Fractionation of Pneumocystis carinii antigens used in an enzyme-linked immunosorbent assay for antibodies and in the production of antiserum for detecting Pneumocystis carinii antigenemia.

Cyst-rich suspensions of Pneumocystis carinii were obtained by differential and gradient centrifugation from heavily infected rat lungs. After preparation of an aqueous-soluble extract of the cyst-rich material, the insoluble residue was extracted with 8 M urea. Small amounts of infected human lung tissue and uninfected rat and human lung were processed similarly. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis showed that both human and rat infected lung extracts contained a large protein (greater than 200,000 daltons). This component was not present in extracts of uninfected lung. In addition, an HCl-soluble extract was prepared from the cyst-rich suspension from infected rat lung. The urea-extracted antigen was most reactive in an enzyme-linked immunosorbent assay. Rabbit antiserum against the HCl-soluble antigen detected circulating antigen in patients' sera in a counterimmunoelectrophoresis assay.

Detection of specific antibody by enzyme-linked immunosorbent assay and antigenemia by counterimmunoelectrophoresis in humans infected with Pneumocystis carinii.

A urea-soluble extract of cyst-rich material from rat lung heavily infected with Pneumocystis carinii was evaluated in an enzyme-linked immunosorption assay for antibody in 461 human sera. The highest level of reactivity occurred in sera submitted for serodiagnosis from proved or highly suspect cases. However, the range of reactivities in these groups, many of whom were on immunosuppressive therapy, was very wide. A more restricted lower range of reactivity was observed in both hospital-family contacts and healthy Serum Bank donors. Because of the overlap in levels of reactivity between the pneumocystosis and control groups, no concise cutoff value to separate infected from noninfected individuals could be made. Specificity of the reactions was shown by absorption of patients' and control sera with uninfected and P. carinii-infected human and rat lung tissue. The data support the concept that P. carinii is highly prevalent as a latent agent in the general population and is provoked to cause clinically manifest disease in the compromised host. Detection of circulating antigen appeared to be specific and possibly a useful adjunct to diagnosis, as 10 of the 14 proved or highly suspect patients with antigenemia did not have measurable antibody to P. carinii.