Crossover inhibition as an indicator of convergent evolution of enzyme mechanisms: A β-lactamase and a N-terminal nucleophile hydrolase

Article abstract of DOI:10.1016/j.febslet.2012.10.019

O-Aryloxycarbonyl hydroxamates and 1,3,4-oxathiazol-2-ones have been identified as covalent inhibitors of β-lactamases and proteasomes, respectively. The products of these inhibition reactions are remarkably similar, involving carbonyl cross-linking of the active sites. We have cross-checked these inhibitors, showing that the former inhibit proteasomes and the latter β-lactamases, to form the same inactive carbonyl adducts. These results are discussed in terms of similarities of the active site structures and catalytic mechanisms. It is likely that a mechanistic imperative has led to convergent evolution of these enzyme active sites, of a β-lactam-recognizing enzyme and a N-terminal protease belonging to different amidohydrolase superfamilies.

Full text of DOI:10.1016/j.febslet.2012.10.019

FEBS Letters 586 (2012) 4186–4189
journal homepage: www.FEBSLetters.org
Crossover inhibition as an indicator of convergent evolution of enzyme
mechanisms: A b-lactamase and a N-terminal nucleophile hydrolase
S.A. Adediran a, G. Lin b, R.B. Pelto a, R.F. Pratt a,
⇑
a Department of Chemistry, Wesleyan University, Middletown, CT 06459, USA
b Department of Microbiology and Immunology, Weill Cornell Medical College, New York, NY 10065, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
O-Aryloxycarbonyl hydroxamates and 1,3,4-oxathiazol-2-ones have been identified as covalent
inhibitors of b-lactamases and proteasomes, respectively. The products of these inhibition reactions
are remarkably similar, involving carbonyl cross-linking of the active sites. We have cross-checked
these inhibitors, showing that the former inhibit proteasomes and the latter b-lactamases, to form
the same inactive carbonyl adducts. These results are discussed in terms of similarities of the active
site structures and catalytic mechanisms. It is likely that a mechanistic imperative has led to conver-
gent evolution of these enzyme active sites, of a b-lactam-recognizing enzyme and a N-terminal
protease belonging to different amidohydrolase superfamilies.
Received 29 August 2012
Revised 1 October 2012
Accepted 2 October 2012
Available online 23 October 2012
Edited by Stuart Ferguson
Keywords:
Enzyme
Ó 2012 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.
b-Lactamase
Proteasome
Inhibitors
Convergent evolution
1. Introduction
report the reaction of the proteosome with 1, serine b-lactamases
with 2, and discuss the results in terms of the similarity of the
To a large degree, the resistance of bacteria to the b-lactam anti-
biotics derives from bacterial expression of b-lactamases [1]. These
enzymes catalyze the hydrolysis of the b-lactam ring and thus
inactivate the antibiotic. b-Lactamase inhibitors are of interest be-
cause they can be administered in combination with a b-lactam
and thereby extend the clinical lifetime of the antibiotic [2].
We recently described the O-aryloxycarbonyl hydroxamates 1
as novel serine b-lactamase inhibitors [3,4]. They were found to
first acylate the nucleophilic active site serine to generate 3. Subse-
quently, the active site became irreversibly inactivated by cross-
linking of the serine to an active site lysine as in 4 (Scheme 1).
Another potential antibiotic target is the protease of the bacte-
rial proteasome [5]. This enzyme is inhibited by a series of 1,3,4-
oxathiazol-2-ones 2 [6]. We noticed that the proposed mechanism
of inhibition, first involving acylation of the nucleophilic threonine
of the active site to form 3 and followed by formation of an inert
complex 4, appeared to have much in common with that of the
b-lactamase with 1 (Scheme 1). Here also, in the inert complex 4,
the active site nucleophile is cross-linked to an adjacent amine,
in this case to the N-terminal threonine amine.
active site structure and mechanism between these two classes
of enzymes.
2. Materials and methods
The Enterobacter cloacae P99 and TEM-2 b-lactamases were pur-
chased from the Centre for Applied Microbiology and Research
(Porton Down, Wiltshire, U.K.) and used as received. The Mycobac-
terium tuberculosis proteasome was over-expressed and purified as
reported [7]. O-Aryloxycarbonyl hydroxamates and the 1,3,4-oxa-
thiazol-2-ones were synthesized as described [4,6]. 5-Phenyl-
1,3,4-dioxazolone (9) was prepared from the reaction of
benzohydroxamic acid with phenyl chloroformate in the presence
of imidazole [8]. All kinetics measurements were carried out in
20 mM MOPS buffer, pH 7.5 at 25 °C (b-lactamases) and 20 mM
HEPES, 0.5 mM EDTA, pH 7.5 at 37 °C (proteasome). Second order
rate constants for inactivation (ki) were determined as described
[4,6] from measurements of enzyme activity (initial velocity)
against 0.2 mM cephalothin (b-lactamase) or 50 lM Suc-LLVY-
AMC (proteasome) as a function of time. Figs. 1A and B were gen-
erated from crystal structure files from the RCSB Protein Data Bank
(1MXO, b-lactamase and 2FHH, proteasome) with the aid of Insight
II (Accelrys) software. The b-lactamase electrospray mass spectra
were obtained from the Mass Spectrometry Laboratory, School of
Chemical Sciences, University of Illinois and the proteasome
Because of the similarity of the intermediates 3 and final prod-
ucts 4, we decided to cross-check these inhibitors. In this paper, we
⇑
Corresponding author. Fax: +1 860 685 2010.
E-mail address: rpratt@wesleyan.edu (R.F. Pratt).
0014-5793/$36.00 Ó 2012 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.febslet.2012.10.019

S.A. Adediran et al. / FEBS Letters 586 (2012) 4186–4189
4187
Mass spectra and crystal structures suggested that these inhib-
itors covalently modify both enzymes as described above
(Scheme 1) [3,4]. A mass spectrum of the P99 b-lactamase after
inactivation by 6 (
tide of the proteasome by 10 (
D
m/s = 26 amu) and of the N-terminal heptapep-
m/s = 26.0 amu) showed carbonyl
D
addition, presumably by the same cross-linking mechanism.
4. Discussion
It is interesting, given this crossover of inhibitors with analo-
gous mechanisms, described above, to consider a possible overlap
of functional groups at the active sites of a b-lactamase and the
proteasome. Could this reveal a common active site architecture
and a common mechanism of catalysis?
Crystal structures of class C b-lactamases [2,9] and the yeast or
M. tuberculosis proteasomes [7,10] show the active site compo-
nents of 13 and 14, respectively [7,11,12]. Hydrogens have been
added to give structures poised for the first step of catalysis, nucle-
ophilic attack on a substrate carbonyl, by the active site serine
(b-lactamase) or threonine (proteasome), as prescribed by several
of the commonly-held mechanisms [4,5]; in each case, nucleophilic
Scheme 1. b-Lactamase and proteasome inhibition mechanisms.
peptide spectra with the assistance of Dr. Haiteng Deng at the Pro-
teomics Resource Center of the Rockefeller University.
3. Results
The class C Enterobacter cloacae P99 and Class A TEM-2 b-lacta-
mases were indeed inactivated by the 1,3,4-oxathiazol-2-ones 5–8
and the 1,3,4-dioxazol-2-one 9. Second-order rates constants for
loss of activity were determined as described in methods and the
results are presented in Table 1. For comparison, rate constants
for the O-aryloxycarbonyl hydroxamates 10–12 are included. Fur-
ther, all of 5–12 also inhibit the proteasome (Table 1).
These data show that the 1,3,4-oxathiazol-2-ones are compara-
bly effective to O-aryloxylcarbonyl hydroxamates as b-lactamase
and proteasome inhibitors. Neither class of inhibitor, of course, is
attack by the alcohol is aided by a lysine e-amine as general base. It
is interesting to note that the tyrosine hydroxyl of the b-lactamase
is replaced in the proteasome structure by that of a water mole-
cule. Structures similar to 13 are seen in the DD-peptidases, the
evolutionary parents of b-lactamases [13,14] and in all protea-
somes. On the other hand, other major groups of N-terminal hydro-
lases, such as the penicillin acylases [15], do not appear to contain
Lys2 [16]. These two groups of enzymes, the b-lactam-recognizing
enzymes and the N-terminal hydrolases, belong to two very
different superfamilies, from the point of view of tertiary structure
optimized for the particular enzymes of Table 1. It is striking, how-
ever, that 6 and 7 are so much better inhibitors of the P99 b-lacta-
mase, and most likely of the TEM-2 enzyme also, than 5 and 8. This
cannot be due to intrinsic reactivity since the same difference is
not seen in the background hydrolysis rates ko, also presented in
Table 1. It may reflect the benefit of an electrophilic aromatic ring
in binding to the b-lactamase active site. Compounds 6 and 7 are
also effective proteasome inhibitors, possibly for similar reasons.
Incorporation of either a better (11) or more hydrophobic (12)
leaving group into 10 leads to a more effective inhibitor of both
the b-lactamases and the proteasome. It is striking that the 1,3,4-
dioxazolone 9, a direct oxa analog of 5, is a weak/non-inhibitor
of both types of enzyme, despite its higher reactivity (see k0, Ta-
ble 1). This probably results from partitioning of a tetrahedral
intermediate in favor of free enzyme rather than 3 and thus 4 (Path
a of Fig. 3b in Ref. [6]).
[16–18]. It may be noted, however, that another superfamily of
serine amidohydrolases, that containing the amidase signature
enzymes, has a very similar array of active site functional groups
[19].

4188
S.A. Adediran et al. / FEBS Letters 586 (2012) 4186–4189
Table 1
Rate constants of b-lactamase and proteasome inhibition.
Inhibitor ki (MÀ1 sÀ1
P99 b-Lactamase
)
ko  105 sÀ1
TEM-2 b-
Mtb
Lactamase
proteasome
5
6
7
8
31
2
<0.7
0.7 0.1
ND
409b
2130b
730b
1.2
2.0
1.1
2800 300
2200 200
13.0 0.6
ND
4.3b 1.9
9
0.23 0.05 ND
NI
300 20
1160 120
2040 160
239
25a
98a
45a
10
11
12
6.1 Â 103,a
2.2 Â 104,a
1.8 Â 104,a
13.4 0.2
66.2 0.2
25.0 0.2
ND = not determined, NI = no inhibition.
a
Ref. [4].
Ref. [6].
b
Scheme 2. A possible common enzyme acylation mechanism.
The tetrahedral intermediates formed from reaction of normal
substrates with each enzyme in Scheme 2 are stabilized by an oxy-
anion hole, composed of backbone NH hydrogen bond donors
[7,14]. The tetrahedral intermediates in the reactions of 10 with
the class C b-lactamase, however, were proposed to be stabilized
by Tyr 150, the tyrosine depicted in 13, as an ‘‘alternative oxyanion
hole’’ [4]. The proteasome analog of the tyrosine hydroxyl would
be the water molecule of 14.
A significant difference, however, evident from the boronate
crystal structures [7,14], between class C b-lactamases and DD-
peptidases on one hand and the proteasome on the other, reflect-
ing their very different evolutionary heritage, is that the former
attack the substrate (peptide) carbonyl on the Re face of the sub-
strate carbonyl, generating a S tetrahedral intermediate, while
the proteasome threonine attacks on the Si face to generate an R
intermediate (Scheme 3). For this reason, the active site compo-
nents of 13 and 14 (otherwise seen as the red and green elements
of Scheme 3) are differently arranged in space and cannot be di-
rectly superimposed. Separately, however, the functional groups
relevant to the two steps of Scheme 2 can be superimposed. Thus,
the Lys2s of 13 and 14, each putatively participating as a general
base in the first step of the acylation reaction (Scheme 2; B of
Fig. 1. (A) Superimposition of active site elements of a boronate transition state
analog structure of the class C AmpC b-lactamase (atomic coloring) with the C
subunit of the M. tuberculosis proteasome (turquoise, with important heteroatoms
atomic colored). The central boron (B) is shown in purple in each case. Only heavy
atoms are shown except for hydrogens on the proteasome water molecule.
Superimposed in this diagram (see text) are Ser/ThrOcBLys2N (RMSD 0.33 Å), the
tetrahedral intermediate formation elements (see Scheme 3), from each structure.
(B) As above, with OBO(TyrO)Lys1N of the b-lactamase superimposed on OBO(H2-
O)ThrN of the proteasome (see text) (RMSD 0.68 Å), the tetrahedral intermediate
breakdown elements (see Scheme 3).
A common acylation mechanism (Scheme 2) can then be
proposed for the b-lactam recognizing enzymes (class C b-lacta-
mases and DD-peptidases) and the proteasome (Scheme 2). [Note
that class A and D b-lactamases, also serine enzymes, have
additional features in the active site to assist turnover of b-lactams;
they are therefore not included in this analysis]. These mechanisms
are supported by crystal structures of complexes of the respective
enzymes with boronic acid transition state analogs [7,14]. It should
be noted, however, that a mechanism of proteasome catalysis
employing the N-terminal threonine amine directly as a general
base, i.e. omitting the intervening water molecule, has also been
suggested and there is a boronic acid complex structure supporting
this idea [20]. It is possible that both mechanisms may be available
to the enzyme, depending on the structure of the peptide substrate.
Scheme 3), can be superimposed [OcBLys2N atoms of the two
structures, RMSD 0.33 Å] as shown in Fig. 1A. It is then noticeable
that Tyr 150 of the b-lactamase and the H2O of the proteasome are
on opposite sides of the boronate, hydrogen-bonded to stereo-
chemically distinct boronate oxygens. The simplicity of 13 and
14 remains, however, since the general acid elements of Scheme 2
(HA of Scheme 3) can also be superimposed. Thus, OBO(TyrO)Lys1N
of the b-lactamase can be superimposed on OBO(H2O)ThrN of the
proteasome (RMSD 0.68 Å) (Fig. 1B). These are the functional
groups putatively involved in the second step of the acylation reac-
tion, facilitating departure of the leaving group (Scheme 2). It
seems likely therefore, that a mechanistic imperative has forced
convergent evolution of the active site functional groups of the

S.A. Adediran et al. / FEBS Letters 586 (2012) 4186–4189
4189
Scheme 3. Participants in the two steps, acylation, red, and deacylation, green, cannot be superimposed all together but each can be seperately, as in Fig. 1A and B.
b-lactam-recognizing enzymes and the proteasome, despite their
different stereochemistry.
References
[1] Frére, J.-M., Ed., (2012). Beta-Lactamases, Nova, NY.
The stereochemistry of nucleophilic attack at an amide sub-
strate carbonyl group varies among amidohydrolases. Two en-
zymes with the same array of active site functional groups but
with different stereochemistries of attack (and usually with differ-
ent protein folds) are usually thought to arise from convergent
evolution. For example, the serine carboxypeptidases of Clan SC,
Family S10 [21] attack the Si face while the classical serine prote-
ases of Clan PA, Family S1 attack the Re [22]. The catalytic triad is,
however, believed to act in the same way in each family. A number
of serine/lysine amidohydrolases, including the proteasome, as
noted above, attack the Si face [23] while others, including the
b-lactam-recognizing enzymes and amidase signature enzymes
approach the Re face [19].
With respect to Scheme 2, it should be noted, as has been previ-
ously discussed for b-lactamases [14], that alternative mechanisms
of hydrolysis can be proposed on the basis of crystal structures of
boronate complexes. For example, the roles of the two lysines in
Scheme 2 can be reversed, so that Lys1 (RNH2) is the indirect general
base for nucleophilic attack by Ser1 (Thr) and a protonated Lys2
facilitates leaving group departure. The superimpositions suggested
by this mechanism, however, are not as close as those in Fig. 1.
A cross-check of inhibitors, the O-aryloxycarbonyl hydroxa-
mates, 1, and the 1,3,4-oxathiazol-2-ones, 2, shows that they have
a very similar mechanism of inhibition of b-lactamases and the
proteasome. These enzymes are representative members of two
distinct protein superfamilies, the b-lactam-recognizing enzymes
and N-terminal hydrolases, respectively. Examination of relevant
protein crystal structures has demonstrated the presence of a very
similarly placed array of analogous active site functional groups,
two amines and two hydroxyl groups, in these two groups of en-
zymes. A very similar catalytic mechanism for amide hydrolysis
is thus indicated. We now see that these common elements of ac-
tive site structure and mechanism are available to three serine
amidohydrolase superfamilies, the two above and the amidase sig-
nature enzymes [19]. The relationship between the active sites of
the above enzymes and those of other serine amidohydrolase
superfamilies, for example the classical serine proteases, has been
previously discussed [19].
[2] Drawz, S.R. and Bonomo, R.A. (2009) Three decades of beta-lactamase
inhibitors. Clin. Microbiol. Rev. 1, 160–201.
[3] Wyrembak, P.N., Babaoglu, K., Pelto, R.B., Shoichet, B.K. and Pratt, R.F. (2007)
O-Aryloxycarbonyl hydroxamates: new b-lactamase inhibitors that cross-link
the active site. J. Am. Chem. Soc. 129, 9548–9549.
[4] Pelto, R.B. and Pratt, R.F. (2008) Kinetics and mechanism of inhibition of a
serine b-lactamase by O-aryloxycarbonyl hydroxamates. Biochemistry 47,
12037–12046.
[5] Borissenko, L. and Groll, M. (2007) 20S proteasome and its inhibitors:
crystallographic knowledge for drug development. Chem. Rev. 107, 687–
717.
[6] Lin, G., Li, D., de Carvalho, L.P.S., Deng, H., Tao, H., Vogt, G., Wu, K., Schneider, J.,
Chidawanyika, T., Warren, J.D., Li, H. and Nathan, C. (2009) Inhibitors selective
for mycobacterial versus human proteasomes. Nature 461, 621–626.
[7] Hu, G., Lin, G., Wang, M., Dick, L., Xu, R.-M., Nathan, C. and Li, H. (2006)
Structure of the Mycobacterium tuberculosis proteasome and mechanism of
inhibition by a peptidyl boronate. Mol. Microbiol. 59, 1417–1428.
[8] Wyrembak, P.N. (2003) N,O-Diacyl hydroxylamines as b-lactamase inhibitors.
BA Thesis, Wesleyan University.
[9] Kelly, J.A. and Kuzin, A.P. (1995) The refined crystallographic structure of a DD-
peptidase penicillin-target enzyme at 1.6 Å resolution. J. Mol. Biol. 254, 233–
236.
[10] Groll, M., Ditzel, L., Löwe, J., Stock, D., Bochtler, M., Bartunik, H.D. and Huber, R.
(1997) Structure of 20S proteasome from yeast at 2.4 Å resolution. Nature 386,
463–471.
[11] Matagne, A., Dubus, A., Galleni, M. and Frére, J.-M. (1999) The b-lactamase
cycle: a tale of selective pressure and bacterial ingenuity. Nat. Prod. Rep. 16,
1–19.
[12] Fernollar-Ferrer, C., Frau, J., Donoso, J. and Munoz, F. (2008) Evolution of class
C
b-lactamases: factors influencing their hydrolysis and recognition
mechanism. Theor. Chem. Acc. 121, 209–218.
[13] Sauvage, E., Kerff, F., Terrak, M., Ayala, J.A. and Charlier, P. (2008) The
penicillin-binding proteins: structure and role in peptidoglycan biosynthesis.
FEMS Microbiol. Rev. 32, 234–258.
[14] Dzhekieva, L., Rocaboy, M., Kerff, F., Charlier, P., Sauvage, E. and Pratt, R.F.
(2010) Crystal structure of a complex between the Actinomadura R39 DD-
peptidase and a peptidoglycan-mimetic boronate inhibitor: interpretation of a
transition state analogue in terms of catalytic mechanism. Biochemistry 49,
6411–6419.
[15] Duggleby, H.J., Tolley, S.P., Hill, C.P., Dodson, E.J., Dodson, G. and Moody, P.C.E.
(1995) Penicillin acylase has a single amino acid catalytic centre. Nature 373,
264–268.
[16] Oinonen, C. and Rouvinen, J. (2000) Structural comparisons of Ntn-hydrolases.
Protein Sci. 2, 2329–2337.
[17] Galperin, G.Y. and Koonin, E.V. (2012) Divergence and convergence in enzyme
evolution. J. Biol. Chem. 287, 21–28.
[18] Goffin, C. and Ghuysen, J.-M. (1998) Multimodular penicillin-binding proteins:
an enigmatic family of orthologs and paralogs. Microbiol. Mol. Biol. Rev. 62,
1079–1083.
[19] Pratt, R.F. and McLeish, M.J. (2010) Structural relationship between the active
sites of b-lactam-recognizing and amidase signature enzymes: convergent
evolution? Biochemistry 49, 9688–9697.
Acknowledgments
[20] Groll, M., Berkers, C.R., Ploegh, H.L. and Ovaa, H. (2006) Crystal structure of the
boronic acid-based proteasome inhibitor bortezomib in complex with the
yeast 20S proteasome. Structure 14, 451–458.
[21] Rawlings, N.D., Barrett, A.J. and Bateman, A. (2010) MEROPS: the peptidase
database. Nucleic Acids Res. 38, D227–D233.
[22] Bullock, T.A., Breddam, K. and Remington, S.J. (1996) Peptide aldehyde
complexes with wheat serine carboxypeptidase II: implications for the
catalytic mechanism and substrate specificity. J. Mol. Biol. 255, 714–725.
[23] Paetzel, M., Dalbey, R.E. and Strynadka, N.C.G. (1998) Crystal structure of a
bacterial signal complex with a b-lactam inhibitor. Nature 396, 186–190.
We acknowledge the support and encouragement of Dr Carl F.
Nathan of the Department of Microbiology and Immunology, Weill
Cornell Medical College, New York, New York. This research was
supported by the National Institutes of Health (RFP, AI017986
and CN, AI080618). We also acknowledge the assistance of Dr. Hui-
lin Li and Dr. Shaoqing Yang at the Brookhaven National Laboratory
with structural studies.