The 1,3-1,4-beta-glucanases from Bacillus macerans and Bacillus lichen
iformis, as well as related hybrid enzymes, are stable proteins compri
sed of one compact jellyroll domain. Their structures are studied in a
n effort to reveal the degree of redundancy to which the three-dimensi
onal structure of protein domains is encoded by the amino acid sequenc
e. For the hybrid 1,3-1,4-beta-glucanase H(A16-M), it could be shown r
ecently that a circular permutation of the sequence giving rise to the
variant cpA16M-59 is compatible with wildtype-like enzymatic activity
and tertiary structure (Hahn et al., Proc. Natl. Acad. Sci. USA 91:10
417-10421, 1994). Since the circular permutation yielding cpA16M-59 mi
micks that found in the homologous enzyme from Fibrobacter succinogene
s, the question arose whether de novo circular permutations, not guide
d by molecular evolution of the 1,3-1,4-beta-glucanases, could also pr
oduce proteins with native-like fold. The circularly permuted variants
cpA16M-84, cpA16M-127, and cpA16M-154 were generated by PCR mutagenes
is of the gene encoding H(A16-M), synthesized in Escherichia coli and
shown to be active in beta-glucan hydrolysis. CpA16M-84 and cpA16M-127
were crystallized in space groups P2(1) and P1, respectively, and the
ir crystal structures were determined at 1.80 and 2.07 Angstrom resolu
tion. In both proteins the main parts of the beta-sheet structure rema
in unaffected by the circular permutation as is evident from a root-me
an-square deviation of main chain atoms from the reference structure w
ithin the experimental error. The only major structural perturbation o
ccurs near the novel chain termini in a surface loop of cpA16M-84, whi
ch becomes destabilized and rearranged. The results of this sturdy are
interpreted to show that: (1) several circular permutations in the co
mpact jellyroll domain of the 1,3-1,4-beta-glucanases are tolerated wi
thout radical change of enzymatic activity or tertiary structure, (2)
the three-dimensional structures of simple domains are encoded by the
amino acid sequence with sufficient redundancy to tolerate a change in
the sequential order of secondary structure elements along the sequen
ce, and (3) the native N-terminal region is not needed to guide the fo
lding polypeptide chain toward its native conformation. (C) 1998 Wiley
-Liss, Inc.