After tryptic digestion of intact Staphylococcus aureus the residual portion of protein A that was still bound to the cell wall was cleaved off with lysostaphin. From the two digests Fc‐binding fragments were isolated and the following characteristics observed. (a) There are four Fc‐binding, highly homologous regions, each consisting of 58–62 amino acid residues. (b) These regions are consecutively arranged from the N‐terminal part of the protein. (c) The residual C‐terminal part, approximately 150‐residues long, differs to a great extent with respect to primary and secondary structures from the four active regions. Furthermore, it is suggested that the protein is bound to the bacterial cell wall structure through this C‐terminal part.

The sequence of region X of staphylococcal protein A has been determined. The hypothesis has been put forward that this region spans the Staphylococcus aureus cell wall and is responsible for the binding to the peptidoglycan. The primary amino acid sequence of region X was determined for two strains exhibiting cell-wall-bound protein A, Cowan I and 8325 -4. The sequence determination of the Cowan I material is partial and was performed by Edman degradation, in contrast to the sequence of the 8325 -4 material which was completely analyzed by nucleotide sequencing of the corresponding gene. The region consists of two structurally different domains, a highly repetitive region (X,), with an octapeptide structure repeated approximately 12 times, and a C-terminal domain (X,) with an unique sequence. A comparison between the two strains reveals a high mutual homology as well as a high internal homology between the octapeptide structures. Six out of eight amino acids are identical in the repetition of this structure throughout region X, in both proteins and the other two are changed in a rather regular pattern.

No abstract

The covalent structures of the four highly homologous Fc-binding regions in protein A, regions D, A, B, and C, have been studied by enzymic fragmentations of previously isolated fragments originating from these regions and subsequent isolation of the generated peptides by ion-exchange chromatography, molecular-sieve chromatography, high-voltage paper electrophoresis and paper chromatography. The complete sequence of region B was elucidated by combining the results of Edman degradations on isolated fragment B peptides with the previously reported N-terminal sequence of the same fragment. Furthermore, Edman degradations of fragment D, A and C peptides differing from the region B sequence provided the structures of subregions not identical to corresponding subregions within region B. Thus, it is possible to propose a highly probable covalent structure for the N-terminal 27 000-molecular-weight portion of protein A responsible for the IgG-Fc-binding activities. However, it was not possible to assign the activities to specific structures within the regions.The sequence data indicate that not only mutual homology between the four regions exists, but also internal homologies within the regions. Furthermore, the data strongly supports the hypothesis of a stepwise gene fusion procedure being involved in the evolution of the protein.

To study the active site(s) in protein A, partial tryptic digestions of the protein and of intact Staphylococcus aureus were performed. Fragments which bind to the Fc‐part of human IgG were isolated by affinity chromatography on IgG‐Sepharose 4B and purified by ion‐exchange chromatography on phosphocellulose. From a partial tryptic digest of pure protein A at 30°C, pH 8.2 for 30 min we have isolated and characterized six active fragments with molecular weights ranging from 6000 to 8000. Two active fragments, obtained in high yields by digestion at pH 7.2 of intact protein‐A‐containing bacteria, were shown to be similar to two of the six characterized fragments from the digest of pure protein A. All fragments appeared to have similar amino acid sequences, judged by peptide mapping, specific staining and amino acid analysis; some are very possibly overlapping peptides. Each fragment probably contains only one active site region since all are monovalent in the Fc‐reaction when studied with a hemagglutination technique. The maximal molar yield of active fragments obtained from the digestion of pure protein A accounts for about 210% of the amount of protein A used. Thus protein A, suggested to consist of repeating units, should exhibit at least three similar if not identical active regions.