The 2.4 A resolution crystal structure of boar seminal plasma PSP-I/PSP-II: a zona pellucida-binding glycoprotein heterodimer of the spermadhesin family built by a CUB domain architecture.

The crystal structure of porcine seminal plasma spermadhesin PSP-I/PSP-II heterodimer has been determined in two crystal forms by multiple isomorphous replacement in an hexagonal crystal (space group P6(1)22) and molecular replacement in a trigonal crystal of space group P3(2)21. The crystal structure has been refined at 2.4 A resolution to an R-factor of 20.0% (Rfree = 25.9%) for 14,809 independent reflections with intensities greater than 2 sigma (I), with root-mean-square deviations of 0.009 A and 1.657 degrees from ideal bond lengths and bond angles, respectively. The final model includes 1688 non-hydrogen protein atoms of 221 amino acids and 79 water molecules. PSP-I/PSP-II represents the first crystal structure of a mammalian zona pellucida-binding protein. PSP-II displays a putative carbohydrate-recognition site located around its Asn50. This region shares structural features with sugar binding sites of known lectin structures of the leguminous and galectin families. PSP-I and PSP-II are N-glycosylated at asparagine residues 50 and 98, respectively, and show site heterogeneity. Only the innermost N-acetylglucosamine of PSP-I is defined in the crystal structure. Both subunits of the PSP-I/PSP-II heterodimer are built by a single CUB domain architecture. The CUB domain displays a novel fold, which consists of a compact ellipsoidal beta-sandwich structure (42 A x 27 A x 23 A) organized into two 5-stranded beta-sheets. Each sheet contains parallel and antiparallel beta-strands. Two disulphide bridges, which are conserved in all spermadhesin molecules and many CUB domains, crosslink loop LA and strand beta 4 and loops LE and LG, respectively, at opposite edges of the same face of the domain. The four highly conserved aromatic residues and 15 out of 17 invariant hydrophobic residues, which define the CUB domain signature, display an interior location, suggesting that this hydrophobic core may be essential for maintaining the overall folding of the domain. Most of the hydrophobic core residue characteristics are conserved in the jellyroll topology of certain icosahedral virus capsid proteins, indicating that the CUB domain and the viral proteins share a minimal structural core.

The 2.15 A crystal structure of CG-16, the developmentally regulated homodimeric chicken galectin.

Differential developmental regulation of expression, fine-specificity differences in ligand recognition and disparate capacity for homodimerization are characteristics of the two currently known proto-type chicken galectins. The X-ray crystal structure of the first avian galectin, the homodimeric agglutinin from chicken liver (CG-16), has been solved in the absence of ligand in two crystal forms. Although the arrangement of lectin dimers in the two crystals is different, the structure of the monomers and their association into the extended beta-sandwich that characterises the dimer are virtually identical. The fold establishes a beta-sandwich motif composed of a five-stranded and a six-stranded beta-sheet evocative of proto-type mammalian galectins. The carbohydrate-binding site is occupied by six water molecules that take the place of the sugar in the complex. They help to stabilise in the absence of the ligand the spatial arrangement of the amino acid side-chains involved in sugar recognition. Docking of N-acetyllactosamine into the binding site reveals that three of these water molecules, which are in direct contact with the protein, occupy positions equivalent to the key sugar hydroxyl groups, namely the hydroxyls at positions 4 and 6 of the galactose unit and at position 3 of the N-acetylglucosamine unit. Crystallographic data are fully consistent with the binding features in solution previously derived from chemical mapping with deoxy, fluoro and O-methyl derivatives and laser photo-CIDNP (chemically induced dynamic nuclear polarisation) studies. The possible molecular basis for the monomeric character of the chicken intestinal galectin as well as potential mechanisms of oxidative inactivation by disulphide bridging are evaluated on the basis of the given structural information concerning the CG-16 dimer interface and the cysteine residues, respectively.

The 2.15 A crystal structure of a triple mutant plastocyanin from the cyanobacterium Synechocystis sp. PCC 6803.

The crystal structure of the triple mutant A42D/D47P/A63L plastocyanin from the cyanobacterium Synechocystis sp. PCC 6803 has been determined by Patterson search methods using the known structure of the poplar protein. Crystals of the triple mutant A42D/D47P/A63L, which are stable for days in its oxidized form, were grown from ammonium sulfate, with the cell constants a = b = 34.3 A and c = 111.8 A belonging to space group P3(2)21. The structure was refined using restrained crystallographic refinement to an R-factor of 16.7% for 4070 independent reflections between 8.0 and 2.15 A with intensities greater than 2 sigma (I), with root mean square deviations of 0.013 A and 1.63 degrees from ideal bond lengths and bond angles, respectively. The final model comprises 727 non-hydrogen protein atoms within 98 residues, 75 water molecules and a single copper ion. The overall tertiary fold of Synechocystis plastocyanin consists of a compact ellipsoidal beta-sandwich structure made up of two beta-sheets embracing a hydrophobic core. Each sheet contains parallel and antiparallel beta-strands. In addition to the beta-sheets, the structure contains an alpha-helix from Pro47 to Lys54 that follows beta-strand 4. The three-dimensional structure of Synechocystis plastocyanin is thus similar to those reported for the copper protein isolated from eukaryotic organisms and, in particular, from the cyanobacterium Anabaena variabilis, the only cyanobacterial plastocyanin structure available so far. The molecule holds an hydrophobic region surrounding His87, as do other plastocyanins, but the lack of negatively charged residues at the putative distant remote site surrounding Tyr83 could explain why the Synechocystis protein exhibits a collisional reaction mechanism for electron transfer to photosystem I (PSI), which involves no formation of the transient plastocyanin-PSI complex kinetically observed in green algae and higher plants.

Crystal structure of acidic seminal fluid protein (aSFP) at 1.9 A resolution: a bovine polypeptide of the spermadhesin family.

We report the three-dimensional crystal structure of acidic seminal fluid protein (aSFP), a 12.9 kDa polypeptide of the spermadhesin family isolated from bovine seminal plasma, solved by the multiple isomorphous replacement method and refined with data to 1.9 A resolution with a final R-factor of 17.3%. aSFP is built by a single CUB domain architecture, a 100 to 110 amino-acid-residue extracellular module found in 16 functionally diverse proteins. The structure of aSFP reveals that the CUB domain displays a beta-sandwich topology organised into two 5-stranded beta-sheets, each of which contain two parallel and four antiparallel strands. The structure of aSFP is almost identical to that of porcine spermadhesins PSP-I and PSP-II, which in turn show limited structural similarity with jellyroll topologies of certain virus capsid proteins. Essentially, topologically conserved residues in these proteins are those internal amino acids forming the hydrophobic core of the CUB and the jellyroll domains, suggesting their importance in maintaining the integrity of these protein folds. On the other hand, the structure of aSFP shows structural features that are unique to this protein and which may provide a structural ground for understanding the distinct biological properties of different members of the spermadhesin protein family.

Crystallization and preliminary X-ray diffraction studies of aSFP, a bovine seminal plasma protein with a single CUB domain architecture.

Bovine acidic seminal fluid protein (aSFP) is a 1.29 kDa polypeptide of the spermadhesin family built by a single CUB domain architecture. The CUB domain is an extracellular module present in 16 functionally diverse proteins. To determine the three-dimensional structure of aSFP, the protein was crystallized at 21 degrees C by vapor diffusion in hanging drops, using ammonium sulfate, pH 4.7, and polyethyleneglycol 4,000 as precipitants, containing 10% dioxane to avoid the formation of clustered crystals. Elongated prismatic crystals with maximal size of 0.6 x 0.3 x 0.2 mm3 diffract to beyond 1.9 A resolution and belong to space group P2(1)2(1)2(1), with cell parameters a = 52.4 A, b = 41.5 A, c = 48.2 A. There is one aSFP molecule per asymmetric unit, which corresponds to a crystal volume per unit molecular mass of 2.04 A3/Da, and analytical ultracentrifugation analysis show that aSFP is a monomeric protein.

Crystallization and preliminary X-ray diffraction analysis of boar seminal plasma spermadhesin PSP-I/PSP-II, a heterodimer of two CUB domains.

Boar spermadhesin PSP-I/PSP-II (M(r) 29 000-30 000), a non-covalent heterodimer of two CUB domains, was crystallized in two crystal forms. Complete diffraction data sets for hexagonal (space group P6(1,5)22) and trigonal (space group P3(1,2)21) crystals have been collected up to 2.9 and 2.5 angstrom resolution, respectively. Cell constants of the hexagonal and trigonal crystal forms are a=b=87.2 angstrom, c=152.4 angstrom, and a=b=96.2 angstrom, c=70.8 angstrom, respectively. The calculated packing parameters (Vm) are 2.8 and 3.2 angstrom(3)/DA for the hexagonal and trigonal crystal forms, respectively, indicating that, in both cases, the asymmetric unit is constituted by one PSP-I/PSP-II heterodimer. This paper reports the first crystals of a protein built up by a CUB domain architecture.

Crystallization and preliminary X-ray diffraction analysis of bovine seminal plasma PDC-109, a protein composed of two fibronectin type II domains.

PDC-109 is a 13 kDa glycoprotein and the major phosphorylcholine- and heparin-binding protein of bull seminal plasma. It is built by an acidic 23-residue N-terminal sequence followed by a tandem of fibronectin type II domains. Full-length PDC-109 was crystallized in complex with o-phosphorylcholine by vapor diffusion in sitting drops. Crystals grew to maximal size of 0.5 x 0.3 x 0.2 mm3, diffract x-rays beyond 2.6 A resolution, and belong to space group P321 with unit cell dimensions a = b = 93.6 A, c = 52.7 A.

Spermadhesins: a new protein family. Facts, hypotheses and perspectives.

Spermadhesins are a novel family of secretory proteins expressed in the male genital tract of pig, horse and bull. They are major products of the seminal plasma and have been found to be peripherally associated to the sperm surface. The structure and function of spermadhesins have been thoroughly investigated in the pig, which exhibits the greatest diversity of members: AWN, AQN-1, AQN-2, PSP-I and PSP-II and its glycosylated isoforms. They are multifunctional proteins showing a range of ligand-binding abilities, e.g. carbohydrates, sulfated glycosaminoglycans, phospholipids and protease inhibitors, suggesting that they may be involved in different steps of fertilization. Isolated porcine spermadhesins bind the zona pellucida glycoproteins in a cation-dependent manner with a Kd in a low micromolar range, and AWN, AQN-1 and AQN-3 display similar binding affinity for glycoproteins containing Gal beta(1-3)-GalNAc and Gal beta(1-4)-GlcNAc sequences in O-linked and N-linked oligosaccharides, respectively. During sperm passage through the epididymis AQN-3 and AWN have been shown to bind tightly to the sperm surface by interaction with the phospholipids of the membrane bilayer. At ejaculation the spermadhesins form a protective coat around the sensitive acrosomal region of the sperm head, thus possibly preventing premature acrosome reaction. During in vitro capacitation most of these aggregated sperm adhesins are lost, with the exception of phospholipid-bound spermadhesins. AWN and AQN-3 may now serve as a primary receptor for the oocyte zona pellucida, thus contributing to initial binding and recognition between sperm and egg. The amino acid sequence of spermadhesins does not show any discernible similarity with known carbohydrate recognition domains (CRD). However, they belong to the superfamily of proteins with a CUB domain with a predicted all-beta structure. The crystal structure of the heterodimeric complex of the spermadhesins PSP-I/PSP-II has been solved, showing that the overall structure of both spermadhesins consists of a beta-sandwich with five (parallel and antiparallel) beta-strands. It is the first three-dimensional structure of a zona pellucida-binding protein and reveals the architecture of the CUB domain. The spermadhesins represent a novel class of lectins that may be involved in sequential steps of fertilization, at least in the pig.

A procedure for the large-scale isolation of major bovine seminal plasma proteins.

A method for the stepwise purification of bovine seminal plasma proteins aSFP, PDC-109, TIMP-2, RNAse, and BSP-30K/PDC-109 complex is presented. It is based on the use of nondenaturing, classical chromatographic procedures. The isolation protocol is reproducible, inexpensive, easy to carry out in a common laboratory, and yields large amounts of pure proteins for biochemical and structural studies.

The crystal structures of two spermadhesins reveal the CUB domain fold.

Spermadhesins, 12,000-14,000 M(r) mammalian proteins, include lectins involved in sperm-egg binding and display a single CUB domain architecture. We report the crystal structures of porcine seminal plasma PSP-I/PSP-II, a heterodimer of two glycosylated spermadhesins, and bovine aSFP at 2.4 A and 1.9 A resolution respectively.