Physicochemical and biochemical characterization of human alpha 1-microglobulin expressed in baculovirus-infected insect cells.

DNA encoding the signal peptide and the alpha 1-microglobulin part of the human alpha 1-microglobulin-bikunin gene was expressed in baculovirus-infected insect cells. Recombinant alpha 1-microglobulin was secreted and could be purified from the medium with a yield of 20-30 mg/ L. Biochemical and physicochemical characterization showed that the recombinant protein was very similar to alpha 1-microglobulin isolated from human urine and plasma, except that the recombinant protein had smaller N-linked oligosaccharides, lacked the O-linked oligosaccharide, and was devoid of sialic acid. Recombinant alpha 1-microglobulin migrated upon SDS-PAGE as two bands, 27 and 29 kDa, representing alpha 1-microglobulin with one and two N-linked carbohydrates, respectively. An overall structural similarity was indicated as antibodies raised against human urinary alpha 1-microglobulin were found to recognize recombinant, plasma, and urinary alpha 1-microglobulin in a similar manner. CD studies suggested an almost identical secondary structure for recombinant and urinary alpha 1-microglobulin but a slightly different structure for plasma alpha 1-microglobulin. The absorbance spectrum as well as visual examination demonstrated that recombinant, urinary, and plasma alpha 1-microglobulin carried a yellow-brown chromophore, but that plasma alpha 1-microglobulin was slightly less intensely colored. Although it is still a puzzle why the immunosuppressive plasma protein alpha 1-microglobulin and the protease inhibitor bikunin, which have no known function in common, are cotranslated from the same mRNA, it can be concluded that bikunin is not necessary for an adequate translation, folding, and secretion of alpha 1-microglobulin. Furthermore, since recombinant alpha 1-microglobulin was produced in large amounts and found to be very similar to plasma and urinary alpha 1-microglobulin, it may prove to be useful in structural and functional studies of the protein.

Coiled-coil structure of group A streptococcal M proteins. Different temperature stability of class A and C proteins by hydrophobic-nonhydrophobic amino acid substitutions at heptad positions a and d.

M proteins and M-like proteins, expressed on the surface of group A streptococci and binding to human plasma proteins, can be divided into two classes, A and C, depending on the structure of the central repeated regions. The class C proteins have been shown to be dimers with a coiled-coil structure. In this work, we have compared the structure and binding of a class A protein, Mrp4, and a class C protein, Arp4, expressed by the same bacterial strain. Circular dichroism spectra, gel filtration, and binding assays showed that both proteins had a coiled-coil dimer configuration and a high-affinity binding at 20 degrees C. However, striking differences were seen at 37 degrees C. The class A protein, Mrp4, was still a coiled-coil dimer with high affinity binding activity, whereas the class C protein, Arp4, had lost both the coiled-coil structure and binding activity. Raising the temperature even higher, Mrp4 retained the coiled-coil structure up to 70-90 degrees C. Furthermore, a recombinant protein, Mrp(C), in which the A-repeats of Mrp4 were replaced by the C-repeats of Arp4, lost its coiled-coil structure and fibrinogen-binding around 40-45 degrees C. These results suggest a high thermal stability of class A proteins and a low stability of class C proteins and that the structural basis for this can be found partly in the A- and C-repeats. Analysis of the amino acid sequences of the A- and C-repeats, revealed a large difference, 87% and 45%, respectively, in the content of hydrophobic amino acid residues in the positions regarded as important for the formation of the coiled-coil structure. In particular, several alanine residues in the A-repeats were replaced by serine residues in the C-repeats. Our results suggest that important structural and functional changes within the M protein family have evolved by specific hydrophobic-nonhydrophobic amino acid replacements.

Solution structure of the albumin-binding GA module: a versatile bacterial protein domain.

The albumin-binding GA module is found in a family of surface proteins of different bacterial species. It comprises 45 amino acid residues and represents the first known example of contemporary module shuffling. Using 1H NMR spectroscopy we have determined the solution structure of the GA module from protein PAB, a protein of the anaerobic human commensal and pathogen Peptostreptococcus magnus. This structure, the first three-dimensional structure of an albumin-binding protein domain described, was shown to be composed of a left-handed three-helix-bundle. Sequence differences between GA modules with different affinities for albumin indicated that a conserved region in the C-terminal part of the second helix and the flexible sequence between helices 2 and 3 could contribute to the albumin-binding activity. The effect on backbone amide proton exchange rates upon binding to albumin support this assumption. The GA module has a fold that is strikingly similar to the immunoglobulin-binding domains of staphylococcal protein A but it shows no resemblance to the fold shared by the immunoglobulin-binding domains of streptococcal protein G and peptostreptococcal protein L. When the gene sequences, binding properties and thermal stability of these four domains are analysed in relation to their global folds an evolutionary pattern emerges. Thus, in the evolution of novel binding properties mutations are allowed only as long as the energetically favourable global fold is maintained.

The GA module, a mobile albumin-binding bacterial domain, adopts a three-helix-bundle structure.

We present the first study of the secondary structure and global fold of an albumin-binding domain. Our data show that the GA module from protein PAB, an albumin-binding protein from the anaerobic bacterial species Peptostreptococcus magnus, is composed of a left-handed three-helix bundle. The helical regions were identified by sequential and medium range NOEs, values of NH-C alpha H coupling constants, chemical shift indices, and the presence of slowly exchanging amide protons, as determined by NMR spectroscopy. In addition, circular dichroism studies show that the module is remarkably stable with respect to both pH and temperature.