INRA, Unite de Phytopharmacie et Mediateurs Chimiques, 78026 Versailles Cedex, France

Background and objectives
Botrytis cinerea (anamorph of Botryotinia fuckeliana), the causal agent of grey mould, can infect a wide range of crops including grapevine. The control of this disease is mainly achieved by applications of fungicides. The intensive use of benzimidazoles (e.g. carbendazim) and dicarboximides (e.g. procymidone) led to the development of resistant strains which reduced their efficacy. The situation is less critical, for the more recently introduced, phenylpyrroles (e.g. fludioxonil) and anilinopyrimidines (e.g. pyrimethanil), because they are still highly effective in practice. However, monitoring of B. cinerea natural populations allowed us to distinguish three anilinopyrimidine-resistant phenotypes: AniRl phenotype is highly resistant only to anilinopyrimidines whereas the other phenotypes are weakly resistant to anilinoprymidines and to other fungicides like phenylpyrroles (AniR2) or sterol demethylation inhibitors (DMis) (AniR3) [1]. The anilinopyrimidines are known to inhibit the biosynthesis of methionine [2].
Our objective was to study the genetic inheritance of the multiple resistance in AniR2 and AniR3 strains and to understand the mechanisms involved.

Results and discussion
Strains AniR2 and AniR3 were crossed with vaid-type reference strains (AniS). For each apothecium, random sample of about 100 germinated ascospores were collected individually to obtain monoascospore strains. Sensitive and resistant strains in progeny were distinguished according to Leroux and Gredt [1] by testing conidia germination andlor germ tube elongation in presence of fungicides. Ascospore progenies showed a 1:1 segregation of AniS and AniR phenotypes. This suggests that the two phenotypes, AniR2 and AniR3, are each encoded by a single major gene. In our ascospore progenies, all AniR strains had the resistance phenotype of the parental strains. Since AniR2 and AniR3 strains displayed resistance to molecules with different modes of action, the putative mechanism could be a "multidrug resistance" (MDR) which was described in several organisms and which consists in an active excretion of toxic molecules. We carried out experiments on uptake of 14C tebuconazole (DMI) by AniS, AniR2 and AniR3 strains. In sensitive strains, 14C tebuconazole was accumulated in a transient pattern only if the concentration in culture medium was at least 100 pM. This accumulation reached a maximum (2.6 nmol of tebuconazole/mg of dry weight) after 10 to 15 min of incubation. Then, radioactivity decreased, after one hour, to the level of 0.6 nmol/mg. In AniR3 strains, the fungicide uptake was low with a level of incorporation of 0.6 nmol/mg and could explain their tebuconazole resistance. For AniR2 strains, the radioactivity accumulation was intermediate (1.3 nmol/mg). Fungicide incorporation was proportional to the initial concentration in culture medium, suggesting a passive incorporation. It was shown that efflux in sensitive strains can be suppressed by inhibitors of mitochondrial respiration indicating an ATP-dependant mechanism. In sensitive strains, this mechanism could be inducible whereas it could be constitutive in resistant ones. The 14c pyrimethanil initial incorporation was also probably passive but without transient accumulation pattern in susceptible strains and the fungicide uptake was similar in all strains.
The mechanism implicated in the resistance is not clearly identified but we showed a MDR segregation always corresponding to 1:1 ratio in the progeny from crosses between resistant and wild-type strains. The characterisation of the involved gene(s) is under progress.

1. Leroux P and Gredt M, 1995. Agronomie 15, 367-70.
2. Fritz R, Lanen C, Colas V, and Leroux P, 1997. Pesticide Science 49, 40-6.