Protein III (PIII) was originally described by McDade and Johnston (14). In their studies of the gonococcal outer membrane, they observed that all gonococcal strains contained a similar protein whose apparent molecular weight in sodium dodecyl sulfate-polyacrylamide gel electrophoresis increased upon treatment with a reducing agent. Furthermore, they and others (14, 17) provided evidence, using bifunctional cross-linking reagents, that PIII was closely associated with the protein I (PI) porin within the gonococcal outer membrane. It was further shown that PIII was structurally and immunologically conserved among several strains of gonococci (11,12). In studies in which several proteolytic enzymes were used to investigate the surface orientation of gonococcal outer membrane proteins, PIII was found to be resistant to protease treatment (2). However, Shafer and Morse, using lysosomal cathepsin G, showed that PIII in gonococcal outer membranes could be cleaved by this enzyme (20). In addition, the rate at which PIII was degraded by cathepsin G could be correlated to the lipooligosaccharide expressed by a particular strain of gonococcus; i.e., the PIII associated with low-molecular-weight lipooligosaccharide was cleaved more rapidly than was the PIII associated with a higher-molecular-weight lipooligosaccharide. Surface-labeling experiments and the ability of PIII to react in vivo with a monoclonal antibody also indicated that PIII had surface-exposed domains (22). Our interest was piqued by several unanswered questions. We wanted to know (i) how this highly conserved, surface-exposed protein could exist in an organism that expended a great deal of its genome and energy in antigenic variation; (ii) how, if at all, PIII affected the functions of PI; and (iii) how PIII could be eliminated as a contaminant in preparations of PI.
PURIFICATIONTo study PIII more directly, we developed a method for purifying it (13) and began to examine its biochemical and immunochemical characteristics. Early on, we observed that to purify PI, it was essential to separate PI from PIII in the initial extraction step. Otherwise, PIII remained tightly associated with PI throughout the subsequent steps of purification. We found that if the pH of the extraction buffer was altered, different proteins would be solubilized. If the extraction was performed at pH s 4, PI and PII were quantitatively released, but PIII remained with the cell debris (1). PIII was solubilized in rather pure form by reextraction of the cell debris with a pH 10.5 buffer. It was completely released with minimal PI contamination. The PIII was then purified by cation-exchange chromatography on CM-Sepharose, followed by gel filtration on Sephacryl-200 (13). The resulting product retained all the biochemical and immunochemical characteristics of the native PIII, except that, unlike the native molecule, purified PIII was highly susceptible to * Corresponding author.proteases. The amino acid composition and amino-terminal sequences of Pill isolated from several strains of Neis...