Purpose: To determine the direct impact that cyp51A mutations have on triazole resistance among a collection of pan-triazole resistant A. fumigatus clinical isolates from the United States, and utilize whole genome sequencing and transcriptome profiling by RNA-seq to identify other potential effectors of clinical triazole resistance.
Methods: Ten triazole resistant clinical isolates of A. fumigatus from the United States were obtained. Triazole minimum inhibitory concentrations (MICs) were performed for voriconazole, isavuconazole, itraconazole and posaconazole. Whole genome sequencing was preformed using the Ion Torrent platform. Clustered regularly interspaced palindromic repeat (CRISPR) associated endonuclease mediated transformations were used to correct any cyp51A mutations identified, including promoter region alterations. Triazole minimum inhibitory concentrations for each resulting strain were compared to its clinical parent isolate. Transcriptional profiling was performed using RNA-seq for all ten resistant clinical isolates and compared to a composite of five triazole susceptible control isolates.
Results: All ten isolates were resistant to voriconazole, isavuconazole, itraconazole and posaconazole. Whole genome sequencing revealed seven isolates to have cyp51A mutations not found among triazole susceptible isolates (including TR34/L98H, TR46/Y121F/T289A, G138C, M263I, I367F, and G448S). CRISPR mediated correction of any of the identified cyp51A mutations did not restore triazole susceptibility. One isolate without any cyp51A mutation did possess a previously characterized mutation in hapE (P88L). No mutations unique to triazole resistant isolates were found in cyp51B, or the transcriptional regulator srbA. However, numerous mutations in other ergosterol biosynthetic genes, including erg3, erg4, and erg5, were identified. RNAseq transcriptome profiling revealed that two of the three isolates with tandem repeat associated cyp51A mutations overexpressed cyp51A by 3 to 7- fold, relative to the susceptible composite. One of these isolates also overexpressed multiple ergosterol biosynthesis associated genes including cyp51B and srbA by >3-fold. Additionally, the isolate with a P88L hapE mutation overexpressed cyp51A by 3-fold. Of the ten resistant clinical isolates, eight were observed to overexpress abcC, and all but one isolate overexpressed at least one efflux pump gene previously associated with triazole resistance in A. fumigatus (abcC, abcA, atrF, atrI, mdrA, and mdr1). Most notably, one isolate possessing no cyp51A mutation was observed to overexpress five of these six genes by 4 to 36-fold.
Conclusion: In this collection of ten pan-triazole resistant clinical isolates of A. fumigatus, three of the ten isolates possessed no cyp51A mutation, and CRISPR mediated correction of any of the identified cyp51A mutations in the remaining seven isolates did not restore triazole susceptibility. This demonstrates that while cyp51A mutations may contribute to triazole resistance, the resistance observed in this collection of isolates is not dependent on cyp51A mutations. While no mutations unique to resistant isolates were identified in cyp51B or srbA, mutations in other ergosterol biosynthesis associated genes such as hapE, erg3, erg4, or erg5, potentially contribute to the observed triazole resistance. Additionally, overexpression of genes associated with ergosterol biosynthesis (such as cyp51A, cyp51B, srbA), and efflux (such as abcC, abcA, atrF, atrI, mdrA, and mdr1) was observed across this collection and may also contribute to triazole resistance.
Full conference title:
- AAA 8th (2018)