To try to develop highly sensitive, non-invasive diagnostic methods, we have been evaluating polymerase chain reaction (PCR) using primers based on the major surface glycoprotein (MSG) genes of human Pneumocystis. PCR using primers based on this gene is potentially a highly sensitive method since this is a multicopy gene (estimated at approximately equal to 50-100 copies/genome). We have been evaluating the diagnostic potential using a conserved region of the gene family. Our studies have shown that the sensitivity of MSG-based primers is greater than that of previously utilized primers. As part of our collaboration with investigators at UCSF and Makerere University, we have been able to demonstrate that Pneumocystis is uncommon in HIV-infected patients in Uganda, which may be in part a result of the widespread use of trimethoprim-sulfamethoxazole prophylaxis in Uganda. We are currently conducting a study in collaboration with the Microbiology Department and investigators at UCSF and Makerere University to compare matched BAL/induced sputum samples and oral wash samples from San Francisco and Uganda to see if the sensitivity of the two sample types is similar when using PCR. We have developed and evaluated two typing techniques for human Pneumocystis. The first uses tandem repeats that occur in an intron of the MSG gene. By sequencing multiple MSG genes in a number of human Pneumocystis isolates, we have been able to demonstrate that recombination occurs in human Pneumocystis. These results led to studies that demonstrated that the MSG repertoire in human Pneumocystis is very diverse, while the MSG repertoire in rat and mouse Pneumocystis is identical or very limited among different isolates. Based on these studies, we have developed a restriction fragment length polymorphism (RFLP) typing assay for human Pneumocystis, and have been able to demonstrate substantial diversity among human isolates. Using RFLP analysis, we have been able to demonstrate that all the isolates from an outbreak of PCP in renal transplant patients in Germany appear to be the same strain, indicating that recent infection is important, and that either host- to-host transmission has occurred or that all individuals ere infected from a common source. We subsequently showed that the same isolate was responsible for an outbreak of PCP in renal transplant patients in Zurich, Switzerland, raising the possibility that this strain may be particularly virulent in renal transplant patients. To address this further, we examined samples from an outbreak of PCP in Japan. We found that while a single organism was responsible for the outbreak there as well, it was different from the European strain based on RFLP analysis. We have subsequently analyzed samples from an outbreak in Denmark and were able to show that 3 unique strains were responsible for this outbreak. Each strain was identified over a limited period of time, with 2 of the 3 strains overlapping in time. We are currently analyzing samples from Brest, France, and Bern, Switzerland that are again from outbreaks in renal transplant recipients. Preliminary results suggest that one strain is common to the outbreaks in Brest and Bern, though not all cases were caused by this strain. In addition to RFLP, we are using multi-locus sequence typing (MLST). MLST is the most commonly used method for typing Pneumocystis in other laboratories, and this will allow us to compare the results of these two typing methods. We are also exploring more sequencing larger regions of the Pneumocystis genome to be able to better examine the phylogenetic relationships among different isolates. These studies should provide improved diagnostic methods for PCP, and help to better understand the epidemiology of Pneumocystis infection. In collaboration with investigators at the Broad Institute and Leidos Biomedical Research, Inc., we have undertaken a project to sequence the genome of Pneumocystis species from different hosts. The major difficulty with this project is obtaining DNA of sufficient purity to allow next generation sequencing, sincethe organism cannot be cultured. As a first step, we were able to completely sequence the mitochondrial genomes of P. carinii, P. murina, P. jirovecii, and to show that while there is a high level of synteny between the P. carinii and P. murina genomes, P. jirovecii has substantial levels of recombination compared to the other 2 species. We have subsequently been able to purify Pneumocystis DNA from infected rat, mouse, and human lungs sufficiently to allow NGS sequencing to generate high quality sequence data that has permitted development of nearly complete draft genomes of all 3 Pneumocystis species (P. carinii, P. murina, P. jirovecii). RNA sequencing in parallel has allowed identification of P. murina and P. carinii transcripts and characterization of the genes encoded by this organism. We are in the process of analyzing the genome data, comparing the 3 organisms to each other as well as to other related fungal species. Availability of these genome should allow us to better understand the biology of this family of organisms, should potentially allow identification of metabolic pathways that need to be complemented to successfully culture the organism, and should facilitate the identification of important antigens and pathogenic factors.
Showing the most recent 10 out of 13 publications
