The induced-fit effect and rotation of the alpha-helical subdomain may play a role in controlling the nucleotide-dependent change in cassette-cassette interaction affinity believed to represent the power-stroke of ABC transporters. Outward rotation of the alpha-helical subdomain also likely facilitates Mg-ADP release after hydrolysis. The MJ1267 structures therefore define features of the nucleotide-dependent conformational changes that drive transmembrane transport in ABC transporters.

Finally, it is noteworthy that one of the intersubunit packing interactions in the MJ0796 crystal involves antiparallel ␤-type hydrogen bonding interactions between the outermost ␤-strands in the two core ␤-sheets, leading to their fusion into a single extended ␤-sheet, a type of structural interaction that has been proposed to play a role in mediating the aggregation of ␤-sheet-containing proteins.ABC (ATP-binding cassette) transporters are ubiquitously distributed ATP-dependent transmembrane solute pumps and ion channels (1-4). These proteins play a causative role in a variety of fatal genetic diseases, including cystic fibrosis (5, 6) and adrenoleukodystrophy (7). They also play a role in cancer where overexpression of one family member mediates development of multidrug resistance (2, 6), a major barrier to effective chemotherapy of advanced malignancies. The fundamental architecture of an ABC transporter comprises a pair of soluble ATP-binding cassettes attached to a pair of ␣-helical transmembrane domains (each with six to eight transmembrane ␣-helices). There is unambiguous and strong sequence homology in the ϳ200-residue core of the ATP-binding cassettes between even the most remotely related members of the ABC transporter superfamily. However, homology between the transmembrane domains is weak, consistent with their proposed role in determining the diverse transport substrate specificities of the various family members.Previously, the crystal structures of the HisP ATP-binding cassette from Salmonella typhimurium (8) and the MalK ATPbinding cassette from Thermococcus litoralis (9) have been reported, establishing the basic fold of the cassette. A crystal structure of a homodimer of the soluble Rad50 DNA repair enzyme has also been reported (10). Although this protein is distantly related to the cassettes from true ABC transporters and lacks the ABC-specific ␣-helical subdomain, its structure is closely similar to that of HisP in the F 1 -type (11) ATP-binding core and the antiparallel ␤-subdomain, with the two proteins sharing a 1.11-Å root mean square (r.m.s.) 1 deviation for 78 C-␣ atoms in these regions (compared with an ϳ0.40-Å r.m.s. deviation for superposition of the equivalent region in different crystal forms of the same ATP-binding cassette) (see Fig. 1 for nomenclature). Based on the fact that the nearly phylogenetically invariant LSGGQ sequence (1, 12) completes the ATPase active site in the 2-fold related cassette in the MgAMP-PNPbound form of Rad50, this structure established a possible model for the ATP-binding cassette dimer believed to be a conserved feature of all ABC transporters (1-4). Although these crystal structures have provided substantial insight into the structural basis of ABC transporter activity, additional data are needed to define both the structural consequences of the significant sequence variability observed in the N-and C-terminal regions of the cassettes and also the nature of the nucleotide-dependent changes in cassette structure that control the conformational reacti...

SummaryOne of the more natural but less commonly studied forms of colonial bacterial growth is pattern formation. This type of growth is characterized by bacterial populations behaving in an organized manner to generate readily identifiable geometric and predictable morphologies on solid and semi-solid surfaces. In our first attempt to study the molecular basis of pattern formation in Bacillus subtilis, we stumbled upon an enigma: some strains used to describe pattern formation in B. subtilis did not have the phenotypic or genotypic characteristics of B. subtilis. In this report, we show that these strains are actually not B. subtilis, but belong to a different class of Bacilli, group I. We show further that commonly used laboratory strains of B. subtilis can co-exist as mixed cultures with group I Bacilli, and that the latter go unnoticed when grown on frequently used laboratory substrates. However, when B. subtilis is grown under more stringent semiarid conditions, members of group I emerge in the form of complex patterns. When B. subtilis is grown under less stringent and more motile conditions, B. subtilis forms its own pattern, and members of group I remain unnoticed. These findings have led us to revise some of the mechanistic and evolutionary hypotheses that have been proposed to explain pattern growth in Bacilli.