Probing the elusive catalytic activity of vertebrate class IIa histone deacetylases

Article abstract of DOI:10.1016/j.bmcl.2008.02.025

It has been widely debated whether class IIa HDACs have catalytic deacetylase activity, and whether this plays any part in controlling gene expression. Herein, it has been demonstrated that class IIa HDACs isolated from mammalian cells are contaminated with other deacetylases, but can be prepared cleanly in Escherichia coli. These bacteria preparations have weak but measurable deacetylase activity. The low efficiency can be restored either by: mutation of an active site histidine to tyrosine, or by the use of a non-acetylated lysine substrate, allowing the development of assays to identify class IIa HDAC inhibitors.

Full text of DOI:10.1016/j.bmcl.2008.02.025

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 1814–1819
Probing the elusive catalytic activity of vertebrate
class IIa histone deacetylases
Philip Jones,* Sergio Altamura, Raffaele De Francesco, Paola Gallinari, Armin Lahm,
Petra Neddermann, Michael Rowley, Sergio Serafini and Christian Steinkuhler
¨
IRBM/Merck Research Laboratories, Via Pontina km 30,600, 00040 Pomezia, Italy
Received 18 November 2007; revised 7 February 2008; accepted 9 February 2008
Available online 14 February 2008
Abstract—It has been widely debated whether class IIa HDACs have catalytic deacetylase activity, and whether this plays any part
in controlling gene expression. Herein, it has been demonstrated that class IIa HDACs isolated from mammalian cells are contam-
inated with other deacetylases, but can be prepared cleanly in Escherichia coli. These bacteria preparations have weak but measur-
able deacetylase activity. The low efficiency can be restored either by: mutation of an active site histidine to tyrosine, or by the use of
a non-acetylated lysine substrate, allowing the development of assays to identify class IIa HDAC inhibitors.
Ó 2008 Elsevier Ltd. All rights reserved.
The acetylation status of lysine residues on histone
tails, and an increasing number of non-histone sub-
strates (including transcription factors, heat shock and
structural proteins), is tightly controlled by two counter-
acting enzyme families, histone acetyl transferases
(HATs) and histone deacetylases (HDACs).1,2 This
dynamic process has a crucial role in chromatin struc-
ture and hence gene transcription, whereby the presence
of acetyl groups (Ac-) on these lysine residues neutral-
izes the positive charges of the histone tails thereby
decreasing their interaction with DNA, relaxing the
chromatin, and allowing access to transcription factors.
In contrast, removal of the Ac-groups condenses the
chromatin, leading to transcriptional repression. Inter-
est in the HDACs has been stimulated due to discovery
that they control key processes such as skeletal and mus-
cle formation, cardiac hypertrophy, T-cell differentia-
tion and neuronal survival, and are deregulated in
neoplasias.3 Consequently, HDAC inhibitors (HDACi)
are being investigated and vorinostat (ZolinzaÒ, for-
merly known as SAHA) recently became the first clini-
cally approved HDACi.4
350 amino acids harbouring the known or putative
HDAC catalytic domain.5,6 These two classes can be
distinguished from a third, mechanistically distinct class,
the sirtuin family.7 Class II HDACs are further divided
into sub-classes IIa and IIb, the former (HDACs 4, 5,
7 + 9) contain an N-terminal regulatory domain.6
Although the role of class IIa HDACs in tissue-specific
gene regulation is well documented,8–11 the contribution
of the catalytic domain to this activity is controversial.
A substantial body of evidence illustrates that class IIa
HDACs exert transcriptional repression through pro-
tein–protein interactions using the N-terminal domain,
while the catalytic domain may not be required.3,6,12
For instance, a natural HDAC9 splice variant lacking
the catalytic domain retains full transcriptional repres-
sive functions.13 In parallel, several groups have ques-
tioned whether class IIa HDACs possess any intrinsic
deacetylase activity, hypothesizing that any activity
may be due to the presence of associated HDACs in
multi-protein complexes.6,12 Indeed, HDACs 4, 5 + 7
have been shown to associate with HDAC3.14,15 Fur-
thermore, attempts to impair the activity of HDAC4
by site-directed mutagenesis in the catalytic site were
inconclusive as mutants showing impaired deacetylase
activity lost the ability to interact with HDAC3.15 Also,
attempts to obtain active recombinant class IIa HDACs
have been unsuccessful.16
The HDAC superfamily of enzymes can be divided into
two large groups (classes I + II) based on characteristic
conserved sequence motifs within a domain of about
Concurrently with work to develop subtype selective
HDACi’s (see following paper), we became interested
in establishing whether class IIa HDACs demonstrate
Keywords: Histone deacetylase; HDAC.
*
Corresponding author. Tel.: +39 0691093559; fax: +39 0691093654;
e-mail: philip_jones@merck.com
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.02.025

P. Jones et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1814–1819
1815
deacetylase activity, using HDAC4 as a representative
example, with a view to establishing a screening plat-
form if activity was detected.17
vated ester 5 completed the synthesis. The two
compounds displayed nanomolar affinities against class
I HDACs 1 and 3, whilst also inhibiting flag-tagged
HDAC4 preparation with similar potency (Table 1). In
contrast, both compounds lost some activity against
HDAC6, with 4 displaying substantially reduced activity
on this isoform having IC50 = 230 nM. An inactive con-
trol 6 was also prepared and revealed not to inhibit any
isoforms.
Initially, flag-tagged HDACs were prepared and purified
from mammalian cells. C-terminally flag-tagged
HDACs 1, 4 and 6 were expressed and purified from
HEK293 cells, while in order to obtain functional
HDAC3 it was necessary to co-express and co-immuno-
purify flag-tagged HDAC3 together with a GAL4 DBD-
fusion of the N-CoR deacetylase activation domain
from the same cell line.17
With these reagents in hand, two parallel experiments
were conducted: UV cross-linking to the HDAC iso-
forms and analysis of the covalently bound proteins,
and pull-down experiments using streptavidin-coated
beads pre-adsorbed with these biotinylated probes to
capture the HDACs.17,18
In order to demonstrate that indeed we were measuring
and inhibiting purified flag-tagged HDAC isoforms two
structurally distinct activity probes, 4 and 9, were pre-
pared, based on HDACi’s Apicidin and NVP-LAQ824
(Scheme 1). Each probe carried a photo-activable
cross-linking moiety and a biotin residue. The probes
were prepared appending long side chains to portions
of the molecule which were expected to be away from
the active site and solvent exposed. The hydroxamic acid
probe 9 was synthesized by replacing the hydroxylethyl
side chain of NVP-LAQ824 by an aminohexyl chain
and then coupling the hydroxy-succinimide ester 5 bear-
ing both the phenyl azide cross-linking group and a bio-
tin residue. In contrast, the Apicidin based probe 4 was
prepared by firstly alkylating the indole moiety with
methyl bromoacetate, followed by hydrolysis and addi-
tion of the aminohexyl chain. Coupling with the acti-
Unsurprisingly, given their potency both probes cross-
linked HDACs 1 and 3 (Table 1), whereas the negative
control 6 lacking the HDACi failed to cross-link to
either of these deacetylases. In contrast, only the
hydroxamic acid probe 9 cross-linked significantly with
HDAC6. The Apicidin based probe 4 only showed weak
cross-linking, in line with its reduced potency on
HDAC6, IC50 = 230 nM compared to 9. Instead, the
cross-linking and pull-down data obtained with flag-
tagged HDAC4 expressed in HEK293 cells did not re-
flect the inhibition constants. While both probe com-
pounds showed similar inhibition of this enzyme
preparation, only the hydroxamic acid probe 9 effec-
1) H2, Pd/C
1) H2N(CH2)6NHCbz
O
2) NaH, Bu4NI,
MeO2CCH2Br, 80oC
3) LiOH
O
O
EDC, HOBT, iPr2NEt
O
O
O
2) H2, Pd/C, DMF
N
N
H
N
H
H
N
NH
N
NH
N
NH
H
H
H
N
N
N
N
N
N
H
O
OMe
O
O
CO2H
O
N
NH2
O
O
O
O
O
O
1
2
3
O
O
O
N
H
NH
HN
N
NH
H
N
O
O
N
H
H
5, iPr2NEt, DMF
N
S
N
O
S
N
S
N
O
H
H
O
O
HN
O
O
4
O
O
N3
NH
HN
HN
NH
O
O
O
O
H
H
S
N
N
S
S
S
X
S
N
H
N
S
NH
H
HN
O
O
O
O
NH
O
O
N
5: X =
O
SO3Na
OH
O
N
6: X = OMe
N3
N3
H
O
N
BocHN
1) NaBH(OAc)3
2) LiOH
H2N
HN
5, DMF
9
3) PyBOP, THPO-NH2
4) HCl, MeOH
5) TFA/DCM
O
O
OH
OMe
N
H
H
N
N
7
8
HN
HN
Scheme 1. Activity probes 9 and 4, based on NVP-LAQ824 (hydroxyethyl has been replaced by an aminohexyl chain) and Apicidin (1), respectively.

1816
P. Jones et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1814–1819
Table 1. IC50’s of HDACi probes on HDAC isoforms from mammalian cells, and results of UV cross-linking and pull-down experiments17
IC50 (nM)
UV cross-linking
Pull-down of HDAC4
HDAC1b HDAC3b HDAC4c HDAC6b HDAC1 HDAC3 HDAC4 HDAC6
LAQ
Probe 9
73
20
42
6
4
5
98
Yes
Yes
Yes
Yes
Yes Competed by
NVP-LAQ824, but not
Apicidin or MS-275
No
Apicidin
Probe 4
230
Yes
No
Yes
No
No
No
Weak
No
Negative Control 6 NA at
1 lM
NA at
1 lM
NA at
1 lM
NA at
1 lM
No
aValues are means >2 experiments (std. dev. were within 30% of the IC50 values) measured with flag-tagged enzyme from HEK293 cells, using b‘Fluor
de Lys’ substrate21 or [ H]-Ac histones.
c 3
tively cross-linked HDAC4.17 No cross-linking was seen
with the Apicidin probe 4. Similarly, the hydroxamic
acid probe 9, but not 4, was able to pull-down HDAC4.
Pre-incubation of the HDAC4 preparation with a 10-
fold excess of NVP-LAQ824 was able to prevent pull-
down of HDAC4 by 9, but this could not be competed
with an excess of either Apicidin or MS-275. These data
suggest that only the hydroxamic acid 9 binds to
HDAC4. The lack of correlation between cross-linking,
pull-down and inhibition experiments also points to the
conclusion that HDAC4 does not, or does only to a
marginal extent, contribute to the deacetylase activity
of the HDAC4 complex from HEK293 cells.
ing processes. Potentially two scenarios could be envis-
aged to account for this inefficiency in deacetylase
activity: lack of an optimal transition state in the active
site of the enzyme, or the use of an inappropriate
substrate.17
Beside from the presence of N-terminal extension of
class IIa HDACs, another significant difference is the
substitution of a tyrosine residue present in the catalytic
domain of class I HDACs with a histidine residue
(Fig. 1a). This tyrosine residue is crucial to the proposed
deacetylase mechanism and is believed to act as a tran-
sition-state stabilizer.19 Indeed, in crystal structures with
bound inhibitors this residue makes a H-bond with the
carbonyl of the HDACi.19,20 This Y–H substitution is
surprising as molecular modeling suggests that class
IIa enzymes would be unable to make a similar H-bond
without substantial structural reorganization. Indeed,
such an H-bond would be substantially longer, and
therefore the class IIa HDACs should be expected to
be less effective in processing Ac-lysines substrates
(Fig. 1b).
Having demonstrated that purified HDAC class IIa iso-
forms could not be isolated from mammalian cells atten-
tion turned to Escherichia coli, that lack histones and
endogenous HDACs. Accordingly the N-terminally
truncated HDAC4 catalytic domain (CD) (starting at
Thr653) was expressed, purified to homogenity and
tested for activity on [3H]acetyl histones cores.
Although deacetylase activity was seen with this enzyme,
it was only modest, requiring 1 lM enzyme to see sub-
stantial conversion (Table 2). Intriguingly this activity
could be inhibited with NVP-LAQ824 but not Apicidin,
substantiating the previous findings in the cross-linking
and pull-down experiments.
In an effort to restore activity to HDAC4 it was
decided to reintroduce the tyrosine residue in the
hope of developing a ‘gain of function’ mutation.
Accordingly, HDAC4 CD containing this H976Y
mutation was generated in E. coli and its activity
tested on a [3H]acetyl histone substrate. A 1000-fold
gain in activity was observed and activity of this mu-
tant could be observed with nanomolar levels of en-
zyme (Table 2). Furthermore, optimization of the
Following this observation that HDAC4 does indeed
have catalytic activity, albeit weak, interest arose in
developing a screening platform to be able to identify
isoform selective HDACi’s. However, routine screening
with high enzyme concentrations would not be practical
and necessitated the evolution of more efficient screen-
protocol allowed use of the commercial ‘Fluor de
21
Lys’ substrate,
and this HDAC4 ‘gain of function’
(GOF) permitted IC50’s to be determined, albeit using
a mutated enzyme but a natural Ac-lysine substrate
(Table 3).22
Table 2. Deacetylation of [3H]acetyl histones cores by HDAC4WT or
HDAC4(H976Y) ‘Gain of function’ catalytic domains from E. coli
The second hypothesis to explain the weak of activity of
HDAC4 is that Ac-lysine residues are not the natural
substrates. Indeed HDACs have been proposed and/or
reported to process a number of other substrates.23,24
Consequently a small library of potential substrates
was prepared from Boc-L-Lys-MCA by routine acyla-
tion reaction (Fig. 2), and the conversion of these by
HDACs 1, 3, 4, 5, 6 and 7 was measured. The acyl
groups selected were based on those known to be acti-
vated by coenzyme A (CoA), the rationale being that
HATs use Ac-CoA to acetylate lysine residues. Incuba-
Conditions
% Deacetylation in 4 ha
1 lM HDAC4
1 lM HDAC4WT + 10 lM
NVP-LAQ824
12
3
1 lM HDAC4WT + 10 lM
Apicidin
12
1 nM HDAC4(H976Y)
10 nM HDAC4(H976Y)
100 nM HDAC4(H976Y)
33
45
65
a Values are means >2 experiments.

P. Jones et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1814–1819
1817
Figure 1. Active site geometries of (a) class I and (b) class IIa HDACs showing impact of the Y–H substitution and lengthening of H-bond to bound
inhibitor.
Table 3. IC50’s of HDACi probes on HDAC isoforms
IC50 (nM)
HDAC1b,d
HDAC3b,d
HDAC4GOFb,e
HDAC4WTc,e
HDAC6b,d
Vorinostat
NVP-LAQ824
PXD101
30
57
540
4.4
13% inhibition at 10 lM
420
43
8.3
2.6
18
3.6
46
24
20% inhibition at 2 lM
NA at 10 lM
NA at 10 lM
15
MS-275
MGCD-0103
120
130
400
610
NA at 10 lM
45% inhibition
at 5 lM
3600
NA at 10 lM
32% inhibition
at 2 lM
4300
Apicidin
23
9.7
8700
33
O
O
F
F
F
N
H
1400
120
200
1700
O
O
11
O
N
H
NHMe
12
10
NA at 10 lM
NA at 10 lM
ND
12
c
aValues are means >2 experiments, standard deviations were within 30% of the IC50 values. Measured using the bFluor de Lys or 10a substrates.
e
HDACs expressed in dHEK293 cells or E. coli.
tions of these substrates revealed that only 10b, 10c, 10g,
10n, 10q and 10r where deacylated to a minor extent by
HDACs 1, 3 and 6. However, trifluoroacetamide 10a
was efficiently processed not just by sub-nanomolar lev-
els of HDAC4, but also by the other two class IIa
HDACs 5 and 7 (Fig. 2). This substrate proved to be
poorly processed by class I HDACs demonstrating the
subtle differences between the classes. This finding per-
mitted development of a second HDAC4 screening as-
say: HDAC4WT, albeit using the wild-type enzyme
but an ‘unnatural’ substrate.25
the other hydroxamic acid HDACi’s (vorinostat and
PXD101) show lack of selectivity but interestingly ap-
pear to be more potent inhibitors of the H976Y mu-
tant than the wild-type enzyme. The basis for this
selectivity between HDAC4GOF and WT is not
understood, but the rank-order of the inhibitors is
maintained. Both Apicidin and aminobenzamides, like
MS-275 and MGCD-0103, failed to inhibit either
HDAC4 preparation in line with the cross-linking
and pull-down experiments. However, an interesting
observation is that the trifluoromethyl ketone class
of HDACi, exemplified by 11,26 appears to inhibit
both wild-type and ‘gain of function’ HDAC4. Inter-
estingly, this appears to be a specific feature of the tri-
fluoromethyl ketones, as a related class of electron
deficient ketones, the ketoamides exemplified by 12,27
failed to inhibit either form of HDAC4.
The development of these two assays allowed the
benchmark compounds to be profiled (Table 3).
Unsurprisingly based on the previous data NVP-
LAQ824 is able to inhibit both HDAC4 WT and ‘gain
of function’, along with the other HDACs. Similarly

1818
P. Jones et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1814–1819
O
H
N
R
(S)
O
O
N
H
NHBoc
O
F
F
F
R
CH3
C2H5
C3H7
CO2H
10f
=
O
CH3
10e
O
O
O
O
O
O
O
10a
10b
10c
10d
10g
CH3
CH3
CH3
CH3
CH3
CH3
OH
CH3
CH3
CH3
OH
CH3
CH3
10i
O
O
O
O
O
O
O
O
O
10j
10h
10k
10l
10m
10n
CH3
CH3
CH3
CH3
CH3
CH3
CO2H
CO2H
10r
CO2H
O
O
O
OH
10q
O
O
O
10p
10o
10s
10t
60000
50000
40000
30000
20000
10000
0
HDAC1
HDAC3
HDAC4
HDAC5
HDAC6
HDAC7
Figure 2. (a) Synthetic substrates from Boc-L-Lys-MCA; (b) conversion of 10a by HDAC isoforms showing class IIa selectivity.
7. Blander, G.; Guarente, L. Annu. Rev. Biochem. 2004, 73,
417.
8. Zhang, C. L.; McKinsey, T. A.; Chang, S.; Antos, C. L.;
Hill, J. A.; Olson, E. N. Cell 2002, 110, 479.
9. Vega, R. B.; Matsuda, K.; Oh, J.; Barbosa, A. C.; Yang,
X.; Meadows, E.; McAnally, J.; Pomajzl, C.; Shelton, J.
M.; Richardson, J. A.; Karsenty, G.; Olson, E. N. Cell
2004, 119, 555.
10. Chang, S.; McKinsey, T. A.; Zhang, C. L.; Richardson, J.
A.; Hill, J. A.; Olson, E. N. Mol. Cell Biol. 2004, 24, 8467.
11. Dequiedt, F.; Kasler, H.; Fischle, W.; Kiermer, V.;
Weinstein, M.; Herndier, B. G.; Verdin, E. Immunity
2003, 18, 687.
12. Yang, X. J.; Gregoire, S. Mol. Cell Biol. 2005, 25, 2873.
13. Zhou, X.; Richon, V. M.; Rifkind, R. A.; Marks, P. A.
Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 1056.
14. Fischle, W.; Dequiedt, F.; Fillion, M.; Hendzel, M. J.;
Voelter, W.; Verdin, E. J. Biol. Chem. 2001, 276, 35826.
15. Fischle, W.; Dequiedt, F.; Hendzel, M. J.; Guenther, M.
G.; Lazar, M. A.; Voelter, W.; Verdin, E. Mol. Cell 2002,
9, 45.
In conclusion, it has been demonstrated that the enzy-
matic activity of class IIa HDACs expressed in mamma-
lian cells is due to the presence of contaminating
deacetylases, likely to be endogenous class I HDACs
present in the class IIa complex. When pure class IIa
HDACs can be isolated from bacteria they possess a
weak but measurable intrinsic deacetylase activity, the
low catalytic efficiency of which is due to the presence
of a unique H residue. Finally the low efficiency can
be restored and measured either by an H–Y mutation
in the active site to give a ‘gain of function’, or by the
use of an non-Ac-lysine ‘unnatural’ substrate and the
wild-type enzyme.
References and notes
1. Cheung, W. L.; Briggs, S. D.; Allis, C. D. Curr. Opin. Cell
Biol. 2000, 12, 326.
16. Wang, A. H.; Bertos, N. R.; Vezmar, M.; Pelletier, N.;
Crosato, M.; Heng, H. H.; Th’ng, J.; Han, J.; Yang, X. J.
Mol. Cell Biol. 1999, 19, 7816.
2. Minucci, S.; Pelicci, P. G. Nat. Rev. Cancer 2006, 6,
38.
3. Martin, M.; Kettmann, R.; Dequiedt, F. Oncogene 2007,
26, 5450.
4. Grant, S.; Easley, C.; Kirkpatrick, P. Nat. Rev. Drug
Discov. 2007, 6, 21.
17. Lahm, A.; Paolini, C.; Pallaoro, M.; Nardi, M. C.; Jones,
P.; Neddermann, P.; Sambucini, S.; Bottomley, M. J.; Lo
´
Surdo, P.; Carfı, A.; Koch, U.; De Francesco, R.;
Steinkuhler, C.; Gallinari, P. Proc. Natl. Acad. Sci.
¨
U.S.A. 2007, 104, 17335.
5. De Ruijter, A. J. M.; van Gennip, A. H.; Caron, H. N.;
Kemp, S.; van Kuilenburg, A. B. P. Biochem. J. 2003, 370,
737.
6. Verdin, E.; Dequiedt, F.; Kasler, H. G. Trends Genet.
2003, 19, 286.
18. In the former, flag-tagged HDACs were incubated with the
probe moiety for 2 h followed by UV irradiation. The
biotinylated compounds were then separated by SDS–

P. Jones et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1814–1819
1819
PAGE and analyzed by Western blot using alkaline
phosphatase conjugated ExtrAvidin. While in the latter,
streptavidin-coated magnetic beads pre-treated with the
probes were incubated for 3 h at RT with flag-tagged
HDAC4. The bound proteins were detached from the bead,
separated and visualized by Coomassie-Blue staining.
19. Finnin, M. S.; Donigian, J. R.; Cohen, A.; Richon, V. M.;
Rifkind, R. A.; Marks, P. A.; Breslow, R.; Pavletich, N. P.
Nature 1999, 401, 188.
23. Riester, D.; Wegener, D.; Hildmann, C.; Schwienhorst, A.
Biochem. Biophys. Res. Commun. 2004, 324, 1116.
24. Chen, Y.; Sprung, R.; Tang, Y.; Ball, H.; Sangras, B.;
Kim, S. C.; Falck, J. R.; Peng, J.; Gu, W.; Zhao, Y. Mol.
Cell. Proteom. 2007, 6, 812.
25. DMSO/compound solution was incubated for 10 min with
His-tagged HDAC4 CD(653–1084) from E. coli in assay
buffer (25 mM Tris/HCl, pH 8, 137 mM NaCl, 2.7 mM
KCl, 1 mM MgCl2, 0.1 mg/ml BSA), 10a substrate solu-
tion was added and left for 1 h at 37 °C and the reaction
stopped by adding developer21/TSA solution. Measure the
fluorescence at ex.360 nM/em.460 nM.
20. Vannini, A.; Volpari, C.; Filocamo, G.; Casavola, E. C.;
Brunetti, M.; Renzoni, D.; Chakravarty, P.; Paolini, C.;
De Francesco, R.; Gallinari, P.; Steinkuhler, C.; Di
¨
Marco, S. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 15064.
21. HDAC Fluorescent Activity Assay, BioMol Research
Laboratories (Plymouth Meeting, PA).
26. Frey, R. R.; Wada, C. K.; Garland, R. B.; Curtin, M.
L.; Michaelides, M. R.; Li, J.; Pease, L. J.; Glaser, K.
B.; Marcotte, P. A.; Bouska, J. J.; Murphy, S. S.;
Davidsen, S. K. Bioorg. Med. Chem. Lett. 2002, 12,
3443.
27. Wada, C. K.; Frey, R. R.; Ji, Z.; Curtin, M. L.; Garland,
R. B.; Holms, J. H.; Li, J.; Pease, L. J.; Guo, J.; Glaser, K.
B.; Marcotte, P. A.; Richardson, P. L.; Murphy, S. S.;
Bouska, J. J.; Tapang, P.; Magoc, T. J.; Albert, D. H.;
Davidsen, S. K.; Michaelides, M. R. Bioorg. Med. Chem.
Lett. 2003, 13, 3331.
22. DMSO/compound solution were incubated for 10 min
with His-tagged HDAC4GOF (653-1084, H976Y) from
E. coli in assay buffer (20 mM Hepes, pH 7.5, 137 mM
NaCl, 2.7 mM KCl, 1 mM MgCl2, 0.1 mg/ml BSA), Fluor-
de-Lys substrate21 solution was added and left for 1 h at
37 °C and the reaction stopped by adding developer21/
TSA solution. Measure the fluorescence at ex.360 nM/
em.460 nM.