A fluoroacetamidine-based inactivator of protein arginine deiminase 4: Design, synthesis, and in vitro and in vivo evaluation

Article abstract of DOI:10.1021/ja0576233

Protein arginine deiminase 4 (PAD4) is a calcium-dependent transcriptional corepressor that has been implicated in the onset and progression of rheumatoid arthritis. Herein we describe the synthesis and in vitro evaluation of a fluoroacetamidine-containing compound, N-α-benzoyl-N⁵-(2-fluoro-1-iminoethyl)-l-ornithine amide, 1, hereafter referred to as F-amidine, that is the most potent PAD4 inhibitor ever described. Additional studies described herein indicate that F-amidine can also inhibit PAD4 activity in vivo. The bioavailability of this compound suggests that F-amidine will be a powerful chemical probe of PAD4 function that can be used to dissect the roles of this enzyme in both rheumatoid arthritis and transcriptional control. The fact that inhibition is of an irreversible nature suggests that, with appropriate functionalization, F-amidine analogues will be robust activity-based protein-profiling and proteomic capture reagents. Copyright

Full text of DOI:10.1021/ja0576233

Published on Web 01/07/2006
A Fluoroacetamidine-Based Inactivator of Protein Arginine Deiminase 4:
Design, Synthesis, and in Vitro and in Vivo Evaluation
Yuan Luo, Bryan Knuckley, Young-Ho Lee, Michael R. Stallcup, and Paul R. Thompson*
Department of Chemistry & Biochemistry, UniVersity of South Carolina, 631 Sumter Street,
Columbia, South Carolina 29208, and Department of Biochemistry & Molecular Biology, UniVersity of Southern
California, 1333 San Pablo Street, MCA 51A, Los Angeles, California 90089
Received November 8, 2005; E-mail: thompson@mail.sc.edu
Rheumatoid arthritis (RA) is a chronic autoimmune disease
affecting ∼1% of the worldwide population, and while a number of
pharmaceuticals have been used to treat this disease, most only target
its symptoms and not the underlying causes of the disease. Recent
biochemical and genetic evidence has strongly suggested that a dys-
Figure 1. Reaction catalyzed by PAD4. R ) peptide backbone.
regulated protein arginine deiminase (PAD) activity (Figure 1), in
particular protein arginine deiminase 4 (PAD4), contributes to the
onset and progression of RA.1-3 A role for PAD2 in the patho-
physiology of multiple sclerosis has also been described.4 Ad-
ditionally, PAD4 activity has been implicated in the control of gene
transcription, primarily through its ability to catalyze the deimination
of Arg residues present in histones H2A, H3, and H4, as well as
p300, a transcriptional coactivator.5-7 However, the underlying
Figure 2. Structure of F-amidine and benzoyl arginine amide (BAA), a
nonphysiological PAD4 substrate.
mechanisms of this activity are only beginning to be understood.
Because of its potential as a therapeutic target for the treatment
of RA, we initiated studies to characterize the molecular mechanism
of PAD4 and to develop PAD4 inhibitors. Herein, we describe the
N-R-benzoyl-N5-(2-fluoro-1-iminoethyl)-L-ornithine amide, 1, here-
after referred to as F-amidine (Figure 2), the most potent PAD
inhibitor described to date.
Although the catalytic mechanism of PAD4 is not fully estab-
lished, it is clear from structural studies8 and precedents from related
systems9,10 that PAD4 utilizes an active site Cys (Cys645) to
catalyze the hydrolytic deimination of Arg residues through a
proposed amidino-Cys intermediate.11 Because the proposed mech-
anism is analogous to that employed by Cys proteases and because
fluoromethyl ketones have long been used as covalent inactivators
of Cys proteases, we hypothesized that such compounds designed
to mimic the basic structure of small-molecule PAD substrates (i.e.,
benzoylated arginines)11 would be reasonable inactivators of PAD4.
However, these compounds have several less than desirable char-
acteristics, including the fact that they would not be isosteric, they
lack H-bond donors (vide infra), and they lack positive charge.12
Therefore, we hypothesized that a fluoroacetamidine-containing
compound based on the structure of a benzoylated Arg (Figure 2)
would react with the active site Cys in PAD4 (Cys645) analogously
to a fluoromethyl ketone (Figure 3), but would in fact provide super-
ior inactivation kinetics because F-amidine is positively charged,
it closely mimics the structure of Arg, and it possesses potential
H-bond donors for both Asp350 and Asp473, two active-site
residues that are important for substrate recognition and catalysis.
The synthesis of F-amidine, which is described in detail in the
Supporting Information, utilized a solid-phase synthetic methodol-
ogy that involved the on-resin coupling of an ethyl fluoroacetimidate
hydrochloride to N-R-benzoyl ornithine (Scheme S1). The com-
pound was obtained in >90% purity after cleavage from the resin
and was further purified by reverse-phase HPLC.
Figure 3. Potential mechanism of PAD4 inactivation by F-amidine.
IC50 of F-amidine is 21.6 ( 2.1 µM when preincubated with Ca2+
versus an IC50 of >200 µM when preincubated in the absence of
Ca2+, thereby indicating that F-amidine inactivation, vide infra, of
PAD4 is Ca2+-dependent. This observation is consistent with the
hypothesis that F-amidine reacts with an active-site residue because
PAD4 undergoes a conformational change upon Ca2+ binding that
reorients active-site residues into positions that are competent for
catalysis.8,11 In contrast to the potent inhibition observed for PAD4,
F-amidine does not inhibit the methyltransferase activity of either
PRMT1 or PRMT4, two Arg methyltransferases whose regiospeci-
ficity overlaps with that of PAD4 (not shown); thereby demonstrat-
ing that F-amidine selectively inhibits this Arg-modifying enzyme.
To determine whether F-amidine possesses time-dependent
inhibition properties, time courses of product formation were
monitored in either the absence or presence of this compound. The
results of these experiments indicate that F-amidine does in fact
display time-dependent inhibition properties (Figure 4A); that is,
the progress curves are nonlinear and reach a plateau value, where
Vs ) 0. On the basis of these results, rapid dilution time course
experiments (Figure 4B) were used to differentiate between
reversible and irreversible inhibition. The results of these experi-
ments indicate that there is no recovery13 of activity upon dilution
of the preformed PAD4‚F-amidine complex into assay buffer
containing high concentrations of substrate (∼7.5 × Km), thereby
indicating that F-amidine is an irreversible inactivator of PAD4.
Consistent with this proposal is the fact that activity is not recovered
even after dialyzing inactivated PAD4 for 20 h.
To evaluate the potency of F-amidine, IC50’s were determined
by preincubating this compound with PAD4 in the absence or
presence of Ca2+ for 15 min prior to assaying (Figure S1). The
9
1092
J. AM. CHEM. SOC. 2006, 128, 1092-1093
10.1021/ja0576233 CCC: $33.50 © 2006 American Chemical Society

C O M M U N I C A T I O N S
Figure 4. F-Amidine is an irreversible inactivator of PAD4. (A) Plots of
product formation versus time in the absence and presence of increasing
concentrations of F-amidine. (B) Rapid dilution of preformed PAD4‚F-
amidine complexes into assay buffer containing excess substrate. (C)
Figure 5. F-Amidine inhibits the p300GBD-GRIP1 interaction in CV-1
cells in a concentration-dependent manner. A mammalian two-hybrid assay
was used to monitor interactions between GRIP1 and the p300GBD and
the enhancement afforded by either wild-type or mutant PAD4. F-Amidine
was added to the cell culture medium at the concentrations (in µM) indicated
in the figure.
Concentration dependence of kobs
.
F-Amidine inactivates PAD4 in both a time- (as noted above)
and concentration-dependent manner (Figure 4A). The KI and kinact
values for this process are 330 ( 90 µM and 1.0 ( 0.1 min-1
,
groups, F-amidine derivatives will be robust activity-based protein-
profiling and proteomic capture reagents. Such compounds will be
useful for identifying the in vivo conditions under which this
enzyme is activated and isolating activated PAD4, thereby enabling
the identification of the numbers and types of post-translational
modifications that occur to this enzyme during PAD4 activation in
vivo. Finally, the fact that the synthesis of F-amidine relies on facile
chemistry involving the coupling of an amine to an acetimidate
hydrochloride indicates the ease of generating large, structurally
diverse libraries containing this warhead, thereby leading to the
identification of additional PAD4 inactivators with improved
potency and selectivity.
Acknowledgment. This work was supported by startup funds
provided to P.R.T. from the USC Research Foundation and to
M.R.S. from NIH Grant No. DK55274.
Supporting Information Available: Additional references and
complete author lists for refs 1, 5, and 7, materials, methods, and
spectral characterization of F-amidine. This material is available free
of charge via the Internet at http://pubs.acs.org.
respectively, yielding a robust kinact/KI of 3000 M-1 min-1 (Figure
4C). Note that the rate of PAD4 inactivation by F-amidine can be
decreased by increasing the concentration of substrate in the
inactivation assays (Figure S2). Substrate protection of this type is
consistent with our observation that inactivation is Ca2+-dependent
and strongly suggests that F-amidine inactivates PAD4 by co-
valently modifying an active-site residue.
Although the site of modification in PAD4 has yet to be es-
tablished, the fact that inactivation is substrate- and Ca2+-dependent
is consistent with the modification of an active-site residue, and
on the basis of similarities between the fluoroacetamidine warhead
described herein and 2-chloroacetamidine, which covalently modi-
fies the active-site Cys in a PAD4 paralogue, dimethylarginine
dimethylamino hydrolase (DDAH),14 the most likely site of
modification in PAD4 is Cys645, the active-site nucleophile.
The inhibitory properties of F-amidine were also evaluated in
vivo using a mammalian two-hybrid assay that uses a luciferase
reporter assay to monitor the interaction between the p300 GRIP1
binding domain (p300GBD) and GRIP1, an estrogen receptor
transcriptional coactivator6-cotransfection of PAD4, which results
in the deimination of the p300GBD and enhances its interaction
with GRIP1.6 Treatment of the cells with F-amidine reduces the
p300GBD and GRIP1 interaction as the concentration of this
compound is increased, with maximal inhibition being noted at 200
µM (Figure 5). In contrast, control experiments in which the
Cys645Ser mutant was transfected into cells in place of wild-type
enzyme show minimal inhibition, thereby indicating that the
decrease in luciferase activity is not a nonspecific effect. And, while
the mechanism of transport into these cells is not known, these
results are particularly impressive because they indicate that
F-amidine is bioavailable.
References
(1) Suzuki, A.; et al. Nat. Genet. 2003, 34, 395-402.
(2) Vossenaar, E. R.; Zendman, A. J.; van Venrooij, W. J.; Pruijn, G. J.
Bioessays 2003, 25, 1106-1118.
(3) Hill, J. A.; Southwood, S.; Sette, A.; Jevnikar, A. M.; Bell, D. A.; Cairns,
E. J. Immunol. 2003, 171, 538-541.
(4) Moscarello, M. A.; Pritzker, L.; Mastronardi, F. G.; Wood, D. D. J.
Neurochem. 2002, 81, 335-343.
(5) Wang, Y.; et al. Science 2004, 306, 279-283.
(6) Lee, Y. H.; Coonrod, S. A.; Kraus, W. L.; Jelinek, M. A.; Stallcup, M. R.
Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 3611-3616.
(7) Cuthbert, G. L.; et al. Cell 2004, 118, 545-553.
(8) Arita, K.; Hashimoto, H.; Shimizu, T.; Nakashima, K.; Yamada, M.; Sato,
M. Nat. Struct. Mol. Biol. 2004, 11, 777-783.
(9) Lu, X.; Galkin, A.; Herzberg, O.; Dunaway-Mariano, D. J. Am. Chem.
Soc. 2004, 126, 5374-5375.
(10) Das, K.; Butler, G. H.; Kwiatkowski, V.; Clark, A. D., Jr.; Yadav, P.;
Arnold, E. Structure 2004, 12, 657-667.
The development of inhibitors targeting PAD4 is critical for
validating this enzyme as a drug target; therefore we designed and
synthesized F-amidine, an irreversible inactivator of PAD4. In
comparison to other known PAD inhibitors, for example, taxol (IC50
(11) Kearney, P. L.; Bhatia, M.; Jones, N. G.; Luo, Y.; Glascock, M. C.;
Catchings, K. L.; Yamada, M.; Thompson, P. R. Biochemistry 2005, 44,
10570-10582.
(12) IC50 for N-R-benzoyl-N5-(2-fluoro-1-ketoethyl)-L-ornithine amide is . 500
µM.
≈ 5 mM15) and 2-chloroacetamidine (KI ) 20 ( 5 mM; kinact
)
(13) Because the line representing the F-amidine rapid dilutions experiments
fits well to a linear fit, the small amount of product formation observed
after extended incubations is most likely due to the presence of a small
amount of active PAD4 (<5%, based on comparing the slopes of the two
lines) that was not inactivated during the preincubation of PAD4 with
Ca2+ and F-amidine.
0.7 ( 0.1 min-1; kinact/KI ) 35 M-1 min-1), F-amidine is the most
potent PAD4 inhibitor described to date.14 The bioavailability of
F-amidine suggests that this compound will be a powerful chemical
probe that can be used to discern the role of PAD4 in the various
signaling pathways (e.g., transcription control and differentiation)
in which its activity has been implicated.
(14) Stone, E. M.; Schaller, T. H.; Bianchi, H.; Person, M. D.; Fast, W.
Biochemistry 2005, 44, 13744-13752.
(15) Pritzker, L. B.; Moscarello, M. A. Biochim. Biophys. Acta 1998, 1388,
154-160.
Additionally, the fact that F-amidine forms a covalent linkage
to PAD4 indicates that, with the addition of appropriate functional
JA0576233
9
J. AM. CHEM. SOC. VOL. 128, NO. 4, 2006 1093