Equivalence of the projected structure of thin catalase crystals preserved for electron microscopy by negative stain,glucose or embedding in the presence of tannic acid

Thin crystals of beef liver catalase have been examined by electron microscopy following various preservation procedures. In the first part of this investigation, micrographs of three principal projections were obtained from thin sections of micro-crystals embedded in the presence of tannic acid. Computer reconstructions confirmed the space group assignment of P2₁2₁2₁ and permitted the packing arrangement of the catalase tetramers to be deduced to a resolution of about 20 Å. These results corroborate the packing model for this crystal form proposed by Unwin (1975) on the basis of molecular modeling of one projection. In the second part of this investigation, the projected structures of the thin crystals in various preserving media were compared. The negative contrasting of crystals embedded in the presence of tannic acid was confirmed by direct comparison with nonembedded, negatively stained thin platelet crystals. In addition, good agreement at 20 Å resolution was observed between the structure of negatively stained crystals and the structure of crystal platelets preserved in glucose and examined by lowdose methods, while moderate agreement was established with the published data of Taylor (1978) for crystals embedded in thin ice films. Tannic acid alone was also found to serve as a suitable medium for preserving catalase crystals to a resolution of 3.7 Å as judged by electron diffraction. Overall, we demonstrate that projections obtained from thin sections of catalase crystals embedded in the presence of tannic acid can provide a reliable, negatively contrasted representation of the protein structure to 20 Å resolution. Examination of sectioned crystals could thus provide a useful adjunct to X-ray crystallographic studies of protein crystals and three-dimensional reconstruction of crystal thin sections should ultimately be possible.     

Two-dimensional crystallization conditions of human leukotriene C4 synthase requiring adjustment of a particularly large combination of specific parameters

Human leukotriene C₄ synthase (LTC₄S) forms highly ordered two-dimensional (2D) crystals under specific reconstitution conditions. It was found that control of a larger number of parameters than is usually observed for 2D crystallization of membrane proteins was necessary to induce crystal formation of LTC₄S. Here, we describe the parameters that were optimized to yield large and well-ordered 2D crystals of LTC₄S. Careful fractioning of eluates during the protein purification was essential for obtaining crystals. While the lipid-to-protein ratio was critical in obtaining order, four parameters were decisive in inducing growth of crystals that were up to several microns in size. To obtain a favorable diameter, salt, temperature, glycerol, and initial detergent concentration had to be controlled with great care. Interestingly, several crystal forms could be grown, namely the plane group symmetries of p2, p3, p312, and two different unit cell sizes of plane group symmetry p321.     

Trimers, dimers of trimers, and trimers of trimers are common building blocks of annexin a5 two-dimensional crystals

Annexin A5 is a member of a family of homologous proteins sharing the ability to bind to negatively charged phospholipid membranes in a Ca(2+)-dependent manner. Annexin A5, as well as other annexins, self-assembles into two-dimensional (2D) ordered arrays upon binding to membranes, a property that has been proposed to have functional implications. Electron microscopy and atomic force microscopy experiments have revealed that annexin A5 forms two types of 2D crystals-with either p6 or p3 symmetry-that are both based on annexin trimers. In this study, we describe three other crystal forms that coexist with the p6 crystals. All crystal forms are made of the same building blocks, namely, dimers of trimers and trimers of trimers. A mechanistic model of the formation of the annexin A5 2D crystals is proposed.     

Structure of cytochrome c oxidase in deoxycholate-drived two-dimensional crystals.

The structure of the two-dimensional crystals of cytochrome oxidase prepared with deoxycholate has been investigated. The crystals have space group p12₁ and contain a monomer (two heme-two copper complex) in the asymmetric unit. They are in the form of sheets and contain no continuous bilayer; the entire surface of the molecule seems to be visible in negatively stained samples. The monomer is roughly 110 Å long and resembles a lopsided “Y”. The domains which form the arms of the Y are 55 Å in length and have a center to center separation of 40 Å. These domains are on the matrix side of the molecule and are thought to be buried in the bilayer of the inner mitochondrial membrane. The cytoplasmic side of the molecule is composed of the single large domain which is the stem of the Y. The overall structure matches that of cytochrome oxidase seen in the p22₁2₁ crystals derived by Triton X100 treatment of mitochondria.     

Three-Dimensional Electron Microscopy of the Photosystem II Core Complex

A three-dimensional image of the spinach photosystem II core complex composed of CP47, D1, D2, cytochromeb-559, andpsbI gene product was reconstructed at 20-Å resolution from the two-dimensional crystals negatively stained with phosphotungstate. Confirming the previous proposal, the crystal had ap22₁2₁symmetry. One PSII core complex was measured to be 80 × 80 Å in the membrane plane and 88 Å normal to it. The mass distribution was asymmetric about the lipid bilayer, consistent with predictions from the amino acid sequences. The lumenal mass consisted of three domains forming a characteristic triangular platform with another domain on top of it. Three stromal domains were smaller and linearly arranged. Due to strong stain exclusion in the hydrophobic core part of the lipid bilayer, the transmembrane region appeared to be imaged with a reversed contrast. Inverting the contrast resulted in a reasonable density distribution for that part. Thus, though the information on the transmembrane region is limited, the domain structure of the PSII core complex was revealed and allowed us to propose a model for the arrangement of subunits in the PSII core complex.     

Helical channels in crystals of gramicidin A and of a cesium--gramicidin A complex: an x-ray diffraction study.

X-ray crystallographic studies of gramicidin A crystallized from methanol (P2₁) and ethanol (P2₁2₁2₁), and of a Cs⁺ gramicidin A complex crystallized from methanol (P222₁, P2₁2₁2 or P2₁2₁2₁) are reported here. The asymmetric unit consists of two molecules of gramicidin A in the native crystals and four molecules in the cesium complex crystal. Patterson analyses show that gramicidin A in these crystals forms a cylindrical helical channel. In the two types of native gramicidin crystals, the diameter of this channel is about 5 å and its length is about 32 å. Cesium ions are bound inside this channel in crystals of the cesium-gramicidin A complex. The channel in this complex is considerably shorter (26 Å) and wider (6·8Å) than in the native forms. The Patterson maps of these three crystal forms are compatible with either the single-stranded β-helix (Urry, 1971) or the double-stranded parallel or anti-parallel, β-helix (Veatch et al., 1974).     

Effect of Lipid Head Groups on Double-Layered Two-Dimensional Crystals Formed by Aquaporin-0

Aquaporin-0 (AQP0) is a lens-specific water channel that also forms membrane junctions. Reconstitution of AQP0 with dimyristoyl phosphatidylcholine (DMPC) and E. coli polar lipids (EPL) yielded well-ordered, double-layered two-dimensional (2D) crystals that allowed electron crystallographic structure determination of the AQP0-mediated membrane junction. The interacting tetramers in the two crystalline layers are exactly in register, resulting in crystals with p422 symmetry. The high-resolution density maps also allowed modeling of the annular lipids surrounding the tetramers. Comparison of the DMPC and EPL bilayers suggested that the lipid head groups do not play an important role in the interaction of annular lipids with AQP0. We now reconstituted AQP0 with the anionic lipid dimyristoyl phosphatidylglycerol (DMPG), which yielded a mixture of 2D crystals with different symmetries. The different crystal symmetries result from shifts between the two crystalline layers, suggesting that the negatively charged PG head group destabilizes the interaction between the extracellular AQP0 surfaces. Reconstitution of AQP0 with dimyristoyl phosphatidylserine (DMPS), another anionic lipid, yielded crystals that had the usual p422 symmetry, but the crystals showed a pH-dependent tendency to stack through their cytoplasmic surfaces. Finally, AQP0 failed to reconstitute into membranes that were composed of more than 40% dimyristoyl phosphatidic acid (DMPA). Hence, although DMPG, DMPS, and DMPA are all negatively charged lipids, they have very different effects on AQP0 2D crystals, illustrating the importance of the specific lipid head group chemistry beyond its mere charge.     

Surface topography of the p3 and p6 annexin V crystal forms determined by atomic force microscopy

Annexin V is a member of a family of structurally homologous proteins sharing the ability to bind to negatively charged phospholipid membranes in a Ca(2+)-dependent manner. The structure of the soluble form of annexin V has been solved by X-ray crystallography, while electron crystallography of two-dimensional (2D) crystals has been used to reveal the structure of its membrane-bound form. Two 2D crystal forms of annexin V have been reported to date, with either p6 or p3 symmetry. Atomic force microscopy has previously been used to investigate the growth and the topography of the p6 crystal form on supported phospholipid bilayers (Reviakine et al., 1998). The surface structure of the second crystal form, p3, is presented in this study, along with an improved topographic map of the p6 crystal form. The observed topography is correlated with the structure determined by X-ray crystallography.     

Improved Model of a LexA Repressor Dimer Bound to recA Operator

Abstract

A complete three dimensional model for the LexA repressor dimer bound to the recA operator site consistent with relevant biochemical and biophysical data for the repressor was proposed from our laboratory when no crystal structure of LexA was available. Subsequently, the crystal structures of four LexA mutants Δ_1–67 S119A, S119A, G85D and Δ_1-67 quadruple mutant in the absence of operator were reported. It is examined in this paper to what extent our previous model was correct and how, using the crystal structure of the operator-free LexA dimer we can predict an improved model of LexA dimer bound to recA operator. In our improved model, the C-domain dimerization observed repeatedly in the mutant operator-free crystals is retained but the relative orientation between the two domains within a LexA molecule changes. The crystal structure of wild type LexA with or without the recA operator cannot be solved as it autocleaves itself. We argue that the ‘cleavable’ cleavage site region found in the crystal structures is actually the more relevant form of the region in wildtype LexA since it agrees with the value of the pre-exponential Arrhenius factor for its auto- cleavage, absence of various types of trans-cleavages, difficulty in modifying the catalytic serine by diisopropyl flourophosphate and lack of cleavage at Arg 81 by trypsin; hence the concept of a ‘conformational switch’ inferred from the crystal structures is meaningless.     

Raman spectra and conformation of the glycerophosphorylcholine headgroup

The Raman and infrared spectra of L-α-glycerophosphylcholine (GPC) and its cadmium complex (GPC·CdCl₂) have been measured for both the anhydrous and hydrate crystals. The Raman spectra of GPC and GPC·CdCl₂ dissolved in water are also reported. All the spectra fall into three distinctive patterns, demonstrating the existence of at least three different kinds of rotational isomers; the first is stable in the anhydrous GPC crystal, the second is stable in the GPC hydrate and GPC·CdCl₂ hydrate crystals and the third is stable in the anhydrous GPC·CdCl₂ crystal and also in aqueous solutions of GPC·CdCl₂. The latter form is reported for the first time.Raman spectra of dimyristoyl-L-α-phosphatidylcholine (L-DMPC) and dipalmitoyl-(L and DL)-α-phosphatidylcholine (L and DL-DPPC) hydrates have been measured below, around and above the transition temperature T₁. The spectra below T₁ indicate that the conformation of the glycerophosphorylcholine (GPC) headgroup in the L-DMPC crystal is similar to that in the L-DPPC crystal. The headgroup conformation in the DL-DPPC crystal is quite different from that in the L-DPPC crystal, although they became the same aftera crystalline trasnformation of DL-DPPC. The changes of the Raman spectra through the T₁ transition for the above crystals show that the headgroup conformation remains unchanged at the beginning o the transition.     

Crystallization of tRNAs as Cetyltrimethylammonium Salts

VARIOUS species of transfer RNA have been crystallized by controlled precipitation from aqueous solutions containing organic solvents or ammonium sulphate (reviewed in refs. 1 and 2). These methods have produced a great variety of crystal forms which, with a few exceptions^3,4, are usually poorly ordered as judged by X-ray diffraction. This is probably because the interactions between molecules are few and rather nonspecific, making the crystal structure extremely sensitive to the crystallization conditions. For this reason, attempts have been made to crystallize tRNA as the cetyltrimethylammonium (CTA-) salt. The additional interaction between hydrophobic cetyl cations bound to the different molecules may stabilize the crystal lattice and have a positive effect on the crystallization process and therefore on the order of the crystals. We report here the production of crystals of CTA-salts of five different tRNAs; tRNA^Met_f, tRNA^Glu, tRNA^Phe, tRNA^Tyr from E. coli and tRNA^Phe from yeast. In the case of tRNA^Met_f, different crystal forms were obtained in the presence of different cations.     

A maximum likelihood approach to two-dimensional crystals

Maximum likelihood (ML) processing of transmission electron microscopy images of protein particles can produce reconstructions of superior resolution due to a reduced reference bias. We have investigated a ML processing approach to images centered on the unit cells of two-dimensional (2D) crystal images. The implemented software makes use of the predictive lattice node tracking in the MRC software, which is used to window particle stacks. These are then noise-whitened and subjected to ML processing. Resulting ML maps are translated into amplitudes and phases for further processing within the 2dx software package. Compared with ML processing for randomly oriented single particles, the required computational costs are greatly reduced as the 2D crystals restrict the parameter search space. The software was applied to images of negatively stained or frozen hydrated 2D crystals of different crystal order. We find that the ML algorithm is not free from reference bias, even though its sensitivity to noise correlation is lower than for pure cross-correlation alignment. Compared with crystallographic processing, the newly developed software yields better resolution for 2D crystal images of lower crystal quality, and it performs equally well for well-ordered crystal images.     

Polarized Raman scattering from oriented single microcrystals of d(A5T5)2 and d(pTpT)

Polarized Raman spectra have been obtained from single microcrystals of the duplex of the decamer d(A₅T₅)₂ using a Raman microscope. This is the first report of Raman spectra from a crystal of a deoxyoligomer that contains only long, nonalternating sequences of adenine and thymine. Sequences containing d(A)_n and d(T)_n are of interest in view of recent suggestions that they induce bends in DNA and that they might exist in a nonstandard B-conformation. Polarized Raman spectra of a crystal of d(pTpT) have also been obtained. Both crystals display Raman bands whose intensities are very sensitive to the orientation of the crystal with respect to the direction of polarization of the incident laser beam. These spectra indicate that the helical axes of the oligonucleotides are parallel to the long axes of the crystals and that the d(A₅T₅)₂ is not appreciably bent in the crystal. The Raman spectrum from the d(pTpT) crystal indicates that all of the furanose ring puckers are in a C2′-endo configuration since only the C2′-endo marker band at 835 ± 5 cm^?1 is present. Crystals of d(A₅T₅)₂ show measurable Raman intensities in both the 838- and 816-cm^?1 bands. This indicates the presence of both the C2′-endo and C3′-endo, or possibly other non-C2′-endo, furanose conformations. The 816-cm^?1 band is weak so that only a small fraction of the residues are estimated to be in the non-C2′-endo conformation. In both the d(pTpT) and d(A₅T₅)₂ crystals the intensity of the bands due to vibrations of the backbone show only a small dependence on orientation of the crystals. This result is explained by the low symmetry of the puckered sugar rings. It is concluded that Raman spectra obtained from oligonucleotide crystals in which the orientation of the crystal axes to the laser polarization is not carefully controlled may contain intensity artifacts that are due to polarization effects.     

Recrystallization of Ice Crystals in Trehalose Solution at Isothermal Condition

An isothermal ice recrystallization behavior in trehalose solution was investigated. The isothermal recrystallization rate constants of ice crystals in trehalose solution were obtained at ?5 °C, ?7 °C, and ?10 °C. Then the results were compared to those of a sucrose solution used as a control sample. Simultaneous estimation of water mobility in the freeze-concentrated matrix was conducted by ¹H spin–spin relaxation time T₂ to investigate mechanisms causing the different ice crystal recrystallization behaviors of sucrose and trehalose. At lower temperatures, lower recrystallization rates were obtained for both trehalose and sucrose solutions. The ice crystallization rate constants in trahalose solution tended to be smaller than those in sucrose solution at the same temperature. Although different ice contents (less than 3.6%) were observed between trehalose and sucrose solutions at the same temperature, the recrystallization behaviors of ice crystals were not markedly different. The ¹H spin–spin relaxation time T₂ of water components in a freeze-concentrated matrix for trehalose solution was shorter than in a sucrose solution at the same temperature. Results show that the water mobility of trehalose solutions in freeze-concentrated matrix was less than that of sucrose solutions, which was suggested as the reason for retarded ice crystal growth in a trehalose solution. Results of this study suggest that the replacement of sucrose with trehalose will not negatively affect deterioration caused by ice crystal recrystallization in frozen foods and cryobiological materials.     

Crystallization and preliminary crystallographic analysis of the N-terminal actin binding domain of human fimbrin

We have crystallized the N-terminal actin binding domain (ABD1) of human fimbrin, a representative member of the largest class of actin crosslinking proteins. Diffraction from these crystals is consistent with the orthorhombic space group P2₁2₁2₁ (a = 50.03 Å, b = 61.24 Å, c = 102.30 Å). These crystals contain one molecule in the asymmetric unit and diffract to at least 1.9 Å resolution. The crystal structure of ABD1 will be the first structure of an actin crosslinking domain. Proteins 28:452–453, 1997. © 1997 Wiley-Liss, Inc.     

Crystal Structure Prediction based on Statistical Potentials

Organic molecule crystals are becoming more and more important in applications like piezoelectricity, ferroelectricity and pigments. These properties depend on the molecule and on the crystal structure. For this reason much effort is being made to predict the crystal structure of organic molecules. We have developed a new algorithm differing mainly in three features from other approaches (simulated annealing, Monte Carlo etc.). First, we analyze just one molecule for proper symmetry operations building up the crystal; second, the program works in a discrete space; and finally the scoring function (energy function) is derived statistically from known crystal structures and tabulated. Our program computes a list of crystal structures weighted according to our scoring function. The new algorithm FlexCryst is currently implemented for the four space groups P1, Pbar1, P2₁, and P2₁2₁2₁. The three latter space groups are widespread in nature. The algorithm computes structural models of acceptable quality and shows excellent time performance. During our validation we found the experimental structure among the structures proposed by the algorithm in 123 of 129 cases for P1, in 66 of 95 cases for Pbar1, in 73 of 100 cases for P2₁, and in 94 of 98 cases for P2₁2₁2₁. The performance depends on the space group. In the case of P1 the run time per molecule is about two minutes and increases up to roughly one hour for the space group P2₁.     

Additional criterion for the determination of the handedness of 21 helices in crystals of bile acids: Crystal structure of a tert‐butylphenyl derivative of cholic acid

[3β,5β,7α,12α]‐3‐(4‐t‐Butylbenzoilamine)‐7,12‐dihydroxycholan‐24‐oic acid was synthesized and recrystallized from chlorobenzene and acetone. Orthorhombic P2₁2₁2₁ and monoclinic P2₁ crystals were obtained, respectively, and both crystals include solvent and water molecules with a 1:1:1 stoichiometry. In the second case, there are two nonequivalent molecules of the steroid in the unit cell. In both crystals, a crossing structure results for the molecular packing, stabilized by hydrogen bonds between the steroid molecules. In the second crystal, water links steroid molecules of the same type, which are not directly connected through hydrogen bonds. The steroid molecules define helicoidal assemblies along 2₁ screw axes. The handedness in the crystal in chlorobenzene is unambiguously defined by following the criterion proposed by Miyata et al., as the steroids are in a belly‐to‐belly disposition around the axis. This motif does not appear in the crystal in acetone, and other criterion is required. On the basis of the fact that a staircase and its banister have the same handedness and using the crystal in chlorobenzene as reference, the additional criterion has been established. According to it, in the absence of a belly‐to‐belly orientation, the handedness must be defined by keeping the bile acid with its right side oriented toward the axis. Pitch angles of the helices and tilt angles of the molecules are also determined. Chirality, 2011. © 2011 Wiley‐Liss, Inc.     

2‐Thiopheneformamidinium‐Based 2D Ruddlesden–Popper Perovskite Solar Cells with Efficiency of 16.72% and Negligible Hysteresis

Formamidinium (FA)‐based 3D perovskite solar cells (PSCs) have been widely studied and they show reduced bandgap, enhanced stability, and improved efficiency compared to MAPbI₃‐based devices. Nevertheless, the FA‐based spacers have rarely been studied for 2D Ruddlesden–Popper (RP) perovskites, which have drawn wide attention due to their enormous potential for fabricating efficient and stable photovoltaic devices. Here, for the first time, FA‐based derivative, 2‐thiopheneformamidinium (ThFA), is successfully synthesized and employed as an organic spacer for 2D RP PSCs. A precursor organic salts‐assisted crystal growth technique is further developed to prepare high quality 2D (ThFA)₂(MA)_n?1Pb_nI_3n+1 (nominal n = 3) perovskite films, which shows preferential vertical growth orientations, high charge carrier mobilities, and reduced trap density. As a result, the 2D RP PSCs with an inverted planar p‐i‐n structure exhibit a dramatically improved power conversion efficiency (PCE) from 7.23% to 16.72% with negligible hysteresis, which is among the highest PCE in 2D RP PSCs with low nominal n‐value of 3. Importantly, the optimized 2D PSCs exhibit a dramatically improved stability with less than 1% degradation after storage in N₂ for 3000 h without encapsulation. These findings provide an effective strategy for developing FA‐based organic spacers toward highly efficient and stable 2D PSCs.     

Characterization of crystals of a cytochrome oxidase (nitrite reductase) from Pseudomonas aeruginosa by x-ray diffraction and electron microscopy

Cytochrome oxidase from Pseudomonas aeruginosa has been crystallized from 2 m-ammonium sulfate. The crystals occur principally as thin diamond-shaped plates of space group P2₁2₁2 with unit cell dimensions of 92 Å × 115 Å × 76 Å. Determination of the density of glutaraldehyde-fixed, water-equilibrated crystals (1.167 g/cm³), coupled with the unit cell volume (804,000 ų), indicates that there is one subunit (~63,000 Mr) per asymmetric unit. X-ray diffraction data which were limited to 12 Å resolution due to small crystal size were obtained for the hk0 and 0kl zones using precession photography. Amplitude and phase data for the hk0, 0kl, and h0l zones were obtained from computer-based Fourier analysis of appropriate micrographs recorded from negatively stained microplates and thin sections of larger crystals using minimal beam electron microscopy. For crystals embedded in the presence of tannic acid it was possible to achieve 20 Å resolution which is comparable to the resolution achieved with negative staining of thin crystalline arrays. In addition, unstained electron diffraction on glutaraldehyde-fixed, glucose-stabilized plates was recorded to a resolution of 9 Å. The three-dimensional packing of the cytochrome oxidase dimer in the unit cell has been deduced from computer reconstructed images of the three principal projections along the crystallographic axes. The cytochrome oxidase dimer is located in the unit cell with the dimer axis coincident with a crystallographic 2-fold axis; thus within the resolution of the present data in projection (9 Å) the two subunits are identical, in agreement with biochemical evidence. The crystals have been prepared with the enzyme in the fully oxidized state and upon reduction a progressive cracking of the crystals is observed, possibly due to a conformational change dependent on the oxidation state of the heme iron.     

The Phospholamban Pentamer Alters Function of the Sarcoplasmic Reticulum Calcium Pump SERCA

The interaction of phospholamban (PLN) with the sarco-endoplasmic reticulum Ca²⁺-ATPase (SERCA) pump is a major regulatory axis in cardiac muscle contractility. The prevailing model involves reversible inhibition of SERCA by monomeric PLN and storage of PLN as an inactive pentamer. However, this paradigm has been challenged by studies demonstrating that PLN remains associated with SERCA and that the PLN pentamer is required for the regulation of cardiac contractility. We have previously used two-dimensional (2D) crystallization and electron microscopy to study the interaction between SERCA and PLN. To further understand this interaction, we compared small helical crystals and large 2D crystals of SERCA in the absence and presence of PLN. In both crystal forms, SERCA molecules are organized into identical antiparallel dimer ribbons. The dimer ribbons pack together with distinct crystal contacts in the helical versus large 2D crystals, which allow PLN differential access to potential sites of interaction with SERCA. Nonetheless, we show that a PLN oligomer interacts with SERCA in a similar manner in both crystal forms. In the 2D crystals, a PLN pentamer interacts with transmembrane segments M3 of SERCA and participates in a crystal contact that bridges neighboring SERCA dimer ribbons. In the helical crystals, an oligomeric form of PLN also interacts with M3 of SERCA, though the PLN oligomer straddles a SERCA-SERCA crystal contact. We conclude that the pentameric form of PLN interacts with M3 of SERCA and that it plays a distinct structural and functional role in SERCA regulation. The interaction of the pentamer places the cytoplasmic domains of PLN at the membrane surface proximal to the calcium entry funnel of SERCA. This interaction may cause localized perturbation of the membrane bilayer as a mechanism for increasing the turnover rate of SERCA.