In-host microevolution of Aspergillus fumigatus: A phenotypic and genotypic analysis

Abstract

In order to survive, Aspergillus fumigatus must adapt to specific niche environments. Adaptation to the human host includes modifications facilitating persistent colonisation and the development of azole resistance. The aim of this study is to advance understanding of the genetic and physiological adaptation of A. fumigatus in patients during infection and treatment. Thirteen A. fumigatus strains were isolated from a single chronic granulomatous disease patient suffering from persistent and recurrent invasive aspergillosis over a period of 2 years. All strains had identical microsatellite genotypes and were considered isogenic. Whole genome comparisons identified 248 non-synonymous single nucleotide polymorphisms. These non-synonymous mutations have potential to play a role in in-host adaptation. The first 2 strains isolated were azole susceptible, whereas later isolates were itraconazole, voriconazole and/or posaconazole resistant. Growth assays in the presence and absence of various antifungal stressors highlighted minor changes in growth rate and stress resistance, with exception of one isolate showing a significant growth defect. Poor conidiation was observed in later isolates. In certain drug resistant isolates conidiation was restored in the presence of itraconazole. Differences in virulence were observed as demonstrated in a Galleria mellonella infection model. We conclude that the microevolution of A. fumigatus in this patient has driven the emergence of both Cyp51A-independent and Cyp51A-dependent, azole resistance mechanisms, and additional phenotypes that are likely to have promoted fungal persistence.

Keywords: Aspergillus fumigatus; Azole-resistance; Fungal growth; In-host microevolution; Whole genome sequencing.

Fig. 1

Observed colony morphology of the series. Sabouraud dextrose agar plates were spot inoculated with 5 × 10² conidia and incubated at 37 °C for 96 h.

Fig. 2

Phylogenetic tree based on whole genome sequences of the A. fumigatus series. (A) Single nucleotide polymorphism based phylogenetic tree was constructed using the SNPhylo pipeline and the whole genome sequences of the entire series as well as unrelated isolates IFM59361-1, 09-7500806, 08-19-02-61 and Afu 1042/09. (B) Unrooted phylogenetic tree of the series constructed using the SNPhylo pipeline. Tree scale represents nucleotide substitutions per site.

Fig. 3

Comparison of mycelial growth of A. fumigatus isolates V157-40 and V157-80 in the presence and absence of zinc at pH 4.5 and 7.5. Isolates were pre-cultured on glucose minimal media lacking zinc for 7 d at 37 °C. Conidia were harvested via immersion in 30 mL PBS containing 0.05% Tween-80 and counted. Glucose minimal media plates lacking zinc or containing 1 mM zinc at pH 4.5 and 7.5 were spot inoculated with 5 × 10² conidia. Every 24 h for 96 h colony diameter was measured; results at 96 h are shown. Data was obtained in triplicate and mean values ± SD are shown (^*p < 0.05; two-tailed Students T-test).

Fig. 4

Comparison of mycelial growth of selected A. fumigatus isolates on solid media with increasing concentrations of itraconazole. Sabouraud dextrose agar plates were spot inoculated with 5 × 10² conidia and incubated at 37 °C. Colony diameter was measured every 24 h for 96 h; results for 96 h are shown. Data was obtained in triplicate and mean values ± SD are shown (^*p = 0.003 compared to mean of V130-15, V157-39, V157-47 and V157-59; two-tailed Students T-test).

Fig. 5

Growth kinetics of selected A. fumigatus isolates in liquid media. Flat-bottomed 96-well plates were seeded with 1 × 10⁵ conidia in RPMI with or without voriconazole (A) or posaconazole (B) in various concentrations. Plates were incubated at 37 °C for 48 h inside a spectrophotometric plate reader; the optical density at 450 nm was automatically measured every 20 min with 5 s shaking before every reading. Optical density at 48 h is shown. Data was obtained in duplicate, mean values ± SD are shown. No significant differences were observed between isolates under the same condition.

Fig. 6

Comparison of amount of conidia produced by the A. fumigatus strains throughout the series. T75 culture flasks containing Sabouraud dextrose agar with or without the addition of 4 mg/L itraconazole, were inoculated with 1x10⁵ conidia and incubated at 37 °C for 7 d. Conidial suspensions were prepared via immersion in 30 mL PBS containing 0.05% Tween-80 and counted. Data was obtained in duplicate and mean values ± SD are shown (^*p = 0.023, ^**p = 0.0009; two-tailed Students T-test).

Fig. 7

Survival of Galleria mellonella larvae infected with specific isolates. Groups of 10 G. mellonella larvae were infected with 6 × 10³ conidia in the last pro-leg using a 0.33 mm Micro-Fine needle. Two control groups of larvae were included in each experiment; 10 unmanipulated larvae and 10 larvae injected with phosphate-buffered saline. Larvae were monitored for 6 d; larval death was characterised by lack of movement and melanisation. Survival after infection with isolate V157-62 or V135-39 was significantly higher in comparison to isolate V130-15 (^*p < 0.05; two-tailed Students T-test).