Project #2: Peptide Deformylase
Peptide
deformylase (PDF) catalyzes the hydrolytic cleavage of the formyl group at the
N-terminus of nascent eubacterial proteins during protein synthesis. As PDF is essential for bacterial survival
but absent in higher animals, PDF constitutes a promising target for a new
class of antibacterial agents, making investigation of the structure and
function of this enzyme an important endeavor.
PDF is also of great interest from the bioinorganic viewpoint, as it is
the only example of an iron metalloamidase.
Other metallopeptidases use ZnII as the metal ion and the
structural and functional characteristics of PDF would otherwise indicate that
the enzyme belongs in the mononuclear ZnII enzyme family. The choice of iron by nature is intriguing
considering the inherent instability of FeII towards oxidation and
that PDF catalyzes a non-redox-active reaction.
Crystal
structures of the full-length protein indicate the metal ligands to be Cys90,
His132, and His136 in the HEXXH motif and a water molecule or hydroxide
ion. The presence of the cysteinate
sulfur ligand as part of the mixed N,S environment is rather uncommon; an N or O
donor from His or Asp/Glu is more typical in “zinc” metallopeptidases. A network of hydrogen bonds aligns the
metal-coordinating ligands, and substitution of FeII by ZnII or
NiII leads to only minimal structural changes. High-resolution crystal structures of PDF in
complex with formate show differences in formate binding modes when FeII,
CoII, and ZnII are present.
The proposed
deformylation mechanism (Figure 1) begins with nucleophilic attack of the
hydroxide ligand on the carbonyl carbon of the N-terminal formyl group of the
peptide. This is followed by the
transition from a tetrahedral to a five-coordinate metal center and subsequent
formation of an enzyme-formate complex.
Hydrogen bonds from spatially close residues stabilize charge moieties
throughout the mechanism. The metal
functions as an electron-withdrawing group to favor deprotonation of the
metal-bound water to yield a hydroxide ligand and as a Lewis acid to activate
the bound carbonyl substrate, the latter function of which is likely affected
by the presence of the cysteinate ligand.
Interestingly, catalytic activity is lost in ZnII-substituted
forms, but maintained with NiII and CoII. Tighter binding of the ZnII ion is
proposed to lead to the inability of zinc to change between tetrahedral and
five-fold coordination during catalysis.
Highly
promising biomimetic models for PDF have recently been developed by Goldberg
and co-workers. An N2Sthiolate
ligand (Figure 2), 2-methyl-1-[methyl-(2-pyridin-2-yl-ethyl)amino]propane-2-thiol
(or PATH, pyridine-amine-thiolate system), has been demonstrated to bind ZnII and
CoII, bears an aliphatic thiolate as a mimic of the cysteine donor
in PDF, and yields monomeric metal complexes.
The hydrolysis of 4-nitrophenyl acetate by the (PATH)ZnII-hydroxide
complex has also been observed. Another
possible biomimetic system is a series of pseudotetrahedral Zn(II) complexes of
the heteroscorpionate ligands (3-tert-butyl-2-thio-5-methylphenyl)bis(3,5-dimethylpyrazolyl)methane
(L2SH) and bis(3,5-dimethyl-pyrazolyl)(1-methyl-1-sulfanylethyl)methane (L3SH)
(Figure 3).
Computational
experiments will commence with characterizing the hydroxide and formate
complexes with the PATH ligand and the metals FeII, ZnII,
CoII, and NiII.
Oxidation potentials for the complexes will be calculated to assess the
redox stability of the metal. The tendency
of the metal to favor a hydroxide ligand over an aqua ligand will be addressed
by computing pKb values for the hydroxide ligand.
Determination
of the mechanism for the 4-nitrophenyl acetate hydrolysis by the (PATH)MII-hydroxide
complexes will follow. Analysis of these
results will be used to gauge the effect of the thiolate ligand in facilitating
catalysis. These calculations will also
provide insight as to the cause of the inhibition in ZnII-PDF – for
example, whether it results from an inability to change coordination geometry
or the inability of water to replace formate as a ligand to ZnII.
The next
stage of the investigation will involve answering analogous questions for QM
models of the enzyme active sites, which will be derived from available crystal
structures. Comparison of these results
with those from the biomimetic models can be used to suggest modifications
which might lead to greater reactivity in the biomimetic systems.
Finally,
QM/MM methods can be utilized in order to include explicitly the
hydrogen-bonding environment (as well as other protein environment effects)
around the coordinating cysteine and histidine residues in the
calculations. The roles of
non-coordinating residues near the PDF active site can then be elucidated, with
implications towards predicting the results of mutations and designing enzyme
inhibitors.
(1) Chan, M. K.; Gong,
W. M.; Rajagopalan, P. T. R.; Hao, B.; Tsai, C. M.; Pei, D. H. Biochemistry 1997, 36, 13904-13909.
(2) Hao, B.; Gong, W.
M.; Rajagopalan, P. T. R.; Zhou, Y.; Pei, D. H.; Chan, M. K. Biochemistry 1999, 38, 4712-4719.
(3) Lipscomb, W. N.;
Strater, N. Chem. Rev. 1996, 96, 2375-2433.
(4) Chang, S. C.;
Karambelkar, V. V.; diTargiani, R. C.; Goldberg, D. P. Inorg. Chem. 2001, 40, 194-195.
(5) Chang, S. C.;
Sommer, R. D.; Rheingold, A. L.; Goldberg, D. P. Chem. Commun. 2001,
2396-2397.
(6) Hammes, B. S.;
Carrano, C. J. Inorg. Chem. 1999, 38, 4593-4600.