The sphingolipid inhibitor myriocin increases Candida auris susceptibility to amphotericin B

The emergence of the pathogenic yeast Candida auris is of global concern due to its ability to cause hospital outbreaks and develop resistance against all antifungal drug classes. Based on published data for baker's yeast Saccharomyces cerevisiae, sphingolipid biosynthesis, which is essential for maintaining membrane fluidity and formation of lipid rafts, could offer a target for additive treatment.

7][18][19] The Erg11 F126L mutation can only be found in clade III isolates while the Y132F and K143R mutations are frequently found in clade I and IV isolates. 14,16Interestingly, clade II isolates are often susceptible to fluconazole. 14Resistance to echinocandins is emerging in some countries, which is mainly caused by S639 modifications in the 1,3-ß-D-glucan synthase subunit encoding gene FKS1. 14,18,20Such mutations have been detected in isolates from clades I, III and IV, but not from clade II. 14Elevated minimum inhibitory concentrations (MICs) for amphotericin B are rare for most fungi but are frequently found in C. auris isolates, especially those from clade I.A global study defined nearly 50% of all analysed clade I isolates as "resistant" based on phenotypic testing. 14Multidrug resistant isolates are frequently found in C. auris, especially in clade I, where 45% of all known isolates were resistant to two antifungal drugs. 14 cell membrane components, sphingolipids are involved in a variety of biological processes of the fungal cell and might be an interesting target for new treatment approaches. 214][25][26] Such disruptions of the sphingolipid imbalance can also lead to a mis-sorting of the efflux pump Cdr1 in the cell membranes of C. albicans which correlates with increased azole sensitivity. 22,230][31] Interestingly, C. auris was found to be highly susceptible to the sphingolipid biosynthesis inhibitor myriocin. 29,30Based on these observations, we tested the susceptibility of different clinical C. auris isolates to myriocin and consequently examined the effects of a combination of the inhibitor together with antifungal drugs.We could show that presence of myriocin synergistically increases the susceptibility of C. auris to amphotericin B.

| C. auris is highly susceptible to myriocin
It was previously shown that C. auris is susceptible to myriocin. 29,30 could confirm this observation in our settings.Additionally, we found that C. auris was much more affected by the compound in comparison to other medically important Candida species (Figure 1).After a 3-day incubation at 37°C, C. auris already displayed some growth defects at 250 nM myriocin (Figure 1).With increasing concentrations of the compound, a dramatically decrease in C. auris growth ability was observed, leaving it the Candida species which was most susceptible to myriocin in our screening (Figure 1).
In contrast, C. tropicalis, Clavispora lusitaniae and Meyerozyma guilliermondii were highly resistant against myriocin as their growth was not even impaired at 5 μM (Figure 1).

| Myriocin increases activity of amphotericin B against resistant isolates
We further examined if the addition of myriocin could have beneficial effects on the activity of amphotericin B against C. auris.

F I G U R E 1
Candida auris is highly susceptible to the sphingolipid inhibitor myriocin.The indicated medically important Candida spp.were grown on YPD medium with or without different myriocin concentrations.Methanol was used as solvent control for myriocin.Candida cell concentrations are from left to right: 1 × 10 6 cells/mL, 1 × 10 4 cells/mL and 1 × 10 2 cells/mL.Pictures were taken after 3 days of growth at 37°C.
We therefore selected the susceptible reference strain B8441 and two amphotericin B resistant clinical isolates, NRZ-2017-505 and NRZ-2019-731.The influence of myriocin on the amphotericin B susceptibility of these strains was tested with BioMerieux E-tests on RPMI1640 agar plates.These plates additionally contained methanol as solvent control and three different concentrations of myriocin (2.5 nM, 25 nM and 250 nM).Initially, C. auris B8441 had an amphotericin B MIC-value of 0.5 μg/mL (Figure 2).While this remained unaffected in presence of 2.5 nM myriocin, the susceptibility of B8441 strongly increased at 25 nM, leading to a MIC of 0.008 μg/ mL (Figure 2).At the highest myriocin concentration (250 nM), the strain was completely susceptible with a MIC below 0.002 μg/mL (Figure 2).A similar dynamic was observed in the amphotericin B resistant strains.The isolate NRZ-2017-505 held an initial MIC of 4 μg/mL, which could already be reduced to 0.25 μg/mL in presence of 2.5 nM myriocin.It further decreased with increasing myriocin concentrations, reaching a MIC of 0.002 μg/mL at 250 nM myriocin (Figure 2).For the isolate NRZ-2019-731 we observed an initial MIC of 24 μg/mL.Surprisingly even the lowest tested myriocin concentration of 2.5 nM dramatically increased the susceptibility of the clinical isolate, leading to a beyond 7-fold MIC reduction to 0.25 μg/ mL (Figure 2).With increasing myriocin concentrations the MICs further declined.The MIC of 0.006 μg/mL in presence of 250 nM was comparable to the other two test strains (Figure 2).second method.With the help of an amphotericin B broth microdilution approach in the presence of 25 nM or 250 nM myriocin 12 out of the 24 tested strains showed amphotericin B MICs of 0.03 μg/mL at 250 nM (Table 1).Ten isolates had a MIC of 0.06 μg/mL and only two of 0.125 μg/mL (Table 1).All isolates would be therefore regarded as susceptible to amphotericin B, even those which initially showed resistance to the drug applying the CDC tentative breakpoint (NRZ-2017-288, NRZ-2017-505, NRZ-2019-207, NRZ-2019-731, NRZ-2021-103 and NRZ-2021-170) (Table 1).

| Synergistic effects of myriocin on the activity of amphotericin B against C. auris
Due to these observations, we once more enhanced the testing scheme.C. albicans, C. glabrata and C. auris were tested with a broth microdilution checkerboard assay in RPMI1640 medium with myriocin and/or amphotericin B. We selected four C. auris strains representing the most prevalent clades: B8441 (clade I), B11220 (clade II), B11221 (clade III) and B11244 (clade IV).In contrast to the initial plate assay, all tested C. auris strains displayed a myriocin MIC of 5 μg/mL while the MICs for amphotericin were between 1 μg/mL and 4 μg/mL or higher (Figure 3).For the combination of amphotericin B and myriocin we used concentrations of the latter between 80 nM to 5 μM while amphotericin concentrations were between 0.008 and 8 μg/mL.Already at 80 nM, two C. auris strains (B11220 and B11221) showed a decrease in their susceptibility to amphotericin B which displayed a synergistic effect according to the calculation of the fractional inhibitory concentration index (FICI) (Figure 3).From 160 nM onwards, all further myriocin concentrations caused a synergistic effect on the amphotericin B susceptibility of the four tested C. auris strains (Figure 3).

| Myriocin has only limited effects on the susceptibility against fluconazole
Due to the identified synergistic effect of myriocin addition to the activity of amphotericin B against C. auris, we further examined if 256 μg/mL (Figure 4).Strikingly, even 2.5 nM myriocin already increased the susceptibility of B8441 against fluconazole, leading to a decrease of the MIC down to 4 μg/mL.This beneficial effect was further strengthened with increasing myriocin concentrations (Figure 4).However, we could not observe the same effect for the other two clinical isolates.Both were still resistant to fluconazole in presence of 2.5 and 25 nM myriocin.A MIC decrease was only noticed in presence of the highest myriocin concentration (250 nM; Figure 4).Therefore, we conclude that myriocin might increase the susceptibility to fluconazole only in some C. auris strains and has no general synergistic effect as observed for amphotericin B activity.

| DISCUSS ION
3][34][35] This is especially interesting for human pathogenic fungi like C. auris which rapidly develop resistance against available antifungal drugs.In accordance with previous works, 29,30 we showed that C. auris is highly susceptible to myriocin.However, due to its high toxicity against all eukaryotic cells the compound itself is not a direct treatment option. 33,36Nonetheless, it can be an interesting starting point to We have also tested the combination of myriocin and fluconazole, but only observed limited effects on the fungal susceptibility.Benefits of a myriocin-mediated disruption of the sphingolipid biosynthesis onto the fluconazole susceptibility might be much more strain specific.This might also depend on the overall susceptibility of a strain to myriocin.[40] Synergistic effects of a combination of sphingolipid inhibitors and antifungal drugs were also reported for C. glabrata, but not for Aspergillus fumigatus. 35,41In addition, a previous work showed that the levels of dihydrosphingosine are higher in so-called CRS-MIS (caspofungin reduced susceptibility-micafungin increased susceptibility) mutants in C. albicans, C. glabrata and A. nidulans, but inhibition of the sphingolipid biosynthesis by myriocin increased susceptibility. 41,42If this is also true for C. auris remains to be tested in future experiments.
Overall, our data suggest that sphingolipids might play an important role in the decreased susceptibility of C. auris against amphotericin B. Therefore, a combination of this drug with sphingolipid biosynthesis inhibition might be a useful approach for improving the treatment of C. auris infections.

| Strains and media
All fungal strains used in this study are listed in Table 2.They were routinely maintained in YPD medium (20 g/L glucose, 20 g/L peptone, 10 g/L yeast extract, with 20 g/L agar if required) at 30°C or 37°C.

| Growth tests with myriocin
2 mg myriocin powder (Sigma Aldrich, Darmstadt, Germany) were resuspend in 1 mL methanol at 37°C to create a 5 mM stock solution which was routinely stored at −20°C.Myriocin from this stock solution was then added to YPD agar prior to pouring the plates, reaching final concentrations between 2.5 nM and 5 μM.As a solvent control, 100 μL methanol were added to YPD agar.Selected Candida strains were then grown overnight in YPD medium in a shaking incubator at 37°C.Cells from these overnight cultures were diluted in fresh YPD medium to reach concentrations of 1 × 10 6 cells/mL, 1 × 10 4 cells/mL and 1 × 10 2 cells/mL.5 μL of these dilutions were then dropped onto YPD with or without 1 μM and 5 μM myriocin or methanol as solvent control.The plates were then incubated for 3 days at 37°C prior to photography.agar.Afterwards, C. auris cells and E-test stripes were applied as described before.

| Broth microdilution with myriocin and / or amphotericin B
We have used a collection of 24 clinical C. auris isolates within this study (Table 2).These strains were mainly obtained by the NRZMyk from German hospitals or from the CDC & FDA Antibiotic Resistance Isolate Bank (Table 2).
The broth microdilution testing for amphotericin B alone or in combination with myriocin was performed in accordance with the guidelines of The European Committee on Antimicrobial Susceptibility Testing (EUCAST). 43For testing the susceptibility to amphotericin B, strains were grown in RPMI1640 medium with drug concentrations ranging from 0.008 to 8 μg/mL for 24 h at 37°C.In case of the combinatory experiment with myriocin, the wells also contained a final concentration of 25 nM or 250 nM myriocin.

| Broth microdilution checkerboard approach with amphotericin B and myriocin
To test the combinatory effects of amphotericin B and myriocin, we have used a broth microdilution checkerboard approach with RPMI1640 medium in accordance with the EUCAST guidelines. 44ugs tested were amphotericin B and myriocin.Final concentrations ranged from 0.008 to 8 μg/mL for amphotericin B and 80 nM to 5 μM for myriocin.The 96-well plates were incubated for 24 h at 37°C.The FICI was calculated as previously described and synergy was defined if the FICI was ≤0.5. 45 In accordance with the results obtained by E-tests, we confirmed a rise in amphotericin B susceptibility of 24 clinical isolates using a F I G U R E 2 Susceptibility of selected Candida auris strains to amphotericin B in presence of myriocin.Cells of C. auris strains B8441, NRZ-2017-505 and NRZ-2019-731 were plated onto RPMI1640 agar with or without the indicated concentrations of myriocin.Methanol was used as a solvent control.Afterwards, amphotericin B E-tests were applied.The plates were incubated at 37°C for 48 h prior to photography.Minimal inhibitory concentrations are displayed below the single pictures.
Shown are the results of broth microdilution assays with 24 clinical C. auris strains grown in RPMI1640 medium with different concentrations of amphotericin B (0-4 μg/mL).If required, 25 or 250 nM myriocin were added to the medium.TA B L E 1 Impact of myriocin addition on the susceptibility of clinical Candida auris isolates to amphotericin B. myriocin could also affect the susceptibility of C. auris against fluconazole.We tested B8441 and the clinical isolates NRZ-2017-505 and NRZ-2019-731 with fluconazole E-tests on RPMI1640 agar plates with different myriocin concentrations.Initially, all three strains were resistant to fluconazole with MICs of more than

F I G U R E 3 37 F I G U R E 4
Synergistic antifungal effects of amphotericin B and myriocin against Candida auris strains from different clades.C. auris strains representing the four major clades were grown in 96-well plates which contained amphotericin B, myriocin or both in a checkerboard approach.Wells with fungal growth are shown in red.Synergism between the two drugs is displayed by red-marked amphotericin B MICs. Synergism was defined if the Fractional Inhibition Concentration Index for both drugs was ≤0.5.further develop and modify a more specific and less toxic substance.The best example is the development of fingolimod, a well-known derivate of myriocin.It can antagonise sphingosine activity and is now used for the treatment of multiple sclerosis patients. 36Clinical C. auris isolates often display a decreased susceptibility against amphotericin B. 14 Within this work we have tested if myriocin-mediated disruption of the sphingolipid biosynthesis could lead to an increased susceptibility of the fungus against amphotericin B. A combination of the drug with myriocin resulted in a clear enhancement of antifungal activity.The synergistic effect could be shown even at very low concentrations of myriocin and significantly decreased the MICs for amphotericin B. This synergism between amphotericin B and myriocin was observed for strains of the four major C. auris clades, indicating a general effect rather than a strain-or clade-specific one.Even phenotypically resistant C. auris strains became susceptible after the addition of sublethal myriocin concentrations between 2.5 and 250 nM.Our results thus indicate an important role of sphingolipids for the ability of C. auris to adapt to and withstand amphotericin B. This is in accordance with previous works showing that induced changes of the sphingolipid composition within cell membranes directly influenced the interaction of amphotericin B with other membrane components such as ergosterol and phospholipids and thus increased fungal susceptibility. 25,33,34,Susceptibility of selected Candida auris strains to fluconazole in presence of myriocin.Cells of C. auris strains B8441, NRZ-2017-505 and NRZ-2019-731 were plated onto RPMI1640 agar with or without the indicated concentrations of myriocin.Methanol was used as a solvent control.Afterwards, fluconazole E-tests were applied.The plates were incubated at 37°C for 48 h prior to photography.Minimal inhibitory concentrations are displayed below the single pictures.
Strains used in this study.
Antifungal drug susceptibility testing for single strains was performed with E-tests (BioMérieux, Marcy-l'Étoile, France) for amphotericin B and fluconazole.Candida cultures were grown overnight in YPD at 37°C.According to the manufacturer's instructions, fungal cells were then diluted to a McFarland of 0.5 and streaked onto RPMI1640 agar (Becton Dickinson, Heidelberg, Germany) before the E-test stripes were applied.To test the effects of myriocin, 100 μL of different dilutions of the compound in methanol were plated on RPMI1640 agar to reach final concentrations between 2.5 nM and 250 nM.As solvent control, 100 μL methanol were plated onto the TA B L E 2