Discovery of highly reactive self-splicing group II introns within the mitochondrial genomes of human pathogenic fungi

Abstract

Fungal pathogens represent an expanding global health threat for which treatment options are limited. Self-splicing group II introns have emerged as promising drug targets, but their development has been limited by a lack of information on their distribution and architecture in pathogenic fungi. To meet this challenge, we developed a bioinformatic workflow for scanning sequence data to identify unique RNA structural signatures within group II introns. Using this approach, we discovered a set of ubiquitous introns within thermally dimorphic fungi (genera of Blastomyces, Coccidioides and Histoplasma). These introns are the most biochemically reactive group II introns ever reported, and they self-splice rapidly under near-physiological conditions without protein cofactors. Moreover, we demonstrated the small molecule targetability of these introns by showing that they can be inhibited by the FDA-approved drug mitoxantrone in vitro. Taken together, our results highlight the utility of structure-based informatic searches for identifying riboregulatory elements in pathogens, revealing a striking diversity of reactive self-splicing introns with great promise as antifungal drug targets.

Figure 1.

Schematic of the overall workflow. (A) The group II intron discovery and annotation flow. Mitochondrial genome data from GenBank are subject to RNAweasel analysis to identify putative intron domain 5 location. The data are subsequently analyzed by MFannot to annotate the intron, predict intron splice sites and identify putative intronic opening reading frame (ORF). The outcome is the putative sequence for a given group II intron. (B) Secondary structure prediction workflow. The putative sequence for the group II intron is subject to folding prediction and refinement. In the pilot constraint-free folding trial, intron domains are identified and processed separately. Subsequent separate domain folding refinements yield folding constraints and validate the splice sites. Once folding constraints have been obtained from separate domain folding trials, the intron is folded in whole and the predicted secondary structure is subsequently visualized.

Figure 2.

Secondary structure prediction of representative mitochondrial pre-rRNA group II introns newly identified in dimorphic fungi. Important tertiary interactions used as folding constraints (EBS1-IBS1, EBS2-IBS2, α-α’ and γ-γ’) are indicated in the secondary structure diagrams. (A–C) The mitochondrial large subunit rRNA introns in Histoplasma capsulatum (H.c.LSU.I1), Blastomyces dermatitidis (B.d.LSU.I2) and Coccidioides immitis (C.i.LSU.I3), respectively. (D) The mitochondrial small subunit rRNA intron found in Coccidioides immitis (C.i.SSU.I1). The intronic domain 4 of B.d.LSU.I2, C.i.LSU.I3 and C.i.SSU.I1 introns, are reduced to a simple stem-loop capped with UUCG tetraloop.

Figure 3.

Representative pre-rRNA group II introns readily self-splice under near-physiological conditions. The reaction is performed in the buffer containing 40 mM NH₄-HEPES pH 7.5, 150 mM NH₄Cl and 10% PEG-8000 under various magnesium ion concentrations as indicated on top of the gel lanes. After incubation at 37°C for 1 h, the reaction is quenched and loaded onto the 5% polyacrylamide/ 8 M urea gel (as described in Materials and Methods). The splicing gels are shown for (A) H.c.LSU.I1 intron, (B) C.i.LSU.I1 intron and (C) C.i.SSU.I1 intron. The middle part of the gel is not shown as no bands are visible within that region.

Figure 4.

Kinetic characterization of representative novel fungal group II introns. (A–C) Precusor depletion time courses without (black circle, •) and with (blue triangle, ▴) 10% PEG-8000. Crowding agent PEG-8000 greatly enhances reaction rate and helps resolving non-converting species under the low-salt condition (40 mM NH₄-HEPES pH 7.5, 150 mM NH₄Cl) at 37°C. (A) Time course of precursor depletion for H.c.LSU.I1 intron under 10 mM MgCl₂. The observed rates without and with PEG-8000 are 0.039 ± 0.001 min^–1 and 0.36 ± 0.04 min^–1, respectively. (B) Time course of precursor depletion for C.i.LSU.I3 intron under 5 mM MgCl₂ without and with 10% PEG-8000. The observed rates without and with PEG-8000 are 0.75 ± 0.03 min^–1 and 2.00 ± 0.22 min^–1, respectively. (C) Time course of precursor depletion for C.i.SSU.I1 intron under 5 mM MgCl₂ without and with 10% PEG-8000. The observed rates without and with PEG-8000 are 0.43 ± 0.02 min^–1 and 2.94 ± 0.18 min^–1, respectively. (D) Two introns from C. immitis, C.i.LSU.I3 (black circle, •) and C.i.SSU.I1 (blue triangle, ▴) rapidly self-splice under extra mild condition (40 mM NH₄-HEPES pH 7.5, 0.5 mM MgCl₂) at 37°C with 10% PEG-8000. The observed rates for C.i.LSU.I3 and C.i.SSU.I1 intron are 0.28 ± 0.02 min^–1 and 0.33 ± 0.02 min^–1, respectively. Data represent the average of n = 3 independent experiments; error bars represent standard errors of the mean (S.E.M.).

Figure 5.

Mitoxantrone potently inhibits the intron self-splicing reaction for all three representative fungal group II introns in vitro. The observed rate constants (k_obs) are plotted against compound concentration to give the inhibition constant K_i. The K_i values are determined to be 0.64 ± 0.08 μM, 0.18 ± 0.03 μM and 0.15 ± 0.01 μM for (A) H.c.LSU.I1, (B) C.i.LSU.I3 and (C) C.i.SSU.I1 intron respectively. Data represent the average of n = 3 independent experiments; error bars represent standard errors of the mean (S.E.M.).