Fig. 8

Alignment of CRISPR arrays with M. sedula reference arrays. Numbers of spacer are denoted. According to the CRISPR database, M. sedula has four CRISPR arrays in its genome. The repeat sequence of Array 647 and 646 (detected in the SHP metagenome by BLAST) matched repeat sequences of the longest CRISPR array (CRISPR ID: NC_009440_4, refer as “reference array” hereafter) in the M. sedula DSM 5348 reference genome. Since the order of spacers could be associated with time of virus infection, older spacers were located near the “ancient” end of the array. With a comparative analysis of the spacer sequences, all 27 spacers in Array 647 perfectly matched the array in the reference genome (with regards to sequence similarity and orientation). The 27 spacers were located at the end of the reference array. Array 646 contained 42 spacers; the last 38 spacers perfectly matched the 83^th to 121^st spacers in the reference array. Sequences of the first four spacers in Array 646 were different from all of the spacers in reference array; therefore, there was divergence of virus infection history between the M. sedula-related species in the SHP and M. sedula DSM5348. Moreover, since no spacers similar to the 1^st to 89^th and 122^nd to 124^th spacers of the reference array were detected, those spacers were designated “unknown”