Regional Differences in Antifungal Susceptibility of the Prevalent Dermatophyte Trichophyton rubrum

In vitro susceptibility testing for Trichophyton rubrum has shown resistance to terbinafine, azoles and amorolfine, locally, but epidemiological cutoffs are not available. In order to assess the appropriateness of current first-line antifungal treatment for T. rubrum in China, we characterized antifungal susceptibility patterns of Chinese T. rubrum strains to nine antifungals and also described the upper limits of wild-type (WT) minimal inhibitory concentrations (MIC) (UL-WT) based on our study and another six studies published during the last decades. Sixty-two clinical isolates originating from seven provinces in China were identified as T. rubrum sensu stricto; all Chinese strains showed low MICs to eight out of nine antifungal drugs. Terbinafine (TBF) showed the lowest MICs of all antifungal classes tested in both the Chinese and global groups, with a 97.5% UL-WT MIC-value of 0.03 mg/L. No non-WT isolates were observed for TBF in China, but were reported in 18.5% of the global group. Our study indicated that TBF was still the most active drug for Chinese T. rubrum isolates, and all strains were within the WT-population. TBF therefore remains recommended for primary therapy to dermatophytosis caused by T. rubrum in China now, but regular surveillance of dermatophytes and antifungal susceptibility is recommended. Supplementary Information The online version of this article (10.1007/s11046-020-00515-z) contains supplementary material, which is available to authorized users.


Introduction
Trichophyton rubrum has been among the most prevalent dermatophytes causing tinea pedis and tinea unguium since the early twenty-first century [1]. The global predominance of T. rubrum suggests that this species has a significantly higher capacity of transmission than other anthropophilic dermatophytes [2]. Numerous authors have noted, as confirmed by wholegenome sequencing [3], that T. rubrum is clonal with a highly conserved gene content, low levels of variation, and little evidence of recombination. On the Indian subcontinent, the species is still common [4][5][6], but seems gradually to be replaced by members of the T. mentagrophytes group [4,5,7], thus having a similar fate as the classical, disappearing dermatophytes Epidermophyton floccosum and Microsporum audouinii [1]. One of the main characteristics promoting global spread is the low virulence of T. rubrum [8]. Its mild, hardly noticeable cutaneous infections with transmission via skin scales released from mild hyperkeratosis does not interfere with transmissionenhancing interaction of human hosts. Infections only are more serious in CARD9-deficient patients or in those with dysfunctional cellular immunity, e.g., with cirrhosis, AIDS, hematological malignancies or solid organ transplants [9][10][11].
Acquired resistance, as observed in numerous fungi, is however a concern. T. rubrum is regarded to have limited capacity to develop resistance to terbinafine even after prolonged exposure [12], but it has been proven that T. rubrum can develop resistance to azoles, amorolfine and terbinafine after prolonged exposure to sub-inhibitory concentrations of these drugs [13][14][15][16][17][18]. Antifungal drug resistance may contribute to treatment failure and lead to persistent and chronic infections [13,14]. The first report on T. rubrum exhibiting resistance to terbinafine was published in 2003 [15], followed by reports from the Americas, Europe and Asia [16][17][18]. Ebert et al. [19] indicated that resistance of T. rubrum for terbinafine was as high as 44%, although lower than that of T. mentagrophytes group in India. This study also showed that the systemic antifungal drug terbinafine, first-line drug for dermatophytosis, had lost its in vitro activity in most parts in India [19].
It has been speculated that the resistance in dermatophytes might be related to drug exposure [5,19]. The management of over-the-counter drug use in China is similar to that in India, but resistance of T. rubrum has not systematically been reported from China. No clinical break points (CBP) or epidemiological cutoffs (ECV) are available to guide antifungal treatment and classify resistance in T. rubrum. The Guizhou Province, located in the southwest of China, is humid and mountainous, with poor transportation and backward economy. Local people and even doctors in basic hospitals pay insufficient attention to dermatophytosis and have poor awareness of diagnosis and treatment, whether it will lead to difference in drug susceptibility of T. rubrum in Guizhou region? Based on that, we compared clinical isolates from Guizhou Province with six regions in China in terms of genomic diversity of strains and in vitro susceptibility to a panel of antifungal drugs. In order to investigate potential trends in drug resistance of T. rubrum, we also reviewed available MIC values from published literature and describe wild-type (WT) MIC distributions of nine drugs for T. rubrum according to the criteria used by CLSI. these, five were isolated from tinea pedis, and six others from tinea cruris, tinea corporis, tinea capitis, tinea faciei and tinea manuum, respectively; the remaining 20 strains were from tinea unguium. All strains were isolated during the period 2016-2019.

Identification
Identification of isolates was done by phenotype [20] and confirmed by rDNA internal transcribed spacer regions (ITS) sequencing. Briefly, isolates were subcultured on Sabouraud's Glucose Agar (SGA, homemade) incubated at 28°C for one week. DNA extraction was by the cetyltrimethylammonium bromide (CTAB) method [21]. ITS of the rDNA operon was amplified with primers ITS5 and ITS4 according to in Jiang et al. [21,22]. PCR products were sequenced by TSINGKE Biological Technology (Beijing, China). GenBank accession numbers for new sequences are given in supplementary were prepared in dimethyl sulfoxide (DMSO) at a concentration of 2 mg/mL (except FCZ, which was dissolved in distilled water at final concentration 102.4 mg/mL). Drug stock solutions were diluted in RPMI 1640 medium buffered with 3-N-(morpholino)propanesulfonic acid (MOPS), in twice the final concentration followed by addition of equal volumes of the pre-adjusted inoculum of fungal isolates, in 96-well microtiter plates. Final concentrations of the antifungals tested ranged from 0.125 to 64 lg/mL for FCZ, 0.0078 to 4 lg/mL for KTZ, 0.004 to 2 lg/mL for BFZ, 0.002 to 1 lg/mL for ITZ, MCZ and AMF, 0.00025 to 0.125 lg/mL for TBF, 0.001 to 0.5 lg/mL for NAF and LLCZ. Drug plates were stored at -70°C.

Inoculum and Quantification
Strains were subcultured from primary SGA plates to potato dextrose agar (PDA) to induce conidiation. Plates were incubated at 28°C for 9-14 days. Conidia were collected by gently flushing 5 mL phosphate buffer saline (pH = 7.4) on colonies and aspirating the suspension into a sterile collection tube. Suspensions were counted on a hemocytometer and diluted in RPMI 1640 to the desired concentration of 1 9 10 3 * 3 9 10 3 CFU/mL.

Reading of Results
Microdilution plates were incubated at 35°C and visually read after 5-7 days. Endpoint MICs for azoles and TBF were considered when prominent inhibition (approximately 80%) was reached compared to the control wells, while for NAF and AMF 100% growth inhibition was required. Ranges and geometric means (GMs) of the MICs were determined for each group and drug after 6 days. If no growth was observed or growth was inadequate, incubation was extended to 7 days. Candida parapsilosis (ATCC 22019) and T. mentagrophytes (ATCC MYA 4439) were included as quality control strains. All experiments were performed using two independent replicates on different days.

Statistics
Statistical analysis was performed by Mann-Whitney U and Kruskal-Wallis tests (Student's t test) using SPSS software v21. One-way analysis of variance (ANOVA) was used to compare the geometric mean (GM) MICs between the groups and within distinct geographic areas. P values of \ 0.05 were considered significant. ECV values were computed by the Microsoft Excel spreadsheet calculator ECOFFinder XL 2010 v2.1 (https://clsi.org/), which follows a methodology established by Turnidge et al. [26] Literature Search

Identification
T. rubrum is reliably identified by the rDNA ITS barcoding marker [27,28]. Phylogenetic analysis resolves species boundaries between the closely related siblings in the T. rubrum complex (Fig. 1). In the alignment (555 bp including gaps), three groups differing by single nucleotide polymorphisms (SNPs) were revealed. Group 1 (T. rubrum sensu stricto) comprised two clusters matching with Haplotypes 5 and 6 [27]. H5 comprised 134 identical isolates and included the neotype strain CBS 392.58 (Group 1A). All strains from our study clustered into this subclade and revealed 100% sequence similarity. Five strains (Group 1B), one of which was from China, deviated by a single SNP matching haplotype H6 [27]. Two remaining clusters in the T. rubrum complex contained reference strains of T. violaceum (Group 2, 7 bp distance) and T. soudanense (Group 3, 4 bp distance).
The nine drugs belong to four categories, and MIC values were compared in the 62 T. rubrum strains accordingly. Significant differences between antifungal classes were observed (Fig. 2b). Azoles were divided into a topical and a systemic category, respectively. The systemic azoles included KTZ, ITZ and FCZ, while the topical azoles included LLCZ, BFZ and MCZ. MIC values of the systemic azoles were significantly higher than those of the topically administered compounds (GM 0.2236 vs. 0.03245 lg/mL; P \ 0.001). Allylamines, including TBF and NAF, which can be used systemically or topically, were associated with lowest MICs (P \ 0.001). The morpholine derivatives included the single drug AMF, which is used only topically. AMF was active against all T. rubrum strains with low MICs, GMs being just slightly higher than that of the topical azole category (GM: 0.05271 vs. 0.03245 lg/ mL, P \ 0.05).
For a third comparison of MIC values, we separated the T. rubrum strains geographically, i.e., a Guizhou group and those from remaining Chinese provinces. There was no difference in drug susceptibility between the two groups for the eight drugs; only for AMF, the 31 strains from Guizhou showed slightly higher MIC values than those from remaining China (GM MIC values: 0.059 vs. 0.047 lg/mL, P \ 0.05) (Fig. 2C).
In addition to our own data, six published studies presented the distribution of MICs values for nine antifungal drugs against T. rubrum, including two studies form China, three studies from India and one from Iran. We compared to the upper limit of wildtype MIC (UL-WT) distribution of T. rubrum between China and the global data set ( Table 3). The number of strains tested exceeded 100 for KTZ, ITZ and TBF. Global data were not available for BFZ, MCZ, NAF and AMF. Both the 95% and 97.5% MICs were calculated to determine the wild-type MIC (UL-WT). For the antifungals which showed a multimodal, nonsymmetrical or truncated MIC frequency distribution,     UL-WT was determined for seven studies (13,14,20,23,24,25,26) Several studies (5,6,15,23) indicated MIC50 values rather than GM MIC values; one study (22) indicated MIC90 values ''-'': Data was not available

Discussion
Dermatophyte infections have received renewed interest during the last five years because of the emergence of recalcitrant, highly virulent species in South Asia [19,51]. The causative species was recently described as T. indotineae (T. mentagrophytes group) [52]. Given the rapid spread of dermatophytes, the potential replacement of mild T. rubrum by virulent T.
mentagrophytes group is a significant public health risk. In China, T. rubrum is still the predominant species among dermatophytes, similar to previously published data [1,3]. Antifungal resistance has also been reported in T. rubrum [16][17][18][19] and thus, a potential public health problem is apparent. T. rubrum is identified phenotypically in the routine laboratory. For confirmation of identity, the rDNA internal transcribed spacer (ITS) barcoding gene is known to be sufficient for the distinction of siblings within the T. rubrum complex [27,28]. In addition to 62 strains from our study, a global set of 71 sequences was found to be identical to T. rubrum haplotype H5 containing the type strain (CBS 392.58) [27]. The 100% match with phenotypic identification validates the IDs of older publications where no sequencing data are available. Five of the sequences, among which one from China, showed one SNP distance and clustered with T. rubrum H6 [27]. The ITS data support the earlier view of a global, largely clonal population structure with low levels of variation and no evidence of recombination [3]. The ITS marker is not epidemiologically associated with resistance, since 21 T. rubrum strains with higher MICs recorded in previous studies belonged to the same population [18,19,53], as confirmed in our study (Fig. 1).
Judging from proposed breakpoints for dermatophytes [49,53], i.e., [ 2 lg/mL for FCZ and [ 1 lg/ mL for ITZ, KTZ and TBF, the tested T. rubrum strains from China should be regarded as in vitro susceptible to all antifungal drugs. FCZ had a GM of  We therefore compared the MIC values in our study to similar studies in China as well as from other countries. Most of the English-language studies came from India (9/26) and Iran (6/26) ( Table 2). Six additional studies from China were included. Although all studies were performed by following the CLSI M38-A2 protocol [29], small differences were observed in the protocols that we used, particularly in the incubation temperature: either 28°C (six studies) or micro-broth dilution requiring incubation at 33-35°C. It has been shown that 33°C was the most suitable growth temperature for T. rubrum in terms of dry weight of mycelium and colony diameter [55]. The prevalence of T. rubrum involving protruding body parts is associated with an optimum growth below 35°C [56]. The slow growth of the fungus requires reading of results after 5-7 days rather than after 72-96 h [57].
The most frequently studied drugs during the last decade globally are ITZ, FCZ, KTZ and TBF.
Although the MIC range of each study is relatively large, the GM value of most antifungal drugs is fixed in a certain range, indicating repeatability and reliability of the research results, which lays a foundation for the interpretation of ECV of T. rubrum for different antifungal drugs. Among the four drugs, the GM values of FCZ are scattered between studies (Fig. 3e), and MIC values are relatively high. In contrast, MIC values against TBF and ITZ were consistent, and GM values in different studies were basically at the same level (Fig. 3f, g). However, two studies on TBF from India published in 2019, which indicated that T. rubrum isolates that are highly resistant to TBF, diverge significantly from remaining studies (Fig. 3g).
ECV analysis was performed simultaneously. We combined global distributions of MIC values from seven studies to distinguish between WT and non-WT populations and calculated upper limit of WT MIC (UL-WT) or epidemiological breakpoints. Since our study still could not totally fulfil the criteria of evaluation of ECV according to CLSI guidelines [58], we propose the UL-WT instead of ECV for T. rubrum of Chinese origin (Table 3).
FCZ was shown to have poor activity against T. rubrum, having the highest 97.5% UL-WT (8 lg/mL) agents. Different from T. mentagrophytes / T. interdigitale complex [59], the 97.5% UL-WT value of global T. rubrum for TBF was very low, i.e., 0.03 lg/ mL. In contrast to China, the global group had a high percentage of isolates above UL-WT (global: 18.5%, China: 0), which indicated that TBF may be still considered as the first choice for treatment in China. Naftifine, another allylamine drug, also has good activity against T. rubrum, with a low 97.5% UL-WT (0.06 lg/mL). In addition to these two drugs, T. rubrum and T. mentagrophytes/T. interdigitale showed similar UL-WT values for the azole and morpholines [59]. LLCZ, BFZ and AMF are good options to treat T. rubrum infection due to low 97.5% UL-WT with a low percentage of non-WT isolates. However, the amount of strains tested for the above three drugs is small, and more verification is required. Based on the classification and comparison of antifungal agents in this study, we preliminarily determined that topical TBF and NAF should be still recommended as first-line therapy against superficial skin infection caused by T. rubrum in China. Antifungal creams should remain without steroids, and an adequate treatment period can be estimated at 2 weeks after the rash disappears [54], avoiding potential development of resistance. The azole drugs KTZ and MCZ, and particularly FCZ, are not recommended.
The limitation of this study is that the number of isolates is small; not every drug included MICs from the required minimum 100 unrelated isolates. However, our study preliminarily described and explored the UL-WT of T. rubrum, understood the trend of its sensitivity to a variety of antifungal agents and recommended first-line treatment for skin infection of this species in China. Regular surveillance of dermatophytes and antifungal susceptibility is recommended, since susceptibility profiles in general seem to be prone to change. At the same time, because of its highly conserved gene content, global prevalence and low virulence, T. rubrum may be a good choice as a research model for the mechanism of dermatophytes resistance.
Acknowledgements The authors thank Xu-Dong Wang from Institute for Molecules and Materials of Radboud University for his important contributions to the Graphic design. Special thanks to Weiwei Wu of Hainan Dermatology Hospital for sharing the original data.
Author Contributions YJ designed this project, carried out strains collection, phylogenetic tree construction, participated in the data analysis and drafted the manuscript. YS participated in the data analysis and revised the manuscript. WL performed DNA extraction and sequencing, carried out the literature search, carried out AFST tests. ZS and BZ carried out the statistical analysis. SA and AMS performed the antifungal susceptibility data analysis, contributed to the discussion and revision of the manuscript. ZW checked the AFST tests and provided the support of AFST methods. YJ, YS, RL, PEV and SdH participated in the design of the study, statistical analysis and manuscript revision and review. All authors read and approved the manuscript.

Compliance with Ethical Standards
Competing interest The authors declare no conflicts of interest.
Ethical Approval The authors confirm that this material is original and has not been published in whole or in part elsewhere and that the manuscript is not currently being considered for publication in another journal and that all authors have been personally and actively involved in substantive work leading to the manuscript and will hold themselves jointly and individually responsible for its content.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.