The molecular epidemiology of pneumocystis jirovecii inCape Town, South Africa

[摘要] ENGLISH ABSTRACT : Pneumocystis jirovecii is an opportunistic fungal pathogen that causes Pneumocystispneumonia (PCP) in immunocompromized hosts. PCP is associated with substantialmorbidity, and mortality rates range from 10% to 40%.The diagnosis of PCP relies on the microscopic detection of P. jirovecii in stained clinicalsamples. Polymerase chain reaction (PCR) may provide better sensitivity than microscopy;therefore, evaluation and implementation of PCR assays are required for the detection ofPneumocystis infection.P. jirovecii is not cultivatable, therefore molecular tools are used for characterizing P.jirovecii genotypes; common targets are the dihydropteroate synthase (DHPS) andmitochondrial large subunit rRNA (mtLSU rRNA) genes. DHPS is a therapeutic target;mutations may be associated with co-trimoxazole prophylaxis and treatment failure.Polymorphisms in mtLSUrRNA have been used for phylogenetic studies.Aims: 1) to evaluate a real time PCR (rtPCR) assay for diagnosis of PCP by comparing theperformance to immunofluorescence (IF) and 2) to describe the molecular epidemiology ofP. jirovecii isolates from Tygerberg Hospital by analyzing DHPS and mtLSU rRNA genes.Methods: Clinical samples from 305 children and adult patients at Tygerberg Hospital werecollected, after testing using IF. DNA was extracted using the NucliSens easyMAG platform(Biomérieux). The rtPCR assay targeting the major surface glycoprotein (MSG) gene wasevaluated to detect P. jirovecii DNA. The DHPS and mtLSU rRNA genes were amplified bynested PCR and analyzed by DNA sequencing.Results: The SYBR Green rtPCR detected P.jirovecii in 57% of samples (175/305)compared to the 7% (21/305) detected by IF. Our rtPCR had a sensitivity of 100% andspecificity of 46%, although this increased if the detection threshold increased. Of the 50negative control samples used in this study, none tested positive for P.jirovecii.There were 237 lower respiratory tract (LRT) and 58 upper respiratory tract (URT) samples.The yield of PCR in LRT samples was 55.3% (131/237) compared to 70.6% (41/58) in URTsamples (p=0.03). In contrast, none of the URT samples were positive using IF, and 8.9%(21/237) of LRT samples were positive on IF.DHPS was successfully amplified in 123 (70.3%) samples; and mtLSU in 126 (72%)samples. Genotype 1 (wild type) was the predominant DHPS genotype, and a mutation rate of42.3% was recorded for this gene. The mtLSU genotype 3 was present in 50.8% of samples,genotype 1 (42%) was the next most common genotype. Mixed genotypes were detected in2.4% of the samples analyzed for each gene. There was no clear association between DHPSpolymorphisms and mtLSU genotype.Conclusions: The SYBR Green rtPCR was more sensitive than IF for detection of P.jirovecii; especially in URT samples, which is comparable to previous studies. The DHPSmutation rate increased to 42% from 27% recorded in 2013 from our division. The increase inDHPS mutation rate may be a result of on-going co-trimoxazole use, for prophylaxis ortreatment of PCP or other infections.Our findings need to be linked to clinical data to better understand transmission dynamics andpotential impact of strain variation on clinical outcome, and further studies are required tobetter describe the local strain diversity.

[发布机构] Stellenbosch University