into large bundles, and macrofibers are not an intermediate of the fiber-to-crystal transition.(ABSTRACT TRUNCATED AT 400 WORDS)     

Molecular packing in a second monoclinic crystal of deoxygenated sickle hemoglobin

A close correspondence has been demonstrated between double filaments of deoxygenated hemoglobin S molecules as found in monoclinic crystals, forms I and II, and in sickle fibers. We have carried out a low resolution study of monoclinic form II by X-ray diffraction analysis. Its structure differs from that of form I solely by a shift along the a-axis of the molecular centers of the asymmetric unit, which forms the double filament. The magnitude of the translation was determined from a minimum residual calculation. The x co-ordinates of the symmetry related molecular centers of antipolar double filaments are approximately the same. This means that the double filaments are nearly in register. A minor component associated with form II crystals proved to be form I. The possible existence of additional forms is discussed.The significance of the molecular arrangement in form II is related to its presence in sickle fibers. We have determined the contacts between antipolar double filaments in this form as well as a number in form I not tabulated previously. These new contacts represent additional stabilizing interactions that might provide targets for the design of stereospecific antisickling agents.     

Intermolecular contacts within sickle hemoglobin fibers

By combining X-ray crystallographic co-ordinates of sickle hemoglobin (HbS) molecules with three-dimensional reconstructions of electron micrographs of HbS fibers we have synthesized a model for the structure of the clinically relevant HbS fiber. This model largely accounts for the action of 55 point mutations of HbS whose effect on fiber formation has been studied. In addition, it predicts locations at which additional point mutations are likely to affect fiber formation. The number of intermolecular axial contacts decreases with radius until, at the periphery of the fiber, there are essentially no axial contacts. We suggest that this observation accounts for the limited radial growth of the HbS fiber and that a similar mechanism may be a factor in limiting the size of other helical particles. The methodology for the synthesis of the fiber model is applicable to other systems in which X-ray crystallographic and electron microscopic data are available.     

X-ray diffraction studies of 14-filament models of deoxygenated sickle cell hemoglobin fibers. II. Models based on the deoxygenated sickle hemoglobin crystal structure

The calculated transforms of a number of crystal-based models of the deoxygenated sickle cell hemoglobin fiber have been compared with X-ray diffraction data of 15 A (1 A = 0.1 nm) resolution. The fiber models consist of 14 single strands of sickle cell hemoglobin (HbS) molecules, which associate into seven protofilaments arranged similarly to those present in the crystal structure. Six of the protofilaments are arranged in three crystallographic until cells extending in the c-axis direction with the seventh protofilament positioned so as to provide an elliptical cross-section when the assemblage is viewed down the fiber axis. Models were generated by systematically and independently translating each of the model's three subcells in steps of 3.5 A along the fiber axis. The seventh protofilament was kept fixed as a point of reference. Each translation of a subcell corresponded to a different fiber model whose transform was then compared with observed data. In all, over 46,000 transforms were computed; of these, three models with minimal residuals were identified. The free energy of packing for all crystal-based models was evaluated to find configurations of protofilaments possessing minimal free energies. The results of the calculations support the subcell configurations of two of the three models with minimal residuals.     

Molecular insights into the irreversible mechanical behavior of sickle hemoglobin

Sickle cell disease is caused by the amino acid substitution of glutamic acid to valine, which leads to the polymerization of deoxygenated sickle hemoglobin (HbS) into long strands. These strands are responsible for the sickling of red blood cells (RBCs), making blood hyper-coagulable leading to an increased chance of vaso-occlusive crisis. The conformational changes in sickled RBCs traveling through narrow blood vessels in a highly viscous fluid are critical in understanding; however, there are few studies that investigate the origins of the molecular mechanical behavior of sickled RBCs. In this work, we investigate the molecular mechanical properties of HbS molecules. A mechanical model was used to estimate the directional stiffness of an HbS molecule and the results were compared to adult human hemoglobin (HbA). The comparison shows a significant difference in strength between HbS and HbA, as well as anisotropic behavior of the hemoglobin molecules. The results also indicated that the HbS molecule experienced more irreversible mechanical behavior than HbA under compression. Further, we have characterized the elastic and compressive properties of a double stranded sickle fiber using six HbS molecules, and it shows that the HbS molecules are bound to each other through strong inter-molecular forces.     

Polymerization of deoxy-sickle cell hemoglobin in high-phosphate buffer

Deoxy-sicklecell hemoglobin (HbS) polymerizes in 0.05 M phosphate buffer to form long helical fibers. The reaction typically occurs when the concentration of HbS is about 165 mg/ml. Polymerization produces a variety of polymorphic forms. The structure of the fibers can be probed by using site-directed mutants to examine the effect of altering the residues involved in intermolecular interactions. Polymerization can also be induced in the presence of 1.5 M phosphate buffer. Under these conditions polymerization occurs at much lower concentrations (ca. 5 mg/ml), which is advantageous when site-directed mutants are being used because only small quantities of the mutants are available. We have characterized the structure of HbS polymers formed in 1.5 M phosphate to determine how their structures are related to the polymers formed under more physiological conditions. Under both sets of conditions fibers are the first species to form. At pHs between 6.7 and 7.3 fibers initially form bundles and then crystals. At lower pHs fibers form macrofibers and then crystals. Fourier transforms of micrographs of the polymers formed in 1.5 M phosphate display the 32- and 64-A(-1) periodicity characteristic of fibers formed in 0.05 M phosphate buffer. The 64-A(-1) layer line is less prominent in Fourier transforms of negatively stained fibers formed in 1.5 M phosphate possibly because salt interferes with staining of the fibers. However, micrographs and Fourier transforms of frozen hydrated fibers formed in high and low phosphate display the same periodicities. Under both sets of reaction conditions HbS polymers form crystals with the same unit cell parameters as Wishner-Love crystals (a = 64 A, b = 185 A, c = 53 A). Some of the polymerization intermediates were examined in the frozen-hydrated state in order to determine whether their structures were significantly perturbed by negative staining. We have also carried out reconstructions of the frozen-hydrated fibers in high and low phosphate to compare their molecular coordinates. The helical projection of the reconstructions in low phosphate shows the expected 14-strand structure. In high phosphate the 14-strand fibers are also formed and their molecular coordinates are the same (within experimental error) as those of fibers formed in 0.05 M phosphate. In addition, the reconstructions of high-phosphate fibers reveal a new minor variant of fiber containing 10 strands. The polymerization products in 1.5 M phosphate buffer were generally indistinguishable from those formed in 0.05 M phosphate buffer. Micrographs of frozen hydrated specimens have facilitated the interpretation of previously published micrographs using negative staining.     

Structural analysis of polymers of sickle cell hemoglobin. III. Fibers within fascicles

We have examined the structure of hemoglobin S fibers, which are associated into large bundles, or fascicles. Electron micrographs of embedded and cross-sectioned fascicles provide an end-on view of the component fibers. The cross-sectional images are rotationally blurred as a result of the twist of the fiber within the finite thickness of the section. We have applied restoration techniques to recover a deblurred image of the fiber. The first step in this procedure involved correlation averaging images of cross-sections of individual fibers in order to improve the signal-to-noise ratio. The rotationally blurred image was then geometrically transformed to polar co-ordinates. In this space, the rotational blur is transformed into a linear blur. The linearly blurred image is the convolution of the unblurred image and a point spread function that can be closely approximated by a square pulse. Deconvolution in Fourier space, followed by remapping to Cartesian co-ordinates, produced a deblurred image of the original micrograph. The deblurred images indicate that the fiber is comprised of 14 strands of hemoglobin S. This result provides confirmation of the fiber structure determined using helical reconstruction techniques and indicates that the association of fibers into ordered arrays does not alter their molecular structure.     

Triclinic crystals associated with fibers of deoxygenated sickle hemoglobin.

Triclinic crystals have been found in capillaries that initially contained deoxygenated sickled erythrocytes, and in solutions of sickle hemoglobin that were stirred during deoxygenation. In both cases these crystals occur as a phase transition from fibers. They have been observed only as twins; the a-axis of one member is related to that of its twin by 180 degrees rotation about the b* direction. The volume of the triclinic crystal unit cell is half that of the monoclinic crystals that have also transformed from fibers. Analysis of X-ray diffraction data indicates that the two molecules in the triclinic unit cell that repeat at an interval of 64 A form double filaments similar to those found in the monoclinic crystals and in the fiber. The existence of the triclinic crystals which contain only one double filament per unit cell removes a postulated requirement that antipolar double filament pairs be the sole unit of the fiber architecture.     

Oxygen equilibrium of emulsified solutions of normal and sickle hemoglobin.

Emulsions containing microdroplets of concentrated solutions of normal or sickle hemoglobin dispersed in a continous oil phase have been prepared, and the aggregation of sickle hemoglobin within microdroplets in a deoxygenated emulsion demonstrated. The equilibrium oxygen saturation of hemoglobin in the emulsions has been measured as a function of the partial pressure of oxygen by a novel spectrophotometric technique which corrects for the scattering of light in the emulsion. The half-saturation oxygen pressure and cooperativity of oxygen binding are substantially greater in concentrated (27 g/dl) solutions of sickle hemoglobin than in solutions of non-aggregating hemoglobin. The shape of the oxygen equilibrium curve of concentrated sickle hemoglobin is qualitatively discussed in terms of a previously proposed model (1).     

Patterns in the quinary structures of proteins. Plasticity and inequivalence of individual molecules in helical arrays of sickle cell hemoglobin and tubulin.

The four recognized levels of organization of protein structure (primary through quaternary) are extended to add the designation quinary structure for the interactions within helical arrays, such as found for sickle cell hemoglobin fibers or tubulin units in microtubules. For sickle cell hemoglobin the main quinary structure is a 14-filament fiber, with a number of other minor forms also encountered. Degenerate forms of the 14-filament fibers can be characterized that lack specific pairs of filaments; evidence is presented which suggests an overall organization of the 14 filaments in pairs, with particular pairs aligned in an antiparallel orientation. For tubulin, a range of quinary structures can be detected depending on the number of protofilaments and whether adjacent protofilaments composed of alternating alpha- and beta-subunits are aligned with contacts between like or unlike subunits and with parallel or antiparallel polarity. Thus, in contrast to quarternary structure, which generally involves a fixed number of subunits, the quinary structures of proteins can exhibit marked plasticity and inequivalence in the juxtaposition of constituent molecules.     

Parallel helical domains in DNA branched junctions containing 5',5' and 3',3' linkages

The Holliday junction is a central intermediate in genetic recombination. It contains four strands of DNA that are paired into four double helical arms that flank a branch point. In the presence of Mg2+, the four arms are known to stack in pairs forming two helical domains whose orientations are antiparallel but twisted by about 60 degrees. The basis for the antiparallel orientation of the domains could be either junction structure or the effect of electrostatic repulsion between domains. To discriminate between these two possibilities, we have constructed and characterized an analogue, called a bowtie junction, in which one strand contains a 3',3' linkage at the branch point, the strand opposite it contains a 5',5' linkage, and the other two strands contain conventional 3',5' linkages. Electrostatic effects are expected to lead to an antiparallel structure in this system. We have characterized the molecule in comparison with a conventional immobile branched junction by Ferguson analysis and by observing its thermal transition profile; the two molecules behave virtually identically in these assays. Hydroxyl radical autofootprinting has been used to establish that the unusual linkages occur at the branch point and that the arms stack to form the same domains as the conventional junction. Cooper-Hagerman gel mobility analyses have been used to determine the relative orientations of the helical domains. Remarkably, we find them to be closer to parallel than to antiparallel, suggesting that the preferred structure of the branch point dominates over electrostatic repulsion. We have controlled for the number of available bonds in the branch point, for gel concentration, and for the role of divalent cations. This finding suggests that control of branch point structure alone can lead to parallel domains, which are generally consistent with recombination models derived from genetic data.     

An X-ray determination of the molecular interactions in hemoglobin C: a disease characterized by intraerythrocytic crystals.

The X-ray structure of cyanomet human hemoglobin C has been solved and refined, R ～27%. The molecular packing can be represented in two dimensions by two sets of parallel strands, one set in the $\text{b}$ direction and the other in the $\text{c}$ direction. Taken together the two sets of strands interconnect the molecules into square nets or layers where each molecule contacts its four nearest neighbors. Molecules in one layer are displaced in $\text{a}$ and $\text{b}$ so that they fit into the “holes” of the square arrays of the adjacent layers (normal to $\text{a}$) resulting in a pseudo body-centered cubic packing. This packing can account for the hemoglobin crystallization in and fragility of the erythrocytes. The aberrant β6A3 Lys residue is in a position to influence the crystal formation.     

The hand of the helix of deoxyhemoglobin S fibers.

Electron micrographs of deoxyhemoglobin S fiber cross sections provide an end-on view of the fiber whose appearance is sensitive to small changes in orientation. We have developed a procedure to exploit this sensitivity in order to determine the hand of these particles. In a sickle hemoglobin fiber the hemoglobin molecules form long pitch helical strands which twist about the particle axis with a pitch of about 3000 A. Tilting a 400-A-thick cross section by a few degrees aligns one of the long pitch helices so that it is nearly parallel to the direction of view. When a strand of hemoglobin molecules in a fiber is aligned in this manner it appears as a strongly contrasted bright spot. It is this spot, rather than the fiber axis, which appears to be the apparent center of rotation of the cross section. The direction of the displacement of the spot from the particle axis depends upon the particle hand and tilt direction. We have used this property to determine that sickle hemoglobin fibers are right-handed particles. This method may be applicable to other particles with long pitch helices as well.     

The solubility of hemoglobin beta 4 S, the mutant subunits of sickle cell hemoglobin

The single subunit hemoglobin β₄^S was found to have a solubility comparable to that of oxygenated rather than deoxygenated Hb S, although it contains twice as many mutant chains as the parent hemoglobin and probably has a quarternary structure similar to deoxyhemoglobin A. This finding supports the assumption that receptor sites in the α chains of sickle hemoglobin are essential for sickling.     

Secondary structure of the r(CUUCGG) tetraloop.

The oligonucleotide r(GGACUUCGGUCC) has been observed to adopt a hairpin conformation by solution NMR and a double helical conformation by X-ray diffraction. In order to understand this apparent conflict, we used time-resolved fluorescence depolarization and 19fluorine NMR to follow the secondary structure of this dodecamer as the solution composition was changed stepwise from the NMR experimental conditions to those used for crystallization. Calculation of the dodecamer concentration in the crystal (180 mM strands) and the cation concentration needed for neutrality (>2 M) prompted investigation of a tethered species, in which two dodecamers are connected by a string of 4 nt, geometrically equivalent to approximately 100 mM strands, in 2.5 M NaCl. The RNA tetraloop and its DNA analog maintain a single-strand hairpin conformation in solution, even under the conditions used to grow the crystal. Under high salt conditions, the tethered RNA and DNA analogs of this sequence yield secondary components which could be the double helical conformation. Crystal contacts in addition to solvent changes and high RNA concentrations are needed to obtain the double helix as the predominant species.     

Clockwise and counterclockwise pinwheel colony morphologies of Bacillus subtilis are correlated with the helix hand of the strain

Helical macrofiber-producing strains of Bacillus subtilis grown on fresh complex medium semisolid surfaces formed "pinwheel"-shaped colonies. Clockwise pinwheel projections arose from colonies of strains that produce right-handed helical macrofibers in fluid cultures. Most strains able to make left-handed helical macrofibers in fluid grew as disorganized wavy colonies without directed projections. A phage-resistant left-handed mutant was found that produces very tight colonies with pinwheel projections that lie counterclockwise relative to the colony. The pinwheel colony morphology is interpreted therefore in terms of the cell surface organization and helical growth.