Genomic structure of an economically important cyanobacterium, Arthrospira (Spirulina) platensis NIES-39

Abstract

A filamentous non-N(2)-fixing cyanobacterium, Arthrospira (Spirulina) platensis, is an important organism for industrial applications and as a food supply. Almost the complete genome of A. platensis NIES-39 was determined in this study. The genome structure of A. platensis is estimated to be a single, circular chromosome of 6.8 Mb, based on optical mapping. Annotation of this 6.7 Mb sequence yielded 6630 protein-coding genes as well as two sets of rRNA genes and 40 tRNA genes. Of the protein-coding genes, 78% are similar to those of other organisms; the remaining 22% are currently unknown. A total 612 kb of the genome comprise group II introns, insertion sequences and some repetitive elements. Group I introns are located in a protein-coding region. Abundant restriction-modification systems were determined. Unique features in the gene composition were noted, particularly in a large number of genes for adenylate cyclase and haemolysin-like Ca(2+)-binding proteins and in chemotaxis proteins. Filament-specific genes were highlighted by comparative genomic analysis.

Figure 1

Schematic representation of the circular chromosome of A. platensis. A scale indicates the coordinates in megabase pairs. From outside to inside: circle 1, the gaps in the genome; circles 2 and 3, predicted protein-coding genes on the forward and reverse strands; circle 4, G+C content; circle 5, GC skew. Eighteen contig gaps (G01-G18) are numbered in the clockwise direction starting from the end of the longest contig. Functional categories were colour-coded according to the standard colours used by COGs. The genome sequence and annotation of A. platensis NIES-39 are available at GenBank/EMBL/DDBJ under accession no. AP011615.

Figure 2

Gene structure of the ribonucleotide reductase (RNR) gene with two group I intron insertions.

Figure 3

Domain architecture of adenylate cyclases in A. platensis. See Table 1 for abbreviations.

Figure 4

Domain architecture of HsdS proteins. The target recognition domain roughly corresponds to Methylase_S (PF01420).

Figure 5

Selective mutations in the nrsS/R/B/A region.