Báo cáo khoa học: Inhibitor-mediated stabilization of the conformational structure of a histone deacetylase-like amidohydrolase pptx

Abbreviations CypX, cyclopentylpropionyl hydroxamate; DGu, free energy of unfolding; Gdn-HCl, guanidine hydrochloride; HDAC, histone deacetylase; HDAH, histone deacetylase-like amidohydr

structure of a histone deacetylase-like amidohydrolase

Stefanie Kern1, Daniel Riester2, Christian Hildmann2, Andreas Schwienhorst2and

Franz-Josef Meyer-Almes1

1 Department of Chemical Engineering and Biotechnology, Darmstadt University of Applied Sciences, Germany

2 Institut fu¨r Molekulare Genetik und Praeparative Molekularbiologie, Institut fuer Mikrobiologie und Genetik, Goettingen, Germany

Nucleosomal histones are subject to a variety of

post-transcriptional covalent modifications, including

acetylation, methylation, phosphorylation and

ubiquiti-nation [1] Reversible histone acetylation has been

shown to facilitate access of the transcriptional

machin-ery to DNA by disruption of nucleosome–nucleosome

and nucleosome–DNA interactions [2–4] Acetylation of

histone proteins occurs at the e-amino group of lysine residues near the N-termini of these proteins The steady-state histone acetylation level is the result of opposing actions of histone acetyltransferases and his-tone deacetylases (HDACs) In particular, HDACs are promising therapeutic targets on account of their involvement in regulating genes involved in cell cycle

Keywords

FB188; HDAH; histone deacetylase; protein

denaturation; protein stablization

Correspondence

F.-J Meyer-Almes, Department of Chemical

Engineering and Biotechnology, University

of Applied Sciences Darmstadt,

Schnittspahnstr 12, 64287 Darmstadt,

Germany

Fax: + 1649 6151168404

Tel: + 1649 6151168406

E-mail: meyer-almes@h-da.de

Website: http://www.h-da.de/cub/

(Received 24 January 2007, revised 11 May

2007, accepted 15 May 2007)

doi:10.1111/j.1742-4658.2007.05887.x

Histone deacetylases are major regulators of eukaryotic gene expression Not unexpectedly, histone deacetylases are among the most promising tar-gets in cancer therapy However, despite huge efforts in histone deacetylase inhibitor design, very little is known about the impact of histone deacety-lase inhibitors on enzyme stability In this study, the conformational stabil-ity of a well-established histone deacetylase homolog with high structural similarity (histone deacetylase-like amidohydrolase from Bordetella⁄ Alcali-genes species FB188) was investigated using denaturation titrations and stopped-flow kinetics Based on the results of these complementary approa-ches, we conclude that the interconversion of native histone deacetylase-like amidohydrolase into its denatured form involves several intermediates possessing different enzyme activities and conformational structures The refolding kinetics has shown to be strongly dependent on Zn2+ and to a lesser extent on K+, which underlines their importance not only for cata-lytic function but also for maintaining the correct conformational structure

of the enzyme Two main unfolding processes of histone deacetylase-like amidohydrolase were differentiated The unfolding occurring at submolar concentrations of the denaturant guanidine hydrochloride was not affected

by inhibitor binding, whereas the unfolding at higher concentrations of guanidine hydrochloride was strongly affected It was shown that the known inhibitors suberoylanilide hydroxamic acid and cyclopentylpropio-nyl hydroxamate are capable of stabilizing the conformational structure of histone deacetylase-like amidrohydrolase Judging from the free energies of unfolding, the protein stability was increased by 9.4 and 5.4 kJÆmol)1upon binding of suberoylanilide hydroxamic acid and cyclopentylpropionyl hydroxamate, respectively

Abbreviations

CypX, cyclopentylpropionyl hydroxamate; DGu, free energy of unfolding; Gdn-HCl, guanidine hydrochloride; HDAC, histone deacetylase; HDAH, histone deacetylase-like amidohydrolase from Bordetella ⁄ Alcaligenes FB188; m eq , equilibrium parameter which reflects the

difference between the exposed surfaces and intermediate I and unfolded state D; SAHA, suberoylanilide hydroxamic acid.

control [5,6] To date, several HDAC inhibitors show

potency as antitumor agents, with several drug

candi-dates currently in phase I–III clinical trials [7]

Eukary-otic histone deacetylases, as well as their bacterial

homologs, have been grouped into four classes,

pri-marily based on sequence similarity [8] Whereas

class 3 enzymes (also termed ‘sirtuins’) are

NAD-dependent, class 1, 2 and 4 HDACs are zinc-dependent

hydrolases [9] To date, crystal structures of two

class 1 enzymes and one class 2 enzyme, as well as of

enzyme–inhibitor complexes thereof, are known

[10–13] However, no information is currently available

on the conformational stability of these enzymes

Fur-thermore, despite the identification of a large number

of HDAC inhibitors, the effect of inhibitor binding on

enzyme structure and conformational stability of the

enzyme has not been analyzed in detail

In general, the stability of proteins is an issue of

utmost interest in biochemistry and biophysics as well

as in industrial enzyme applications The

conforma-tional stability of most proteins is surprisingly low,

generally between 20 and 60 kJÆmol)1 [14,15] This

small overall stability is the result of large

contri-butions from several important converse forces The

major destabilizing force is conformational entropy

The major stabilizing forces are hydrogen bonding and

the hydrophobic effect, which is also responsible for

the large change in heat capacity between the unfolded

and folded conformations [16,17]

For technical applications of enzymes, in most cases

maximal stability is desired without losses in activity

There are mainly two approaches to stabilize proteins

(a) changes in amino acid sequence or (b) specific

binding of ions or compounds to the folded

conforma-tion The largest increase in conformational stability

resulting from a single change in amino acid sequence

is the Asn57>Ile mutant of yeast iso-1-cytochrome c

by 17.64 kJÆmol)1 [18] Some studies report the

stabil-izing effects of inorganic ions that specifically bind to

the folded conformation of a protein [19–22] For

example, Brandts et al used differential scanning

calorimetry to measure protein stabilization by ferric

ions [23] and highly charged ligands They showed that

binding of cytidine 2¢-monophosphate, other

nucleo-tide monophosphates, pyrophosphate and phosphate

shifted the transition temperature for ribonuclease

thermal unfolding [24] In this study, Brandts et al

suggested to use this approach for screening drug

can-didates for the estimation of binding constants or

screening solution conditions to optimize liquid protein

formulations with respect to stability Recently, small

molecules were found to rescue mutant proteins from

degradation and to facilitate trafficking to their site of

action [25–27] These compounds are called chemical chaperones and those compounds which act selectively

on a certain pharmaceutical target protein are called pharmacological chaperones Although the precise mechanism of action is not yet completely understood,

it is generally assumed that chemical chaperones stabil-ize a protein conformation capable of escaping the quality control system of the cell [25–27] However, in most of these studies the stabilization of the protein conformation was not measured directly and quantified

in terms of free energy

Here, we studied the conformational stability of the HDAC class 2 homolog FB188 HDAH, a bacterial HDAC-like amidohydrolase from Bordetella⁄ Alcali-genes species FB188 [28] FB188 HDAH has been shown to be an excellent model system for HDACs, concerning both structure [13] and function [29,30] The main focus of this report was to investigate the impact of HDAC inhibitors as potential chemical chaperones (i.e stabilizers) as well as zinc and potas-sium ions on the conformational stability of HDAH Two main denaturation phases of HDAH were differ-entiated The denaturation occurring at submolar con-centrations of the denaturant guanidine hydrochloride (Gdn-HCl) was not affected by inhibitor binding, whereas the denaturation at higher concentrations of Gdn-HCl was strongly affected The existence of at least one conformational intermediate was confirmed

by the fact that denaturation of HDAH occurs at a slightly higher denaturant concentration than the loss

of enzyme activity Moreover, the investigation of the denaturation and refolding kinetics supports the view that the interconversion between the native and the completely denatured state of HDAH follows a considerably complex mechanism We have shown that the overall conformational stability of HDAH is significantly increased upon binding of the inhibitors cyclopentylpropionyl hydroxamate (CypX) and sube-roylanilide hydroxamic acid (SAHA) Data of refold-ing kinetics demonstrate the strong stabilizrefold-ing impact

of zinc ions, and, to a lesser extent of potassium ions,

on the conformational structure of HDAH

Results and Discussion

Stabilization of conformational structure of HDAH by inhibitors

Taking FB188 HDAH as a model of HDACs, we were interested to see whether small-molecule inhibitors would also act as molecular chaperones To elucidate the molecular mechanism of stabilization of protein structure by inhibitor binding, we performed titrations

using Gdn-HCl as the denaturant and analyzed

stopped-flow kinetics of the denaturation reaction as

well as refolding of HDAH in the absence and the

presence of small organic molecule inhibitors

Denatur-ation experiments were performed in the presence of

0–4.5 m Gdn-HCl HDAH showed a biphasic

denatur-ation curve upon increasing the concentrdenatur-ation of the

denaturant (Fig 1) The protein fluorescence excited

at 295 nm and measured at 350 nm originates from

five tryptophans of HDAH One of the tryptophans

(Trp13) is solvent accessible Two tryptophans

(Trp179, Trp191) are buried in the hydrophobic center

of the protein, and two further tryptophans (Trp7,

Trp192) have limited solvent accessibility The

trypto-phans with no or limited solvent accessibility

presuma-bly possess relatively high quantum yields Upon

unfolding, these tryptophans become more exposed to

water, which leads to a decrease of their quantum yield

and a shift of the maximum of the emission spectrum

from 353 nm in the native state to 360 nm in the

dena-tured state The emission maximum at 353 nm did not

change at concentrations of Gdn-HCl < 1.5 m In

contrast, an unusual decrease in HDAH protein

fluo-rescence intensity was observed at submolar

concentra-tions of Gdn-HCl This decrease in fluorescence was

not caused by an artificial contribution of the

Gdn-HCl solution used in all denaturation experiments, as

confirmed in a control experiment where the dissolved

amino acid tryptophan was titrated with Gdn-HCl

(data not shown) Therefore, we conclude that partial

unfolding of HDAH takes place at submolar concentrations of Gdn-HCl This denaturation phase

at low denaturant concentration contributes about 40% to the overall process This study concentrates mainly on the denaturation effect at higher Gdn-HCl concentrations, which, in contrast to the unfolding at submolar concentrations of Gdn-HCl, is clearly affec-ted by inhibitor binding The complete denaturation curve can be fitted using a model function consisting

of two addends As the denaturation curve at submo-lar denaturant concentration cannot be explained by a simple two-state model where the native state denatur-ates into an intermediate, the first addend just des-cribes the shape of the denaturation phase at submolar denaturant concentration and does not yield parame-ters with thermodynamic meaning However, the second addend describing the major part of the denaturation curve which is affected by binding of inhibitors directly yields the free energy of unfolding,

DGu, of the intermediate in the absence of denaturant and the equilibrium meq value meq is a parameter which reflects the change in compactness of HDAH upon denaturation The parameter is proportional to the surface area buried in the intermediate state I Therefore, the DGu of 17.9 kJÆmol)1 for the major denaturation phase at higher Gdn-HCl concentrations

is a lower estimate of the overall conformational sta-bility of free HDAH, being consistent with the con-formational stability of most other proteins, which is between 20 and 60 kJÆmol)1 [14,15] There are only rare reports about conformational changes of protein structures at submolar denaturation concentrations The structural changes of horseradish peroxidase and spectrin at submolar concentrations of denaturant are examples reported by Ray et al [31] and Chakrabarti

et al [32], although the observed changes in the bio-physical parameters were much smaller as compared with the denaturation of HDAH Saturating concen-trations of SAHA or CypX were used in all experi-ments where inhibitors were present The binding constants of SAHA and CypX to HDAH were deter-mined by Riester et al [33], using a competitive bind-ing assay based on fluorescence energy transfer, and are summarized in Table 1 As only the denaturation phase at higher denaturant concentration is affected by the binding of inhibitors, the difference between the free energy of protein unfolding of free and complexed HDAH (DDGu) is identical to the increase of the con-formational stability of HDAH by 9.4 and 5.4 kJÆ mol)1 upon binding of SAHA or CypX, respectively (Table 1) This large contribution to the conformational stability is at least one-third of the lower estimate of the conformational stability of the whole protein The

0

20

40

60

80

100

c(Gdn-HCl) / M

Norm Fluorescence Intensity 0

20 40 60 80 100

- Inhibitor CypX SAHA Enzyme Activity

Fig 1 Denaturation curves of 250 n M histone deacetylase-like

amidohydrolase from Bordetella ⁄ Alcaligenes FB188 (HDAH) in the

absence of inhibitor (squares) and in the presence of 100 l M

cyclo-pentylpropionyl hydroxamate (CypX) (triangles) and suberoylanilide

hydroxamic acid (SAHA) (circles) The normalized fluorescence

intensity (excitation 295 nm, emission 350 nm) is plotted versus

the concentration of the denaturant guanidine hydrochloride

(Gdn-HCl) The unbroken lines (except that for enzymatic activity) are the

result of fitting the denaturation data to Eqn (1) The crosses

denote the corresponding relative enzyme activity of HDAH in the

absence of inhibitor.

binding of the hydroxamic acid derivatives to the zinc

ion, His142, His143 and Tyr312 within the active site,

as well as hydrophobic interactions of the aliphatic

chains with Phe152 and Phe208, are believed to

con-tribute to the stabilization of the the major part of the

conformational structure of HDAH [13] The meq

value of free and complexed HDAH is about )11 kJÆ

mol)1Æm)1 This is consistent with the assumption that

the intermediate, I, is more compact than the

dena-tured protein, D The larger the meqvalue, the greater

the difference between I and D in exposed surface

area The magnitude of the meqvalue is comparable to

the overall meq values of other proteins, which range

between )2.5 and )18 kJÆmol)1Æm)1 [34] The part of

the protein that unfolds at submolar denaturant

con-centrations is quite labile and is not affected upon

inhibitor binding To obtain more insight into the

mechanism of stabilization by HDAH inhibitors, the

refolding and denaturation kinetics in the presence and

absence of CypX were investigated If not noted

other-wise, 10 mm K+ was used in these experiments The

refolding kinetics in the presence of 0.5 mm Zn2+ and

100 lm CypX was slightly slower when compared with

the kinetics in the absence of CypX At 50 lm Zn2+,

the overall refolding kinetics was markedly slower,

showing a sigmoidal increase Upon the addition of

100 lm CypX, again the refolding kinetics was only

slightly slower than in the absence of CypX This

neg-ative impact of CypX on refolding can be explained by

the Zn2+ dependency of HDAH refolding As pointed

out in the following section, the refolding kinetics is

strongly dependent on the concentration of zinc ions

(Fig 2A,B) CypX is a hydroxamate derivative and

hy-droxamates are known to complex divalent cations,

such like Zn2+ Thus, free CypX is able to bind Zn2+

ions, which otherwise would accelerate the refolding of

HDAH Under these conditions of refolding, the com-petition between HDAH and CypX for Zn2+ binding causes a slightly retarded refolding kinetics

Table 1 Equilibrium parameters of histone deacetylase-like

amido-hydrolase from Bordetella ⁄ Alcaligenes FB188 (HDAH) in the

absence and the presence of inhibitors K2 denotes the binding

constant of the respective inhibitor to HDAH [33] The free energy

of unfolding (DG u ) from the intermediate to the denatured and the

parameter meqwere obtained from fitting the data of the

equilib-rium denaturation curve to Eqn (1) The increase in conformational

stability of HDAH, DDG u , upon inhibitor binding is the difference

between the free energy of unfolding of HDAH in the absence of

inhibitor and in the presence of 100 l M of the noted inhibitor CypX

or SAHA.

Inhibitor

K 2

(106Æ M )1)

m eq (kJÆmol)1Æ M )1)

DG u (kJÆmol)1)

DDG u (kJÆmol)1)

No inhibitor – ) 10.3 ± 1.6 17.9 ± 3.0 –

0.8 1.0

A

B

C

0.6 0.4 0.2 0.0

0.8 1.0

0.6 0.4 0.2

0.0

0.8 1.0

0.6 0.4 0.2 0.0

Time / s

Fig 2 Refolding kinetics of histone deacetylase-like amidohydro-lase from Bordetella ⁄ Alcaligenes FB188 (HDAH) (A) in the presence

of 10 m M KCl and different concentrations of Zn 2+ (0 m M , blue; 0.05 m M , red; 0.5 m M , dark green; 1.0 m M , black) and (B) in the presence of different concentrations of Zn 2+ and K + ions (0 m M

Zn2++ 0 m M K+, brown; 0 m M Zn2++ 10 m M K+, blue; 1 m M

Zn 2+ + 0 m M K + , orange; 1 m M Zn 2+ + 10 m M K + , black) and (C) in the presence of 10 m M K + and in the presence of different concen-trations of Zn2+ ions in the absence or the presence of 100 l M inhibitor cyclopentylpropionyl hydroxamate (CypX) (0.05 m M

Zn 2+ ) CypX, red; 0.05 m M Zn 2+ + Cyp X, magenta; 0.5 m M

Zn2+) CypX, dark green; 0.5 m M Zn2++ CypX, light green) The normalized fluorescence of stopped-flow experiments is plotted versus time First, the enzyme was denatured using 3 M guanidine hydrochloride (Gdn-HCl) Then, refolding was initiated by diluting the denatured enzyme in Tris buffer, pH 8.0, to a final Gdn-HCl con-centration of 0.6 M

In contrast, CypX has a significant impact on the

denaturation kinetics of HDAH in 15 mm Tris⁄ HCl

buffer, pH 8.0 (Fig 3) At 2.8 m Gdn-HCl the

ampli-tudes of the denaturation curves were almost the same

in the absence and the presence of 100 lm CypX, but

the denaturation kinetics in the presence of CypX was

significantly slower (Fig 3A) At 3.2 m, the

denatura-tion curves could not be distinguished (Fig 3B) This

would be expected if the binding of CypX to HDAH

is inhibited in the presence of 3.2 m Gdn-HCl This

would also explain why the effect of CypX on the

unfolding kinetics in the presence of 2.8 m Gnd-HCl is

weak compared with the strong effect of CypX on the

stability of the protein Based on these results, we

conclude that the inhibitor CypX can stabilize the

conformational structure of HDAH by decelerating the

denaturation process Hydroxamate-derived inhibitors

can even be contraproductive in refolding experiments

of Zn2+-dependent enzymes such as HDAH because hydroxamate complexes Zn2+ ions, which accelerate refolding dramatically (see the next section)

Impact of Zn2+and K+ions on the refolding

of HDAH The crystal structure of HDAH contains one Zn2+ion

at the bottom of the active site and two K+ ions in the neighbourhood of the active site [13] The import-ance of Zn2+ and K+ ions for the conformational structure of HDAH was investigated by measuring the refolding kinetics of HDAH (Fig 2A,B) in the pres-ence or abspres-ence of these cations

The kinetics of HDAH refolding in the presence of

10 mm K+ is strongly dependent on Zn2+, which underlines the pivotal role of Zn2+ within the active site of HDAH for the conformational stability of HDAH (Fig 2A and 4A) If Zn2+is present at 0.5 or

1 mm, the folding of the polypeptide chain into the correct orientation of the enzyme is facilitated as a result of the interactions between the zinc ion and the adjacent amino acids Asp180, Asp268 and H182 The kinetics in the presence of 0.5 mm Zn2+, 10 mm K+ and a final concentration of 0.6 m Gdn-HCl at 21C behaves like a single exponential with a time constant

of 2.3 s With decreasing concentrations of Zn2+, the refolding kinetics becomes strongly retarded At lower

Zn2+concentrations the kinetics changes into a sigmo-idal behaviour, indicating a more complex mechanism

of the refolding process with at least one additional intermediate in the absence or at lower concentrations

of Zn2+, which becomes rate limiting Perhaps the mechanism is followed also in the presence of 0.5 mm Zn2+, where the time course of the refolding kinetics can be described by only one exponential In this case it could be assumed that the first process will

be accelerated by Zn2+, such that this step is no longer rate limiting Another explanation would be that the refolding mechanism would change in the presence of

Zn2+ Refolding rates at 1 mm Zn2+, on the other hand, gave rise to clear double-exponential decays with two well-separated phases Both effects – the slow effect and the fast effect – strongly depend on the final Gdn-HCl concentration (Fig 4B) Such additional effects might be ascribed to one of three phenomena: (a) transient aggregation during folding [35,36] (b) cis-trans isomerization (e.g cis-trans isomerization about prolyl-peptidyl bonds) [37–39] or (c) the formation or decay of folding intermediates [40] Aggregation can be ruled out, as the refolding rate constant did not vary significantly with protein concentration over a 10-fold

1.2

1.0

0.8

0.6

0.4

0.2

0.0

-0.2

1.2

B

A

1.0

0.8

0.6

0.4

0.2

0.0

-0.2

Time / s

Time / s

Fig 3 Denaturation kinetics of 500 n M histone deacetylase-like

amidohydrolase from Bordetella ⁄ Alcaligenes FB188 (HDAH) in

Tris-phosphate buffer, pH 8.0, in the presence of 2.8 M (A) or 3.2 M (B)

guanidine hydrochloride (Gdn-HCl) The normalized fluorescence of

the intrinsic tryptophans of HDAH is plotted versus time At the

lower concentration of Gdn-HCl (A) the difference between the

denaturation kinetics in the presence of 100 l M

cyclopentylpropio-nyl hydroxamate (CypX) (red) and in the absence of inhibitor (blue)

becomes visible.

range (data not shown) The fast phase, at

1 mm Zn2+, is also much faster than a conventional

isomerization step, which usually has reaction rates in

the order of 10)2 to 10)4Æs)1[41] And, in contrast to

our observation, cis-trans isomerization processes do

not depend on the final denaturant concentration of

refolding These arguments point again to at least one,

probably more, additional intermediates

In the following, we will concentrate on the slow phase and analyse the data according to the three-state approach outlined in the experimental procedures Additional information about the accessible surface

of the conformational structures during the refolding process can be obtained from the dependence of the refolding rate constants on the final concentration of Gdn-HCl [42,43]

Figure 4A shows plots of ln(k) of the slow refolding process versus Gdn-HCl concentration at different concentrations of Zn2 + and 10 mm K+ In the absence and the presence of 50 lm Zn2 + at Gdn-HCl concentrations below 0.6 m, the rate constants were about one order of magnitude smaller compared with the refolding kinetics in the presence of 0.5 and 1.0 mm Zn2 + (Fig 4A) At final Gdn-HCl concentra-tions higher than 0.6 m, no temporal change in fluores-cence intensity was observed in the presence of 0 or

50 lm Zn2 + This means that no measurable refold-ing occurs at these concentrations of Zn2 + and Gdn-HCl As seen inFig 4A, the logarithmic refolding rate constants in the presence of 0.5 and 1 mm Zn2 + dis-play a linear dependence at Gdn-HCl concentrations higher than 1.0 m This can be taken as a two-state transition from the denatured to an intermediate state, which dominates the folding reaction at concentrations

of > 1 m Gdn-HCl The slope, which is proportional

to the difference between the accessible protein surface after and before refolding, is negative, which is consis-tent with the assumption that the intermediate is more compact than the denatured protein At concentrations below 1 m Gdn-HCl, the plot of ln(kobs) versus denat-urant concentration shows a clear rollover effect in which the slope of the curve decreases significantly (Fig 4) This behaviour of the folding kinetics suggests that an intermediate accumulates transiently during refolding [44] All curves of logarithmic rate constants versus denaturant concentration could satisfactorily be fitted to both the on-pathway or the off-pathway mod-els (Eqns 7 and 9; Fig 4) On the basis of the kinetic data, it cannot be distinguished between productive on-pathway intermediate I:

D,KI I,kf

k u

N;

and off-pathway intermediate C:

C,KCD,kf

k u

N:

The fitted parameters of both models are summarized

in Table 2 Further experiments are required to iden-tify whether there is an on-pathway or an off-pathway intermediate Taking into account the sigmoidal shape

0.0

-4.0

-3.0

-2.0

-1.0

0.0

A

B

-4.0

-3.0

-2.0

-1.0

0.0

1.0

2.0

3.0

0.5 1.0 1.5 2.0 2.5

0.0 0.5 1.0 1.5 2.0 2.5

c(Gdn-HCl) / M

c(Gdn-HCl) / M

Fig 4 Refolding rate constants of histone deacetylase-like

amid-ohydrolase from Bordetella ⁄ Alcaligenes FB188 (HDAH) The

log-arithm of the refolding rate constants is plotted versus the

concentration of guanidine hydrochloride (Gdn-HCl) The kinetics

were measured in the presence of 10 m M KCl and the denoted

concentrations of ZnCl2 The data were fitted to a two-state model

(dashed line, Eqn 5) and to the on-pathway intermediate model

(solid line, Eqn 7) The kinetic parameters are summarized in

Table 2 (A) The concentrations of ZnCl2are 0 m M (open squares),

0.05 m M (filled squares), 0.5 m M (circles) and 1.0 m M (diamonds).

Only the rate constant of the main slow increasing effect is shown

in panel A (B) At 1 m M Zn 2+ an additional fast effect with

increas-ing fluorescence intensity becomes visible Both the rate constants

of the fast (triangles) and the slow (diamonds) processes are

plot-ted versus the concentration of Gdn-HCl.

of refolding timecourses at very low Zn2+

concentra-tions (Fig 2), the additional fast phase at 1 mm Zn2+

(Fig 4B) and the additional equilibrium denaturation

phase at low-denaturant concentration (Fig 1), which

cannot be explained by a simple transition from the

native to one transition state, it is evident that folding

of HDAH is rather complex and must pass through

more than just one intermediate

Without Zn2+ and K+, no increase in fluorescence

can be detected within 60 s (Fig 2B) In the presence

of 10 mm K+, but in the absence of Zn2+, a slow

sig-moidal increase in fluorescence intensity can be

meas-ured (Fig 2B)

In summary, the refolding mechanism appears to be

strongly dependent on Zn2+ and to a lesser extent on

K+, which underlines the importance of both cations,

not only for the function of the enzyme but also for the

correct conformational structure There is strong

evidence that HDAH folding involves more than one

intermediate It was shown that the known inhibitors

SAHA and CypX are capable of stabilizing the

conformational structure of HDAH Judging from free

energies of unfolding, the conformational stability of a

complex between these inhibitors and HDAH is more

than 30% higher than the stability of unbound HDAH

We suspect that hydroxamate-type HDAC-inibitors

such as CypX or SAHA not only hinder substrates to

obtain access to the active site, but rather may even

freeze HDACs in catalytically unproductive

conforma-tions In this connection it is interesting to note that the

association kinetics of N-(2-furyl)acryloyl-hydroxamic

acid and HDAH can only be satisfactorily described by

a biphasic exponential model [31], suggesting a

multistep binding process, including conformational

changes of the enzyme If we now assume a similar

behavior of eukaryotic HDACs upon inhibitor binding,

it is tempting to speculate that inhibitor-induced conformational changes of HDACs are responsible for breaking up corepressor complexes as, for example, described in the case of acute myelocytic leucemia [45] Furthermore, our results support the assumption that specific ligands of proteins within cells may act as molecular chaperones by stabilizing a protein confor-mation capable of escaping the quality control system

of the cell A better understanding of the impact of inhibitor binding on the stability of target proteins (e.g HDAH) may result in new concepts for lead structures

Experimental procedures

Materials His-tagged FB188 HDAH was prepared as described previ-ously [28] SAHA, CYPX and phenylpropionyl hydroxa-mate were synthesized according to standard methods [13,46,47] If not stated otherwise, the denaturation experi-ments were carried out in Tris-phosphate buffer consisting

of 250 mm sodium chloride, 250 lm EDTA, 15 mm Tris-HCl and 50 mm potassium hydrogen phosphate, pH 8.0

Enzyme activity assay FB188 HDAH exhibits amidohydrolase [28] and esterase activity [48] Amidohydrolase activity was assayed in the two-step assay [49,50] As trypsin activity is required in this type of assay, and trypsin rapidly denatures upon addition

of denaturant, the two-step assay was not suited for activity measurements in samples containing Gdn-HCl Esterase activity was monitored using 4-methylcoumarin-7-acetate as

a substrate [48] This type of assay was used for samples containing Gdn-HCl

Table 2 Kinetic parameters of histone deacetylase-like amidohydrolase from Bordetella ⁄ Alcaligenes FB188 (HDAH) refolding in the presence

of noted concentrations of Zn2+, c(Zn2+) Kinetic parameters of HDAH refolding are shown in the presence of 0.5 and 1.0 m M Zn2+ The parameters were obtained by fitting the data of ln(kobs) versus the concentration of Gdn-HCl, c(Gdn-HCl), to different folding models (see Eqns 5, 7 and 9) KC, ratio of off-pathway intermediate C and denatured state D; kf, the folding rate in the absence of denaturant; KI, ratio of on-pathway intermediate I and denatured state D; m f reflects the change in solvent-accessible area in the process of refolding; m i , reflects the change in solvent-accessible area in the transition from the denatured to the intermediate state.

c(Zn 2+ ) Folding model kf(s)1) mf(kJÆmol)1Æ M )1) K

Denaturation experiments using guanidine

hydrochloride

In the denaturation experiments, 250 nm HDAH in

phos-phate buffer, pH 8.0, were titrated with increasing amounts

of a solution of 8 m Gdn-HCl in the same buffer All

experiments were carried out at 21 ổ 0.2C The

denatura-tion of the protein was followed by measuring its

trypto-phan fluorescence The tryptotrypto-phans of HDAH were excited

at 295 nm and their fluorescence emission was measured at

350 nm in a Hitachi (Tokyo, Japan) F-4000

spectrofluo-rometer using 5 and 10 nm slits, respectively After each

addition of Gdn-HCl and thorough mixing, the

fluores-cence was measured until the signal was constant within ổ

1% for at least 30 s This value was considered to represent

the unfolding equilibrium and was plotted against the

cor-responding Gdn-HCl concentration The resulting graph is

called the denaturation curve

Data analysis of equilibrium protein denaturation

curves

The fitting of equilibrium unfolding curves is described in

detail by Santoro & Bolen [51] and directly gives

thermo-dynamic parameters of the corresponding denaturation

curves The fitting function was slightly complemented to

account for the additional denaturation phase at submolar

denaturant concentrations The fluorescence signal as a

function of denaturant concentration was fitted to the

fol-lowing expression:

1ợ exphb  đơGdn  HCl  IC50ỡi

ợFIợ FDexphđDGuợ meqơGdn  HClỡ=RTi

1ợ exphđDGuợ meqơGdn  HClỡ=RTi ; đ1ỡ

where DGu is the free energy of unfolding of an

interme-diate in the absence of denaturant; meq is the equilibrium

m-value, which is proportional to the difference between

the exposed surfaces of intermediate I and unfolded state

D; FI and FD are the fluorescence signals of I and D;

A, b and IC50 are the amplitude, the steepness and the

inflection point, respectively, and used as arbitrary

parameters to describe the contribution of the additional

denaturation process at submolar denaturant

concentra-tion; [Gdn-HCl] is the concentration of Gdn-HCl and

RT is the product of the gas constant and temperature

All equilibrium and kinetic data were fitted using the

program scientist from micromath (St Louis, MO,

USA)

Stopped-flow kinetics

All measurements of displacement and renaturation kinetics

were carried out on a Bio-Logic (Claix, France) MOS-250

Stopped-Flow instrument equipped with a 150 W xenon mercury light source attached to a manual monochromator

on an optical bench The connection to the Bio-Logic Stopped-Flow instrument was performed through a fiber optic specially designed to match the stopped-flow cuvette dimensions The signal detection was performed by a pho-tomultiplier directly mounted on the stopped-flow and con-nected to its control unit The photomultiplier was attached

at 90 of the light source allowing for fluorescence measure-ments The HDAH tryptophans were excited at 295 nm A polystyrene filter was installed in front of the photomulti-plier tube to reject scattered light The photomultiphotomulti-plier con-trol unit was connected to a 16-bit A⁄ D board installed in

a PC driven by the acquisition and analysis software bio-kine32 (Claix, France)

The core unit of the instrument is a temperature-con-trolled metal block containing three syringes and a mixing chamber The syringes are driven by precise and robust high-speed stepping-motors The dead time of the appar-atus was calculated to be below 2 ms The temperature was controlled at 21 ổ 0.2C

Stopped-flow data were fitted to either a monophasic

Fđtỡ Ử A1 1 exp t

s1

 

ợ B đ2ỡ

or a biphasic Fđtỡ Ử A1 1 exp  t

s1

 

ợ A2 1 exp t

s2

 

ợ B đ3ỡ exponential model by using a nonlinear least-square fit-ting procedure integrated in the analysis software bio-kine32 F(t) is the observed fluorescence of the protein at time t after the start of the reaction and B is the back-ground signal A1 and A2 are the amplitudes of two exponential changes, and s1 and s2 are their respective kinetic time constants The refolding kinetics were initi-ated by mixing buffer consisting of 15 mm Tris⁄ HCl,

pH 8.0, and denoted concentrations of Zn2+, K+ ions or CypX, and completely denatured HDAH dissolved in the same buffer in the presence of 3 m Gdn-HCl The dena-turation kinetics were carried out by mixing 500 nm HDAH (final concentration) in 15 mm Tris⁄ HCl, pH 8.0,

in the absence or presence of 100 lm CypX and 15 mm Tris⁄ HCl buffer, pH 8.0, containing denoted concentra-tions of Gdn-HCl

Data analysis of refolding kinetics The analysis of the kinetic data is based on a linear rela-tionship between the log of microscopic rate constants and the denaturant concentration The following equations were adapted from Mogensen et al [52] and slightly modified to

fit rate constants determined from stopped-flow experi-ments Under the condition of the experiments the contri-bution of unfolding could be disregarded

Two-state folding:

D,kf

k u

ln kobs¼ ln exp ln kf

þ mf½Gdn HCl

=RT

ð5Þ where in this simple two-state case D denotes the denatured

state and N denotes the native state of the protein; kobsis

the observed folding rate; kf is the folding rate in the

absence of denaturant and mf is the corresponding m-value

which reflects the change in solvent-accessible area in the

process of refolding For a simple two-state folding

mech-anism a plot of ln kobs versus the concentration of

Gdn-HCl, c(Gdn-HCl), is expected to be linear over the whole

range of denaturant concentration The accumulation of

on- or off-pathway intermediates during folding will give

rise to deviations from linearity, particularly at low

denatu-rant concentrations The models for folding mechanisms

over in- or off-pathway intermediates follow

(A) Folding over an on-pathway intermediate:

D,KII,kf

k u

ln kobs¼ ln exp ln kf

þ mf½Gdn HCl=RT

1þ exp  ln Kð ð ð Þ þ mI I½Gdn HCl=RTÞÞ

; ð7Þ where I is an on-pathway intermediate between unfolded

and native protein and KI¼ [I] ⁄ [D] in the absence of

denat-urant

(B) Folding with an off-pathway intermediate:

C,Kc D,kf

k u

ln kobs¼ ln exp ln kf

þ mf½Gdn HCl=RT

1þ exp ln Kð ð Þ þ mc I½Gdn HCl=RTÞ

; ð9Þ

where C is an off-pathway folding intermediate and KC¼

[C]⁄ [D] in the absence of denaturant

Acknowledgements

This work was in part supported by grants to A.S

(BioFuture 0311852 from the Bundesministerium fu¨r

Forschung und Technologie, Germany and Human

Frontier Science Program RGY0056⁄ 2004-C)

References

1 Hildmann C, Riester D & Schwienhorst A (2007)

His-tone deacetylases an important class of cellular

regula-tors with a variety of functions Appl Microbiol

Biotechnol 75, 487–497

2 Struhl K (1998) Histone acetylation and transcriptional

regulatory mechanisms Genes Dev 12, 599–606

3 Zhang Y & Reinberg D (2001) Transcription regulation

by histone methylation: interplay between different covalent modifications of the core histone tails Genes Dev 15, 2343–2360

4 Cosgrove MS & Wolberger C (2005) How does the his-tone code work? Biochem Cell Biol 83, 468–476

5 Ferreira R, Naguibneva I, Mathieu M, it-Si-Ali S, Robin P, Pritchard LL & Harel-Bellan A (2001) Cell cycle-dependent recruitment of HDAC-1 correlates with deacetylation of histone H4 on an Rb-E2F target pro-moter EMBO Report 2, 794–799

6 Richon V & Berrie C (2002) Histone Deacetylase Inhibi-tors – Redefining Pharmaceutical Approaches to the Treatment of Cancer LeadDiscovery, Bodiam, UK Ref Type: Generic

7 Riester D, Hildmann C & Schwienhorst A (2007) His-tone deacetylase inhibitors turning epigenic mechanisms

of gene regulation into tools of therapeutic intervention

in malignant and other diseases Appl Microbiol Biotech-nol 75, 499–514

8 Gregoretti IV, Lee YM & Goodson HV (2004) Molecu-lar evolution of the histone deacetylase family: func-tional implications of phylogenetic analysis J Mol Biol

338, 17–31

9 Grozinger CM & Schreiber SL (2002) Deacetylase enzymes: biological functions and the use of small-mole-cule inhibitors Chem Biol 9, 3–16

10 Finnin MS, Donigian JR, Cohen A, Richon VM, Rif-kind RA, Marks PA, Breslow R & Pavletich NP (1999) Structures of a histone deacetylase homologue bound to the TSA and SAHA inhibitors Nature 401, 188–193

11 Vannini A, Volpari C, Filocamo G, Casavola EC, Bru-netti M, Renzoni D, Chakravarty P, De Paolini CFR, Gallinari P, Di Steinkuhler C et al (2004) Structural snapshots of human HDAC8 provide insights into the class I histone deacetylases Structure 12, 1325–1334

12 Somoza JR, Skene RJ, Katz BA, Mol C, Ho JD, Jen-nings AJ, Luong C, Arvai A, Buggy JJ, Chi E et al (2004) Structural snapshots of human HDAC8 provide insights into the class I histone deacetylases Structure

12, 1325–1334

13 Nielsen TK, Hildmann C, Dickmanns A, Schwienhorst A

& Ficner R (2005) Crystal structure of a bacterial class 2 histone deacetylase homologue J Mol Biol 354, 107–120

14 Ahmad F, Yadav S & Taneja S (1992) Determining sta-bility of proteins from guanidinium chloride transition curves Biochem J 287, 481–485

15 Pace CN (1990) Conformational stability of globular proteins Trends Biochem Sci 15, 14–17

16 Privalov PL & Gill SJ (1988) Stability of protein struc-ture and hydrophobic interaction Adv Protein Chem 39, 191–234

17 Baldwin RL (1986) Temperature dependence of the hydrophobic interaction in protein folding Proc Natl Acad Sci USA 83, 8069–8072

18 Das G, Hickey DR, McLendon D, McLendon G &

Sherman F (1989) Dramatic thermostabilization of yeast

iso-1-cytochrome c by an asparagine—isoleucine

replacement at position 57 Proc Natl Acad Sci USA 86,

496–499

19 Wilson CJ, Apiyo D & Wittung-Stafshede P (2004) Role

of cofactors in metalloprotein folding Q Rev Biophys

37, 285–314

20 Bushmarina NA, Blanchet CE, Vernier G & Forge V

(2006) Cofactor effects on the protein folding reaction:

acceleration of alpha-lactalbumin refolding by metal

ions Protein Sci 15, 659–671

21 Pace CN & Grimsley GR (1988) Ribonuclease T1 is

sta-bilized by cation and anion binding Biochemistry 27,

3242–3246

22 Cashikar AG & Rao NM (1996) Unfolding pathway in

red kidney bean acid phosphatase is dependent on

lig-and binding J Biol Chem 271, 4741–4746

23 Lin LN, Mason AB, Woodworth RC & Brandts JF

(1994) Calorimetric studies of serum transferrin and

ovotransferrin Estimates of domain interactions, and

study of the kinetic complexities of ferric ion binding

Biochemistry 33, 1881–1888

24 Plotnikov V, Rochalski A, Brandts M, Brandts JF,

Williston S, Frasca V & Lin LN (2002) An

autosam-pling differential scanning calorimeter instrument for

studying molecular interactions Assay Drug Dev

Technol 1, 83–90

25 Bernier V, Lagace M, Lonergan M, Arthus MF, Bichet

DG & Bouvier M (2004) Functional rescue of the

con-stitutively internalized V2 vasopressin receptor mutant

R137H by the pharmacological chaperone action of

SR49059 Mol Endocrinol 18, 2074–2084

26 Petaja-Repo UE, Hogue M, Bhalla S, Laperriere A,

Morello JP & Bouvier M (2002) Ligands act as

pharma-cological chaperones and increase the efficiency of delta

opioid receptor maturation EMBO J 21, 1628–1637

27 Chaudhuri TK & Paul S (2006) Protein-misfolding

dis-eases and chaperone-based therapeutic approaches

FEBS J 273, 1331–1349

28 Hildmann C, Ninkovic M, Dietrich R, Wegener D,

Ri-ester D, Zimmermann T, Birch OM, Bernegger C, Loidl

P & Schwienhorst A (2004) A new amidohydrolase from

Bordetella or Alcaligenes strain FB188 with similarities

to histone deacetylases J Bacteriol 186, 2328–2339

29 Riester D, Wegener D, Hildmann C & Schwienhorst A

(2004) Members of the histone deacetylase superfamily

differ in substrate specificity towards small synthetic

substrates Biochem Biophys Res Commun 324, 1116–

1123

30 Hildmann C, Wegener D, Riester D, Hempel R,

Scho-ber A, Merana J, Giurato L, Guccione S, Nielsen TK,

Ficner R et al (2006) Substrate and inhibitor specificity

of class 1 and class 2 histone deacetylases J Biotechnol

124, 258–270

31 Ray S, Bhattacharyya M & Chakrabarti A (2005) Conformational study of spectrin in presence of submolar concentrations of denaturants J Fluoresc 15, 61–70

32 Chakrabarti A & Basak S (1996) Structural alterations

of horseradish peroxidase in the presence of low concen-trations of guanidinium chloride Eur J Biochem 241, 462–467

33 Riester D, Hildmann C, Schwienhorst A & Meyer-Almes FJ (2007) Histone deacetylase inhibitor assay based on fluorescence resonance energy transfer Anal Biochem 362, 136–141

34 Maxwell KL, Wildes D, Zarrine-Afsar A, De Los Rios

MA, Brown AG, Friel CT, Hedberg L, Horng JC, Bona

D, Miller EJ et al (2005) Protein folding: defining a

‘standard’ set of experimental conditions and a prelim-inary kinetic data set of two-state proteins Protein Sci

14, 602–616

35 Silow M & Oliveberg M (1997) Transient aggregates in protein folding are easily mistaken for folding interme-diates Proc Natl Acad Sci USA 94, 6084–6086

36 Otzen DE, Kristensen O & Oliveberg M (2000) Designed protein tetramer zipped together with a hydrophobic Alzheimer homology: a structural clue to amyloid assembly Proc Natl Acad Sci USA 97, 9907–9912

37 Kiefhaber T, Kohler HH & Schmid FX (1992) Kinetic coupling between protein folding and prolyl isomeriza-tion I Theoretical models J Mol Biol 224, 217– 229

38 Kiefhaber T, Grunert HP, Hahn U & Schmid FX (1990) Replacement of a cis proline simplifies the mech-anism of ribonuclease T1 folding Biochemistry 29, 6475–6480

39 Jackson SE & Fersht AR (1991) Folding of chymotryp-sin inhibitor 2 2 Influence of proline isomerization on the folding kinetics and thermodynamic characterization

of the transition state of folding Biochemistry 30, 10436–10443

40 Fersht AR (1999) Structure and Mechanism in Protein Science A Guide to Enzyme Catalysis and Protein Fold-ing Freeman, New York, NY

41 Nall B (1994) Mechanism of Protein Folding IRL Press, Oxford

42 Matouschek A, Kellis JT Jr, Serrano L & Fersht AR (1989) Mapping the transition state and pathway of protein folding by protein engineering Nature 340, 122– 126

43 Denton ME, Rothwarf DM & Scheraga HA (1994) Kinetics of folding of guanidine-denatured hen egg white lysozyme and carboxymethyl (Cys6,Cys127)-lyso-zyme: a stopped-flow absorbance and fluorescence study Biochemistry 33, 11225–11236

44 Baldwin RL (1996) On-pathway versus off-pathway folding intermediates Fold Des 1, R1–R8