Tài liệu Báo cáo khoa học: The effect of small molecules in modulating the chaperone activity of aB-crystallin against ordered and disordered protein aggregation pdf

In this study, we have explored the potential for small molecules such as arginine and guanidine to affect the chaperone activity of aB-crystallin against disordered amorphous and ordere

activity of aB-crystallin against ordered and disordered protein aggregation

Heath Ecroyd and John A Carver

School of Chemistry and Physics, The University of Adelaide, Australia

Protein aggregation is the result of the mutual

associa-tion of partially folded intermediate states of a protein,

most likely via predominately hydrophobic

interac-tions Protein aggregation can proceed via disordered

or ordered mechanisms: which mechanism

predomi-nates is thought to be determined by a number of

fac-tors, including the rate of unfolding, the amino acid

sequence of the protein, the experimental conditions

and the nature of the intermediate state(s) that form

[1,2] Disordered aggregation results in amorphous

aggregates of protein, whilst ordered aggregation

pro-duces amyloid fibrils, long threadlike protein structures

that are rich in b-sheet and resistant to proteolytic

deg-radation Protein misfolding, and in particular amyloid

fibril formation, is associated with a range of diseases, including Alzheimer’s, Parkinson’s and Creutzfeldt-Jakob diseases, type II diabetes and possibly cataracts [3–5] Protein aggregation is also responsible for inclu-sion body formation, and therefore the ability to pre-vent it would be of enormous benefit in recombinant protein production, avoiding the need for resolubiliza-tion of the aggregated and precipitated protein Thus, studies aimed at preventing protein aggregation are of interest due to both their biomedical and biotechnolog-ical applications

In terms of biotechnological applications, small mol-ecules such as guanidine and urea are well-established suppressors of aggregation, and are often used to

Keywords

amyloid fibril; arginine; protein aggregation;

small heat-shock protein; aB-crystallin

Correspondence

H Ecroyd, School of Chemistry and

Physics, The University of Adelaide,

Adelaide, SA 5005, Australia

Fax: +61 8 830 34358

Tel: +61 8 830 35505

E-mail: heath.ecroyd@adelaide.edu.au

(Received 12 November 2007, revised 16

December 2007, accepted 20 December

2007)

doi:10.1111/j.1742-4658.2008.06257.x

Protein aggregation can proceed via disordered or ordered mechanisms, with the latter being associated with amyloid fibril formation, which has been linked to a number of debilitating conditions including Alzheimer’s, Parkinson’s and Creutzfeldt-Jakob diseases Small heat-shock proteins (sHsps), such as aB-crystallin, act as chaperones to prevent protein aggre-gation and are thought to play a key role in the prevention of protein-mis-folding diseases In this study, we have explored the potential for small molecules such as arginine and guanidine to affect the chaperone activity

of aB-crystallin against disordered (amorphous) and ordered (amyloid fibril) forms of protein aggregation The effect of these additives is highly dependent upon the target protein undergoing aggregation Importantly, our results show that the chaperone action of aB-crystallin against aggrega-tion of the disease-related amyloid fibril forming protein a-synucleinA53T

is enhanced in the presence of arginine and similar positively charged com-pounds (such as lysine and guanidine) Thus, our results suggest that target protein identity plays a critical role in governing the effect of small mole-cules on the chaperone action of sHsps Significantly, small molemole-cules that regulate the activity of sHsps may provide a mechanism to protect cells from the toxic protein aggregation that is associated with some protein-misfolding diseases

Abbreviations

ANS, 8-anilino-1-naphthalene sulphonate; DTT, 1,4-dithiothreitol; Gdn, guanidine; RCMj-CN, reduced and carboxymethylated j-casein; sHsp, small heat-shock protein; ThT, thioflavin T.

inhibit aggregation of expressed proteins or to

resolubilize proteins that have already aggregated into

inclusion bodies [6,7] In suppressing aggregation, these

small molecules act by weakening the hydrophobic

in-termolecular interactions between unfolded or partially

folded protein intermediates that are responsible for

the aggregation process The amino acid arginine is

also often employed as a suppressor of aggregation,

and is thought to facilitate correct folding of proteins

by destabilizing incorrectly folded structures [8,9]

However, high concentrations of guanidine, urea

and⁄ or arginine are usually required for this purpose

and must be removed during purification of the

recom-binant protein

In vivo, protein aggregation is prevented through the

action of a broad range of highly specialized proteins

known as molecular chaperones One such chaperone is

a-crystallin, a small heat-shock protein (sHsp) that acts

to prevent protein aggregation intracellularly [10]

a-Crystallin is present in large concentrations in the eye

lens, where it is thought to provide stability and

struc-tural support to the other proteins present It is made

up of two closely related subunits, aA- and

aB-crystal-lin, which exist at an approximate molar ratio of 3 : 1

in the mammalian lens Moreover, aB-crystallin is

found at significant levels in other tissues, such as the

heart, kidney, muscle and brain, and its expression is

up-regulated in response to stress and pathological

con-ditions [11,12] Recent studies have shown that

signifi-cant levels of aB-crystallin are found in protein

deposits such as those associated with disease [13,14]

The molecular chaperone action of aA- and

aB-crystal-lin is manifested by binding to partially unfolded or

misfolded target proteins, thus inhibiting their

aggrega-tion and precipitaaggrega-tion Whilst the chaperone acaggrega-tion of

aB-crystallin against amorphously aggregating target

proteins has been well established, it is only recently

that studies have shown that aB-crystallin also acts to

prevent ordered amyloid fibril assembly [15–18]

Some studies have shown that structural

perturba-tion of a-crystallin and⁄ or its two subunits (e.g

through heating) enhances its chaperone activity

against amorphously aggregating target proteins [19–

21], presumably due to increased exposure of its

hydrophobic surfaces that facilitate target protein

binding [22] In addition to temperature, other

treat-ments (e.g reduction) [23,24] and post-translational

modifications (e.g phosphorylation) [18,25,26] that

slightly perturb the structure of a-crystallin have been

shown to enhance the chaperone activity of the protein

against amorphously aggregating target proteins Of

particular note, low concentrations of denaturant, such

as guanidine hydrochloride (Gdn-HCl) enhance the

chaperone activity of a-crystallin against reduction-induced amorphous aggregation of the insulin B-chain [27] Moreover, it was also shown that millimolar con-centrations of arginine hydrochloride (Arg-HCl) had a similar effect on the chaperone activity of aB-crystallin [27], which was reported to occur via enhancement of the dynamics of subunit assembly [28] However, to date there have been no reports of the effects of such compounds on the chaperone activity of aB-crystallin against ordered protein aggregation leading to fibril formation

In this study, we have explored the potential for small molecules such as Arg-HCl and Gdn-HCl to affect the chaperone activity of aB-crystallin against disordered (amorphous) and ordered (amyloid fibril) forms of protein aggregation We report that the effect

of these additives on the chaperone action of aB-crys-tallin is dependent on the target protein used, and therefore the results highlight the need to assess the activity of chaperone proteins against a variety of tar-get proteins before drawing conclusions about their generic effects Of particular note, the results from this study show that the chaperone action of aB-crystallin against aggregation of the disease-related amyloid fibril forming protein, a-synucleinA53T, is enhanced in the presence of Arg-HCl and similar positively charged compounds (such as Lys-HCl and Gdn-HCl) Fibril formation by a-synuclein is causally linked to Lewy body formation that occurs in diseases such as Parkin-son’s, and the A53T mutant is associated with early-onset Parkinson’s disease Thus, our results suggest that small molecules that act on sHsps in a similar manner to Arg-HCl may provide a mechanism to pro-tect cells from the toxic protein aggregation that is associated with some protein-misfolding diseases

Results

The effect of Arg-HCl on the chaperone activity of aB-crystallin is target protein-specific

In order to investigate the effect of Arg-HCl on the chaperone activity of aB-crystallin, we examined a variety of model target proteins to determine the generic effects of Arg-HCl In particular, we used both ordered (amyloid fibril-forming) and disordered (amorphous) target protein aggregation systems In investigating the effect of Arg-HCl on the chaperone action of aB-crystallin, we also looked at related mole-cules, to investigate whether any observed effects were specific to Arg-HCl Thus, we also investigated the effects of (a) glycine (Gly), to test whether any effects were attributable to addition of an amino acid to the

solution; (b) lysine hydrochloride (Lys-HCl), to test

whether any effects were attributable to adding a basic

amino acid; and (c) Gdn-HCl, to test whether any

effects of Arg-HCl were attributable to the

guanidini-um group of the molecule We tested each of these

compounds at low (10 mm), intermediate (100 mm)

and high (250 mm) concentrations unless otherwise

indicated At these concentrations, the additives were

found to change the pH of the buffers used in these

aggregation assays by < 0.1 units However, at very

high concentrations (e.g > 500 mm), some of the

compounds had significant effects on the pH of these

buffers (i.e increasing the pH by > 0.2 units) In

addition, for each assay we used concentrations of

aB-crystallin that only partially inhibited aggregation

of the target protein in order to enable the effects of

the compounds on the chaperone activity to be readily

interpreted

Disordered (amorphous) aggregation systems

Reduction-induced aggregation of a-lactalbumin

Upon addition of 1,4-dithiothreitol (DTT), aggregation

and precipitation of a-lactalbumin commenced after

25 min and reached a plateau after 90 min The

amount of DTT-induced aggregation of a-lactalbumin was increased in a concentration-dependent manner by the addition of Gly, such that, at 250 mm, light scat-tering due to its precipitation had increased by

50 ± 7% [mean ± standard error of the mean (SEM)], i.e the calculated percentage protection value was negative because this treatment increased the amount of precipitation compared to that observed when a-lactalbumin was incubated alone (Fig 1A,C) Lys-HCl had a similar concentration-dependent effect However, Arg-HCl had the opposite effect whereby increasing concentrations of Arg-HCl decreased the amount of precipitation, such that, at high concentra-tions, it had decreased by 60 ± 3% compared to that observed when a-lactalbumin was incubated alone Gdn-HCl had a more complex effect, whereby concen-trations up to 100 mm increased the amount of light scattering, but the high concentration (i.e 250 mm) decreased it (Fig 1A,C) Whilst Gly, Lys-HCl and Arg-HCl had no significant effect on the lag phase of precipitation of a-lactalbumin (approximately 25 min), Gdn-HCl decreased it to 15 min (Fig 1A)

Addition of aB-crystallin at a 1.0 : 1.0 w⁄ w ratio

of a-lactalbumin : aB-crystallin decreased the precipi-tation of a-lactalbumin by 81 ± 8% The ability of

A

C

B

Fig 1 The effect of additives on the ability

of aB-crystallin to prevent the DTT-induced

aggregation of a-lactalbumin a-Lactalbumin

( , 0.5 mgÆmL)1) was incubated at 37 C in

50 m M phosphate buffer, pH 7.2, containing

100 m M NaCl with 20 m M DTT in (A) the

absence or (B) the presence of aB-crystallin

(0.5 mgÆmL)1), and the change in light

scat-tering at 340 nm was monitored over time.

For both (A) and (B), the additives were

250 m M of Gly (d), Lys-HCl ()), Arg-HCl

( ) or Gdn-HCl (h) The buffer-only control

(r) is also shown in (A) and (B) (C)

Percent-age protection (mean ± SEM of four

inde-pendent experiments), calculated 90 min

after the start of the assay, when

a-lactalbu-min was incubated with increasing

concen-trations of the additives, in the absence ( )

or presence ( ) of aB-crystallin The

per-centage protection that would be expected

assuming no influence of the additives on

the chaperone activity of aB-crystallin,

calcu-lated as described in Experimental

proce-dures, is also shown (j) The asterisks

indicate a significant (P < 0.05) decrease in

the chaperone ability of aB-crystallin in the

presence of that concentration of the

addi-tive.

aB-crystallin to protect against this precipitation was

significantly decreased in the presence of Gly,

Lys-HCl and Gdn-Lys-HCl, such that, when they were present

at high concentrations, aB-crystallin had no effect on

the amount of light scattering compared to that

observed when the additives were present alone

(Fig 1C) In contrast, the chaperone action of

aB-crystallin against a-lactalbumin was maintained in the

presence of intermediate concentrations of Arg-HCl,

but was not further enhanced by it (Fig 1C) The

sig-nificant decrease in the amount of precipitation in

the presence of high concentrations of Arg-HCl

in the absence of the chaperone precluded analysis of

the effect of this concentration on the protective

ability of aB-crystallin

Reduction-induced aggregation of the insulin B-chain

Light scattering due to DTT-induced amorphous

aggregation and precipitation of the insulin B-chain

commenced after 10 min and reached a plateau after

45 min (Fig 2A) The amount of precipitation was

increased in a concentration-dependent manner by Gly

and Lys-HCl compared to that observed when insulin

was incubated alone (Fig 2A,C) Addition of Arg-HCl

(up to 250 mm) had a negligible effect on the amount

of precipitation Similarly, low and intermediate con-centrations of Gdn-HCl had no effect on the precipita-tion of insulin; however, high concentraprecipita-tions (i.e

250 mm) had a protective effect, decreasing the amount of light scattering by 48 ± 2% (Fig 2A,C) None of the additives used affected the lag phase of the aggregation

When incubated in the presence of aB-crystallin alone (at a 1.0 : 1.0 w⁄ w ratio of insulin : aB-crystal-lin), the precipitation of insulin was inhibited by

40 ± 4% (Fig 2B,C) Only Arg-HCl significantly (P < 0.05) enhanced this protective activity of aB-crystallin, such that, at 250 mm Arg-HCl, the light scattering due to precipitation of insulin was decreased by 65 ± 8% Low and intermediate con-centrations of Gly had no effect on the chaperone activity of aB-crystallin against this target protein, but it was significantly reduced at 250 mm A similar trend was observed for Lys-HCl, with high concen-trations significantly inhibiting the ability of aB-crys-tallin to prevent precipitation (Fig 2C) Gdn-HCl had no effect on the chaperone activity of aB-crystal-lin against the DTT-induced aggregation and precipi-tation of insulin

A

C

B

Fig 2 aB-crystallin protects against the DTT-induced aggregation of insulin, and this activity is enhanced by Arg-HCl Insulin ( , 0.25 mgÆmL)1) was incubated at 37 C

in 50 m M phosphate buffer, pH 7.2, with

10 m M DTT in (A) the absence or (B) the presence of aB-crystallin (0.25 mgÆmL)1) For other details, refer to the legend to Fig 1 In addition, the hash symbol (#) indi-cates a significant (P < 0.05) increase in the chaperone ability of aB-crystallin in the pres-ence of that concentration of the additive.

Heat-induced aggregation of catalase

We used bovine catalase as the model substrate to test

the effect of the small molecules on the chaperone

abil-ity of aB-crystallin against a target protein undergoing

heat-stressed induced aggregation and precipitation

Aggregation of catalase occurs at high temperatures,

i.e 55C, and these studies aimed to investigate

whether these small molecules could further enhance

the well-characterized increase in the chaperone

activ-ity of aB-crystallin at high temperatures due to

changes in its tertiary structure [20,21] The

precipita-tion of catalase commenced after 20 min, and the

increase in light scattering due to precipitation of the

protein reached a maximum after 90 min (Fig 3A)

All of the additives tested increased the amount of

light scattering due to precipitation of catalase

com-pared to that observed when it was incubated alone

Of these, Gdn-HCl had the most dramatic effect, with

250 mm Gdn-HCl increasing the amount of

precipita-tion of catalase by 190 ± 5% (Fig 3A) The presence

of aB-crystallin at a 1.0 : 0.5 w⁄ w ratio of

lase : aB-crystallin inhibited the precipitation of

cata-lase by 71 ± 7% (Fig 3B) This chaperone activity

was not affected by increasing concentrations of Gly,

but was completely abolished by intermediate and high concentrations of Lys-HCl, and was inhibited by Gdn-HCl in a concentration-dependent manner (Fig 3B,C) Intermediate concentrations (i.e 100 mm) of Arg-HCl significantly inhibited the ability of aB-crystallin to prevent the precipitation of catalase; however, this effect was not seen at high concentrations of Arg-HCl, i.e the chaperone activity of aB-crystallin was main-tained in the presence of 250 mm Arg-HCl

Ordered aggregation leading to amyloid fibril formation

We employed two models to examine the effect of the small molecules on the ability of aB-crystallin to pre-vent amyloid fibril formation – a familial mutant of the disease-related protein a-synuclein (i.e a-synuclein-A53T) and reduced and carboxymethylated j-casein (RCMj-CN), both of which are natively disordered proteins [29] We employed these systems as they both form fibrils at physiological pH and temperature [30,31], and so can be used to examine the activity of aB-crystallin without confounding factors such as low

pH or the presence of other denaturants, which are often required in other amyloid fibril-forming systems

A

C

B

Fig 3 Heat-induced amorphous

aggrega-tion of catalase is increased by increasing

concentrations of the additives Catalase

( , 0.5 mgÆmL)1) was incubated at 55 C in

50 m M phosphate buffer, pH 7.2, in (A) the

absence or (B) the presence of aB-crystallin

(0.25 mgÆmL)1) For other details, refer to

the legend to Fig 1.

In both systems, fibril formation was assessed by an

in situthioflavin T (ThT) fluorescence assay

Amyloid fibril formation by RCMj-CN

Fibril formation by RCMj-CN, as monitored by an

increase in ThT binding, showed a gradual increase

over the time course of the assay (Fig 4A) At the end

of the assay, electron micrographs of negatively stained

RCMj-CN fibrils showed them to be thread-like

structures, approximately 100–700 nm in length

(Fig 6A,B), similar to those reported previously [30]

Addition of Gly slightly increased the degree of ThT

binding in a concentration-dependent manner, such

that, at 250 mm, there was an increase of 10 ± 1%

compared to that observed when RCMj-CN was

incubated alone (Fig 4A,C) Lys-HCl, Arg-HCl and

Gdn-HCl all decreased the change in ThT fluorescence

associated with amyloid fibril formation by RCMj-CN

in a concentration-dependent manner, such that, at

250 mm of Arg-HCl and Gdn-HCl, the increase in

ThT was almost completely abolished (Fig 4A),

pre-cluding analysis of the effect of these concentrations

on the chaperone activity of aB-crystallin (Fig 4B,C)

None of the compounds had an effect on the

morphol-ogy of the amyloid fibrils formed (data not shown) When incubated in the presence of aB-crystallin, the change in ThT fluorescence associated with amyloid fibril formation by RCMj-CN decreased by 30 ± 3% (1.0 : 0.5 w⁄ w ratio of RCMj-CN : aB-crystallin) (Fig 4B) The amino acids had no significant effect on the chaperone activity of aB-crystallin against this fibril-forming target protein (Fig 4C) At 100 mm, Gdn-HCl had a negative effect on the chaperone acti-vity of aB-crystallin in preventing amyloid fibril forma-tion by RCMj-CN

Amyloid fibril formation by a-synucleinA53T

At 37 C, the increase in ThT fluorescence associated with fibril formation by a-synucleinA53T reached a plateau after 140 h (Fig 5A) Electron micrographs of a-synucleinA53T at the end of the assay confirmed the formation of fibrils, which were long (between 1 and

5 nm), straight and unbranched (Fig 6C,D) Addition

of Gly and Lys-HCl at 250 mm increased both the rate and magnitude of the change in ThT fluorescence asso-ciated with fibril formation by a-synucleinA53T (Fig 5A,C) Overall, Arg-HCl had little effect on fibril formation by a-synucleinA53T, whereas Gdn-HCl at

A

C

B

Fig 4 Ordered aggregation of RCMj-CN into amyloid fibrils is inhibited by aB-crystal-lin but this activity is not affected by Arg-HCl The change in ThT fluorescence at

490 nm was used to monitor amyloid fibril formation by RCMj-CN ( , 0.5 mgÆmL)1) in (A) the absence or (B) the presence of aB-crystallin (0.25 mgÆmL)1) For both (A) and (B), RCMj-CN was incubated at 37 C in

50 m M phosphate buffer, pH 7.2, without shaking for 15 h in the presence of 250 m M

of Gly (d), Lys-HCl ()), Arg-HCl ( ) or Gdn-HCl (h) The buffer-only control (r) is also shown (C) Percentage protection data (mean ± SEM of three independent experi-ments), calculated 15 h after the start of the assay, for RCMj-CN incubated with increas-ing concentrations of the additives in the absence ( ) or presence ( ) of aB-crystallin The percentage protection that would result

if there was no influence of the additives on the chaperone activity of aB-crystallin, as described in the Experimental procedures, is also shown (j) The asterisk indicates denotes a significant (P < 0.05) decrease in the chaperone ability of aB-crystallin in the presence of 100 m M Gdn-HCl.

250 mm inhibited it by 53 ± 5% This significant

decrease in the amount of aggregation in the presence

of high concentrations of Gdn-HCl precluded analysis

of the effect of this concentration when aB-crystallin

was also present Therefore, we also tested Gdn-HCl at

100 mm in these studies (Fig 5), and this concentration was found to inhibit fibril formation by a-synuclein-A53T by 21 ± 2% None of the compounds were found to have an effect on the morphology of the fibrils formed by a-synucleinA53T (data not shown), and thus

A

C

B

Fig 5 Amyloid fibril formation by

a-synucle-inA53T is inhibited by aB-crystallin, and this

chaperone activity is enhanced by Lys-HCl,

Arg-HCl and Gdn-HCl Fibril formation was

induced by incubating a-synucleinA53T ( ;

2.0 mgÆmL)1) with constant shaking at

37 C in 50 m M phosphate buffer,

contain-ing 100 m M NaCl, pH 7.4, for 5 days either

in (A) the absence or (B) the presence of

aB-crystallin (0.4 mgÆmL)1) and either

250 m M of Gly (d), 250 m M of Lys-HCl ()),

250 m M of Arg-HCl ( ) or 100 m M of

Gdn-HCl (h) The buffer-only control (r) is also

shown (C) Percentage protection (mean ±

SEM of three independent experiments) for

a-synucleinA53T incubated with the

addi-tives in the absence ( ) or presence of

aB-crystallin ( ) was calculated using data from

the 160 h time point The percentage

pro-tection that would result if there was no

influence of the additives on the chaperone

activity of aB-crystallin is also shown (j),

the hash symbol (#) denotes a significant

(P < 0.05) increase in the chaperone ability

of aB-crystallin in the presence of the

addi-tive Note that the concentration of Gdn-HCl

used in this experiment is 100 m M

Fig 6 Amyloid fibrils formed by the

ordered aggregation of RCMj-CN and

a-synucleinA53T Electron micrographs of

RCMj-CN (0.5 mgÆmL)1, A and B) and

a-synculeinA53T (2.0 mgÆmL)1, C and D)

500 lgÆmL)1) following incubation at 37 C

in 50 m M phosphate buffer, pH 7.2, for 15 h

and 50 m M phosphate buffer containing

100 m M NaCl, pH 7.4, for 5 days,

respec-tively The scale bars represent 1 lm (A, C)

and 0.2 lm (B, D).

the change in ThT fluorescence is interpreted to be

directly attributable to a change in the number of fibrils

formed in the presence of these additives In the

presence of aB-crystallin (1.0 : 0.2 w⁄ w ratio of

a-synucleinA53T : aB-crystallin), the increase in ThT

fluorescence associated with fibril formation by

a-synu-cleinA53T was decreased by 46 ± 3% (Fig 5B,C) Gly

had no significant effect on the chaperone activity of

aB-crystallin in preventing the increase in ThT

fluores-cence associated with fibril formation by

a-synuclein-A53T, but both Lys-HCl and Arg-HCl were found to

significantly increase its chaperone activity, such that,

at 250 mm, the percentage protection was increased

to 27 ± 3% (Lys-HCl) and 99 ± 4% (Arg-HCl)

(Fig 5C) Similarly, at 100 mm, Gdn-HCl also

signifi-cantly increased the chaperone activity of aB-crystallin

(84 ± 4%) against this target protein

The effect of Arg-HCl on the structure and

assembly of aB-crystallin

We investigated whether the effects of these additives

on the chaperone action of aB-crystallin were

attribut-able to changes in the quaternary structure and

oligo-merization of the protein We found that, at 250 mm,

none of the compounds had a significant effect on the

oligomeric size of aB-crystallin as assessed by

size-exclusion chromatography (Fig 7A) (i.e in either the

absence or presence of the compounds, aB-crystallin

was found to elute with an apparent mass of 580 kDa,

which corresponds to the mass of the oligomer

reported previously [32]) We also found no significant

differences in the accessibility of exposed hydrophobic

clusters, as assessed by ANS fluorescence (Fig 7B), or

solvent accessibility of the N-terminal tryptophan

resi-dues (Trp9 and Trp60), as assessed by intrinsic

fluores-cence (data not shown), in the presence of these

compounds Thus, it appears that the additives may

cause subtle changes in the structure of both the target

protein and aB-crystallin that lead to changes in the

chaperone activity of aB-crystallin for some target

pro-teins but not others

Discussion

We have investigated the effect of Arg-HCl on the

chaperone activity of aB-crystallin against various

tar-get proteins undergoing either disordered (amorphous)

or ordered (i.e amyloid fibril formation) aggregation

We show that the effect of these compounds on the

chaperone activity of aB-crystallin is dependent on

the target protein undergoing aggregation Thus, our

results highlight the need to consider a number of

aggregation systems in order to assess the effect of var-ious additives and⁄ or modifications on the overall activity of chaperone proteins Of the target proteins tested, Arg-HCl was found to specifically increase the activity of aB-crystallin against DTT-induced precipi-tation of insulin at intermediate and high concentra-tions, and it also increased the activity of aB-crystallin

in preventing the aggregation leading to amyloid fibril formation by a-synucleinA53T when used at high concentrations With regard to the latter result, the increase in chaperone activity resulting in the inhibi-tion of fibril formainhibi-tion by a-synucleinA53T was not specific for Arg-HCl as Lys-HCl and Gdn-HCl showed similar effects (Fig 5C)

A number of studies have indicated that small mole-cules, including common metabolites such as pante-thine and glutathione [33], can increase the chaperone activity of a-crystallin We confirm here previous results showing that high concentrations of Arg-HCl

Fig 7 The additives have no effect on the oligomeric size of aB-crystallin (A) or its ability to bind ANS (B) (A) aB-aB-crystallin (1.0 mgÆmL)1), in the absence or presence of 250 m M of the addi-tives, was loaded on to a Superdex 200HR 10 ⁄ 30 column and eluted in 50 m M phosphate buffer, pH 7.2, at a flow rate of 0.4 mLÆmin)1 Calibration of the column was performed using (1) blue dextran, void; (2) thyroglobulin, 670 kDa; (3) c-globulin,

158 kDa; (4) ovalbumin, 44 kDa; (5) myoglobulin, 17 kDa (B) ANS fluorescence of aB-crystallin (0.1 mgÆmL)1) in 50 m M phosphate buffer, pH 7.2, alone ( ) or in the presence of 250 m M of Gly (d), Lys-HCl ()), Arg-HCl ( ) or Gdn-HCl (h), monitored following exci-tation at 350 nm The samples were maintained at 37 C for

30 min before the fluorescence spectra were obtained.

(> 100 mm) increase the chaperone activity of

aB-crystallin against the DTT-induced precipitation of

insulin [27,28] These studies also showed that 100 mm

Arg-HCl increases the chaperone activity of

a-crystal-lin against the thermally induced aggregation of

f-crys-tallin at 43C [27] Our results indicate that this effect

of Arg-HCl is not limited to proteins undergoing

dis-ordered (amorphous) aggregation, as Arg-HCl also

increases the ability of aB-crystallin to reduce amyloid

fibril formation by a-synucleinA53T This result is

sig-nificant due to the association of this type of protein

aggregation with disease Lys-HCl and Gdn-HCl also

enhanced the chaperone activity of aB-crystallin

against this fibril-forming protein, implying that it is

the common positively charged group that plays a role

in increasing the activity of aB-crystallin against this

target protein To our knowledge, this is the first study

that has investigated the effects of small molecules,

such as amino acids and Gdn-HCl, on the chaperone

function of sHsps against amyloid fibril-forming target

proteins Whilst the concentrations used in these

stud-ies are high, the results suggest that small molecules

such as these may represent important therapeutic

leads for increasing the protective ability of chaperone

proteins against disease-related amyloid fibril

forma-tion

Interestingly, none of the compounds tested

increased the chaperone activity of aB-crystallin

against amyloid fibril formation by RCMj-CN, a

milk-derived protein that readily forms fibrils under

conditions of physiological pH and temperature The

differences in the effect of the small molecules on the

chaperone activity of aB-crystallin against the two

amyloid fibril-forming target proteins may be

attribut-able to differences in the rate of fibril formation

(RCMj-CN forms fibrils much more rapidly than

a-synucleinA53T) or the nature of the amyloidogenic

intermediate(s) with which aB-crystallin interacts

Moreover, we found no generic effect of each

com-pound on the chaperone activity of aB-crystallin

We have previously shown that phosphorylation of

aB-crystallin, which occurs under conditions of cellular

stress [34,35], also has a differential effect on its

chap-erone activity, increasing the activity against some

target proteins, but decreasing it against others [18]

Thus, we conclude that aB-crystallin most likely

employs various methods of binding (or binding

modes) in order to prevent the aggregation of stressed

proteins Some of these binding modes (or binding

sites) are favoured by phosphorylation or interaction

with compounds such as Arg-HCl, whilst others are

either not affected or are perturbed Studies using

destabilized T4 lysozyme mutants have shown that

both aA- and aB-crystallin possess at least two binding modes, and that these are influenced by both external factors (e.g changes in temperature and pH) and intrinsic factors (e.g mutation and phosphorylation) [23,26,36] Various binding modes may facilitate the interaction of aB-crystallin with the various intermedi-ates formed during the aggregation process of diverse targets It may also enable the chaperone protein to better cope with the various types of stresses experi-enced by cells that cause proteins to unfold

Of course, the effect of compounds such Arg-HCl and Gdn-HCl may be also due to changes that they induce in the stability and⁄ or intermediate states of the target protein itself The denaturant effect of guanidine

on proteins is well established; it decreases the stability

of the native protein but also suppresses aggregation

by weakening the hydrophobic intermolecular interac-tions between the unfolded states of a protein (i.e increasing the solubility of the unfolded state) In con-trast, arginine has been shown to suppress aggregation

of some proteins by acting on the unfolded state of the protein and increasing the reversibility of unfolding [37] Arginine had no effect on the stability of the pro-tein’s native state, although it may also interact with it [37] This effect of arginine on protein aggregation has been attributed to the guanidinium group of the compound, which, through electrostatic interactions, prevents the intermolecular interactions leading to aggregation [37–39] However, its effects vary from protein to protein [9] This is clearly evident from our studies in which, even at low concentrations, the aggre-gation of target proteins examined was affected by the compounds used, and this varied for different target proteins (e.g whilst Arg-HCl at 250 mm had little effect on the aggregation of insulin or a-synucleinA53T alone, it dramatically increased the aggregation of cat-alase and a-lactalbumin but significantly decreased the ordered aggregation leading to fibril formation by RCMj-CN) As such, consideration not only for the effect of compounds on the activity of the chaperone protein, but also its destabilized target, must be taken into account when examining the effect of an additive

on the activity of chaperone proteins

We have shown that the mechanism by which the tested molecules influence the activity of aB-crystallin

is not through gross quaternary structural changes (as assessed by size-exclusion chromatography; see Fig 6A) or changes in exposure of the tryptophan resi-dues or clustered regions of exposed hydrophobicity (as assessed by intrinsic and ANS fluorescence) of the protein With regard to the effect of Arg-HCl on the mass of aB-crystallin, a previous study [27], using glyc-erol sedimentation, reported that 300 mm Arg-HCl

resulted in a decrease in the size of aB-crystallin, which

implies that, at higher concentrations than used in this

study, Arg-HCl may have a significant effect on the

quaternary structure of aB-crystallin However, at

250 mm, we found that the effect of these additives on

the mass of aB-crystallin is negligible, and these data

are in agreement with previous work using Gdn-HCl

at the same concentration [40,41] Previous studies

employing both near and far-UV circular dichroism

have also reported that there is little effect of Arg-HCl

on the overall secondary or tertiary structure of

a-crys-tallin, but that Arg-HCl mediates an increase in

sub-unit exchange and destabilization of the overall

structure of a-crystallin (as assessed by denaturation

with urea) [28] Arginine’s side chain, the guanidinium

group, is able to interact with a number of functional

groups, including the aromatic side chains of some

amino acids, through a stacking mechanism [42] The

interaction of arginine with aromatic amino acids of

aB-crystallin may facilitate its effects Our results

sug-gest that an increase in subunit exchange in the

pres-ence of Arg-HCl may only be important in enhancing

the chaperone activity of sHsps against certain target

proteins Moreover, these are likely to be limited to

those situations in which the chaperone forms only a

transient complex with the target protein, such as

has been described for the amorphous aggregation of

a-lactalbumin [43] and amyloid fibril formation by

apoC-II [16], as we found no evidence that the overall

ability of aB-crystallin to suppress the aggregation of

these target proteins was the same after extended time

periods

In summary, our results show that the effect of

small compounds (such as Arg-HCl) on the chaperone

activity of aB-crystallin is highly dependent on the

aggregating target protein Significantly, we found that

Arg-HCl, Lys-HCl and Gdn-HCl increased the ability

of aB-crystallin to prevent the ordered aggregation

leading to amyloid fibril formation of a mutant form

of the Parkinson’s disease-related protein a-synuclein

(i.e a-synucleinA53T) These results suggest that, due

to their action on molecular chaperone proteins,

bio-logically compatible small molecules, such as Arg-HCl,

may be potential candidates as therapeutic agents in

the treatment of protein-misfolding diseases

Experimental procedures

Materials

Bovine j-casein was obtained from Sigma Chemical Co

(St Louis, MO, USA), and was reduced and

carboxymethy-lated (RCMj-CN) prior to use as described previously [44]

Thioflavin T (ThT), 8-anilino-1-napthalene sulfonate (ANS) and b-mercaptoethanol, Arg-HCl, Gdn-HCl, Lys-HCl and Gly were also obtained from Sigma The vector pET24d(+) (Novagen, Madison, WI, USA) containing the gene for expression of human aB-crystallin was a kind gift from

W de Jong and W Boelens (University of Nijmegen, Neth-erlands), and the vector pRSETB (Invitrogen, Carlsbad,

CA, USA) containing the human a-synucleinA53T gene was a kind gift from R Cappai (University of Melbourne, Australia) The aB-crystallin and a-synucleinA53T proteins were expressed and purified as described previously [45,46] SDS–PAGE analysis of the purified proteins indicated that they contained < 5% contaminating proteins The concen-trations of proteins used in these studies were determined

by spectrophotometric methods using a Cary 5000 UV-Vis-NIR spectrophotometer (Varian, Melbourne, Australia), and calculated extinction coefficients based on amino acid sequences All the buffers in these experiments were passed through a 0.2 lm filter prior to use

Intrinsic and extrinsic fluorescence Intrinsic tryptophan fluorescence spectra of aB-crystallin (0.1 mgÆmL)1 in 50 mm phosphate buffer, pH 7.2), in the presence or absence of the amino acids or Gdn-HCl, were recorded using a Cary Eclipse fluorescence spectrophotome-ter (Varian) equipped with temperature control and using a cuvette with a 1 cm path length The excitation wavelength was set at 295 nm, and fluorescence emission was moni-tored between 300 nm and 400 nm The excitation and emission slit widths were set at 5 nm Samples were main-tained at 37C for 30 min before being assayed

For the ANS binding studies, a stock solution of meth-anolic ANS (100 mm) was diluted 1000-fold into a 0.1 mgÆmL)1 protein solution in 50 mm phosphate buffer,

pH 7.2 Emission fluorescence spectra were monitored (400–600 nm) following excitation at 350 nm The excita-tion and emission slit widths were set at 5 nm Samples were maintained at 37C for 30 min before being assayed

Chaperone activity assays

To test the relative chaperone activity of aB-crystallin in the presence or absence of the additives, we monitored the aggregation and⁄ or precipitation of various target proteins using either ThT fluorescence or turbidity assays (see below) The effect of the additives on aggregation of the target protein (in the absence and presence of aB-crystallin) was assessed at the end of each assay by calculating the percentage protection using the formula:

%protection¼ 100 ðDIc DIsÞ

DIc

where DIc and DIs represent the change in absorbance or ThT fluorescence for the target protein in the absence