第一部分 须癣毛癣菌复合体的分类

目的：探讨须癣毛癣菌复合体的分类及其进化历程。

方法：本实验共收集了182株菌，分别来自欧洲（n=63）、印度（n=46）、中国（n=49）、日本（n=2）、澳大利亚（n=21）和美国（n=1），其中162株来自病人，17株来自动物，3株来自土壤。采用ITS构建系统进化树对所有菌株进行分型，再从每个分枝中挑选113株菌进行Tef1-α、HMG测序并构建系统进化树。采用R语言建立纠缠树，并结合多基因位点比较ITS、 Tef1-α、HMG树的稳定性。采用FastANI计算3个来自不同基因型的菌的平均核苷酸相似度。最后分析来自不同基因型菌株的形态学及生理学特征，其中包括菌落的形态、Tween 80试验、毛发穿孔试验、尿素酶分解试验和角蛋白降解试验。

结果：基于ITS系统进化树显示，各基因型的突变位点较少，支持率较低，而HMG基因系统进化树的支持率较高，且纠缠树显示ITS、Tef1-α和HMG基因的系统进化树的一致性也较高，但全基因组显示3个来自不同基因型菌株的平均核苷酸相似度大于95%。生理学实验显示趾间毛癣菌和须癣毛癣菌的尿素酶分解、脂肪降解及毛发穿孔能力强于印度毛癣菌。而在角蛋白降解方面，趾间毛癣菌降解角蛋白的能力强于须癣毛癣菌和印度毛癣菌。

结论：在分子学上，须癣毛癣菌和趾间毛癣菌难以区分，但印度毛癣菌与须癣毛癣菌和趾间毛癣菌可以采用HMG基因区分。基于分子学的分类，可以观察到须癣毛癣菌、趾间毛癣菌和印度毛癣菌的形态学及生理学特征的差异具有统计学意义。须癣毛癣菌可能是祖先，具有亲土性，而趾间毛癣菌和印度毛癣菌可能是亲人型的克隆分支。

第二部分 须癣毛癣菌复合体体外药物敏感试验及特比萘芬耐药机制研究

目的：探讨除印度外其他国家耐药情况、特比萘芬耐药的分子机制。

方法：选取了来自印度、中国、澳大利亚、德国和荷兰共计135株菌，基于上述新的分类标准，分为35株须癣毛癣菌（Trichophyton. mentagrophytes s. str.）、64株印度毛癣菌（Trichophyton. indotineae）和36株趾间毛癣菌（Trichophyton. interdigitale s. str.），选取了皮肤科常用的8种药物（氟康唑、灰黄霉素、伊曲康唑、酮康唑、咪康唑、萘替芬、舍他康唑、特比萘芬）进行药物敏感试验，筛选耐药菌株，对敏感和耐药的菌株进行角鲨烯环氧化物酶（SQLE）基因测序，发现突变位点，同时在解除药物压力作用的情况下，评价耐药的稳定性。

结果：通过药物敏感试验共检测37株耐药的印度毛癣菌，对所有耐药的菌株和15株敏感的印度毛癣菌、10株须癣毛癣菌、10株趾间毛癣菌进行SQLE测序，发现34株最小抑菌浓度（MIC）>16 mg/L的菌株（53%），均存在Phe³⁹⁷Leu位点的氨基酸替换（1189 T>C,1191 C>A）,或合并有Ala⁴⁴⁸Thr氨基酸的替换（1342 G>A）。其中2株(3%)MIC=0.5 mg/L的印度毛癣菌存在Phe415Val和Leu393Ser替换。除了耐药菌株有突变位点，在敏感的须癣毛癣菌中发现两个新的氨基酸替换位点（Lys276Asn和Leu419Phe）。通过分析SQLE不同氨基酸替换的菌株，发现存在双重突变的菌株(Phe377Leu/Ala448Thr)，其氟康唑的MIC高于其他单个氨基酸替换或者野生型菌株。耐药菌株和敏感菌株经过10次无药物压力培养基传代，前后药敏结果一致，提示耐药相对稳定。

结论：特比萘芬MIC值高的菌株仅局限于印度毛癣菌，其似乎不能通过基因交换在须癣毛癣菌复合体之间交换，且这种耐药相对比较稳定。

First chapter：Taxonomy of the Trichophyton mentagrophytes / T. interdigitale species complex

Objective: In order to define species limits, a taxonomic study was undertaken combining molecular, morphological, and physiological characteristics as evidence of classification.

Methods: The study included 182 strains in total, of which 162 strains originated from patients, 17 from animals and 3 from soil. Geographically, strains were obtained from Europe (n=63), India (n=46), China (n=49), Japan (n=2), Australia (n=21) and U.S.A. (n=1). Multilocus data comparison using tanglegrams, combining ITS, Tef1-α, and HMGloci, was performed with random selection of 113 strains representing all genotypes. FastANI was used to compute pairwise Average Nucleotide Identity (ANI) values among all genomes of three genotypes. Differences between groups in anonymous markers need to be supported by phenotypic characteristics of morphology and physiology including urea hydrolysis, lipolytic activity, hair perforation, keratin degradation were performed.

Results: Bootstrap support of the branches remained low due to the small number of mutations basing on ITS phylogenetic tree, however, bootstrap support of the branches high in HMG gene. Multilocus data comparison using tanglegrams, combining ITS, Tef1-α and HMG loci, it is suggested that phylogenetic trees own a good consistency. The FastANI values of all pairs exceeded 95%, which is generally is taken to indicate conspecificity. Physiologically, T.interdigitale and T. mentagrophytes are similar, and different from T. indotineae. Urea hydrolysis was found mostly positive in T. interdigitale and T. mentagrophytes, while most strains of T. indotineae strains were weakly positive or negative. The lipolytic abilities and hair perforation of T. mentagrophytes and T. interdigitale were very similar, and were higher than those of T. indotineae. Keratin degradation was significantly larger in T. interdigitale than in T.mentagrophytes and T. indotineae (p < 0.01).

Conclusion：Molecular characteristics show that T. mentagrophytes s. str. and T. interdigitale s. str. can be distinguished with difficulty from each other, but are unambiguously different from the Indian genotype, T. indotineae by sequences of the HMG gene. The entities were confirmed by multilocus analysis using tanglegrams. Phenotypic characters of morphology and physiology are not diagnostic, but statistically significant differences are observed between the molecular siblings. Trichophyton mentagrophytes represents the ancestral, homothallic cloud of genotypes with a probable geophilic lifestyle, while T. indotineae and T. interdigitale behave as anthropophilic, clonal offshoots.

Second Chapter: Antifungal susceptibility and mutations in the squalene epoxidase gene in dermatophytes of the Trichophyton mentagrophytes species complex

Objective: To investigate the molecular background of the squalene epoxidase (SQLE) gene in order to understand the risk of emergence and spread of multi-resistance in dermatophytes.

Methods: Antifungal susceptibility for fluconazole, griseofulvin, itraconazole, ketoconazole, miconazole, naftifine, sertaconazole, and terbinafine was tested in 135 isolates from India, China, Australia, Germany and The Netherlands. Based on the latest taxonomic insights, strains were identified as three species: T. mentagrophytes s. str. (n=35), T. indotineae (n=64 representing the Indian clone) and T. interdigitale s. str. (n=36). The terbinafine resistance strains were screened and detected the mutation of the squalene epoxidase (SQLE).13 strains of terbinafine resistance were choosed for evaluate the stability by serial transfer on drug-free medium.

Results: High minimum inhibitory concentrations (MICs) of terbinafine (>16 mg/L) were found in 34 (53%) T. indotineae isolates. These isolates showed an amino acid substitution in the 397th position of the SQLE gene. Elevated MICs of terbinafine (0.5 mg/L) were noted in 2 (3%) T. indotineae isolates; these isolates lead to Phe415Val and Leu393Ser of the SQLE gene. Stability of the effect of the mutations was proven by serial transfer on drug-free medium. Substitutions of Lys276Asn and Leu419Phe were found in susceptible T. mentagrophytes strains. The double mutant Phe377Leu / Ala448Thr showed higher MIC values for triazoles.

Conclusions: High MICs of terbinafine are as yet limited to T. indotineae, and are unlikely to be distributed through the T. mentagrophytes species complex by genetic exchange.