The bacterial strain Spiroplasma apis B31^T was isolated from a honeybee (Apis mellifera) in southwestern France affected by May disease (1). Previous experiments demonstrated that the artificial infection of honeybees by S. apis through injection or food ingestion leads to classical symptoms of May disease, such as nervous disorder, loss of flight ability, and eventually, high mortality in a few days (2). Moreover, high titers of S. apis cells were found in the hemolymph of the dead bees. The lethal effect of S. apis infection on bees might be prevented by tetracycline, but not by penicillin (2). To investigate the virulence factors of S. apis and to improve the taxon sampling of available Spiroplasma sequences for comparative analysis, we obtained S. apis B31^T from the American Type Culture Collection (ATCC 33834) for complete genome sequencing.

The procedures for genome sequencing, assembly, and annotation are based on those described in our previous studies (3, 4). Briefly, the Illumina HiSeq 2000 platform was used to generate 101-bp reads from one paired-end library (~160-bp insert, 6,146,318 reads) and one mate-pair library (~3.8-kb insert, 8,552,782 reads). The de novo assembly was performed using ALLPATHS-LG release 42781 (5). The initial assembly was iteratively improved by mapping the raw reads to the contigs by the Burrows-Wheeler Aligner (BWA) version 0.6.2 (6) for visual inspection by IGV version 2.1.24 (7). All gaps were filled by using reads overhanging at the contig margins or by reassembling the reads mapped to the gap regions with Phrap version 1.090518 (http://www.phrap.org/). The programs RNAmmer (8), tRNAscan-SE (9), and Prodigal (10) were used for gene prediction. For each protein-encoding gene, the gene name and product description were annotated based on the orthologous genes identified by OrthoMCL (11) in other Spiroplasma genomes (4, 12, 13), including those of S. chrysopicola (GenBank accession no. CP005077), S. diminutum (GenBank accession no. CP005076), S. melliferum (GenBank accession no. AMGI01000001 to AMGI01000024), S. syrphidicola (GenBank accession no. CP005078), and S. taiwanense (GenBank accession no. CP005074 to CP005075). The genes that do not have identifiable orthologs in other Spiroplasma genomes were manually curated based on BLASTp (14) searches against the NCBI nonredundant (nr) protein database (15). The Kyoto Encyclopedia of Genes and Genomes (KEGG) database (16) was used to assist annotation.

The circular chromosome of S. apis B31^T is 1,160,554 bp in size, with a 28.3% G+C content, and no plasmid was found. The first version of annotation includes one set of 16S-23S-5S rRNA genes, 29 tRNA genes (covering all 20 amino acids), and 997 protein-encoding genes.