Adenylate kinase from Bacillus subtilis, like the enzyme from Bacillus
stearothermophilus, contains a structural zinc atom. Cys153 in the en
zyme from B. stearothermophilus, which is involved in the zinc coordin
ation, is replaced in the adenylate kinase from B. subtilis by an aspa
rtic acid residue. Therefore, we were interested in establishing wheth
er this difference has an impact on the structure, the metal chelation
, and the overall stability of these proteins. We also were interested
in determining whether His 138, which is conserved in many adenylate
kinases, can act as a fourth partner in the metal chelation and, in ge
neral, whether His can successfully replace Cys or Asp in coordinating
zinc in the adenylate kinase from B. subtilis. The adk gene from B. s
ubtilis was cloned by polymerase chain reaction. The wild-type protein
, together with several variants obtained by site-directed mutagenesis
, were expressed in Escherichia coil and analyzed by biochemical and p
hysicochemical methods. The H138N and D153C mutants of adenylate kinas
e from B. subtilis exhibited properties similar to those of the wild-t
ype protein, indicating that His138 is not involved in metal coordinat
ion and that Asp153, just like Cys in the analogous position in the en
zyme from B. stearothepmophilus, can participate in zinc chelation. Th
is is the first experimental evidence indicating that aspartic acid ca
n be involved in the coordination of a structural zinc atom. On the ot
her hand, the D153H and D153T variants showed significant changes in t
heir zinc-binding properties. Dialysis of the latter proteins against
buffer (in both the presence and the absence of 2 mM EDTA) resulted in
removal of the metal ion and loss of enzymatic activity. These mutant
s reacted readily with 5,5-dithiobis(2-nitrobenzoic acid) under ''nati
ve'' conditions, unlike the wild-type protein and the H138N and D153C
variants of the enzyme. A fifth modified form of adenylate kinase from
B. subtilis probed in this study, C130H, showed characteristics simil
ar to those of the D153T mutant with respect to both metal chelation a
nd reactivity toward 5, 5'-dithiobis(2-nitrobenzoic acid). Differences
between the circular dichroism spectra of wild-type enzyme and those
of the C130H mutant suggest a less compact structure of the mutant, wh
ich also explains its decreased stability against denaturation by temp
erature or guanidinium hydrochloride or against inactivation by trypsi
n. In conclusion, the zinc-chelating property of adenylate kinase from
B. subtilis, and in general from the Gram-positive bacteria, is compa
tible only with the presence of three or four Cys residues in the foll
owing sequence: Cys-X(2)-Cys-X(16)-Cys-X(2)-Asp/Cys.