The nucleotide sequences offic-1 involved in the cell filamentation induced by cyclic AMP (5), and the number of the cell division genes is increasing, but their interaction is not still clear because of the complexity of the interaction. We have focused on cyclic AMP (cAMP) as a possible regulatory factor in cell division. Although cAMP is not essential for growth in Escherichia coli because the cya mutant is viable, much evidence supports the idea that cAMP controls cell growth or cell division in E. coli (3,9,22,23,26). We isolated cAMP-requiring mutants with double mutations of cya and cid (24, 25). We have found that an increase in intracellular cAMP levels induced cell filamentation in afic-i mutant at high temperature (14,23,27). cAMP is required for anucleate cell production (9). The filamentation of ftsZ(Ts) is repressed in the cya or crp mutant (26). Interestingly, Hughes et al. showed that cAMP binds DnaA protein and has a regulatory role in DNA replication (8).It is important to know the structure of the fic gene to understand its function in relation to cAMP. In this study we sequenced the fic and fic-i genes and found that thefic gene has some sequence-similar regions to other cell division genes.Plasmid pHB3 (1, 13), containing the fic gene; pMK108 (13), containing thefic-i gene; and other subcloned plasmids used as the source of DNA for sequencing were purified by the procedure of Birnboim and Doly (2). For preparation of plasmids on a small scale, the alkaline sodium dodecyl sulfate method of Davis et al. (4, 16) was used. Restriction fragments digested by AluI, EcoRV, HaeIII, HincII, HpaII, RsaI, Sau3A, TaqI, and ThaI were subcloned into M13 vector mpl8 or mpl9 (28). We sequenced a 2.5-kilobase-pair (kb) AvaI-AvaI region which includes the fic-i gene and its surroundings, as reported previously (13). The sequence strategy is shown in Fig. 1. Almost all double-stranded DNAs were sequenced by the dideoxy-chain termination method (20) (shown by arrows in Fig. 1). The Maxam and Gilbert method (17) was used for the parts (dashed arrows in * Corresponding author. Fig. 1) which are difficult to analyze by the M13 chain termination method. In the base sequence (Fig. 2) Fig. 2. The calculated protein size from the sequence of fic-i was 22,930 daltons, coinciding well with the protein product of 21 kilodaltons identified by the maxicell method, as reported previously (13).To find where in the DNA the mutation was generated in thefic-i gene, we sequenced, by the dideoxy-chain termination method, a 1.3-kb MluI-EcoRV region containing thefic gene. The fic DNA was isolated from pHB3. An enzymatically digested DNA fragment was inserted into an mpl8 or mpl9 vector, and sequencing was performed (detailed sequence strategy is not shown). G at position 1574 was changed to C in the fic-i mutant (Fig. 2). By this mutation,