Proposal for a unified nomenclature for target‐site mutations associated with resistance to fungicides

Abstract Evolved resistance to fungicides is a major problem limiting our ability to control agricultural, medical and veterinary pathogens and is frequently associated with substitutions in the amino acid sequence of the target protein. The convention for describing amino acid substitutions is to cite the wild‐type amino acid, the codon number and the new amino acid, using the one‐letter amino acid code. It has frequently been observed that orthologous amino acid mutations have been selected in different species by fungicides from the same mode of action class, but the amino acids have different numbers. These differences in numbering arise from the different lengths of the proteins in each species. The purpose of the present paper is to propose a system for unifying the labelling of amino acids in fungicide target proteins. To do this we have produced alignments between fungicide target proteins of relevant species fitted to a well‐studied ‘archetype’ species. Orthologous amino acids in all species are then assigned numerical ‘labels’ based on the position of the amino acid in the archetype protein. © 2016 The Authors. Pest Management Science published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.


BACKGROUND
Evolved resistance to fungicides is a major problem limiting our ability to control agricultural, medical and veterinary pathogens. 1,2 Research over the last 30 years has often defined the mechanism conferring reduced sensitivity to the fungicide. Many cases of resistance have been ascribed to the activity of efflux pumps 3 or to overexpression of target genes, 4 but the majority are due, at least partly, to substitutions (or indels) in the amino acid sequence of the target protein.
The convention for describing amino acid substitutions is to cite the wild-type amino acid, the codon number and the new amino acid, using the one-letter amino acid code (see Oliver and Hewitt, 5 box 6.1, p. 138 for a description of the system). A wellknown example is the alanine (A) for glycine (G) substitution in the cytochrome b gene at position 143 conferring resistance to strobilurin fungicides, referred to as G143A. 6 Further alterations can be amino acid deletions designated with a Δ and insertions with an 'ins'.
Target-site amino acid substitutions have been described for seven fungicide groups (named here according to the FRAC convention 7 ) and their target proteins. These are C3 and cytochrome b (Cytb) ( Table 1); G1 and two sterol C14-demethylases (paralogues Cyp51A and Cyp51B) (Tables 2  (Table 9); G3 and the 3-keto reductase (Erg27). Where more than one species has been studied, it has frequently been observed that orthologous amino acid mutations have been associated with resistance to fungicides with the same mode of action. In cases where the proteins are strongly conserved between species, the mutations have identical numbers. For example, the orthologous Cytb G143A mutation has been found in 22 species (Table 1). However in other cases orthologous mutations have different numbers -e.g. Cyp51B amino acid Y137 in Zymoseptoria tritici is orthologous to amino acids numbered from 131 to 145 in different species (Table 3). Similarly, SdhB amino acid H277 in Pyrenophora teres is orthologous to amino acids numbered from 249 to 278 (Table 5). These differences in numbers create unnecessary confusion and obscure the relationships between mutations in different species. Resistance caused by insertions in promoters and in efflux pumps have a much lower level of homology and so are not considered here.

THE PROPOSAL
The differences in numbering arise from the different lengths of the fungicide target protein in each species. The purpose of the present paper is to propose a system for unifying the labelling of mutant amino acids in fungicide target proteins. We propose that orthologous amino acids (i.e. ones presumed to be descended from the same amino acid in the common ancestor of these species) are given the same number in all species regardless of the actual position. The advantages of a unified system is that it would be easier to memorise common changes, to determine whether the changes were novel or were repetitions of what has already been seen in other species and to link changes to particular active ingredients. Orthologous mutations would be assigned the same 'mutation label'.
We distinguish between 'mutation labels', which refer to the orthology between proteins from different species, and 'amino wileyonlinelibrary.com/journal/ps S508T in PYRPBR S509T in ERYSGH acid numbering', which remains the order of the amino acids in each protein in each species. To avoid confusion, we propose that mutation labels should be italicised and mutation numbers should use regular lettering.
In several cases, amino acid substitutions have been found in the target protein but have not been definitively associated with any change in sensitivity either in vitro or in the field. It may be that the mutation underlying the amino acid substitution is a random event and of no obvious relevance. Definitively linking a mutation to a sensitivity change can be technically very demanding. If resistance to the same class of fungicide is linked to mutations affecting orthologous codons in different species, this is strong, if still circumstantial, evidence of the importance of the mutation. Unifying the mutant labelling system will make it much easier to identify important codon changes. This would assist the prioritisation of research aiming functionally to characterise mutations.

OPTIONS FOR PRODUCING THE ALIGNMENTS
We have produced a set of draft alignments of each target protein for which resistance to multiple species has been reported (Figs 1 to 9) and tables of putatively orthologous amino acids in other species where fungicide resistance has been reported (Tables 1 wileyonlinelibrary.com/journal/ps I269V in SEPTTR a The amino acid position number for this substitution is unknown as only a 176 bp fragment of SdhB gene has been sequenced in both sensitive and resistant isolates.   wileyonlinelibrary.com/journal/ps  to 9). The species included in these tables and alignments have been referred to by their European and Mediterranean Plant Protection Organisation (EPPO) codes 9 as listed in Table 10.
The alignments for b-tubulin and Cytb are essentially colinear in fungi studied to date, and hence there are no changes to be made to the current nomenclature. For the other genes, we have considered four possible methods to generate the alignment. The alignment could be: (1) fitted to the longest gene in the gene set; (2) fitted to a strict consensus alignment; (3) fitted to the gene from the species that is currently the most researched species for wileyonlinelibrary.com/journal/ps  wileyonlinelibrary.com/journal/ps  wileyonlinelibrary.com/journal/ps The aim is to create a set of alignments that would be stable into the foreseeable future and would invoke the least relabelling of mutations that have already been described and published. We favour method 3 (basing the alignment on the species with the most currently described resistance mutations), but also taking into account method 1 (using the longer gene) when alternative species are candidates. We propose that Cyp51A is fitted to ASPEFU (Aspergillus fumigatus), Cyp51B and Cytb are fitted to SEPTRI (Zymoseptoria tritici), b-tubulin to ASPEND (Aspergillus nidulans), the SDH proteins to PYRNTE (Pyrenophora teres), CesA3 to PHYTIN (Phytophthora infestans) and Os-1 to BOTCIN (Botrytis cinerea). For Erg27, mutations associated with resistance have currently been described only in BOTCIN, and thus we propose this species as the archetype. The alignments have been summarised and fungicide resistance associated mutations are given in the tables. By way of example, in Cyp51B the mutation Y136F in ERYSGH would be given the label Y137F. In CANDAL (Candida albicans) the orthologous amino acid is Y132 and has been mutated to both F and H. The Y132H mutation would therefore be given the label Y137H. V151 in SEPTRI is clearly demonstrated to be orthologous to I145F in PHAKPA (Phakopsora pachyrhizi). This mutation would be labelled I151F in PHAKPA and V151F in SEPTRI. The other proposed relabellings are listed in Tables 1 to 9.
By examining the species that have amino acid mutations with common labels, we can infer that positions 137, 148, 461, 476, 483 and 524 in Cyp51B are especially important in conferring resistance to triazole fungicides. This is consistent with numerous functional studies. 10,11 We expect that the alignments should assist the identification of key amino acids in target proteins of newer fungicide classes.

THE PROPOSAL IN PRACTICE
The system must also allow for mutations to be discovered in new species. The parameters used to make the alignments are described below and can be applied to an alignment between the new species and the archetype. We envisage regularly updating the alignments based on new published knowledge.
A potential problem with the system we propose might occur if an amino acid in a newly described mutant gene corresponded to a gap in the archetype protein's sequence. In such a case, the mutation could be labelled as X50.2Y if it concerned the second extra amino acid after number 50 in the archetype sequence. To our knowledge, no examples of mutations of such poorly conserved amino acids causing resistance have been described, but the possibility remains.
wileyonlinelibrary.com/journal/ps  wileyonlinelibrary.com/journal/ps wileyonlinelibrary.com/journal/ps We hope that future studies will refer to the archetype by indicating that the mutation X123Y in the target protein associated with resistance corresponds to the archetype X145Y and refer to this paper or a related web page for support.
We suggest that other target genes from medically important fungi (e.g. the FKS1/2 genes that are targets of Echinocandins) and from herbicide-and insecticide-resistant weeds and insects might also benefit from this approach.
We commend this scheme to the community and seek comment and support. And we urge journal editors to encourage authors to use this new system.

NOTE ON THE ALIGNMENTS
Amino acid sequences were downloaded from NCBI GenBank and annotated with reported amino acid substitutions 8,12 -15 using Geneious 6.1.8 software (Biomatters). Alignments of sequences were generated using the ClustalW 16 algorithm with Blosum scoring matrix, gap opening penalty 10, gap extension penalty 0.5 and free end gaps.
The alignments are available as .doc files and as fasta files in the supporting information.

SUPPORTING INFORMATION
Supporting information may be found in the online version of this article.