Supplemental FIG. 1. Phylogenetic relationship of GGDEF domains from predicted diguanylate cyclases in the genome of Anabaena sp. strain PCC 7120. Genetic loci designations for Anabaena sp. strain PCC 7120 are from CyanoBase, http://bacteria.kazusa.or.jp/cyanobase/. The GGDEF domain from the PleD protein of Caulobacter crescentus was also included in the analysis. The unrooted tree was generated with the PHYLIP software package version 3.65 available through the BioBike system (see reference 1 below) using the neighbor-joining program to calculate distance matrices from a multiple sequence alignment of amino acid positions corresponding to each GGDEF domain produced by ClustalW software. Bootstrap values were obtained with 500 replications. Two distinct phylogenetic groups are indicated. The scale bar represents 0.02 amino acid substitutions per residue.
Supplemental FIG. 2. The all2874 mutant strain was deficient in maintaining a normal heterocyst frequency when grown at high light intensity. The bar graph shows the distribution of the number of vegetative cells in the intervals between heterocysts at 24 h and 96 h after nitrogen step-down of the wild type and the all2874 mutant grown at low light or high light.
Supplemental FIG. 3. Growth of the all2874 mutant strain AMC1572 is slower than that of the wild type after nitrogen step-down at medium and high light intensities. Growth under low light conditions was similar for the two strains. Wild-type and all2874 mutant filaments from nitrate-containing BG-11 cultures were washed and then transferred to BG-110 at time zero to induce heterocyst development. Three independent cultures of the wild type and the all2874 mutant were used for each light condition. Diazotrophic growth of each culture was followed for 7 days by measuring the optical density at 750 nm (OD). Error bars indicate standard deviation.
Supplemental FIG. 4. (A) Domain organization of the 14 genes predicted to encode diguanylate cyclases in Anabaena sp. strain PCC 7120 determined by the SMART program (http://smart.embl-heidelberg.de). (B) Predicted domain organization of the All2875 protein. (C) Diagram of the chromosomal context of all2874.
Supplemental FIG. 5. SDS-PAGE gel stained with Coomassie brilliant blue showing purified proteins: lane 1, prestained molecular weight markers; lane 2, purified PleD* (~49.5 kDa); lane 3, purified All2874-6His (~38.2 kDa); and lane 4, purified HetL-6His (~ 27 kDa).
1. Massar, J. P., M. Travers, J. Elhai, and J. Shrager. 2005. BioLingua: a programmable knowledge environment for biologists. Bioinformatics 21:199-207.