Rho2-dependent cell wall remodeling boosts the fungistatic activity of manogepix against Aspergillus fumigatus

Abstract

Fosmanogepix is a new antifungal agent currently undergoing phase 3 clinical trials. Its active moiety, manogepix, inhibits Gwt1 which is an enzyme essential for the assembly and attachment of glycosylphosphatidylinositol anchors to cell wall proteins. Manogepix has a strong fungistatic activity against most human pathogenic fungal species. Here we characterized the activity of manogepix against the major human pathogenic mold Aspergillus fumigatus. We show that manogepix susceptibility is linked to the expression of gwt1, the gene encoding Gwt1, thereby demonstrating that overexpression of gwt1 can confer resistance. In agreement with a previous study conducted with a different fungal pathogen, we observed an increase of cell well chitin after manogepix treatment. Using a luciferase-based reporter assay, we show that manogepix activates the cell wall integrity stress response pathway. However, manogepix only occasionally causes cell lysis. Mutants that lack the cell wall stress sensor Wsc1 or the Rho GTPase Rho4, which is important for septum formation, are slightly more susceptible to manogepix. In contrast, mutants lacking the Rho GTPase Rho2 or, to a lesser extent, the cell wall stress sensor MidA, both of which are essential for A. fumigatus to survive attacks from granulocytes and other forms of cell wall stress, grow better than wild-type when exposed to manogepix. We show that both mutants, especially Δrho2 but also ΔmidA, fail to upregulate cell wall chitin to wild-type-like levels in response to manogepix. These results implicate a role for the Rho2-dependent stress response in the fungistatic activity of manogepix against A. fumigatus.

1. Introduction

Aspergillus fumigatus is a mold and airborne opportunistic fungal pathogen. It causes infections in humans, of which invasive aspergillosis is the most feared. Invasive aspergillosis is a severe systemic infection that occurs primarily in immunocompromised patients and is associated with a high mortality rate even when treated. There is currently only a limited number of drugs that are approved for the treatment of infections caused by Aspergillus species. These belong to essentially three drug classes: Azoles, which inhibit sterol C14-demethylase CYP51, a key enzyme in the synthesis of the primary fungal sterol ergosterol, resulting in the accumulation of toxic ergosterol precursors (Elsaman et al., 2024; Rybak et al., 2024). Polyenes, which bind to ergosterol having detrimental effects on the membrane integrity, including the formation of fungicidal ion channels (Lewandowska et al., 2021). And, as a second-line medication, echinocandins, which inhibit β-1,3-glucan synthase required for cell wall biogenesis and which exert only a fungistatic activity against Aspergillus species (Dichtl et al., 2015). The emergence of resistance to azole antifungals, frequent problems with drug-drug interactions and tissue penetration, and the high toxicity of polyene antifungals highlights the need for new drugs for the treatment of fungal infections.

Fosmanogepix is a new first-in-class antifungal for the use in humans. It shows activity against most human pathogenic fungi, including Aspergillus species (Gebremariam et al., 2019), and is currently being investigated in phase 3 clinical trials for the treatment of patients with invasive fungal infections (ClinicalTrials.gov, NCT05421858, NCT06925321, 2025). Fosomanogepix is the pro-drug of manogepix (APX001), which inhibits the fungal enzyme Gwt1, located in the endoplasmic reticulum (ER). Gwt1 is required for the assembly and attachment of glycosylphosphatidylinositol (GPI) anchors to proteins destined for the cell surface (Tsukahara et al., 2003; Umemura et al., 2003). Gwt1 catalyzes the palmitoylation of glucosamine-phosphatidylinositol (GlcN-PI) at the second position of the inositol ring using palmitoyl coenzyme A (palmitoyl-CoA) as the substrate (Dai et al., 2024; Tsukahara et al., 2003; Umemura et al., 2003). Inhibition of this enzyme causes an accumulation of immature cell surface proteins in the ER which has fungistatic effects (Mann et al., 2015; McLellan et al., 2012; Miyazaki et al., 2011). Since many GPI-anchored proteins have an integral role in maintaining the structure and organisation of the fungal cell wall (Ibe and Munro, 2021; Samalova et al., 2020), detrimental effects on the cell wall stability are to be expected.

Here we studied the antifungal effects of manogepix against A. fumigatus. We show that overexpression of gwt1 can act as a mechanism to overcome the growth inhibitory effect of manogepix in this mold. Furthermore, we show that manogepix triggers activation of a cell wall salvage response, which results in an increase of cell wall chitin and involves the Rho GTPase Rho2. Surprisingly, this activation contributes to the fungistatic effect of manogepix.

2. Results

2.1. Construction and phenotypic characterisation of a conditional A. fumigatus gwt1 mutant

Manogepix targets the enzyme Gwt1. While homologs of this enzyme have not been directly studied in Aspergillus species so far, a recent study predicted that the structural basis for Gwt1 inhibition is conserved between different fungal species, including Aspergillus (Dai et al., 2024). A. fumigatus encodes one Gwt1 homolog which shares an identity of 38 % and similarity of 55.5 % with its S. cerevisiae homolog. To characterise the role of Gwt1 for growth and viability of A. fumigatus, we constructed a conditional mutant by replacing the endogenous promoter of gwt1 with doxycycline-inducible Tet-On promoter (Helmschrott et al., 2013; Neubauer et al., 2015). This conditional gwt1_tetOn mutant showed significantly reduced gwt1 expression and reduced radial growth under repressive conditions, while addition of doxycycline achieved gwt1 overexpression and fully restored wt-like growth (Fig. 1 A and Supplementary Fig. 1).

Fig. 1.

Phenotypic characterisation of a conditional A. fumigatus gwt1 mutant (A) 1.5 × 10³ conidia of wild-type (wt) or three independent conditional gwt1_tetOn mutants (#1–3) were spotted onto Sabouraud agar (SAB) supplemented with the indicated concentrations of doxycycline (Doxy) and incubated for 28 h at 37 °C. (B) In a series of 10-fold dilutions derived from a starting suspension of 5 × 10⁷ conidia per ml of the indicated strains, aliquots of 3 μl were spotted onto SAB agar plates. Where indicated, media was supplemented with 15 μg ml⁻¹ Doxy, sodium dodecyl sulfate (SDS) 0.02 % (w/v) or calcofluor white (CFW) 40 μg ml⁻¹. Plates were incubated at 37 °C for 30 h, except for CFW-supplemented plates which were incubated for 48 h. (C) A total of 1.0 × 10⁶ conidia of the indicated strains were spread on SAB agar plates. When indicated, SAB was supplemented with 15 μg ml⁻¹ Doxy (+d). Caspofungin Etest strips were applied, and the agar plates were incubated at 37 °C. Representative photos were taken after 30 h. Minimum effective concentration (MEC) readings are indicated in μg ml⁻¹. (D) 4 × 10⁵ conidia of the indicated strains were spread on SAB agar plates. When indicated, medium was supplemented with 0.05 μg ml⁻¹ of manogepix (MGX) or 15 μg ml⁻¹ Doxy (+d). Fifty microliters of caspofungin (100 μg ml⁻¹) were applied in the punch holes of each agar plate. The inhibition zone diameters of three independent replicates per condition were measured, and images were acquired after 40 h of incubation at 37 °C. Statistical significance (*p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001) was calculated using the one-way ANOVA test with Dunnett's post hoc test. Data are representative of 3 experiments conducted under similar conditions.

Gwt1 inhibition would be expected to impair the cell wall integrity of fungi. We therefore tested the susceptibility of the mutant to the cell wall perturbing agents calcofluor white, sodium dodecyl sulfate (SDS) and the echinocandin caspofungin. As shown in Fig. 1 B, repression of gwt1 resulted in a dramatically increased susceptibility to the chitin-binding agent calcofluor white. The susceptibility to the detergent SDS was not increased under repressive conditions. However, under induced conditions, the conditional mutant appeared slightly more resistant (Fig. 1 B). Surprisingly, reduced gwt1 expression resulted in a decreased susceptibility to caspofungin (Fig. 1 C). Similarly, a decreased susceptibility to caspofungin was observed when Gwt1 was chemically inhibited with manogepix in the wild-type (Fig. 1 D).

2.2. Overexpression of gwt1 causes manogepix resistance

Overexpression of the drug target can promote resistance to antimicrobial agents. This is well-demonstrated by the most common azole resistance mechanism in A. fumigatus, which is a tandem repeat (TR) in the promoter of cyp51A, resulting in overexpression of CYP51 (Nywening et al., 2020). In contrast, overexpression of the echinocandin-target β-1,3-glucan glucan synthase is not an established resistance mechanism to echinocandins and rather increases susceptibility to echinocandins (Lee et al., 2023; Loiko and Wagener, 2017). Treatment of A. fumigatus wild-type with manogepix resulted in a marked change of the morphology of hyphae (Fig. 2), reminiscent of the growth of Aspergillus in the presence of fungistatic concentrations of echinocandins (Dichtl et al., 2015). Interestingly, while repression for gwt1 resulted in increased susceptibility to manogepix, induction of the conditional promoter with doxycycline concentrations that facilitate overexpression of gwt1 (Supplementary Fig. 1) significantly increases the resistance to the drug (Fig. 2). This is consistent with similar findings in Candida albicans where gwt1 overexpression also resulted in an increased resistance to Gwt1 inhibitors (Liston et al., 2022; Trzoss et al., 2019; Tsukahara et al., 2003). Our results suggest that overexpression of gwt1 can be a mechanism of manogepix resistance in A. fumigatus.

Fig. 2.

Overexpression of gwt1 increases resistance to manogepix. Conidia of wild-type (wt) or the conditional gwt1_tetOn mutant were inoculated into wells of a 96-well plate in RPMI_MOPS supplemented with the indicated amount of manogepix (MGX). 15 μg ml⁻¹ of doxycycline (Doxy) was supplemented when indicated. Representative images were acquired after 20 h incubation at 37 °C using an automated microscope. Scale bars represent 100 μm and are applicable to all images.

2.3. Manogepix triggers activation of the cell wall salvage response and upregulation of cell wall chitin

Given that GPI-anchored proteins are translocated to the cell membrane to fulfil roles in cell wall remodeling (Ibe and Munro, 2021; Samalova et al., 2020), one may expect that fungi treated with manogepix experience cell wall stress. This would be consistent with the high susceptibility to calcofluor white of the conditional gwt1_tetOn mutant under repressive conditions observed above as well as the occasional cell lysis phenomena observed with the wild type after extensive incubation (Supplementary Fig. 2 A). In fungi, cell wall stress results in activation of a conserved cell wall integrity (CWI) pathway (Dichtl et al., 2016). The promoters of the α-1,3-glucan synthase genes ags1 in A. fumigatus and agsA in Aspergillus niger are established targets of the canonical CWI pathways in these species (Rocha et al., 2016). To analyse the activation of this pathway by manogepix, a mutant that expresses a reporter construct where firefly luciferase is placed under the control of the A. niger agsA promoter was constructed, essentially as described before (Geißel et al., 2018; Ruf et al., 2025). As shown in Fig. 3 A, exposure to the azole antifungal voriconazole resulted in a strong activation of the reporter after approximately 3 h, which is in good agreement with our previous findings that activation of the reporter follows the excessive build-up of cell wall (cell wall patches), which is the consequence of the accumulation of the toxic sterol eburicol triggered by CYP51 inhibition (Elsaman et al., 2024; Geißel et al., 2018). The response following manogepix exposure is much faster than the response induced by voriconazole. In contrast to the voriconazole-induced response, which disappears within the following hours presumably due to the death of the hyphae, the manogepix-induced response steadily increases over time (Fig. 3 A). This demonstrates that manogepix causes cell wall stress and is consistent with a low number of cell lysis events (Supplementary Fig. 2 A). This suggests that this antifungal is rather fungistatic and not imminently lethal for A. fumigatus.

Fig. 3.

Manogepix-induced stress triggers activation of the cell wall integrity pathway and an increase in cell wall chitin. (A) Conidia of an A. fumigatus strain that harbors a luciferase-based cell wall salvage reporter were inoculated in a white 96-well plate in RPMI_MOPS and incubated at 37 °C. After 7 h, medium was supplemented with luciferin and either 0.125 μg ml⁻¹ manogepix (MGX), 2 μg ml⁻¹ voriconazole (Vori) or no drug (control). Luminescence was monitored in a temperature-controlled microplate reader. The graph depicts the mean luminescence of three technical replicates per condition over time. Error bars represent standard deviations of the three replicates. Data were representative of 3 experiments conducted under similar conditions. (B) Wells of an eight-well microscopy slide were co-inoculated with similar numbers of wild-type conidia expressing a cytosolic GFP variant (wt) and conditional gwt1_tetOn mutant conidia (upper bar graph), or inoculated with wild-type conidia without cytosolic GFP (lower bar graph) in RPMI_MOPS. When indicated, medium was supplemented with 15 μg ml⁻¹ of doxycycline (+d; upper bar graph) or 0.125 μg ml⁻¹ MGX (lower bar graph). Samples were then incubated at 37 °C for 16 h if MGX was added, or 9.5 h if no MGX was added. Hyphae were then fixed and stained with calcofluor white. Samples were subsequently washed three times with sterile water. Chitin levels (calcofluor white fluorescence signal intensity) were then analysed using a confocal laser scanning microscope. At least 10 cell wall cross sections were measured for each condition. Statistical significance (*p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001) was calculated with a two-tailed unpaired Student t-test. The error bars indicate standard deviations. These results are representative of 3 experiments performed under similar conditions. Depicted are representative bright field images (BF) and z-stack projections of optical stacks of the fluorescence signal (chitin) that cover the entire hyphae in focus (CFW). Scale bars represent 50 μm.

It was previously reported that manogepix treatment may result in upregulation of cell wall chitin (Liston et al., 2022). To assess this possibility, we analysed the chitin levels of the manogepix-treated wild-type and the conditional gwt1_tetOn mutant under repressive conditions using the chitin-specific dye calcofluor white (Loiko and Wagener, 2017). As shown in Fig. 3 B and Supplementary Fig. 5 B, reduced expression of gwt1 as well as manogepix treatment resulted in A. fumigatus hyphae with significantly higher chitin levels in their cell walls.

2.4. Cell wall stress response regulators contribute the antifungal activity of manogepix

Our results demonstrated that manogepix treatment induces activation of the CWI pathways, suggesting that the response is important for survival and growth of A. fumigatus. To characterise the importance of the CWI pathway-mediated stress response (Dichtl et al., 2016), we analysed the manogepix susceptibility of various mutants with impaired CWI and CWI signaling: Δrho2, a mutant lacking Rho GTPase Rho2 which is involved in cell wall biosynthesis in response to cell wall stress and important for survival of A. fumigatus when challenged by human granulocytes or chemical and physical stress conditions such as calcofluor white, Congo red and heat (Dichtl et al., 2012; Ruf et al., 2025). ΔmidA, a mutant lacking the cell wall stress sensor MidA that is important for survival under very similar stress conditions (Dichtl et al., 2012; Ruf et al., 2025). Δwsc1, a mutant lacking the cell wall stress sensor Wsc1, the absence of which reduces the minimum effective concentration (MEC) of echinocandins (Dichtl et al., 2012). Δrho4, a mutant lacking Rho GTPase Rho4 which is important for septum formation and for survival under various cell wall stress conditions, including when challenged by human granulocytes or azoles antifungals (Dichtl et al., 2015; Geißel et al., 2018; Ruf et al., 2025). And rho2-OE, the complemented Δrho2 mutant that contains multiple copies of the rho2 gene due to multiple integration of the complementation plasmid and is more resistant to cell wall stress (Ruf et al., 2025). Susceptibility testing of the wild-type and the CWI mutants was performed following European Committee on Antimicrobial Susceptibility Testing (EUCAST) antifungal susceptibility testing guidelines, with several modifications to allow for a more detailed evaluation of the impact on growth and morphology, such as microscopic imaging after 40 h and using a lower inoculum of spores (conidia), or adding the metabolic indicator resazurin, to estimate the biomass. The MEC of manogepix, the concentration at which a significant morphological switch occurred and hyphal clusters formed, was 0.125 μg ml⁻¹ for wild-type A. fumigatus (Fig. 4 A). Surprisingly, a mutant lacking Rho2 seemed to be significantly more tolerant to manogepix, showing increased hyphal growth above the MEC compared to wild-type both microscopically and based on resazurin metabolism (Fig. 4 A and B and Supplementary Fig. 3). Notably, distinct morphological differences between the wild-type and Δrho2 hyphae were observed when exposed to manogepix, with the Δrho2 mutant showing clearly longer and more expanded hyphae, an effect that varied slightly in strength depending on the culture media (Fig. 4 A and C). This is in marked contrast to growth of the Δrho2 mutant in the presence of caspofungin, where the opposite effect was observed (Supplementary Fig. 4). Similarly to the Δrho2 mutant, a mutant lacking MidA appeared to form slightly longer and more expanded hyphae than to wild-type in the presence of manogepix, although this microscopic finding was not reflected in enhanced overall resazurin metabolism (Fig. 4 A-C and Supplementary Fig. 3). Again, the ΔmidA mutant appeared to grow slightly worse than the wild-type in the presence of caspofungin (Supplementary Fig. 4). While the Δrho2 and ΔmidA mutants were less constrained in growth in the presence of manogepix, they tended to form more rounded swellings at an earlier stage compared to wild-type at high manogepix concentrations after prolonged incubation, which were prone to cell lysis (Fig. 4 D). Interestingly, osmotically stabilizing the medium with sorbitol partially rescued this growth impairment of the Δrho2 mutant at high manogepix concentrations (Supplementary Fig. 2 B).

Fig. 4.

Cell wall stress response regulators Rho2 and MidA enhance the antifungal efficacy of manogepix. (A and B) Conidia of wild-type (wt) and the indicated mutants were inoculated in RPMI_MOPS into wells of a 96-well plate (1 × 10³ conidia per well). Medium was supplemented with the indicated concentrations of manogepix (MGX). After 72 h incubation at 37 °C, representative images were acquired using an automated microscope (A) or medium was additionally supplemented with 0.002 % (w/v) resazurin, and representative photos taken after a further 16 h incubation at 37 °C (B). (C) Conidia of the indicated strains were inoculated into wells of a 96 well plate, in either RPMI_MOPS (RPMI), Sabouraud (SAB), or Aspergillus minimal media (AMM). Medium was additionally supplemented with either 0.125 or 2 μg ml⁻¹ of MGX. Representative images were acquired after 48 h incubated at 37 °C using an automated microscope. (D) Conidia of the indicated strains were inoculated in RPMI_MOPS supplemented with 1 μg ml⁻¹ propidium iodide and the indicated concentrations of MGX. After 20 h incubation at 37 °C, representative images were acquired using an automated microscope. Depicted are representative images of overlays of the RFP and brightfield channels. (A, C and D) Scale bars represent 100 μm and are applicable to all images in the respective panels.

Overexpression of rho2 coincided with an increase in susceptibility to manogepix (1-fold reduced MEC), less tolerance and much smaller, more compact colonies, and reduced resazurin metabolism compared to wild-type (Fig. 4 A-C and Supplementary Fig. 3). As shown in Fig. 4 A and Supplementary Fig. 3, manogepix impaired the growth but also the survival of the Δrho4 mutant compared to wild-type, suggesting that manogepix causes a partial fungicidal effect, similarly as it was previously observed for the echinocandin caspofungin (Supplementary Fig. 4 and Dichtl et al., 2015). The Δwsc1 mutant was slightly more susceptible to manogepix (1-fold reduced MEC), and the mycelium of the emerging microcolonies appeared slightly more compact compared to wild-type (Fig. 4 A-C).

Taken together, these results demonstrated that the lack of Rho2 and – to some extent - the lack of MidA improves growth of A. fumigatus in the presence of manogepix. In contrast, the lack of Wsc1 or Rho4 as well as the overexpression of Rho2 slightly increase the susceptibility of this pathogen to manogepix.

2.5. The manogepix-induced increase in cell wall chitin depends on Rho2 and MidA

We have recently shown that Rho2 and MidA are both important for A. fumigatus hyphae to withstand attacks from human granulocytes and that activation of Rho2 triggers cell wall remodeling and an increase in cell wall chitin (Ruf et al., 2025). Furthermore, we found that overexpression or constitutive activation of Rho2 significantly reduces the growth rate of A. fumigatus (Ruf et al., 2025). We therefore speculated that the lack of Rho2 and (possibly) of MidA hampers the ability of A. fumigatus to upregulate cell wall chitin in response to manogepix-induced stress. To test this hypothesis, we exposed CWI mutants to manogepix. Chitin levels were subsequently analysed by staining the cell wall chitin with calcofluor white. The fluorescence signal intensity of numerous cell wall cross sections were measured and compared relative to the wt. As shown in Supplementary Fig. 5, the CWI mutants had chitin levels similar to wild-type under normal growth conditions. However, after manogepix exposure, the lack of Rho2 (p-values of n = 3 experiments: 4.49 × 10⁻⁶, 3 × 10⁻⁶, 1.26 × 10⁻⁴) and to a lesser degree the lack of MidA (p-values of n = 3 experiments: 0.0445, 0.0475, 0.38) resulted in a significant decrease of cell wall chitin in Aspergillus hyphae compared to the wild-type (Fig. 5). In contrast, neither the lack of Wsc1 or Rho4 seemed to influence the cell wall chitin levels in hyphae after manogepix exposure (Fig. 5). Compared to wild-type hyphae exposed to manogepix, Rho2-overexpressing hyphae showed comparable chitin levels; a minor observed increase was not statistically significant (p-values of the n = 3 experiments: 0.159, 0.034, 0.145; Fig. 5).

Fig. 5.

Rho2 and MidA regulate the manogepix-induced increase in cell wall chitin. (A and B) Wells of an eight-well microscopy slide were co-inoculated with similar numbers of wild-type conidia expressing a cytosolic GFP variant (wt) and conidia of the Δrho2, ΔmidA, Δwsc1, Δrho4, or rho2 mutant strains in RPMI_MOPS supplemented with 0.125 μg ml⁻¹ manogepix (MGX). After 9.5 h (B; no MGX) or 16 h (A and B; with MGX) incubation at 37 °C, hyphae were fixed and stained calcofluor white. (A) Chitin levels (calcofluor white fluorescence signal intensity) were then analysed using a confocal laser scanning microscope. At least 10 cell wall cross sections were measured for each condition. Statistical significance (*p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001) was calculated with a two-tailed unpaired Student t-test. The error bars indicate standard deviations. Data are representative of 3 experiments conducted under similar conditions. p-values for Δrho2: 4.49 × 10⁻⁶, 3 × 10⁻⁶, 1.26 × 10⁻⁴; p-values for ΔmidA: 0.0445, 0.0475, 0.38; p-values for Rho2-overexpressing strain (rho2): 0.159, 0.034, 0.145. The graph corresponds to the experiment with the following p-values: 4.49 × 10⁻⁶ (Δrho2), 0.0445 (ΔmidA), 0.159 (rho2). (B) Depicted are representative z-stack projections of optical stacks of the calcofluor white fluorescence signal (chitin; CFW) and overlays of these together with z-stack projections of optical stacks of the GFP fluorescence signal that cover the entire hyphae in focus (CFW). Scale bars represent 50 μm and are applicable to all images. (C) Proposed mechanisms that could promote resistance to MGX in A. fumigatus. In a wild-type, MGX inhibits Gwt1 which disrupts the biosynthesis of GPI-anchored proteins and results in accumulation of immature precursors in the ER. The lack of GPI-anchored proteins in the cell wall causes cell wall stress, which triggers the activation of MidA and Rho2, which in turn upregulates cell wall biogenesis and leads to additional growth inhibition. If Rho2 or MidA are lacking (Δrho2, ΔmidA), MGX-induced cell wall stress fails to trigger the Rho2-dependent upregulation of cell wall biogenesis (e.g., chitin), resulting in improved growth compared to wild-type. The impaired cell wall stress response and MGX-induced cell wall damage occasionally cause the lysis of cell compartments. Hyphae septa, the formation of which is dependent on Rho4, mitigate the impact on the mycelium. Overexpression of gwt1, leads to an increased abundance of Gwt1 and insufficient Gwt1 inhibition. GPI-anchored proteins successfully translocate to the cell membrane. This results in an increase of the minimum effective concentration (MEC) of MGX. Created in BioRender. Brazil, S. (2025) https://BioRender.com/s8o7rfn.

3. Discussion

Fosmanogepix and its active moiety, manogepix, have demonstrated promising activity against many clinically relevant fungal pathogens. Here, we characterized its activity against A. fumigatus, which, alongside Candida species, is the most relevant cause of life-threatening invasive fungal infections in Western industrialized countries. Both manogepix treatment as well as reduced expression of the gwt1 gene encoding the manogepix target showed a strong fungistatic activity against A. fumigatus. In contrast, inducing expression of gwt1 in the gwt1_tetOn mutant resulted in increased resistance to manogepix. This is similar to what was previously observed in C. albicans and resembles the CYP51 overexpression-mediated resistance mechanism commonly found in azole-resistant A. fumigatus (Liston et al., 2022; Nywening et al., 2020; Trzoss et al., 2019; Tsukahara et al., 2003).

In agreement with previous observations (Miyazaki et al., 2011), manogepix treatment and reduced expression of gwt1 caused an aberrant growth morphology of the emerging fungal colonies, which was reminiscent of the effects induced by echinocandins. Echinocandins inhibit the biosynthesis of the fungal cell wall polymer β-1,3-glucan which leads to a weakened cell wall. While A. fumigatus is able to survive without β-1,3-glucan, frequent cell lysis phenomena of germinating conidia and compartments of hyphae can be observed (Dichtl et al., 2015; Loiko and Wagener, 2017; Moreno-Velásquez et al., 2017). Since some GPI-anchored proteins are important for cell wall biogenesis, manogepix treatment may also lead to an impaired cell wall. In agreement with this, both our study and the literature provide ample evidence that inhibition of Gwt1 function leads to significant cell wall stress and cell wall alterations: 1) Reduced expression of gwt1 as well as manogepix treatment led to build-up of cell wall chitin, similar as it was observed recently in manogepix-treated C. albicans (Liston et al., 2022), a cellular response typically seen under cell wall stress (Ibe and Munro, 2021). 2) Reduced expression of gwt1 resulted in increased susceptibility to the cell wall-perturbing agent calcofluor, which is in agreement with the increased calcofluor white susceptibility of a S. cerevisiae strain with impaired Gwt1 function (temperature sensitive gwt1–20 mutant) (Umemura et al., 2003). Finally, 3) manogepix exposure triggered a constant activation of the luciferase-based cell wall stress reporter. Based on these results, we investigated how the cell wall integrity and its determinants affects growth and survival of A. fumigatus when exposed to manogepix.

Interestingly, the manogepix-induced cell wall alterations and cell wall stress alone did not translate into many cell lysis phenomena in the wild-type. While we could observe occasional death of fungal compartments, similar as recently shown by others (Liston et al., 2022), these events were very rare compared to the frequent death of fungal compartments that is commonly observed after treatment with echinocandins (Dichtl et al., 2015; Moreno-Velásquez et al., 2017). Accordingly, the lack of Rho4, which is important for septum formation, did not have such a detrimental impact on survival of the mutant in the presence of manogepix as it has in the presence of echinocandins (Dichtl et al., 2015, 2012; Thorn et al., 2024). Surprisingly, our results demonstrated that a lack of Rho2 or MidA did not result in increased susceptibility to manogepix. In contrast, we observed increased growth, especially of the Δrho2 mutant and to a lower extent of the ΔmidA mutant, in the presence of manogepix when compared to wild-type. In line with this, overexpression of Rho2 had the opposite effect and resulted in reduced growth in the presence of manogepix when compared to wild-type. Notably the increased growth of the ΔmidA and especially the Δrho2 mutant came at the cost of a markedly higher frequency of swollen hyphal compartments that were prone to lysis, suggesting that overall cell stability of these mutants is impaired and close to collapse. Very recently, we have shown that Rho2 upregulates cell wall biosynthesis in response to cell wall stress (Ruf et al., 2025). This is essential for A. fumigatus to survive cell wall stress, for example when exposed to calcofluor white or when attacked by granulocytes (Ruf et al., 2025). Furthermore, we have shown that overexpression of Rho2 (using a doxycycline-inducible promoter) reduces the growth rates of A. fumigatus and that constitutive activation of Rho2 triggers a full growth arrest accompanied by inappropriate excessive formation of cell wall, including chitin (Ruf et al., 2025).

Taken together, these results suggest that manogepix/Gwt1 inhibition leads to a strong activation of the cell wall stress response of A. fumigatus, but the actual accompanying cell wall damage does not pose a significant risk for the cell integrity of the mold. However, the cell wall stress response itself leads to an enhanced fungistatic effect through the activation of Rho2, which enhances the overall antifungal activity. The upregulation of the cell wall stress response induced by manogepix could also explain the slight antagonism we observed with caspofungin: while manogepix certainly affects the cell wall and associated membrane in many ways, the increase of cell wall chitin is known to improve survival of fungi exposed to echinocandins (Ibe and Munro, 2021; Lee et al., 2023; Wagener and Loiko, 2017). This antagonistic drug interaction might raise concerns regarding the future use of manogepix-caspofungin combination therapy for the treatment of IA. However, the increase in caspofungin MEC was observed at a relatively low and narrow range of manogepix concentration, calling into question its impact under in vivo treatment conditions.

Could mutations that alter or even disrupt the Rho2-dependent cell wall stress response constitute a clinically relevant resistance mechanism against manogepix? Possible, but on the other hand, conditions in the host are probably more harsh than in vitro growth conditions. We have previously shown that a functioning cell wall stress response, including Rho2, is important for the virulence of A. fumigatus (Dirr et al., 2010; Ruf et al., 2025). Furthermore, a functioning cell wall stress response is important to withstand other antifungals, e.g., azoles and echinocandins (Dichtl et al., 2012; Dirr et al., 2010; this study). Thus, the benefit of increased resistance to manogepix would probably come at the cost of decreased virulence and decreased resistance to other antifungals. Other mechanisms, such as the increased expression of gwt1 as shown in a previous study (Liston et al., 2022) and above, might therefore represent a more promising strategy for overcoming manogepix treatment.

In summary, based on our findings, we propose the following mechanisms that can undermine the antifungal activity of manogepix against A. fumigatus in vitro, and potentially in vivo (Fig. 5 C). 1) Overexpression of gwt1, where an increased abundance of the manogepix target Gwt1 diminishes the inhibitory activity of manogepix, very similar to the common CYP51 overexpression-based azole resistance mechanism (Nywening et al., 2020). 2) Alterations of the manogepix-triggered cell wall stress response, thereby preventing the Rho2-dependent upregulation of cell wall biogenesis and growth inhibition. These mechanisms should be considered as possible causes of resistance, which may occur naturally or under treatment and could be observed in the future.

4. Material and methods

4.1. Strains, culture conditions and chemicals

A non-homologous end joining-deficient derivative of the A. fumigatus strain D141 (AfS35) served as wild-type in this study (Krappmann et al., 2006; Wagener et al., 2008). All strains used in this study are listed in Table 1. The conditional gwt1 mutant (gwt1_tetOn) was constructed by inserting a doxycycline-inducible promoter system (pkiA-tetOn; pYZ002) before the coding sequence of gwt1, essentially as described before (Helmschrott et al., 2013; Neubauer et al., 2015). The cell wall salvage reporter strain was constructed by transforming wild-type with pBG005-phleo as described before (Geißel et al., 2018). Experiments were performed in Sabouraud medium [2 % (w/v) agar, 4 % (w/v) d-glucose, 1 % (w/v) peptone (#LP0034; Thermo Fisher Scientific; Rockford, IL, US), pH 7.0], in RPMI 1640 (with l-glutamine, but without bicarbonate and without phenol red; R8755; Sigma-Aldrich) supplemented with glucose to a final concentration of 2 % (w/v) and with 3-(N-morpholino) propanesulfonic acid (MOPS) at a final concentration of 0.165 mol/l, pH 7.0 (RPMI_MOPS; medium based on the European Committee on Antimicrobial Susceptibility Testing (EUCAST) antifungal susceptibility testing guidelines (Guinea et al., 2022), but without phenol red) or in Aspergillus minimal medium (AMM; (Hill and Kafer, 2001)). Solid media were supplemented with 2 % (w/v) agar (214,030; BD Bioscience, Heidelberg, Germany). Doxycycline was obtained from Clontech (631,311; Mountain View, CA, USA). Resazurin (#R7017), calcofluor white (#F3543), propidium iodide (#P4170), SDS (#L37671), sorbitol (#1.03557), phosphate-buffered saline (#806552) and caspofungin diacetate (#SML0425) were obtained from Sigma-Aldrich (St. Louis, MO, USA). Voriconazole was purchased from Apexbt Technology LLC (#A4320; Houston, TX, USA). Etest strips were purchased from bioMérieux (Marcyl'Etoile, France). Manogepix (#HY18233) was purchased from MedChemExpress (Monmouth Junction, New Jersey, USA). Luciferin was purchased from Promega (E1601; Fitchburg, WI, USA). Stock solutions for caspofungin diacetate (1  mg ml⁻¹), manogepix (5 mg ml⁻¹) and voriconazole (10 mg ml⁻¹) were prepared in dimethyl sulfoxide (DMSO). Working stocks were freshly prepared by dilution in sterile water. Equivalent volumes of DMSO were added to control samples when appropriate.

Table 1.

A. fumigatus strains relevant for or used in this work.

Strain or genotype	Relevant genetic modification	Parental strain	Reference
AfS35 (wt)	akuA::loxP	D141	(Krappmann et al., 2006)
gwt1tetOn	gwt1(p)::ptrA-pkiA(p)-tetOn	AfS35 (wt)	This work
wt + agsA(p)A.niger-Fluc	pBG005-phleo	AfS35 (wt)	This work and (Geißel et al., 2018)
Δrho2	rho2B::loxP-hygroR/blaster	AfS35 (wt)	(Dichtl et al., 2012)
ΔmidA	midA::loxP-hygroR/blaster	AfS35 (wt)	(Dichtl et al., 2012)
Δwsc1	midA::loxP-hygroR/blaster	AfS35 (wt)	(Dichtl et al., 2012)
Δrho4	rho4::loxP-hygroR/blaster	AfS35 (wt)	(Dichtl et al., 2012)
rho2	pSK379-rho2(p)-rho2 (multiple integrations)	Δrho2	(Dichtl et al., 2012)
wt + sGFP	pJW103	AfS35	(Dichtl et al., 2012)

4.2. Microscopy

The chitin content of cell wall cross sections was analysed with a confocal laser scanning microscopy (Leica SP8 microscope; Leica Microsystems, Mannheim, Germany). Conidia were inoculated in Ibidi 15 μ-Slide eight-well slides (#80826; Martinsried, Germany). When indicated, samples were fixed with 3.7 % (v/v) formaldehyde in phosphate-buffered saline for 3 min. Samples were washed three times with sterile water followed by staining of cell wall chitin with calcofluor white (3.33 μg ml⁻¹) for approximately 1 min. Samples were then washed three times with sterile water. The chitin content of cell wall cross sections was analysed using the DAPI channel (excitation 405 nm, emission 410–480 nm). An internal wild-type control expressing cytosolic green fluorescent protein (GFP) variant (wt + sGFP) was inoculated in the same well as the gwt1_tetOn mutant, each of the CWI mutants and the Δrho4 mutant. This internal wt + sGFP control enabled the comparison of chitin content between wild-type and mutant strains in the same well. Data was anonymised before analysing the chitin content of the cell wall cross sections. Additionally, chitin content of the cross sections in image was analysed prior to unmasking the internal wt + sGFP controls using the GFP channel (excitation 488 nm, emission 502–591 nm). At least 10 cross sections were analysed per condition. Statistical significance was calculated using two-tailed unpaired students t-tests.

All other microscopy images were acquired using a Lionheart FX automated microscope (Agilent BioTek; Santa Clara, CA, USA). Conidia were inoculated in clear, flat-bottom, 96-well plates (#3300) purchased from Corning Inc. (NY, USA). When indicated, dead hyphal compartments were stained using 1 μg ml⁻¹ propidium iodide and imaged using the RFP (red fluorescence protein) channel (excitation wavelength 523 nm; filter cube properties: excitation 531/40 nm, emission 593/40 nm, dichroic mirror 568 nm).

4.3. Cell wall stress reporter assay

To analyse activation of the CWI pathway by manogepix, approximately 2 × 10⁴ conidia per well of the luciferase-based cell wall stress reporter strain wt + agsA(p)_A.niger-Fluc were inoculated in RPMI_MOPS in white 96-well polystyrene microplates (#3917) purchased from Corning Inc. (Corning, NY, USA), essentially as described before (Ruf et al., 2025). After 7 h incubation at 37 °C, medium was supplemented with 0.5 mM luciferin and voriconazole, manogepix or no additional drug. Luminescence was measured over time at 37 °C with a Glomax discover microplate reader (#GM3000; Promega).

4.4. Real-time quantitative reverse transcription polymerase chain reaction (RT-qPCR)

RT-qPCR was performed essentially as described recently (Ruf et al., 2025). Briefly, the indicated strains were cultured in Sabouraud medium supplemented with or without 25 μg ml⁻¹ doxycycline. RNA was extracted using the Monarch Total RNA Miniprep Kit (T2010S; New England Biolabs, Ipswich, MA, USA). The RT-qPCR was performed using 1 μg of template RNA per well and the LUNA Universal One Step RT-qPCR Kit (E3005S; New England Biolabs). Expression of gwt1 was normalized to tubA (β-tubulin; AFUA_1G10910) and analysed relative to the expression of gwt1 in the wild type (four biological replicates per condition). Statistical analysis was performed using the unpaired students t-test, comparing strains without doxycycline to each other, and conditions with doxycycline to each other.

4.5. Bioinformatics, information on reproducibility, and calculation of statistical significance

To calculate the percentage identity and similarity between A. fumigatus Gwt1 (Afu1g14870) and S. cerevisiae Gwt1 (YJL091C) the EMBOSS water Pairwise Sequence Alignment tool was used (Madeira et al., 2024). All experiments with qualitative results were performed at least three times under similar conditions with similar results, unless otherwise stated. Statistical significance was calculated using RStudio version 2025.05.0 (Posit Software, Boston, MA, USA).

Declaration of the use of generative artificial intelligence

During the preparation of this work the authors used DeepL (deepl.com; Cologne, Germany) for language optimization, i.e., to identify synonyms, nuance phrases, and to correct errors in spelling, grammar, and punctuation.

CRediT authorship contribution statement

Sean Brazil: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Investigation, Formal analysis, Data curation. Isabel Kopp-Fouquet: Writing – review & editing, Visualization, Validation, Methodology, Investigation, Formal analysis. Samuel Zen Kai Yap: Writing – review & editing, Visualization, Validation, Methodology, Investigation. Michael Carty: Writing – review & editing, Supervision, Investigation. Johannes Wagener: Writing – review & editing, Writing – original draft, Supervision, Resources, Project administration, Investigation, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Johannes Wagener reports financial support was provided by Pfizer Healthcare Ireland. Isabel Kopp-Fouquet reports financial support was provided by Erasmus. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

Supplementary Fig. 1

Supplementary Fig. 2

Supplementary Fig. 3

Supplementary Fig. 4

Supplementary Fig. 5

Data availability

Data will be made available on request.