Mitochondrially encoded acaricide resistance in spider mites

Spider mites (family Tetranychidae) are one of the most economically important phytophagous mite families in several agricultural crops worldwide. Their remarkable biological properties (high reproductive potential, extremely short life cycle, arrhenotoky) combined with the necessary frequent applications of chemical acaricides render the spider mites rapidly resistant to all chemical compounds used in commercial crop protection in the past and today. This is also the case for the relatively recently developed and commercialised carbazate acaricide bifenazate. In this study we have tried to unravel the mechanisms that underlie the resistance against bifenazate at the toxicological, genetic, molecular and biochemical level. Genetic crossing studies revealed a maternal inheritance of the bifenazate resistance trait both in the twospotted spider mite T. urticae and the citrus red mite P. citri. Amplifying and sequencing complete cytb sequences of bifenazate resistant and susceptible strains of T. urticae and P. citri revealed point mutations in 5 sites of the highly conserved cytb cd1- and ef-helices, important both structurally and functionally for the formation of the mitochondrially encoded cytb Qo pocket during the catalysis process of mitochondrial respiration. In T. urticae, a combination of at least 2 cd1-helix amino acid substitutions (G126S + I136T or G126S + S141F) seemed to be necessary to confer a pronounced resistant phenotype, whereas only a single amino acid substitution in the ef-helix resulted in an equally highly resistant phenotype (LC50 > 10,000 ppm, RR > 2780). In P. citri, the fixed cd1-helix amino acid substitutions combination G126S + A133T appeared to be necessary to confer high resistance (LC50 > 500 ppm, RR > 33). The observed parallel bifenazate resistance evolution in T. urticae and P. citri, its maternal inheritance and the presence of altered cytb uniquely found in resistant mite strains initially suggested the mitochondrially encoded cytb Qo as bifenazate’s potential target site. In a survey to investigate the potential role of a mitochondrial bottleneck in the development of bifenazate resistance, it appeared a low resistant T. urticae strain, consisting of heteroplasmic individuals, transmitted its resistant and susceptible haplotypes to progeny in highly variable ratios consistent with a sampling bottleneck of about 190 mtDNA copies. Indeed, the influence of random drift on changes in mutation frequency between generations could be demonstrated most perspicuously by examining the relationship between mutation frequency and expression of a bifenazate resistance phenotype. Bifenazate pressurising favoured heteroplasmic individuals carrying at least 60% mutated mtDNAs for surviving the field application rate of 100 ppm. Consequently, the role of mitochondrial heteroplasmy in the development of bifenazate resistant phenotypes was investigated and it was illustrated how heteroplasmy in an early developmental stage together with genetic bottlenecking and sustained selection can drive populations rapidly to fixation of the resistant trait, as was found in most field collected resistant strains of T. urticae and P. citri. Moreover, these results confirmed indirectly cytb Qo being the target site for bifenazate in spider mites. Cross-resistance between bifenazate and acequinocyl, a known complex III inhibiting pro-acaricide at the cytb Qo site, caused by mutations in the mitochondrial cytb gene of T. urticae and P. citri was detected in bifenazate resistant field and laboratory selected strains and provides further evidence of the interaction of bifenazate with the mitochondrial cytb Qo. The cd1-helix amino acid substitutions combination G126S + S141F or the single amino acid substitution P262T in T. urticae, and the cd1-helix amino acid substitutions combination G126S + A133T in P. citri conferred high cross-resistance to acequinocyl suggesting the active metabolite of bifenazate acts as a cytb Qo inhibitor (QoI). Although not all documented bifenazate resistance linked amino acid substitutions resulted in acequinocyl cross-resistance, bifenazate and acequinocyl should be carefully used in control programs in order to avoid resistance building up. Acequinocyl and bifenazate should not be used in combination or direct alternation, as they appear to be acaricides with a similar mode of action, although currently classified in different IRAC mode of action groups. Resistance to another QoI acaricide, fluacrypyrim, was investigated in several field and laboratory selected spider mite strains. Its resistance was highly dominant and monogenic, and was unexpectedly not linked with the cytb genotype. Moreover, no evidence of increased metabolism of fluacrypyrim by well known detoxification routes was found leading to the conclusion that resistance to fluacrypyrim is due to an altered target site protein, encoded by a nuclear gene. Fluacrypyrim, only registered in Japan today, is not a likely alternative for European crop protection, since several strains from different origin already exhibited high resistance, without the chemical ever being used. In order to assess the current bifenazate/acequinocyl resistance status in field, several field populations of different spider mite species were sampled and screened for resistance mechanisms. Resistance to bifenazate caused by mutations in the mitochondrial cytb gene occurred in T. urticae field strains collected from Belgium and the Netherlands. All resistant strains harboured at least one fixed amino acid substitution or a combination of at least 2 non fixed amino acid substitutions on population level. All susceptible strains had the wild type cytb or harboured a single non fixed G126S amino acid substitution. Since resistance amino acid substitutions were found in 14 out of 20 field strains, bifenazate should be carefully used in control programs in order to avoid further resistance building up. Resistance in the field might be avoided when carefully applying bifenazate on a rare basis, as stability tests showed that resistant strains, harbouring initially non fixed amino acid substitutions on population level, converted into susceptible populations in absence of pressurising. The recently introduced QoI acaricide acequinocyl caused cross-resistance only in one field strain harbouring a single fixed P262T amino acid substitution. Field populations of T. urticae, collected from Australian pear orchards, and the congeneric T. pacificus, collected from Californian vineyards, were reported to cause failure of control by bifenazate. Surprisingly, the bifenazate phenotypical resistance could not be linked with previously described cytb genotypes. None of the cytb point mutations resulting in single amino acid substitutions, assumed to confer resistance to bifenazate in T. urticae or P. citri, were present in bifenazate resistant T. urticae or T. pacificus. Instead, it is likely another resistance mechanism, such as metabolic sequestration and detoxification by specific enzymes, is responsible for bifenazate resistance in these strains of T. urticae and T. pacificus, which may also explain the rather low resistance ratios for T. pacificus compared to those for T. urticae and P. citri bifenazate resistant strains previously described. Additional (biochemical) studies, including cross-resistance research, are needed to reveal the basis of bifenazate resistance in T. pacificus. In this study, first biochemical evidence is presented for a physical interaction between [14C]bifenazate and mitochondrial complex III of T. urticae by using two-dimensional BN/SDS-PAGE separating the different mitochondrial complexes and their subunits. When BN-PAGE gels for bifenazate susceptible mites were examined for [14C] activity, it was clear that two distinct bands of the pattern were labeled. The specificity of this labeling was further confirmed by running identical samples derived from bifenazate resistant mites. Although identical BN-PAGE band patterns were noticed, none of the bands displayed any [14C] activity. In the light of previous evidence, this suggests that amino acid substitutions in the cytb of bifenazate resistant mites interact with binding of the active bifenazate metabolite. When subunits of native complexes were separated by SDS-PAGE, individual proteins spots supported the assignment of complexes to BN-PAGE bands and further confirmed that the radiolabel is retained by mitochondrial complex III. We further showed in TLC assays that bifenazate is differentially metabolized in insects and mites, possibly underlying the high selectivity of the compound. Overall, in spider mites most of the metabolites were found in the water soluble fraction, of which mainly bifenazate-diazene was identified, together with a metabolite of high mobility, that was also abundantly present in mitochondrial extracts. This metabolite was not found in samples of the beetle L. decemlineata, a beetle not effected by high doses of bifenazate, and - although speculatively - might be the bifenazate derived active metabolite in spider mites. In addition, major metabolites in beetles were absent in the spider mite samples. In contrast to the proposed mitochondrial mode of action, it was recently shown that bifenazate can act as a synergist or allosteric modulator of functionally expressed T. urticae GABA receptor homologues, suggesting a neurotoxic mode of action (Hiragaki et al. 2012). We therefore sequenced and determined the expression level of all three T. urticae GABA receptors in bifenazate susceptible and highly resistant strains. We found no mutations linked with resistance in these receptors, and their expression level was not overall changed between strains, contradicting the proposed neurotoxic mode of action. An additional molecular screening of several bifenazate resistant and susceptible T. urticae strains did not reveal any data to support the involvement of the nuclearly encoded proteins ISP and cytc1 in bifenazate resistance and thus indirectly confirm that the cytb Qo is the only target core protein of the mitochondrial complex III for bifenazate. We propose that resistance to bifenazate is most likely procured by the substitution of one or more cytb Qo residues and we consequently strongly argue for the Qo site of the mitochondrially encoded cytb being the target site, taking into account the findings of this study: (1) the observed parallel evolution of bifenazate resistance in T. urticae and the closely related P. citri and its maternal inheritance, (2) the involvement of a mitochondrial bottleneck influencing the resistance pattern as was shown for a heteroplasmic low resistant T. urticae strain, (3) the cross-resistance to the known QoI acaricidal acequinocyl both in T. urticae and P. citri, (4) the presence of altered cytb uniquely found in resistant mite strains, (5) the biochemical evidence for a physical interaction between [14C]bifenazate and mitochondrial complex III of T. urticae, (6) the unlikelihood of involvement of genes encoding GABA receptors together with the lack of amino acid substitutions in the other complex III core proteins ISP and cytc1. Moreover, bifenazate was shown to be differentially metabolised in insects and mites, possibly explaining the highly selective action as a pro-acaricide. In conclusion, when applied to spider mites, bifenazate appears to act as a cytb Qo inhibitor, and its activity can be severely compromised by amino acid substitutions at the binding site. The precise molecular mitochondrial mode of action of bifenazate - or its active metabolite - via inhibition of complex III at the cytochrome b Qo site has yet to be elucidated. Interestingly, bifenazate resistance is the first reported maternally inherited resistant trait associated with cytb Qo in arthropods, and - contrasting with T. urticae – it is one of the first reports of target site resistance in P. citri. This interspecific parallel evolution of bifenazate resistance may be an intriguing equivalent of QoI field resistance developed in protozoal/fungal human pathogens (resistance to atovaquone) and fungal plant pathogens (resistance to strobilurin fungicides). Better understanding of the biological basis of QoI resistance in pest species will facilitate the development of both resistance diagnostic tools and proper anti-resistance strategies aimed at maintaining the high efficacy of these insecto-acaricides