A multicentre study of antifungal susceptibility patterns among 350 Candida auris isolates (2009–17) in India: role of the ERG11 and FKS1 genes in azole and echinocandin resistance

Abstract

Background

Candida auris has emerged globally as an MDR nosocomial pathogen in ICU patients.

Objectives

We studied the antifungal susceptibility of C. auris isolates (n = 350) from 10 hospitals in India collected over a period of 8 years. To investigate azole resistance, ERG11 gene sequencing and expression profiling was conducted. In addition, echinocandin resistance linked to mutations in the C. auris FKS1 gene was analysed.

Methods

CLSI antifungal susceptibility testing of six azoles, amphotericin B, three echinocandins, terbinafine, 5-flucytosine and nystatin was conducted. Screening for amino acid substitutions in ERG11 and FKS1 was performed.

Results

Overall, 90% of C. auris were fluconazole resistant (MICs 32 to ≥64 mg/L) and 2% and 8% were resistant to echinocandins (≥8 mg/L) and amphotericin B (≥2 mg/L), respectively. ERG11 sequences of C. auris exhibited amino acid substitutions Y132 and K143 in 77% (n = 34/44) of strains that were fluconazole resistant whereas WT genotypes, i.e. without substitutions at these positions, were observed in isolates with low fluconazole MICs (1–2 mg/L) suggesting that these substitutions confer a phenotype of resistance to fluconazole similar to that described for Candida albicans. No significant expression of ERG11 was observed, although expression was inducible in vitro with fluconazole exposure. Echinocandin resistance was linked to a novel mutation S639F in FKS1 hot spot region I.

Conclusions

Overall, 25% and 13% of isolates were MDR and multi-azole resistant, respectively. The most common resistance combination was azoles and 5-flucytosine in 14% followed by azoles and amphotericin B in 7% and azoles and echinocandins in 2% of isolates.

Introduction

Candida auris, a recently recognized MDR yeast found in healthcare settings, is considered a major threat to ICU patients. The propensity of this yeast to cause nosocomial bloodstream infections (BSIs) with a high crude in-hospital mortality rate of up to 60% is worrisome.¹ Worldwide reports of C. auris, which was initially described in 2009, have considerably increased in less than a decade, witnessing its spread in five continents.^1,2 Several countries namely Japan, South Korea, India, Kuwait, Oman, South Africa, Pakistan, the UK, Spain, Germany, Norway, Kenya, Israel, Venezuela, Colombia, Panama, the USA and more recently Canada, have reported emergence of C. auris in the last 5 years.^2–24 In 2016, international health agencies including the US CDC,²⁵ ECDC¹⁷ and Pan American Health Organization (PAHO)/WHO²⁶ warned that MDR C. auris is causing persistent colonization and invasive infections among hospitalized patients requiring strict vigilance against cases. However, identification of C. auris in routine microbiology laboratories remains a challenge as many current commercial biochemical identification methods remain inadequate primarily because of a lack of the yeast in their databases.^27,28 The real burden of C. auris is uncertain owing to a lack of availability of diagnostic methods for its identification, particularly in resource-limited countries. Knowledge of antifungal susceptibility data for C. auris is of primary concern as C. auris exhibit consistently high fluconazole MICs and variable susceptibility to the other azoles, echinocandins and amphotericin B.¹ However, by and large, antifungal susceptibility data reported so far in previous studies are limited to low numbers of isolates (<100). A collaborative study undertaken by the US CDC analysed 54 isolates during 2012–15 from four countries including Pakistan (n = 18), India (n = 19), South Africa (n = 10) and Venezuela (n = 5), and reported that 93% of isolates were resistant to fluconazole, 35% to amphotericin B and 7% to echinocandins.² In addition, the report highlighted that 54% of the isolates from four countries with numbers of isolates ranging from 5 to 19 exhibited high voriconazole MICs (>1 mg/L).² That study compared ERG11p amino acid sequences among C. auris isolates and detected various alterations corresponding to well-known azole hot spots within the genome of Candida albicans, including Y132F, K143R and F126T.² Indeed, several mutations in the hot-spot regions (HS) of the ERG11 gene, which lead to amino acid substitutions that alter target protein structures, reduce drug binding affinity and increase azole resistance, have been described in C. albicans, the most frequently isolated human fungal pathogen.^29,30 The substitutions in C. auris reported by Lockhart et al.² were strongly associated with geographic clades, specifically in clonal isolates from South Africa and Venezuela that shared the ERG11p alterations, i.e. F126T in South Africa and Y132F in Venezuela. However, the ERG11p sequences of the 15 Indian isolates analysed showed the presence of substitutions Y132F or K143R, although the isolates were clonal and obtained from three hospitals suggesting that Indian isolates harboured significant resistance alterations.² In this study, we conducted a comprehensive analysis of antifungal susceptibility in 350 clinical isolates of C. auris from 10 hospitals in India collected over a period of 8 years (2009–17) and investigated in vitro activities of 13 antifungal drugs [six azoles (including sertaconazole), amphotericin B, three echinocandins, 5-flucytosine, terbinafine and nystatin] using the CLSI M27-A3 protocol. Further, to determine azole and echinocandin resistance linked to the ERG11 and FKS1 (HSI) genes, respectively, we sequenced the C. auris ERG11 and FKS1 genes to investigate alterations and compared these with the ERG11 and FKS1 genes of azole- and echinocandin-resistant C. albicans. In addition, the relative expression of the C. auris ERG11 gene was studied in a set of isolates exhibiting varied susceptibility to fluconazole.

Materials and methods

Fungal isolates and identification

A total of 350 clinical C. auris isolates collected from nine tertiary care hospitals in Delhi, and the adjoining National Capital Region in the north, and a single centre in south India, Kochi, over a period of 8 years (2009–17) were analysed. The isolates were mainly from patients with candidaemia (blood; n = 267, 76.2%), and other isolates (9.2%) from invasive Candida infections were obtained from tissue (n = 25), pus (n = 6) and pericardial fluid (n = 1). In addition, 14.6% isolates from urine (n = 28), sputum (n = 12) and skin swabs (n = 9) and single isolates each from faeces and the ear canal obtained from colonized patients were included. Of 350 isolates, 120 collected during 2009–14 were identified by sequencing of the internal transcribed spacer region of the ribosomal subunit as described previously²⁶ followed by GenBank basic local alignment search tool (BLAST) pairwise sequence alignment (http://www.ncbi.nlm.nih.gov/BLAST/Blast.cgi). The remaining isolates (n = 230), collected after 2014, were identified as C. auris using MALDI-TOF MS (Bruker Daltonics, Bremen, Germany) with a score value of >2 against the C. auris database (in-house and from Bruker Daltonics).²⁷

Ethics

The study was approved by the V. P. Chest Institute’s (VPCI) Ethics Committee.

Antifungal susceptibility testing

CLSI broth microdilution method

Antifungal susceptibility testing (AFST) was performed using the CLSI broth microdilution method (BMD), following M27-A3/S4.^31,32 The antifungals tested were amphotericin B (Sigma, St Louis, MO, USA), fluconazole (Pfizer, Groton, CT, USA), itraconazole (Lee Pharma, Hyderabad, India), voriconazole (Pfizer), posaconazole (Merck, Whitehouse Station, NJ, USA), isavuconazole (Basilea Pharmaceutica, Basel, Switzerland), 5-flucytosine (Sigma), caspofungin (Merck), micafungin (Astellas, Toyama, Japan) and anidulafungin (Pfizer). In addition, the new topical antifungal drugs, sertaconazole (Optimus, Hyderabad, India), terbinafine (Synergene India, Hyderabad, India) and nystatin (Sigma) were included to test their efficacy against C. auris. All antifungals were dissolved in DMSO. RPMI 1640 medium with glutamine without bicarbonate (Sigma) buffered to pH 7 with 0.165 M MOPS (Sigma) was used. Drug- and yeast-free controls were included, and microtitre plates were incubated at 35 °C and read visually after 24 h, as validated by Pfaller et al.^33,34 CLSI-recommended Candida krusei ATCC 6258 and Candida parapsilosis ATCC 22019 were used as quality control strains. The MIC endpoints for all the drugs except amphotericin B were defined as the lowest drug concentration that caused a prominent decrease in growth (50%) in relation to the controls and for amphotericin B, the MIC was defined as the lowest concentration at which there was 100% inhibition of growth compared with the drug-free control wells. The modal MIC, geometric mean (GM) MIC with 95% CI, MIC₅₀, MIC₉₀, median and range were calculated using Prism version 6.00 (GraphPad Software).

ERG11 gene amplification and sequencing

A total of 44 isolates that showed varied MICs of fluconazole (from 1 to ≥64 mg/L), which included 24 isolates comprising 4-fold differences in MICs (n = 10; range, 1–8 mg/L; n = 14, range, 32 to ≥64 mg/L) were subjected to ERG11 gene sequencing. The remaining 20 isolates selected for ERG11 gene sequencing had elevated MICs for two or more azoles (fluconazole range, 32 to ≥64 mg/L; voriconazole range, 4–16 mg/L; posaconazole range, 4–8 mg/L; isavuconazole, 4 mg/L). The amplification primers for ERG11, i.e. CauErg11F, 5′-GTGCCCATCGTCTACAACCT-3′ and CauErg11R, 5′-TCTCCCACTCGATTTCTGCT-3′ yielding an amplicon of ∼1500 bp and sequencing primers, i.e. CauERG11dF, 5′-TGGGTKGGYTCWGCTGTTG-3′ and CauERG11dR, 5′-TTCWGCTGGYTCCATTGG-3′ were designed on the basis of the C. auris strain XM_018315289 sequence (https://www.ncbi.nlm.nih.gov/nucleotide/1069613591), deposited in the NCBI database. PCR was carried out in a 50 μL reaction volume and the conditions included initial denaturation for 5 min at 95 °C followed by 34 cycles of 30 s at 95 °C, 30 s at 59 °C and 180 s at 72 °C. DNA sequencing was performed using the PCR primers at 2.5 mM concentration. All sequencing reactions were carried out in a 10 μL reaction volume using BigDye Terminator Kit v3.1 (Applied Biosystems, Foster City, CA, USA) according to the manufacturer’s recommendations and analysed on an ABI3130xL Genetic Analyzer (Applied Biosystems). DNA sequences were analysed using Sequencing Analysis software version 5.3.1 (Applied Biosystems). Consensus sequences were made using BioEdit software (version 7.0.5.3) and were aligned with a reference C. auris ERG11 sequence (GenBank accession no. XM_018315289). For mutation analysis, Candida albicans SC5314 ERG11 from assembly orthologous sequences (Candida Genome Database; http://www.candidagenome.org/) were used.^35–39

Quantification of ERG11 gene expression by real-time PCR

The expression of the ERG11 gene was analysed in a total of eight C. auris isolates exhibiting fluconazole MICs ≥64 mg/L (n = 7) and 4 mg/L (n = 1).

RNA extraction

The isolates were grown on an azole-free Sabouraud dextrose agar plate for 48 h at 37 °C. The yeast inoculum was prepared in saline and the inoculum was adjusted spectrophotometrically to an OD of 0.09–0.11 at 530 nm. C. auris was inoculated in 250 mL Erlenmeyer flasks, containing YG medium supplemented with 1.2 g each of uracil and uridine per litre (YG + UU) and incubated in a rotatory shaker at 37 °C for ∼17 h. At a regular interval of time, the cultures were checked for the maximum OD₆₀₀ of 0.3. The total RNA was extracted using Trizol reagent (Sigma–Aldrich, St Louis, MO, USA). RNA integrity was assessed by determination of the OD₂₆₀/OD₂₈₀ absorption ratio, and the integrity was considered maintained if the ratio was >1.95. The RNA was then treated with RNase-free DNase I (NEB, MA, USA) according to the manufacturer’s recommendations. Further, the RNA was quantified and quality was assessed based on the A_260/280 ratio as stated above. The absence of DNA contamination after the RNase-free DNase treatment was verified by PCR amplification of the internal transcribed spacer region.

cDNA synthesis

cDNA was synthesized by using MultiScribe^TM RT (Applied Biosystems, CA, USA) according to manufacturer’s recommendations. Briefly, a 20 μL volume containing 2 μg RNA, 2× RT buffer, 0.8 μL dNTP (100 mM) and 2× random primers in RNase-free water was processed.

RT–PCR

The RT–PCRs were performed on ABI Prism 7500 Fast system (Applied Biosystems, CA, USA). The specific primers for ERG11 (ERG11RTF-5′-CGCTAAGCTTGCGGATGTTT-3′; ERG11RTR-5′-ACTGGAGTGGTCAAGTGGGAAT-3′) and tubulin (TUBRTF-5′- GAGAGAGGCCGAAGGTTGTG-3′; TUBRTR-5′-CCACCCAAGGAGTGAGTAATCTG-3′) of C. auris were designed by the Primer Express software using C. auris sequences of ERG11 (XM_018315289) and tubulin (XM_018311526) deposited in the NCBI database. The tubulin gene was taken as an internal control for comparison of the expression of ERG11 gene in the tested strains. RT–PCR was performed in a 10 μL volume containing the following reagents: 1× Power SYBR^® Green PCR Master Mix (Thermo Fisher Scientific, MA, USA), each primer pair and 0.5 μL cDNA and RNase-free water up to the final volume. Samples were subjected to the initial step of a holding stage at 50 °C for 20 s and 95 °C for 10 min followed by 40 cycles of 15 s at 95 °C and 1 min at 60 °C. Fluorescence data were collected during the 60 °C annealing and extension step and were analysed with the 7500 software v2.0.6 (Applied Biosystems). Independent assays were performed with three replicates for each strain, and expression levels were normalized to the tubulin mRNA level. The expression of the target gene, i.e. ERG11 in the fluconazole-resistant isolates (MIC ≥64 mg/L), was evaluated relative to the expression in C. auris strains with low MICs of fluconazole (4 mg/L). The 2^−ΔΔCT (where C_T is the threshold cycle) analysis method was used to determine the n-fold change in gene transcription. A 2-fold change in the gene expression was considered significant. The experiments were conducted twice by two individuals on different days. The two-tailed Student’s t-test was used to compare categorical variables. P < 0.05 was considered statistically significant.

Further, in two isolates ERG11 gene expression was investigated at 5 and 17 h of fluconazole exposure at the final concentration of 64 mg/L. For fluconazole exposure, two fluconazole-resistant C. auris strains (MIC ≥64 mg/L) were suspended in 50 mL YG + UU medium at an OD₆₀₀ of 0.1. Fluconazole at the final concentration of 64 mg/L was added to the culture when the cells reached an OD₆₀₀ of 0.3 and the mixture was further incubated at 37 °C for 5 h with shaking. Another experimental set-up with same isolates was studied for the constitutive or transient expression of ERG11 gene in the presence of fluconazole for ∼17 h. The control strains received an equivalent amount of DMSO.

FKS1 HSI gene amplification and sequencing

A total of 38 isolates with echinocandin MIC ranges 0.25–16 mg/L were selected for FKS1 HSI sequencing. The whole FKS1 gene was amplified with primers CauFKS1_F (5′-ATGTCTTACGATAACAATC-3′) and CauFKS1_R (5′-TTAGAATGCCTTTGTAGTATAG-3′), and then HSI was sequenced by using primer CauFKS1_F1256-77 (5′-AGAGATACATGAGATTGGGTG-3′). Primers were designed based on XM_018312389 sequence analysis by SeqMan Pro 14 (DNASTAR Lasergene, DNASTAR Inc., WI, USA). Amplification was performed directly from colonies grown on yeast extract peptone dextrose agar plates without prior DNA extraction. Briefly, a single colony was gently removed with a sterile toothpick and transferred directly to a PCR reaction mastermix (30 μL) consisting of 15 μL of EmeraldAmp MAX PCR Master Mix (TaKaRa Bio Inc., Shiga Prefecture, Japan) and 1 μL of each primer (CauFKS1_F and CauFKS1_R) at 10 μM. PCR was performed in a T100 thermal cycler (Bio-Rad Laboratories, Inc., CA, USA). The thermal profile included an initial denaturation for 3 min at 95 °C followed by 40 cycles of 30 s at 95 °C, 30 s at 40 °C and 4 min at 72 °C. Amplicons were visualized on 1% agarose gel stained with GelStar Nucleic Acid Gel Stain (Lonza, Basel, Switzerland), purified using the ZR DNA Sequencing Clean-up Kit (Zymo Research, CA, USA) and sequenced by Genewiz (NJ, USA). Sequencing results were analysed by SeqMan Pro 14 (DNASTAR Lasergene, DNASTAR Inc.).

Results

In vitro susceptibility testing analysis

The MIC data determined for C. auris against 13 antifungals are detailed in Table 1. The MIC distributions spread over 10 dilutions for itraconazole, voriconazole, posaconazole, anidulafungin, micafungin, sertaconazole and 5-flucytosine, and five to nine dilutions for the remaining drugs except for nystatin. Among the azoles, posaconazole (GM 0.05 mg/L) and isavuconazole (GM 0.07 mg/L) exhibited potent activity followed by itraconazole (GM 0.12 mg/L). Notably, 90% (n = 316) of isolates had fluconazole MICs between 32 and ≥64 mg/L and 10% fell between 1 and 16 mg/L. For itraconazole, a modal MIC of 0.125 mg/L was noted and 6.3% (n = 22) had an MIC range of 1–16 mg/L. The MIC₅₀ and MIC₉₀ of voriconazole showed >2-fold dilution difference (MIC₅₀ 0.25 and MIC₉₀ 2 mg/L) and its MIC distribution exhibited an additional peak at 16 mg/L in 2.3% (n = 8) of isolates. Similarly, the MIC distributions of isavuconazole and posaconazole showed an additional peak at 0.25 and 2 mg/L in 10.6% (n = 37) and 3 isolates, respectively. For amphotericin B, 92% (n = 323) of isolates had MIC values of ≤1 mg/L and 8% (n = 27) had MICs ranging from 2 to 8 mg/L. Among echinocandins, modal MICs of micafungin, anidulafungin and caspofungin were 0.125, 0.5 and 1 mg/L, respectively. The modal MIC ±  1 2-fold dilution comprised 87% (n = 305) of the strains for micafungin. Notably, 2% of the isolates had an MIC value of ≥8 mg/L for all the echinocandins. The modal MIC of 5-flucytosine was 0.125 mg/L (n = 205) and three additional peaks at MICs 1, 4 and 64 mg/L in 6.8% (n = 24), 3.1% (n = 11) and 13.4% (n = 47) of isolates respectively were recorded. The topical antifungal sertaconazole showed good activity (GM MIC 0.63 mg/L). However, terbinafine and nystatin had high GM MICs (14.7 and 3.1 mg/L). Although the modal MIC of sertaconazole was 0.25 mg/L, 42% (n = 67 of 160) of isolates had MIC ranges of 1–16 mg/L. Both terbinafine and nystatin MIC distributions spanned within three to five dilutions and for the latter drug all the isolates were within ± 1 2-fold dilution of the modal MIC (4 mg/L).

Table 1.

MIC distribution of C. auris isolates (n = 350)^a against 13 antifungal drugs tested using the CLSI-BMD method

Drug	MIC (mg/L)														GMb	MIC50c	MIC90d
	≤0.016	0.032	0.064	0.125	0.25	0.5	1	2	4	8	16	32	≥64	Range
ITC		66	69	102e	69	22	10	8		1	3			0.03–16	0.12	0.125	0.5
VRC		20	21	94	90	41	32	26	16	2	8			0.03–16	0.31	0.25	2
ISA	44	137	53	35	37	23	9	7	5					≤0.016–4	0.07	0.03	0.5
POS	95	89	83	46	26	2	2	3	2	2				≤0.016–8	0.05	0.03	0.125
AMB				4	27	122	170	19	6	2				0.125–8	0.74	1	1
FLC							5	1	4	1	23	61	255	1–≥64	43.2	64	64
5-FC				205	42	2	24	1	11	5	4	9	47	0.125–≥64	0.51	0.125	64
CAS				2	42	49	175	57	19	5	1			0.125–16	0.96	1	2
MFG	8	18	109	139	57	11	1			3	4			≤0.016–16	0.11	0.125	0.25
AFG	2	1	28	99	71	109	33			7				≤0.016–8	0.27	0.25	1
TRBf								9	10	20	16	55		2–32	14.7	16	32
SERg		12	6	20	32	21	16	15	15	6	15			0.03–16	0.63	0.5	8
NYTh								34	43	3				2–8	3.1	4	4

ITC, itraconazole; VRC, voriconazole; ISA, isavuconazole; POS, posaconazole; AMB, amphotericin B; FLC, fluconazole; 5-FC, 5-flucytosine; CAS, caspofungin; MFG, micafungin; AFG, anidulafungin; TRB, terbinafine; SER, sertaconazole, NYT, nystatin.

a

MIC data from the first 123 isolates and 8 antifungal drugs have been reported previously in comparison with the EUCAST method.⁴⁰

b

GM, geometric mean MICs.

c

MIC₅₀, MIC at which 50% of test isolates were inhibited.

d

MIC₉₀, MIC at which 90% of test isolates were inhibited.

e

Modal MICs are indicated with underlined numbers.

f

110 isolates tested.

g

160 isolates tested.

h

80 isolates tested.

ERG11 mutation and gene expression analysis

The comparison of C. auris ERG11p amino acid sequences with C. albicans showed 15 amino acid substitutions; all of them except E291K, I62V and L282T have been previously listed in the HS regions I, II and III of ERG11p of C. albicans (Table 2).^40,41 Of 15 found substitutions in ERG11p of C. auris, substitutions at positions F105, K119, Y132, K143 and R267 have been previously reported to be associated with azole resistance in C. albicans.

Table 2.

In vitro azole susceptibility and amino acid substitutions in the ERG11 gene of C. auris (n = 44) strains

Isolate	Site of isolation	Hospital	MIC (mg/L)
			ITC	VRC	ISA	POS	FLC	Amino acid substitution corresponding to C. albicans ERG11a
VPCI 1/14	blood	hospital A	0.5	16	0.25	0.25	>64	Y132F
VPCI 2/14	blood	hospital A	0.125	16	4	4	>64	Y132F
VPCI 3/14	blood	hospital B	0.5	4	2	1	64	Y132F
VPCI 4/14	blood	hospital B	0.5	8	2	2	>64	Y132F
VPCI 5/14	blood	hospital C	0.5	4	0.25	0.06	>64	Y132F
VPCI 6/14	tissue	hospital D	0.5	4	0.25	0.06	64	Y132F
VPCI 7/15	blood	hospital D	0.125	4	0.25	0.015	>64	Y132F
VPCI 8/15	tissue	hospital D	0.125	4	0.25	0.015	32	Y132F
VPCI 9/15	pus	hospital D	0.125	4	0.25	0.015	64	Y132F
VPCI 10/15	blood	hospital D	0.125	4	0.125	0.015	>64	K143R
VPCI 11/15	tissue	hospital D	0.125	4	0.06	0.015	>64	Y132F
VPCI 12/14	blood	hospital C	0.125	16	4	2	32	Y132F
VPCI 13/13	tissue	hospital D	0.5	4	4	8	64	Y132F
VPCI 14/13	tissue	hospital D	0.25	4	0.06	0.06	>64	Y132F
VPCI 15/13	tissue	hospital D	0.5	4	0.25	0.125	>64	Y132F
VPCI 16/15	blood	hospital D	0.125	4	0.125	0.015	>64	Y132F
VPCI 17/14	blood	hospital C	0.06	16	4	2	64	Y132F
VPCI 18/14	blood	hospital C	0.25	16	0.5	0.015	>64	K143R
VPCI 19/15	tissue	hospital D	0.125	4	0.25	0.015	>64	K143R
VPCI 20/16	sputum	hospital E	0.5	4	1	0.25	64	K143R
VPCI 21/14	blood	hospital C	0.25	1	0.5	0.06	64	K143R
VPCI 22/14	blood	hospital C	0.25	2	0.5	0.06	>64	K143R
VPCI 23/15	tissue	hospital D	0.03	0.5	0.015	0.015	>64	Y132F
VPCI 24/15	tissue	hospital D	0.03	0.125	0.03	0.015	>64	K143R
VPCI 25/15	pus	hospital D	0.125	2	0.125	0.015	64	Y132F
VPCI 26/15	blood	hospital C	0.125	0.5	0.25	0.06	>64	Y132F
VPCI 27/15	blood	hospital B	0.5	0.125	0.06	0.03	32	K143R
VPCI 28/15	pus	hospital D	0.03	0.25	0.015	0.015	64	Y132F
VPCI 29/12	blood	hospital F	0.125	0.25	0.125	0.06	64	K143R
VPCI 30/12	blood	hospital F	0.125	0.125	0.125	0.06	32	K143R
VPCI 31/13	blood	hospital D	0.5	2	0.5	0.25	64	K143R
VPCI 32/14	blood	hospital C	0.25	1	0.125	0.015	64	K143R
VPCI 33/14	blood	hospital G	0.5	0.5	0.125	0.25	64	K143F
VPCI 34/14	blood	hospital G	0.5	1	0.25	0.25	64	K143R
VPCI 35/12	sputum	hospital G	0.125	0.06	0.03	0.125	8	Y132F
VPCI 36/14	blood	hospital G	0.06	0.03	0.03	0.03	4	K143R
VPCI 37/16	blood	hospital G	0.25	0.125	0.03	0.06	4	Y132F
VPCI 38/16	blood	hospital G	0.06	0.125	0.03	0.03	1	Y132F
VPCI 39/14	tissue	hospital G	0.03	0.03	0.015	0.015	4	no mutation at positions 132 and 143
VPCI 40/15	tissue	hospital G	0.06	0.03	0.03	0.015	4	no mutation at positions 132 and 143
VPCI 41/14	blood	hospital G	0.125	0.06	0.06	0.03	1	no mutation at positions 132 and 143
VPCI 42/14	blood	hospital G	0.06	0.03	0.03	0.03	1	no mutation at positions 132 and 143
VPCI 43/16	blood	hospital G	0.25	0.5	0.125	0.06	2	no mutation at positions 132 and 143
VPCI 44/14	blood	hospital G	0.125	0.03	0.03	0.03	1	no mutation at positions 132 and 143
Summary
Range			0.03–0.5	0.03–16	0.015–4	0.015–8	1–64
Median			0.125	2	0.187	0.06	64
MIC90			0.5	16	2	2	64

ITC, itraconazole; VRC, voriconazole; ISA, isavuconazole; POS, posaconazole; FLC, fluconazole.

a

Amino acid substitutions I62V, F105L, S110A, D116A, K119S, R267T, L282T, E291K, A432S, N440V, F487K, D153E and V437T were present in all 44 isolates. Boldface indicates multi-azole- and pan-azole-resistant isolates.

Notably, Y132F and K143R substitutions responsible for azole resistance in C. albicans were observed in all 34 (77%) sequenced strains that were fluconazole resistant (MICs 32 to ≥64 mg/L). In six isolates, no amino acid substitutions at Y132F and K143R were observed and they had low fluconazole MICs ranging from 1 to 4 mg/L. Furthermore, these isolates had low MICs of other azoles (Table 2). Overall, among 45% (n = 20) of isolates that had Y132F and K143R substitutions, 16 showed cross-resistance to one or more azoles namely voriconazole, isavuconazole and posaconazole and four were pan-azole resistant (Table 2). The other substitutions F105L, S110A, D116A, K119S, R267T, E291K, A432S, N440V, F487K, D153E I62V, L282T and V437T were present in all 44 C. auris isolates investigated. Figure 1 illustrates the expression of ERG11 gene obtained for eight C. auris isolates. No significant fold change (<2) in ERG11 gene expression was observed in any of the isolates tested. Of these, seven C. auris isolates were resistant to fluconazole (MICs 32 to ≥64 mg/L) and had K143R (n = 3) or Y132F (n = 4) substitutions and a single isolate had a low fluconazole MIC (4 mg/L) and had no mutation in the ERG11 gene. However, two strains carrying K143R and Y132F substitutions each, when exposed to fluconazole for 5 h, showed a significant increase in ERG11 transcript to ∼5- to 7-fold (P < 0.006). Further, ERG11 was found constitutively expressed at higher levels in these two strains even after 17 h of fluconazole exposure.

Figure 1.

Expression profile of ERG11 gene in fluconazole-resistant C. auris strains compared with the fluconazole-susceptible isolate (VPCI 40/15). ERG11 gene expression of VPCI 26/15 and VPCI 31/16 C. auris strains after 5 h of fluconazole exposure. Amino acid substitutions in the ERG11 gene and fluconazole MICs are given in parentheses. Asterisks indicate a significant increase (P < 0.006) in ERG11 transcript in the presence of fluconazole. FLC, fluconazole.

FKS1 HSI mutation analysis

Of 38 isolates sequenced for the FKS HSI region, four isolates exhibited a serine to phenylalanine amino acid substitution (S639F), which is equivalent to the FKS1 HSI S645 position in C. albicans and has been associated with elevated echinocandin MICs (Table 3). Notably, all of these isolates were pan-echinocandin resistant with MICs of ≥8 mg/L. Significantly, 34 isolates screened had low echinocandin MICs (range 0.125–1 mg/L) and WT genotype.

Table 3.

Amino acid substitutions in FKS1 HSI in C. auris (n = 38) isolates

			MIC (mg/L)
Isolatea	Site of isolation	Hospital	CAS	MFG	AFG	FLC	Amino acid substitution in FKS1 HSI
VPCI 45/13	blood	D	16	16	8	64	S639F
VPCI 46/14	blood	C	8	16	8	64	S639F
VPCI 47/14	blood	C	8	16	8	64	S639F
VPCI 48/14	blood	C	4	16	8	64	S639F

CAS, caspofungin; MFG, micafungin; AFG, anidulafungin; FLC, fluconazole.

a

34 C. auris isolates had low echinocandin MICs (range 0.125–1 mg/L) and presented WT genotype on FKS1 HSI sequencing.

Discussion

The emergence of Candida species with intrinsically reduced susceptibility or resistance requires continuous monitoring with emphasis on antifungal susceptibility testing using reference methods. Candida auris exhibits uniform resistance to fluconazole and has emerged worldwide.¹ Thus, to gain insight into the antifungal resistance profile of C. auris, we examined the susceptibility profile of a large set of strains collected over a period of 8 years from 10 hospitals in Delhi, National Capital Region and in south India by the reference CLSI-BMD method. Interestingly, during 2009–13 only 10% (n = 35) of C. auris isolates were collected and a remarkable 9-fold increase (n = 315) was observed during 2014–17. This increase is attributed to the heightened awareness among clinical microbiologists regarding misidentification of C. auris by commercial identification systems and their inclination for submitting Candida isolates for analysis by MALDI-TOF MS or sequencing. Five bloodstream isolates from two hospitals in Delhi were identified as C. auris in 2009 when the first report of C. auris was described in Japan. Overall, in the present study, 76.2% of C. auris isolates represented BSIs and 9% represented other invasive candidiasis cases. BSIs due to C. auris were mainly diagnosed in patients admitted to the ICU (61.7%, n = 165) followed by surgical (16.5%, n = 44), medical (14.4%, n = 39) and oncology/haematology wards (7.4%, n = 19). The median age of all BSI patients was 51 years. The two major super-specialty private sector hospitals in north India that contributed all BSI isolates to this study had a high isolation frequency of C. auris, i.e. 33% and 29% respectively followed by Candida tropicalis (19.5% and 13%) and Candida glabrata (13.3% and 9%). In contrast, a low isolation frequency (5%) was noted in a cancer hospital in Delhi where C. tropicalis (33%) followed by C. glabrata (23%) remained the most commonly isolated species. Similarly, a low isolation frequency (6.6%) of C. auris in a tertiary care hospital in south India was observed.

In general, C. auris was uniformly non-susceptible to fluconazole, which has also been reported in previous studies from Asia, Europe, South Africa, South America and North America.^2–24 However, regarding voriconazole, variable susceptibility patterns have been reported in previous studies in a limited number of isolates.^{2,4,7,9,15,16,20,21,27,41,42} A study from Colombia observed that only 4 of the 17 isolates were non-susceptible to voriconazole (MICs ≥2 mg/L) whereas all C. auris isolates reported from Spain and Venezuela, i.e. 8 and 18 isolates respectively had voriconazole MICs ≥2 mg/L.^15,20,21 In the present study, 15% (n = 52) of isolates were non-susceptible to voriconazole. Of these 61.5% (n = 32) of isolates were from BSIs. A more recent update of 77 US clinical cases of C. auris from seven states reported 45 bloodstream isolates and the remaining isolates were from urine (n = 11), respiratory tract (n = 8), bile fluid (4), wound (4), central venous catheter tip (2), bone (1), ear (1) and jejunal biopsy (1) specimens.⁴¹ The AFST data of the first 35 clinical isolates showed that 30 (86%) isolates were resistant to fluconazole (MIC >32 mg/L), 15 (43%) to amphotericin B (MIC ≥2 mg/L) and 1 (3%) to echinocandins (MIC >4 mg/L).⁴¹ In contrast to the high amphotericin B resistance rate of 43% (15 of 35) among US isolates, we observed a lower resistance rate (8%) among Indian C. auris. In addition, high amphotericin B resistance rates from Israel and Pakistan were reported in 5 of 6 and 7 of 18 C. auris isolates respectively using CLSI microbroth dilution.^2,19 However, no amphotericin B resistance (MIC range 0.5–1 mg/L) was observed in 8 Spanish, 10 South African and 12 UK isolates reported in previous studies.^2,15,41 Further, in an outbreak of C. auris in a London cardiothoracic centre between April 2015 and July 2016, 50 C. auris cases were recorded and their susceptibility data showed variable MICs of amphotericin B (0.5–2 mg/L).¹⁴

Regarding echinocandins, anidulafungin and micafungin exhibited potent activity (MIC ≤1 mg/L) against 98% of isolates and only 2% of isolates had high MICs (≥8 mg/L) of all echinocandins. Similarly, only 3% (1 of 35) of US isolates were resistant to echinocandins (MIC >4 mg/L).⁴³ Interestingly, a recent pharmacokinetic/pharmacodynamic (PK/PD) study in a murine invasive candidiasis model suggests that echinocandins are likely to be the most efficacious drug class for most C. auris isolates.⁴⁴ Further, micafungin demonstrated a potent cidal effect against almost all C. auris strains with MICs <4 mg/L and the drug exposure targets (AUC/MIC) were significantly lower than other Candida species.⁴⁴ Although species-specific clinical breakpoints have not yet been established for C. auris, findings in this PK/PD study suggest that susceptibility breakpoints are similar to PK/PD-based breakpoints for other Candida species and fluconazole and amphotericin B. This is based on the fact that the drug exposure associated with optimal outcome was similar for C. auris compared with previous Candida species studies.⁴⁴ Among three topical antifungals used, nystatin and terbinafine had no activity against C. auris (MIC >2 mg/L) suggesting that these drugs may not be useful for skin/oral eradication. Sertaconazole, a topical imidazole drug found to be effective for dermatophyte skin infections,⁴⁵ had variable MICs with 54% of isolates having low MICs. This antifungal may have potential for use in skin decontamination and topical management of sites such as venous cannula entry sites provided the isolate's susceptibility is obtained.

To investigate the resistance mechanism with respect to elevated azole MICs we sequenced the ERG11 gene of C. auris using recently published genomic data.^46,47,ERG11 sequences of C. auris exhibited amino acid substitutions that have been previously identified in resistant but not in WT C. albicans isolates.^29,37 The substitutions at positions Y132F and K143R responsible for azole resistance in C. albicans were observed in all 34 (77%) tested strains exhibiting fluconazole resistance (MICs 32 to ≥64 mg/L) whereas WT genotypes, i.e. without substitutions at these positions were observed in all isolates with low fluconazole MICs (1–2 mg/L) except one suggesting that these substitutions confer a phenotype of resistance to fluconazole. Further, we hypothesized that fluconazole resistance in C. auris isolates was at least partly due to the upregulation of the ERG11 gene, which encodes the azole target lanosterol demethylase.^48,49 However, the ERG11 expression in fluconazole-resistant C. auris strains was comparable with the susceptible strain in the absence of fluconazole. Notably, the expression of ERG11 increased 5- to 7-fold in the presence of fluconazole (5 h). It is plausible that in C. auris, stepwise accumulation of mutations and/or alterations resulting in increased efflux pump expression, necessary for stable high-level fluconazole resistance, plays a role that needs to be elucidated in future studies. It is pertinent to mention here that in azole-resistant C. albicans co-occurrence of ERG11 mutations and overexpression of efflux pumps have been recorded previously. Perea et al.²⁹ reported that in 55% of fluconazole-resistant C. albicans isolates, point mutations in the ERG11 gene (including Y132F and K143R) appear to be combined with the up-regulation of efflux pumps namely MDR1 and CDR1. Ben-Ami et al.¹⁹ demonstrated enhanced ABC-type efflux activity compared with that of C. glabrata that may also contribute to azole resistance in C. auris. The genome of C. auris demonstrates multiple ABC- and MDR-type transporter-encoding genes suggesting that drug efflux may also contribute to azole resistance in C. auris warranting in-depth investigations. Echinocandin resistance is widely linked to mutations in the FKS genes in other Candida species, and in the present study, we identified the S639F mutation in FKS1 in echinocandin-resistant isolates. These mutations are novel for C. auris and confer resistance to all three echinocandin drugs tested.

Finally, in the present study 25% (n = 88) and 13% (n = 45) of isolates were MDR (≥2 classes of drugs) and multi-azole-resistant respectively (Table 4). The most common multidrug combination was azole and 5-flucytosine in 14% (n = 48) of isolates followed by azole and amphotericin B in 7% (n = 26) and azole and echinocandin in 2% (n = 7) of isolates. To draw an analogy with other resistant Candida species such as acquired echinocandin-resistant Candida glabrata, C. auris exhibits higher fluconazole and echinocandin resistance rates, i.e. 90% and 2% compared with 6%–8% and 1% in C. glabrata.⁵⁰ In addition, remarkably high amphotericin B resistance rates, i.e. 8% in C. auris warrant attention as amphotericin B resistance is extremely rare in Candida species with the exception of Candida lusitaniae. Over the last 5 years C. auris has established its foothold as a multi-azole-resistant and MDR Candida species in many centres in India, which warrants serious efforts from healthcare personnel and national agencies to prevent its spread.

Table 4.

Distribution of C. auris isolates exhibiting MDR and cross-resistance to azoles

Resistance type	Number of isolates (%)
MDR
azoles + 5-FC	48 (14)
azoles + AMB	26 (7)
azoles + echinocandinsa	7 (2)
azoles + AMB + 5-FC	4
azoles + echinocandins + 5-FC	2
azoles + AMB + echinocandins	1
Multi-azole resistance
FLC + VRC	39 (11)
pan-azolea	14 (4)
FLC + ITC	6 (1.7)

AMB, amphotericin B (≥2 mg/L); 5-FC, 5-flucytosine (≥32 mg/L); FLU, fluconazole (≥32 mg/L); VRC (≥2 mg/L), voriconazole; ITC (≥2 mg/L), itraconazole (≥2 mg/L); ISA, isavuconazole (≥2 mg/L); POS, posaconazole (≥2 mg/L).

a

Includes only micafungin and anidulafungin (≥8 mg/L).

b

FLC + VRC + ISA + POS (n = 5), FLC + VRC + ISA (n = 3), FLC + ITC + VRC (n = 2), FLC + ITC + VRC + ISA (n = 2), FLC + ITC + VRC + ISA + POS (n = 1), FLC + ITC + ISA + POS (n = 1).

Acknowledgements

Part of this work has been presented at ASM MICROBE 2017, New Orleans, USA, poster 259 (Chowdhary et al. In vitro AFST by CLSI and EUCAST broth microdilution methods and elucidation of azole resistant molecular mechanism in 282 Indian Candida auris isolates).

Funding

This work was supported in part by a research grant from University Grants Commission Research Fellowship, India (F.2-15/2003 SA-I to C. S.); Council of Scientific & Industrial Research, India (F. No. 09/174(0068)/2014-EMR-I to A. S.); and Indian Council of Medical Research (ref. 5/3/3/26/2010-ECD-I to P. K. S.), Government of India, New Delhi, India.

Transparency declarations

J. F. M. has received grants from Astellas, Basilea, F2G and Merck; has been a consultant to Astellas, Basilea, Scynexis and Merck; and has received speaker’s fees from Astellas, Merck, United Medical, Teva and Gilead. D. S. P. received grants from the US National Institutes of Health and from Astellas. He serves on scientific advisory boards for, and receives grant support from, Astellas, Cidara, Amplyx, Scynexis and Matinas; and issued a US patent concerning echinocandin resistance. All other authors: none to declare. The authors alone are responsible for the content and writing of the paper.