---

The a-amylase from Bacillus sp. no. 195 (BAA) consists of two domains: one is the catalytic domain similar to a-amylases from animals and Streptomyces in the N-terminal region; the other is the functionally unknown domain composed of an approx. 90-residue direct repeat in the C-terminal region. The gene coding for BAA was expressed in Streptomyces lividans TK24. Three active forms of the gene products were found. The pH and thermal profiles of BAAs, and their catalytic activities for p-nitrophenyl maltopentaoside and soluble starch, showed almost the same behaviours. The largest, 69 kDa, form (BAA-a) was of the same molecular mass as that of the mature protein estimated from the nucleotide sequence, and had raw-starch-binding and -degrading abilities. The second largest, 60 kDa, form (BAA-b), whose molecular mass was the same as that of the natural enzyme from Bacillus sp. no. 195, was generated by proteolytic processing between the two repeat sequences in the C-terminal region, and had lower activities for raw starch binding and degrading than those of BAA-a. The smallest, 50 kDa, form (BAA-g) contained only the N-terminal catalytic domain as a result of removal of the C-terminal repeat sequence, which led to loss of binding and degradation of insoluble starches. Thus the starch adsorption capacity and raw-starch-degrading activity of BAAs depends on the existence of the repeat sequence in the C-terminal region. BAA-a was specifically adsorbed on starch or dextran (a-1,4 or a-1,6 glucan), and specifically desorbed with maltose or b-cyclodextrin. These observations indicated that the repeat sequence of the enzyme was functional in the starch-binding domain (SBD). We propose the designation of the homologues to the SBD of glucoamylase from Aspergillus niger as family I SBDs, the homologues to that of glucoamylase from Rhizopus oryzae as family II, and the homologues of this repeat sequence of BAA as family III.

---

Abbreviations used: BAA, 'animal'-type a-amylase from Bacillus sp. no. 195; BAA-SBD, starch-binding domain of a-amylase from Bacillus sp. no. 195; CBD, cellulose-binding domain; G1-SBD, starch-binding domain of glucoamylase 1 from Aspergillus niger; PPA, a-amylase from pig pancreas; SBD, starch-binding domain; TAA, a-amylase (Taka-amylase) from Aspergillus oryzae.

¹ To whom correspondence should be addressed (e-mail motoo@biochem.osakafu-u.ac.jp).

The nucleotide sequence data reported will appear in DDBJ, EMBL and GenBank Nucleotide Sequence Databases under the accession number AB006823.

Key words: glucoamylase, polysaccharide, post-translational processing.

---

Received 26 January 2000/19 May 2000; accepted 21 June 2000

---

The Biochemical Society, London © 2000