These two authors contributed equally to this work. SummaryWe hypothesized that cytoplasmic male sterility (CMS) in sugar beet may be the consequence of mitochondrial dysfunctions affecting normal anther development. To test the hypothesis, we attempted to mimic the sugar beet CMS phenotype by inhibiting the expression of mitochondrial pyruvate dehydrogenase (PDH), which is essential for the operation of the tricarboxylic acid (TCA) cycle. Screening with a cDNA library of sugar beet¯ower buds allowed the identi®cation of two PDH E1a subunit genes (bvPDH_E1a-1 and bvPDH_E1a-2). bvPDH_E1a-1 was found to be highly expressed in tap roots, whereas bvPDH_E1a-2 was expressed most abundantly in¯ower buds. Green¯uorescent protein (GFP) fusion of bvPDH_E1a revealed mitochondrial targeting properties. A 300-bp bvPDH_E1a-1 cDNA sequence (from 620 to 926) was connected to a tapetum-speci®c promoter in the antisense orientation and then introduced into tobacco. Antisense expression of bvPDH_E1a-1 resulted in conspicuously decreased endogenous bvPDH_E1a-1 transcripts and male sterility. The tapetum in the male-sterile anthers showed swelling or abnormal vacuolation. It is also worth noting that in the sterile anthers, cell organelles, such as elaioplasts, tapetosomes and orbicules were poorly formed and microspores exhibited aberrant exine development. These features are shared by sugar beet CMS. The results thus clearly indicate that inhibition of PDH activity in anther tapetum is suf®cient to cause male sterility, a phenocopy of the sugar beet CMS.

Beet necrotic yellow vein virus (BNYVV) is the causal agent of rhizomania, the most serious sugar beet disease worldwide. Since the first finding in Japan in 1969, BNYVV became widespread throughout Hokkaido in a few decades and led to the introduction of Rz1‐resistant sugar beet cultivars in the 1990s. Here, we report the historical progress of the BNYVV epidemic in Hokkaido from 1969 to 2019. Previous analysis on samples from 1991 showed that BNYVV isolates were classified into three strains (named O, D, and T) based on the RNA3‐encoded p25 gene. The O‐type viruses were widely detected in Hokkaido, while the D‐ and T‐type viruses were detected in limited areas. The RNA5, encoding the p26 gene, was initially contained in some D‐ and O‐type isolates but not in any T‐type isolates. Interestingly, recent sample analysis revealed that RNA5‐containing T‐type viruses, seemingly more virulent than the other two strains, were widely detected in Hokkaido. Additionally, a small group of virus isolates harbouring a new p25 gene (named C) was found in limited areas. These results suggest that the T‐type viruses, which accompanied RNA5, have been preferentially spread from a limited area to other districts over the last few decades and that this spread might be strongly associated with the recent introduction of Rz1‐resistant sugar beet cultivars. BNYVV‐positive samples also contained mainly beet soil‐borne virus and traces of beet virus Q, both of which are the first to be recorded in Japan.