In vivo antifungal screening against triazole-resistant and -susceptible Aspergillus fumigatus using bioluminescence imaging in Galleria mellonella

Author:

Eliane Vanhoffelen (BE)

Abstract:

Background:  The increased prevalence of triazole-resistant Aspergillus fumigatus (AF) isolates and plateaued effectiveness of antifungal therapy call for urgent optimization and expansion of the current treatment options. Galleria mellonella caterpillars are a promising in vivo model to narrow the gap between in vitro and mouse antifungal research. However, a lack of sensitive and reproducible readouts currently limits its full potential. To maximize this model for efficient antifungal screening, we developed a G. mellonella model of azole-resistant and -susceptible aspergillosis based on bioluminescence imaging (BLI) to non-invasively quantify fungal load over time, and benchmarked this model for first-line antifungals.

Methods:  G. mellonella larvae were infected with a triazole-susceptible (WT) or -resistant (TR34/L98H) AF strain expressing a red-shifted firefly luciferase and kept at 37°C for 5 days post infection. Treatment groups received voriconazole (VCZ, 20 mg/kg/day, Vfend) or amphotericin B (AMB, 10 mg/kg/day, Fungizone). Larval AF burden was measured daily by BLI (IVIS Spectrum) using an optimized D-luciferin dose of 40 µg/g allowing dynamic visualization of fungal burden over time without affecting larval health. BLI readouts were compared with daily survival and health scoring, the current standards for longitudinal follow-up in G. mellonella, and with colony forming unit (CFU) counts of larval homogenates for quantitative validation.

Results:  When comparing survival, health scoring and BLI in their capacity to discriminate between log10-fold differences of AF TR34/L98H infections in G. mellonella over time, BLI was the only readout that significantly distinguished all inocula as early as one day post infection, while survival and health scoring were unable to detect differences between the lower fungal burdens even after 5 days post infection (Fig. 1 A-C). Next, we verified the quantitative character of in vivo BLI fluxes as a readout of fungal burden by validating a broad range of in vivo BLI fluxes above imaging baseline (~4.5 x 104 p/s) against gold-standard CFU, showing an excellent correlation (Fig. 1E). BLI showed a superior dynamic range compared to CFU above imaging baseline, while CFU was more sensitive below this threshold (Fig. 1F).

To benchmark our BLI-based model for antifungal screening, WT or TR34/L98H AF-infected larvae were treated with VCZ or AMB and sensitivity of treatment detection was compared between survival, health scoring and in vivo BLI (Fig 2 A-C; E-G). BLI detected treatment efficacy against triazole-resistant and -sensitive AF as soon as 1 or 2 days post infection respectively, thereby largely outperforming survival and health scoring. Moreover, BLI was the only readout to distinguish between the treatment effects of VCZ and AMB against WT AF (Fig 2C). In vivo BLI fluxes of treated and non-treated groups were successfully validated against CFU for fungal quantification in both WT and TR34/L98H AF strains (Fig. 2D,H).

Conclusions: In conclusion, we established the first refined G. mellonella model of azole-resistant and -susceptible aspergillosis that enables the noninvasive real-time quantification of AF by BLI, thereby outperforming the current longitudinal readouts. This model provides a quick and reproducible in vivo system for evaluating treatment options and is in line with 3Rs recommendations

 

Abstract Number: 87

Conference Year: 2024

Conference abstracts, posters & presentations

Showing 10 posts of 17325 posts found.

  • Title

    Author

    Year

    Number

    Poster