CRISPR-mediated detection of Pneumocystis transcripts in bronchoalveolar, oropharyngeal, and serum specimens for Pneumocystis pneumonia diagnosis

Background. Pneumocystis jirovecii pneumonia (PCP) is a leading cause of fungal pneumonia, but its diagnosis primarily relies on invasive bronchoalveolar lavage (BAL) specimens that are difficult to obtain. Oropharyngeal swabs and serum could improve the PCP diagnostic workflow, and we hypothesized that CRISPR could enhance assay sensitivity to allow robust P. jirovecii diagnosis using swabs and serum. Herein we describe the development of an ultrasensitive RT-PCR-coupled CRISPR assay with high active-infection specificity in infant swabs and adult BAL and serum. Methods. Mouse analyses employed an RT-PCR CRISPR assay to analyze P. murina transcripts in wild-type and Rag2-/- mouse lung RNA, BAL, and serum at 2-, 4-, and 6-weeks post-infection. Human studies used an optimized RT-PCR CRISPR assay to detect P. jirovecii transcripts in infant oropharyngeal swab samples, adult serum, and adult BAL specimens from P. jirovecii-infected and P. jirovecii-non-infected patients. Results. The P. murina assays sensitively detected Pneumocystis RNA in the serum of infected mice throughout infection. Oropharyngeal swab CRISPR assay results identified infants infected with P. jirovecii with greater sensitivity (96.3% vs. 66.7%) and specificity (100% vs. 90.6%) than RT-qPCR compared to mtLSU standard marker, and CRISPR results achieved higher sensitivity than RT-qPCR results (93.3% vs. 26.7%) in adult serum specimens. Conclusion. Since swabs are routinely collected in pediatric pneumonia patients, serum is easier to obtain than BAL, and RT-PCR CRISPR results may not detect P. jirovecii colonization, this assay approach could improve pediatric Pneumocystis diagnosis by achieving specificity for active infection and avoiding the requirement for BAL specimens. Overall design: RNA was extracted with the Trizol method from the BAL cells pellets of two patients with Bare Lymphocyte Syndrome who had clinical PCP. Prior to construction of an Illumina total RNA library, DNase treated RNA was quantitated using a Qubit RNA BR assay kit (Thermo Fisher Scientific: Guide MAN0001987 MP10210, Kit #Q10210). Cytoplasmic, mitochondrial, and bacterial rRNA was removed from 2.5 ug of each sample as indicated by the Illumina RiboZero RNA Removal Kit Reference Guide [(Document # 15066012 v02, ScriptSeq Complete Gold (Epidemiology) Kit #BEP1206 (now obsolete)]. Illumina-compatible cDNA libraries were generated according to the instructions of the TruSeq Stranded Total RNA Sample Preparation Guide (Illumina Document #1000000040499v00, Kit #20020596). All libraries were pooled and denatured following the standard normalization method described by the Illumina Denature and Dilute Libraries Guide for the NextSeq System (Illumina Part #15048776), after which denatured libraries were loaded onto an Illumina NextSeq 550. To determine transcript abundance, FASTQ outputs were aligned to the Pneumocystis jirovecii RU7 genome using EdgeR normalization (66)