Echinocandin Resistance in Candida auris Occurs in the Murine Gastrointestinal Tract Due to FKS1 Mutations

ABSTRACT

Candida auris is resistant to multiple antifungal agents. This study investigated its antifungal susceptibility and explored FKS1 mutations across the isolates from mice enterically colonized with wild-type C. auris and treated with echinocandin. Resistant C. auris with FKS1 mutations, including S639F, S639Y, D642Y, R1354H, or R1354Y, were isolated and found to be micafungin- and caspofungin-resistant in vivo; however, the MICs of isolates with mutation in R1354 remained below the micafungin breakpoint in vitro.

TEXT

Candida auris is a multidrug-resistant yeast responsible for causing invasive fungal diseases. Despite this, most C. auris isolates are susceptible to echinocandins, making it the recommended initial treatment for infections caused by C. auris in adults (1). Clinical echinocandin resistance in C. auris is generally associated with amino acid mutations in the hot spot (HS) regions of FKS1, which encode β-1,3-d-glucan synthase (2, 3). The development of echinocandin resistance has been reported in C. auris isolates recovered from patients treated with echinocandins (4–6).

C. auris was reported to mainly colonize mucosal surfaces such as the skin and respiratory and urinary tracts; however, it could also be isolated from stools, indicating that it colonized the intestinal tract (7, 8). Diarrhea and gastrointestinal (GI) decompression were reported to be risk factors for C. auris infection (9), suggesting the possibility of C. auris colonization of the intestinal tract and dissemination via translocation; this was reported in a murine model (10). The GI tract is the reservoir of Candida spp., where they acquire antifungal drug resistance (11); therefore, studies are needed to elucidate the mechanism of C. auris antifungal resistance acquisition in the intestinal tract. This study investigated C. auris antifungal susceptibility and explored the FKS1 mutations across isolates from mice enterically colonized with C. auris and subsequently treated with micafungin or caspofungin.

C. auris NCPF 8971 (clade I) (12), C. auris JCM 15448 (clade II) (13), and Candida albicans SC 5314 (14) were used in this study. Micafungin and caspofungin MICs were 0.12 and 0.25 μg/mL, respectively, for NCPF 8971 and 0.06 and 0.008 μg/mL each for JCM 15448 and SC 5314, respectively.

All animal experiments were performed in compliance with the Guide for the Care and Use of Laboratory Animals (15) and the institutional regulations and guidelines for animal experimentation after pertinent review and were approved by the Institutional Animal Care and Use Committee of Nagasaki University under protocol number 1906121536. Candida colonization of the intestinal tract of a murine model was established, as described previously (16). Briefly, specific pathogen-free, 5-week-old female DBA/2J mice (CLEA Japan Inc., Tokyo, Japan) were fed a low-protein diet before colonization. The mice were inoculated intragastrically with 0.2 mL of Candida cell suspensions adjusted to 5 × 10⁶ cells/mL. Sterile water with antibacterial agents was provided postinoculation. The mice were treated intraperitoneally with 2.71 mg/kg micafungin (Astellas, Tokyo, Japan) or 0.76 mg/kg caspofungin (MSD, Tokyo, Japan) for 11 days (4 to 14 days after inoculation). Echinocandin dose was the standard (AUC₁₀₀), corresponding to human equivalent exposure, and was calculated using the pharmacokinetic/pharmacodynamic study (17). Fifteen days postinoculation, stool from live mice was collected, weighed, and homogenized. The latter was then appropriately diluted and spread on yeast extract peptone dextrose plates with 200 mg/liter imipenem. After 48-h incubation, a single colony per animal was collected from the developed colonies, stored at −80°C, and tested for drug sensitivity to echinocandins. All animal experiments were performed twice, with 10 mice per group in each experiment, and data from the two experiments were combined to get the final value.

While SC 5314 or JCM 15448 had no colony, NCPF 8971 formed colonies at 3 to 4 log₁₀ CFU/mg of stool. Table 1 shows that the echinocandin susceptibilities of C. auris isolate from the stool of NCPF 8971-inoculated mice treated with micafungin or caspofungin. Antifungal susceptibility tests were performed using the Sensititre YeastOne microtiter panel (TREK Diagnostic Systems, East Grinstead, UK) (18). Although breakpoints have not been established for C. auris, the Centers for Disease Control and Prevention (CDC) published tentative MIC breakpoints as follows: micafungin, caspofungin, and anidulafungin ≥ 4, 2, and 4 μg/mL, respectively (https://www.cdc.gov/fungal/candida-auris/c-auris-antifungal.html). Accordingly, 5%, 20%, and 5% of the micafungin-treated isolates and 10%, 35%, and 5% of caspofungin-treated isolates were considered resistant to micafungin, caspofungin, and anidulafungin, respectively. All micafungin or anidulafungin-resistant isolates were found to be caspofungin resistant.

TABLE 1.

Echinocandin susceptibilities of C. auris wild-type and mutant strains isolated from mouse stool^a

Agents administered	Stain no.	DDBJ accession no.	Amino acid mutation		MIC (μg/mL)
			FKS1 HS1	FKS1 HS2	MFG	CAS	AFG
	NCPF 8971		WT	WT	0.12	0.25	0.12
MFG	1	LC736030	S639Y	WT	8	>8	>8
	2	LC736033	WT	R1354H	1	>8	1
	3		WT	WT	1	>8	1
	4		WT	WT	0.5	2	1
	5		WT	WT	0.25	1	0.25
	6		WT	WT	0.25	0.5	0.25
	7		WT	WT	0.12	1	0.25
	8		WT	WT	0.12	1	0.25
	9		WT	WT	0.12	1	0.25
	10		WT	WT	0.12	1	0.25
	11		WT	WT	0.12	1	0.12
	12		WT	WT	0.12	1	0.12
	13		WT	WT	0.12	1	0.12
	14		WT	WT	0.12	0.5	0.25
	15		WT	WT	0.12	0.5	0.25
	16		WT	WT	0.12	0.5	0.12
	17		WT	WT	0.12	0.5	0.12
	18		WT	WT	0.12	0.5	0.12
	19		WT	WT	0.12	0.25	0.25
	20		WT	WT	0.03	1	0.25
CAS	1	LC736029	S639F	WT	>8	>8	1
	2	LC736031	D642Y	WT	4	>8	>8
	3	LC736034	WT	R1354H	2	>8	2
	4	LC736032	WT	R1354Y	1	4	1
	5		WT	WT	0.5	>8	0.5
	6		WT	WT	0.25	>8	0.5
	7		WT	WT	0.25	1	0.25
	8		WT	WT	0.12	1	0.25
	9		WT	WT	0.12	1	0.25
	10		WT	WT	0.12	1	0.25
	11		WT	WT	0.12	1	0.25
	12		WT	WT	0.12	1	0.12
	13		WT	WT	0.12	1	0.12
	14		WT	WT	0.12	1	0.12
	15		WT	WT	0.12	0.5	0.12
	16		WT	WT	0.12	0.5	0.12
	17		WT	WT	0.12	0.5	0.12
	18		WT	WT	0.12	0.25	0.12
	19		WT	WT	0.06	0.25	0.12
	20		WT	WT	0.03	2	0.25

^aTentative MIC breakpoints were as follows: micafungin ≥4 μg/mL, caspofungin ≥2 μg/mL, and anidulafungin ≥4 μg/mL. MFG, micafungin; CAS, caspofungin; AFG, anidulafungin; WT, wild type.

The sequences of FKS1 HS1, HS2, and HS3 regions of all isolates were analyzed next. FKS1 mutations were found in 10% of micafungin-treated mice and 20% of caspofungin-treated mice. Two isolates harboring FKS1 mutation (S639Y and R1354H) were identified from 20 strains across micafungin-treated mice; four isolates harboring S639F, D642Y, R1354Y, or R1354H were identified from 20 strains in caspofungin-treated mice (Table 1). No isolate with mutation in FKS1 HS3 was observed. All micafungin- or anidulafungin-resistant isolates harbored mutations in FKS1 HS1 or HS2, whereas 45% of caspofungin-resistant isolates had no mutation. The FKS1 sequences of all non-wild-type strains were deposited to the DDBJ with the accession numbers shown in Table 1.

All statistical analyses were carried out using Prism software, version 9.0.2 (GraphPad Software, Inc., La Jolla, CA). The survival curves were compared using log-rank tests (Mantel-Cox). The fungal burden in the organs was analyzed using Mann–Whitney U test. A P value < 0.05 was considered statistically significant.

To examine in vivo echinocandin susceptibility, survival curves of mice infected with wild-type and mutant strains evolved in vivo in the above-described experiment were evaluated. The mice were infected intravenously with C. auris NCPF 8971 (wild type) and four isolates, each harboring S639Y, D642Y, R1354Y, and R1354H, and subsequently treated with micafungin or caspofungin (n = 5/group). Specific pathogen-free, 7-week-old female BALB/c mice (Japan SLC Inc., Shizuoka, Japan) were administered 150 and 100 mg/kg cyclophosphamide 4 and 1 day before infection, respectively, and 100 mg/kg 2 and 5 days postinfection. They were infected intravenously through the lateral vein with 0.2 mL of Candida cell suspension adjusted to 3.5 × 10⁵ cells/mL. Micafungin (2.71 mg/kg) and caspofungin (0.76 mg/kg) were administered intraperitoneally for 7 consecutive days commencing 2 h postinfection. The survival rates were not significantly different between wild-type- and mutant strain-infected mice in the no-treatment group. However, the survival rates of the wild-type-infected mice suggested lower mortality than those of mice infected with all types of FKS1 mutants in micafungin or caspofungin treatment groups (P < 0.05) (Fig. 1).

FIG 1.

Survival analyses of mice infected with C. auris wild-type strain and FKS1 mutants. Immunocompromised mice were inoculated with C. auris NCPF 8971 (wild type [WT]) and isolates harboring FKS1 mutations. Micafungin (2.71 mg/kg) and caspofungin (0.76 mg/kg) were administered intraperitoneally for 7 consecutive days. The survival curves of FKS1 mutant infection were compared with those of FKS1 wild-type infection. The data were analyzed by log-rank (Mantel-Cox) test. Asterisks indicate statistically significant differences (*, P < 0.05). MFG, micafungin; CAS, caspofungin.

The therapeutic efficacy of echinocandin was evaluated by comparing the reduction in C. auris CFU burden in mice kidneys. BALB/c mice were infected intravenously with C. auris NCPF 8971 and four isolates harboring FKS1 mutations. The number of Candida cells was adjusted to 5 × 10⁶ CFU/mouse. Micafungin (2.71 mg/kg) and caspofungin (0.76 mg/kg) were administered intraperitoneally for 3 consecutive days commencing 2 h postinfection. To evaluate fungal burden in kidneys, the mice were euthanized 3 days after infection (n = 5/group). In wild-type-infected mice, both micafungin and caspofungin significantly reduced the kidney fungal burden (P < 0.01) compared to the control group; no significant reduction was observed among the four isolates harboring FKS1 mutations (Fig. 2).

FIG 2.

Treatment efficacy of micafungin and caspofungin against C. auris wild-type strain and FKS1 mutants. BALB/c mice were infected with C. auris NCPF 8971 (wild type) and isolates harboring FKS1 mutations. Micafungin (2.71 mg/kg) and caspofungin (0.76 mg/kg) were administered intraperitoneally for three consecutive days. The number of cells recovered from the bilateral kidneys is indicated for individual mice in the plots, and error bars represent SDs. Asterisks indicate statistically significant differences (**, P < 0.01), and ns indicates no significance (P > 0.05). NT, no treatment; MFG, micafungin; CAS, caspofungin.

This is the first report of C. auris acquiring echinocandin resistance in the murine GI tract. Here, resistant strains were isolated from NCPF 8971-inoculated mice; none were isolated from those inoculated with JCM 15448 and SC 5314. Since only a few C. auris clade II strains are echinocandin resistant (19), different C. auris clades may have different capacities to acquire echinocandin resistance.

In C. auris, three mutations (S639Y, S639P, and S639F) in FKS1 HS1 have been reported to be correlated with echinocandin resistance (2, 3); recently, mutations in FKS1 HS1 (ΔF635, F635Y, F635L, and D642Y) and FKS1 HS2 (R1354S and R1354H) have also been reportedly implicated (4, 5, 20). In our study, isolates harbored mutations in FKS1, such as S639F, S639Y, D642Y, R1354H, and R1354Y. All of these mutations, except R1354Y, have previously been described multiple times. All the mutations showed echinocandin resistance both in vitro and in vivo. Our findings show that the strains with mutation in R1354 were the most isolated, and there could be more strains with this mutation clinically. Interestingly, in vitro MICs of all the isolates with mutation in R1354 exceeded caspofungin breakpoints but not micafungin and anidulafungin, as proposed by the CDC; the therapeutic effect of micafungin in vivo was poor. This indicated that although the MICs of echinocandins are below the breakpoints, if close to 1 μg/mL, strains may become echinocandin resistant.

The present study demonstrated that C. auris acquires echinocandin resistance relatively easily in the murine GI tract. Limitations of the present study include the use of only two C. auris strains and the inability to clarify the echinocandin concentrations in the GI tract. Previous studies reported that the GI caspofungin concentrations were significantly lower than plasma levels, and drug levels within the GI tract were not sufficiently maintained for long time in a murine model (11). Similarly, in our model, GI drug concentrations were unlikely to be high enough to inhibit C. auris growth; this may have created a niche that allowed C. auris to acquire echinocandin resistance. To the best of our knowledge, there are no reports examining echinocandin concentrations in human GI tract; however, insufficient echinocandin concentrations may induce C. auris to acquire echinocandin resistance in humans. Further studies are warranted to examine the relationship between drug concentrations in the GI tract and C. auris resistance acquisition, as well as GI echinocandin concentrations in humans. This study may have implications in clinical practice and future validation; therefore, using various strains with different clades might help identify novel FKS1 mutations.