Tyramine was found to be a pure competitive inhibitor of hydrolysis for positively charged substrates with both wild-type butyrylcholinesterase and D70G.. A comparison of the effects of
Trang 1butyrylcholinesterase-catalysed hydrolysis of acetanilides and homologous esters
Patrick Masson1, Marie-The´re`se Froment1, Emilie Gillon1, Florian Nachon1, Oksana Lockridge2and Lawrence M Schopfer2
1 Unite´ d’Enzymologie, De´partement de Toxicologie, Centre de Recherches du Service de Sante´ des Arme´es, La Tronche Cedex, France
2 University of Nebraska Medical Center, Eppley Institute, Omaha, NE, USA
Keywords
aryl acylamidase; benzalkonium;
butyrylcholinesterase; serotonin; tyramine
Correspondence
P Masson, Unite´ d’Enzymologie,
De´partement de Toxicologie, Centre de
Recherches du Service de Sante´ des
Arme´es, BP 87, 38702 La Tronche Cedex,
France
Fax: +33 4 76 63 69 62
Tel: +33 4 76 63 69 59
E-mail: pmasson@unmc.edu
(Received 30 December 2007, revised 27
February 2008, accepted 17 March 2008)
doi:10.1111/j.1742-4658.2008.06409.x
The effects of tyramine, serotonin and benzalkonium on the esterase and aryl acylamidase activities of wild-type human butyrylcholinesterase and its peripheral anionic site mutant, D70G, were investigated The kinetic study was carried out under steady-state conditions with neutral and positively charged aryl acylamides [o-nitrophenylacetanilide, o-nitrotrifluoropheny-lacetanilide and m-(acetamido) N,N,N-trimethylanilinium] and homologous esters (o-nitrophenyl acetate and acetylthiocholine) Tyramine was an acti-vator of hydrolysis for neutral substrates and an inhibitor of hydrolysis for positively charged substrates The affinity of D70G for tyramine was lower than that of the wild-type enzyme Tyramine activation of hydrolysis for neutral substrates by D70G was linear Tyramine was found to be a pure competitive inhibitor of hydrolysis for positively charged substrates with both wild-type butyrylcholinesterase and D70G Serotonin inhibited both esterase and aryl acylamidase activities for both positively charged and neutral substrates Inhibition of wild-type butyrylcholinesterase was hyper-bolic (i.e partial) with neutral substrates and linear with positively charged substrates Inhibition of D70G was linear with all substrates A comparison
of the effects of tyramine and serotonin on D70G versus the wild-type enzyme indicated that: (a) the peripheral anionic site is involved in the non-linear activation and inhibition of the wild-type enzyme; and (b) in the presence of charged substrates, the ligand does not bind to the peripheral anionic site, so that ligand effects are linear, reflecting their sole interaction with the active site binding locus Benzalkonium acted as an activator at low concentrations with neutral substrates High concentrations of ben-zalkonium caused parabolic inhibition of the activity with neutral sub-strates for both wild-type butyrylcholinesterase and D70G, suggesting multiple binding sites Benzalkonium caused linear, noncompetitive inhibi-tion of the positively charged aryl acetanilide m-(acetamido) N,N,N-trime-thylanilinium for D70G, and an unusual mixed-type inhibition⁄ activation (a > b > 1) for wild-type butyrylcholinesterase with this substrate No fundamental difference was observed between the effects of ligands on the butyrylcholinesterase-catalysed hydrolysis of esters and amides Thus,
Abbreviations
AAA, aryl acylamidase; ASCh, acetylthiocholine; ATMA, m-(acetamido) N,N,N-trimethylanilinium; BuChE, butyrylcholinesterase; DFP, diisopropylfluorophosphate; NATAc, N-acetylanthranilic acid; Nbs2, 5,5¢-dithiobis(2-nitrobenzoic acid); o-NA, o-nitroaniline; o-NAC,
o-nitroacetanilide; o-NP, o-nitrophenol; o-NPA, o-nitrophenylacetate; o-NTFNAC, o-nitrotrifluoroacetanilide; o-NTMNPA, o-N-trimethylnitro-phenylaniline; PAS, peripheral anionic site.
Trang 2Cholinesterases are structurally related hydrolases [1].
Acetylcholinesterase (EC 3.1.1.7) plays a key role in
the cholinergic system in terminating the action of
ace-tylcholine, but no clear physiological function has yet
been assigned to butyrylcholinesterase (BuChE; EC
3.1.1.8) [2] BuChE may have physiological functions
related to its esterase activity In particular, it has been
proposed that BuChE may play a role in fatty acid [3]
and lipoprotein [4] metabolism Studies with knock-out
mice for acetylcholinesterase indicate that BuChE can
act in the central nervous system as a surrogate
acetyl-choline-hydrolysing enzyme [5] Both cholinesterases
also display noncholinergic activities Cholinesterase
isoforms may have nonenzymatic roles in axonal
outgrowth, synaptogenesis, cell adhesion, neuronal
migration and developmental neurotoxicity to
organo-phosphates [6–8] Certain nonenzymatic functions of
acetylcholinesterase have been found to depend on the
peripheral anionic site (PAS) [9]; others appear to be
related to a peptide derived from the enzyme
C-termi-nus [10] However, the physiological relevance of these
activities is still unclear [2,11] One noncholinergic
activity displayed by cholinesterases is aryl acylamidase
(AAA; EC 3.5.1.13) activity [12,13] Indeed, there is
some evidence that the AAA activity of BuChE plays
a role in early brain development [14] and in the
formation of amyloid plaques in Alzheimer’s disease
[2,15]
Human plasma BuChE is of toxicological and
phar-macological importance, because it scavenges and
det-oxifies numerous carboxyl ester drugs and prodrugs
[16–18], and carbamyl and phosphoryl esters, including
nerve agents [19] Numerous widely used chemicals are
aryl acylamides (drugs: acetaminophen, phenacetin,
flutamide, isocarboxazid, lidocaine, butanilicaine;
pesti-cide: acephate; herbicides and fungicides: acetochlor,
propanil and butachlor) The AAA activity of BuChE
in plasma and tissues could participate in the
metabo-lism of these aryl acylamide drugs and xenobiotics
However, the potential detoxification role of the AAA
activity of BuChE needs to be addressed
Known AAAs are serine hydrolases that catalyse the deacylation of N-acyl arylamines [20,21] Several AAAs have been identified in mammalian tissues [22,23] Certain acryl acylamidases are identical to carboxylesterases [24] Albumin also displays an AAA activity [25,26] A correspondence between certain molecular forms of AAAs and cholinesterases has been demonstrated in different organs [22] Deacetylation of retinal melatonin into 5-methoxytryptamine is cataly-sed by an eye AAA [27] However, no clear physiologi-cal function has yet been ascribed to most mammalian AAAs At a minimum, AAAs are toxicologically rele-vant because they deacylate arylamide xenobiotics [20,21,28]
The crystal structures of acetylcholinesterase and BuChE reveal that these enzymes have a common architecture, with only one catalytic triad located at the bottom of a deep gorge [29] However, it has been suggested that esterase and amidase active centres are nonidentical, although they are overlapping [29–33] Contrary to this proposal, recent kinetic studies and structure–activity relationships have clearly indicated that BuChE utilizes the same catalytic site to hydrolyse anilides and esters [24,25]
It has been reported that the AAA activity of BuChE, as well as that of acetylcholinesterase, can be either activated or inhibited by various ligands These ligands include: (a) biogenic amines: serotonin (5-hydroxytryptamine), tryptamine and related mole-cules [22,32,36–39]; (b) kynuramine [22] and tyramine [22,36,37,39]; (c) procainamide [39]; (d) anti-Alzheimer drugs: (+)huperzine A, donepezil, galantamine and tacrine [32,40,41]; and (e) a cationic detergent that is
an acetylcholine (nicotinic) agonist: benzalkonium [42] Although most ligands were found to be reversible inhibitors of the BuChE-catalysed hydrolysis of o-nit-roacetanilide (o-NAC), tyramine was found to be an activator However, kinetic analysis of these inhibiting
or activating effects was either incomplete [22,36–39]
or debatable [33,42] In particular, reported results were interpreted in terms of an AAA site distinct from
butyrylcholinesterase uses the same machinery, i.e the catalytic triad S198⁄ H448 ⁄ E325, for the hydrolysis of both types of substrate The differ-ences in response to ligand binding depend on whether the substrates are neutral or positively charged, i.e the differences depend on the function of the peripheral site in wild-type butyrylcholinesterase, or the absence of its function in the D70G mutant The complex inhibition⁄ activation effects of effectors, depending on the integrity of the peripheral anionic site, reflect the allosteric ‘cross-talk’ between the peripheral anionic site and the cata-lytic centre
Trang 3the ester site Moreover, certain of these studies were
performed using partially purified enzymes from sera
[22,36,42], biological fluids [37,43] or commercial
prep-arations [32,42] that very probably contained serum
albumin as a contaminant Human serum albumin has
been found to display intrinsic AAA activity [25,26]
Thus, in order to provide a complete analysis and to
clarify debated issues, we investigated the effects of
tyramine, serotonin and benzalkonium on the
BuChE-catalysed hydrolysis of neutral and charged aryl acetyl
amides [o-NAC, o-nitrotrifluoroacetanilide
(o-NTF-NAC) and m-(acetamido) N,N,N-trimethylanilinium
(ATMA)] and acetyl esters [o-nitrophenylacetate
(o-NPA) and acetylthiocholine (ASCh)] under
steady-state conditions All of these substrates give the same
acyl enzyme intermediate Effects on wild-type human
BuChE and its PAS mutant D70G were compared
Because the presence of contaminating proteins
dis-playing AAA activity, e.g albumin, in the BuChE
preparation could have biased the results, experiments
were carried out on highly purified recombinant
enzymes free of albumin and any other AAAs It was
found that there was no fundamental difference in the
mechanisms of inhibition and activation for either the
AAA or esterase activities by these ligands In
addition, differences between the behaviour of the
wild-type enzyme and D70G were found to reflect
alterations in the binding of positively charged
substrates⁄ ligands on PAS, regardless of the type of
substrate (acetyl amide or acetyl ester)
Results and Discussion
Action of reversible effectors on AAA activity of
BuChE
The investigation of the effects of the ligands
(tyramine, serotonin and benzalkonium) on the AAA
and esterase activities of BuChE was performed in
par-allel on wild-type enzyme and the D70G mutant The
substrates were neutral and positively charged
acetani-lides (o-NAC and ATMA) and esters (o-NPA and
ASCh)
The hydrolysis of neutral substrates by BuChE, in
the absence of effectors, obeys the Michaelis–Menten
model (Scheme 1, boxed mechanism in Scheme 2) that
is described by Eqn (1):
v¼kcat½E½S
with
Km¼ Ksk3
ðk2þ k3Þ¼
Ks
½1 þ ðk2=k3Þ ð2Þ
kcat¼ k2k3
kcat=Km¼ k2=Ks ð4Þ
The hydrolysis of positively charged substrates by BuChE, in the absence of effectors, shows either activa-tion or inhibiactiva-tion by excess substrate The BuChE-catal-ysed hydrolysis of positively charged substrates is conveniently described by Scheme 2 [44] In Scheme 2, the enzyme–substrate complex SpE corresponds to S bound on PAS Once the first substrate molecule has bound to the catalytic binding site (ES), a second sub-strate molecule can bind to PAS to form the ternary complex SpES The kinetics of this scheme are described
by Eqn (5):
v¼ kcat½E
1þ Km=½S
1þ b½S=Kss
1þ ½S=Kss
ð5Þ
where Kss is the dissociation constant of complexes SpE and SpES (Kss> Km) The parameter b reflects the efficiency with which SpES forms products When b > 1, there is substrate activation; when b < 1, there is substrate inhibition; when b = 1, the enzyme kinetics obey the simple Michaelis–Menten model (Eqn 1) BuChE shows substrate activation with ATMA (b = 1.53, Kss = 0.7
mm [35]) and ASCh (b = 2.7, Kss= 0.6 [45])
The ligands act as either inhibitors or activators depending on the nature of the substrate: neutral or charged Homologous pairs of substrates (e.g acetyl anilide⁄ acetyl ester) show the same type of inhibition The binding constants for D70G were generally higher than those for wild-type BuChE, indicating that PAS
is involved in some of these effects Additional com-plexities are seen with benzalkonium The results are summarized in Tables 1 and 2 The following is an analysis of these effects
Scheme 1 General scheme for hydrolysis of neutral substrates by
BuChE.
Scheme 2 General scheme for hydrolysis of positively charged substrates by BuChE.
Trang 4Effects of tyramine
Tyramine was found to be an activator of both
D70G-and wild-type BuChE-catalysed hydrolysis for neutral
substrates (b > 1, a > 0) (Table 1), but was an
inhibi-tor of hydrolysis for charged substrates (a = 0)
(Table 2) The affinity of wild-type BuChE for
tyra-mine was higher than that of D70G with both acetyl
anilides and acetyl esters
Our results with o-NPA and o-NAC confirm the
reports that tyramine is an activator for the hydrolysis
of o-NAC by wild-type BuChE [22,36,37,39] The acti-vating effect of tyramine yields the expected hyperbolic Dixon and Cornish–Bowden plots (Fig 1A,B) for hydrolysis of both o-NPA and o-NAC This nonessen-tial activation can be mathematically treated in a man-ner similar to partial mixed-type inhibition (see Experimental procedures, Scheme 3) Similar activating effects on the BuChE-catalysed hydrolysis of o-NPA have been reported for the positively charged ligands dibucaine [45], amiloride [46] and tetraalkylammonium compounds [47] This activation was interpreted in
Table 1 Effect of ligands (tyramine, serotonin and benzalkonium) on esterase and AAA activities of BuChE with the neutral substrates o-NPA versus o-NAC Values are means ± standard error from three to five independent determinations H, hyperbolic; L, linear; P, parabolic These terms refer to the appearance of the Dixon plots Hyperbolic curves appear when there is partial inhibition or when there is activation Parabolic curves indicate multiple ligand binding A, activation; C, competitive; I, inhibition ND, not determined.
Substrate
a
No binding up to 4 m M tyramine; weak activation beyond 4 m M bNo activation even at low [benzalkonium]. cWeak activation at low [benzalkonium] at low [o-NPA] d Competitive inhibition occurs beyond 0.3 m M benzalkonium at the lowest [o-NAC].
Table 2 Effect of ligands (tyramine, serotonin and benzalkonium) on esterase and AAA activities of BuChE with the positively charged sub-strates ASCh versus ATMA Values are the means ± standard error from three to five independent determinations H, hyperbolic; L, linear These terms refer to the appearance of the Dixon plots Hyperbolic curves appear when there is partial inhibition or when there is activation.
A, activation; C, competitive; I, inhibition; M, mixed; N, noncompetitive; U, uncompetitive.
Substrate
a Benzalkonium chloride precipitated with Nbs2under our assay conditions b At high substrate concentration [42] c Under experimental con-ditions, but theory predicts HA at high substrate concentration because a > b > 1 d Inhibition at substrate concentration lower than [S]cross ([S] cross = 0.66 m M ) and activation at [S] > [S] cross
Trang 5terms of binding of the positively charged ligands to
PAS This would form a ternary complex, LPASES,
that linearly accelerates catalysis (bkcatwith b > 1)
The degree of activation in the presence of tyramine
was higher for the hydrolysis of o-NAC than for the
hydrolysis of o-NPA This was determined from the
nonactivated and asymptotic limits in the nonlinear
hyperbolic acceleration plots (Fig 1A,B), which
pro-vided estimates of b For o-NPA hydrolysis b = 2.8,
and for o-NAC hydrolysis b = 5.5 For o-NAC,
because the hydrolysis kinetics were performed under
first-order conditions, the b⁄ a ratio was determined
using Eqn (14) (see Experimental procedures): b⁄ a =
14 ± 6 and a = 0.4 (Table 1) For o-NPA,
experi-ments were performed at [S] close to Km, so that
Eqn (13) (see Experimental procedures), which
describes velocity, gives inaccurate values for a
(a < 1) and therefore b⁄ a > 3
The difference in the extent of activation can be
explained by differences in the rate-limiting steps
Because the rate-limiting step for the hydrolysis of
o-NAC is acylation [35], it follows that the activating
effects of tyramine take place at the level of acylation
For the hydrolysis of o-NPA, both acylation and
deac-ylation are partly rate limiting [35] If, by analogy with
its effect on o-NAC, the activating effects of tyramine
reflect the acceleration of acylation, the activation of
hydrolysis of o-NPA should become limited by the
deacylation rate This predicts that activation should
result in a modest increase in activity Because the dif-ference between the rates for acylation and deacylation must be greater for o-NAC, the activation would be expected to be greater, matching the observations obtained (Table 1)
The binding of tyramine to D70G, in the presence
of o-NPA, is weaker (9.1-fold) than binding to the wild-type enzyme, and induces an activating effect on o-NPA hydrolysis (Table 1) However, it was not pos-sible to determine the a and b parameters by nonlinear fitting of Eqn (13) The activating effect on o-NAC hydrolysis is apparent only for tyramine concentrations greater than 4 mm It is so small that it cannot be quantified It is clear, however, that there is a reduc-tion in affinity with D70G for this ligand This sup-ports the hypothesis that PAS plays a role in binding
of this ligand to BuChE
It was found that tyramine inhibited the turnover of the positively charged substrates ASCh and ATMA in
a linear competitive manner (Table 2) The affinity of tyramine for wild-type BuChE, in this inhibitory capacity (0.78–2.4 mm), is essentially the same as its affinity for the wild-type enzyme in its activating capacity for neutral substrates (0.8–1.0 mm) The same
is true for the binding of tyramine to D70G, although the affinity of D70G for tyramine is weaker than the affinity of the wild-type enzyme From the data in Tables 1 and 2, it can reasonably be stated that, for each enzyme form, Ki= KaThis strongly suggests that
[Tyramine] m M
Aλ
0
10
20
30
40
50
60
A
B
[Tyramine] m M
0
2
4
6
8
10
12
14
16
[Tyramine] m M
0 1 2 3 4
0
100
200
300
400
500
600
S = o-NAC
Fig 1 Activating effect of tyramine on
wild-type BuChE-catalysed hydrolysis of
o-NPA and o-NAC in 0.1 M phosphate buffer
at 25 C (A) o-NPA (d, 0.1 m M ; s, 0.2 m M ;
, 0.4 m M ; h, 0.6 m M ): left panel, Dixon
plots of v)1versus [tyramine]; right panel,
Cornish–Bowden plots of [S] ⁄ v versus
[tyra-mine] Nonlinear Dixon plots are expected
for activation (B) o-NAC (d, 1 m M ; s,
2 m M ; , 3.5 m M ; h, 5 m M ); Dixon plots of
v)1versus [tyramine].
Trang 6both the competitive inhibition of hydrolysis of
posi-tively charged substrates and activation of hydrolysis
of neutral substrates result from tyramine binding to
PAS Such qualitatively opposite effects can be
tenta-tively interpreted in terms of allosteric inhibition⁄
acti-vation: the binding of tyramine to PAS induces a
conformational change that affects the formation of
the productive enzyme–substrate complex It should be
remembered that PAS and the binding locus (W82) of
the active site are connected through an W loop
[29,44] For positively charged substrates, the
confor-mational change prevents the productive binding of
substrate, probably by disrupting the W82–p-cation
interaction; in contrast, for neutral substrates, the
con-formational change optimizes the enzyme–substrate
orientation in the active site pocket for acylation
Effects of serotonin
It was found that serotonin inhibited both esterase and
AAA activities of BuChE (Tables 1 and 2), in contrast
with previous reports [22] The inhibition of wild-type
BuChE was partially (hyperbolic) competitive with the
neutral substrate o-NPA (Fig 2A) It was linearly
competitive with o-NAC (Fig 2B) and with the
posi-tively charged substrates ASCh and ATMA (data not
shown) The inhibition of D70G was linear with all
four substrates The affinity of D70G for serotonin
was generally lower than that of the wild-type enzyme:
7.3-fold in the presence of o-NPA, 5.2-fold with ASCh,
three-fold with ATMA and unaffected with o-NAC The actual binding site of serotonin cannot be inferred from these results Although serotonin can bind to the active site binding locus (W82), binding to PAS cannot
be ruled out If serotonin binds only to W82, the affin-ity differences between wild-type BuChE and D70G mutant for this ligand could reflect differences in the conformational plasticity of the active site gorge of these enzymes
Effects of benzalkonium on o-NPA and o-NAC hydrolysis
The hydrolysis of both o-NPA and o-NAC by wild-type BuChE displayed complexities in the presence of benzalkonium The hydrolysis of o-NAC was activated
at low benzalkonium concentration, and then inhibited
as the benzalkonium concentration increased (Fig 3; Table 1) There was no activation phase for the hydro-lysis of o-NPA Inhibition was parabolic and partial This biphasic behaviour suggests at least two binding sites for benzalkonium Activation of the BuChE-catal-ysed hydrolysis of o-NAC by low concentrations of benzalkonium has been reported previously [42] With o-NPA and o-NAC, benzalkonium shows parabolic competitive inhibition Parabolic inhibition suggests that the binding of more than one benzalkonium con-tributes to the inhibition (Fig 4) The multiplicity of cation binding sites was revealed with phenox-azine⁄ phenothiazine dyes for wild-type BuChE [48],
[Serotonin] m M
0
10
20
30
40
50
A
B
[Serotonin] m M
0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 16
0
1
2
3
4
5
6
7
S = o-NPA
[Serotonin] m M
Aλ 410mn
0
500
1000
1500
2000
2500
3000
3500
4000
[Serotonin] m M
-10 -5 0 5 10 15 0 0 2 4 6 8 10 12
2000
4000
6000
8000
10000
S = o-NAC
Fig 2 Inhibitory effect of serotonin on wild-type BuChE-catalysed hydrolysis of o-NPA and o-NAC in 0.1 M phosphate buffer at
25 C (A) o-NPA (d, 0.1 m M ; s, 0.2 m M ; , 0.4 m M ; h, 0.6 m M ; , 0.8 m M ): left panel, Dixon plots of v)1versus [serotonin]; right panel, Cornish–Bowden plots of [S] ⁄ v versus [serotonin] Nonlinear plots indicate partial inhibition (B) o-NAC (d, 2 m M ; s,
4 m M ; , 10 m M ): left panel, Dixon plots of
v)1versus [serotonin]; right panel, Cornish– Bowden plots of [S] ⁄ v versus [serotonin] Converging Dixon plots and parallel Cornish–Bowden plots indicate competitive inhibition.
Trang 7and with propidium for a mutant (A277W⁄ G283D)
having PAS similar to that of acetylcholinesterase [49]
The fact that benzalkonium acts as an apparent
activator at low concentrations with o-NAC and not
with o-NPA suggests that the activation of wild-type
BuChE occurs at the level of the acylation step, similar
to the mechanism suggested for tyramine Activation is
observable with o-NAC because acylation is rate
limit-ing (kcat= k2), whereas it is ‘buffered’ with o-NPA
because acylation and deacylation are both partly rate
limiting [35]
With D70G, benzalkonium showed a clear hyper-bolic activation of o-NAC hydrolysis at low concentra-tions, and parabolic inhibition at high concentrations (Fig 3; Table 1) There was slight activation of o-NPA hydrolysis at low [S] and low benzalkonium concentra-tion (data not shown) Under these condiconcentra-tions, the hydrolysis kinetics are first order (cf Eqn 12) with
kcat⁄ Km= k2⁄ Ks At low [S], the ‘buffer’ contribution
of deacylation does not take place, and activation reflects an effect on acylation (k2) The fact that D70G
is slightly activated at low benzalkonium concentra-tions, whereas the wild-type enzyme is not, indicates that b⁄ a ‡ 1 for D70G, whereas b ⁄ a < 1 for the wild-type enzyme This subtle difference in behaviour between the two enzyme forms reflects the higher conformational plasticity of the active site gorge of D70G compared with that of the wild-type enzyme for acylation with neutral ester
Effects of benzalkonium on ATMA hydrolysis Hydrolysis of ATMA by wild-type BuChE in the pres-ence of increasing concentrations of benzalkonium gave unusual Lineweaver–Burk plots (Fig 5A) in which the lines intersected in the upper right quadrant
at 1⁄ [S]cross 2 ± 0.5 mm)1 This is consistent with benzalkonium being an inhibitor at low substrate con-centration and an activator as the substrate concentra-tion is increased The highest ATMA concentraconcentra-tion (0.5 mm) was below Kss= 0.70 mm [35], so that acti-vation by excess substrate did not take place This pat-tern of inhibition has been reported previously for decamethonium inhibition of the hydrolysis of 7-acet-oxy-4-methylcoumarin by acetylcholinesterase [50] The inhibition of wild-type BuChE hydrolysis of ATMA by benzalkonium can be described by Scheme 3 (see Experimental procedures) For this scheme, the Lineweaver–Burk plot is given by Eqn (6) When a > b, a > 1 and b > 1, the Lineweaver–Burk lines intersect in the first quadrant at 1⁄ [S]cross
1
v¼
aKm
Vmax
½L þ Ki b½L þ aKi
1
½Sþ
1
Vmax
½L þ aKi b½L þ aKi
ð6Þ
with the coordinates of the intersecting point:
1=½Scross¼ b 1
1=Vmax;cross¼ a 1
This very rare situation in which ligand L is an inhibitor at low [S] and an activator at high [S],
[benzalkonium] m M
A430nm
0.0 0.1 0.2 0.3 0.4 0.5
0.000
0.002
0.004
0.006
0.008
0.010
S = o-NAC
Fig 3 Concentration-dependent activation and inhibition of
BuChE-catalysed hydrolysis of o-NAC (bottom curves, 1 m M ; top curves,
5 m M ) by benzalkonium Full lines, wild-type enzyme; broken lines,
D70G mutant.
[benzalkonium] m M
0
200
400
600
800
1000
1200
0 0.1 0.2 0.3 0.4 0.5
S = o-NAC
Fig 4 Dixon plot of the inhibitory portion of the effect of
benzalko-nium on the wild-type BuChE-catalysed hydrolysis of o-NAC The
plot shows only benzalkonium concentrations greater than 0.1 m M
The substrate concentration was 1 m M o-NAC Nonlinearity
indicates multiple binding.
Trang 8beyond [S]cross, is symmetrical to system C5 of partial
and mixed inhibition as described by Segel [51]
The values for a and b can be determined from the
re-plots of 1⁄ Dslope versus 1 ⁄ [L] and 1 ⁄ Dintercept
ver-sus 1⁄ [L] [51]; Dslope of the LineweaverỜBurk plot is
the difference between the slope at ligand
concentra-tion [L] and the slope without ligand (Eqn 9):
DslopeỬ aKmđơL ợ Kiỡ
VmaxđbơL ợ aKiỡ
Km
Vmax
đ9ỡ
and
1 DslopeỬ aKiVmax
Kmđa bỡ
1
ơLợ
1
Km
bVmax
đa bỡ đ10ỡ
when [L]fi ầ, the intercept on the Dslope)1 axis in the re-plot is bVmax⁄ Km(a Ờ b), the intercept on the [L])1 axis is Ờ b⁄ aKi and the slope is aKiVmax⁄ Km (a) b)
Dintercept is the difference between the intercept of the LineweaverỜBurk plot at ligand concentration [L] and the intercept without ligand (Eqn 11):
DinterceptỬ đơL ợ aKiỡ
VmaxđbơL ợ aKiỡ
1
Vmax
đ11ỡ
and
1 DinterceptỬaKiVmax
đb 1ỡ
1
ơLợ
bVmax
đb 1ỡ đ12ỡ Parameter b = 3.1 was determined from the inter-cept on the Dinterinter-cept)1axis re-plot, i.e bVmax⁄ (b) 1) Then, parameter a was determined from the intercept
on the Dslope)1 axis in the re-plot of Dslope)1 versus [L])1, and Kiwas determined from the intercept on the [L])1axis of the Dslope)1 re-plot (Fig 5A, inset) This gave a = 5.7 and Ki= 0.05 lm
The high value of a reflects the decreased affinity of benzalkonium for the enzymeỜsubstrate complex This result is consistent with the proposal that the binding site for benzalkonium is either at PAS or in the active site gorge close to PAS The binding of benzalkonium would then have to induce a conformational change at the active site that is responsible for the increase in kcat
at high [S] beyond [S]cross Thus, the activating effect
of benzalkonium produces an effect similar to the acti-vation by excess substrate that has been found to be dependent on the integrity of PAS [44,49,52,53] Because the rate-limiting step for the hydrolysis of ATMA is acylation (k2>k3) [35], it is probable that activation reflects an increase in the acylation rate This is similar to the activating effect of tyramine binding on the hydrolysis of o-NAC and o-NPA The observation that benzalkonium is an inhibitor of BuChE-catalysed hydrolysis of ATMA at low [S] and
an activator at high [S] suggests that the conforma-tional change induced by the occupancy of PAS is dif-ferent at low and high [S] This difference could be the result of the binding of a second substrate molecule in the gorge at high concentration, which causes a differ-ent (activating) conformational change in the active site Because the rate-limiting step of BuChE-catalysed hydrolysis of ATMA is acylation, the second substrate molecule must bind in the active site gorge of the enzyme already complexed with the first substrate mol-ecule In acetylcholinesterase, binding of an additional
1/[ATMA] m M
Aλ290 nm
0
1000
2000
3000
4000
5000
6000
7000
A
B
1/[ATMA] m M
Aλ290 nm
0
1000
2000
3000
4000
5000
0 2 4 6 8 10
0.0 0.2 0.4 0.6 0.8 1.0
1/[benzalkonium] ộ M
0.001
0.00 3
0.004
0.005
0.00 2
wi ld t yp e
D7 0G
Fig 5 LineweaverỜBurk plots for the inhibition of
BuChE-cataly-sed hydrolysis of ATMA (0.1Ờ0.5 m M ) by different concentrations
of benzalkonium in 0.1 M phosphate buffer at 25 C (A)
Wild-type BuChE:d, no benzalkonium;s, 0.05 l M ; , 0.1 l M ; h,
0.2 l M ; ấ, 0.3 l M Inset: re-plot of Dslope)1of the LineweaverỜ
Burk plot as a function of the reciprocal of the benzalkonium
concentration (B) Mutant D70G:d, no benzalkonium; s, 2 l M ; ,
3 l M ; h, 4 l M
Trang 9substrate molecule to PAS and in the gorge has been
shown to inhibit enzyme activity by preventing the exit
of reaction products [54,55] In contrast, the active site
gorge of BuChE is about 200 A˚3 larger than that of
the acetylcholinesterase gorge, and therefore may easily
accommodate several ligands⁄ substrates [29,56]
with-out inhibition of substrate and product traffic
The inhibition of D70G by benzalkonium appears
to be purely noncompetitive, i.e the lines in the
Lineweaver–Burk plot cross on the y-axis (Fig 5B) with
Ki= 13 ± 2 lm Thus, the affinity of D70G for
ben-zalkonium is at least 260-fold weaker than that of the
wild-type enzyme This suggests that PAS is the binding
site for benzalkonium, and supports the proposal that
the complexity encountered with the wild-type enzyme
reflects the binding of benzalkonium to PAS
Effects of benzalkonium on ASCh hydrolysis
Under our experimental conditions, a study of the
inhibition of ASCh hydrolysis was not possible
because 5,5¢-dithiobis(2-nitrobenzoic acid) (Nbs2)
pre-cipitated with benzalkonium However, the inhibition
of BuChE-catalysed hydrolysis of ASCh by
benzalko-nium has been reported [42] It is unclear how these
authors avoided the precipitation problem In that
study, the inhibition of human BuChE was found to
be of the partial mixed type Unfortunately, the
experi-ments were performed at high substrate concentration,
in the concentration range corresponding to substrate
activation (cf Experimental procedures, Scheme 2,
Eqn 10) Thus, the reported Ki value (1.03 lm) [42]
probably reflects the inhibition of substrate activation
That is, benzalkonium is probably competing with the
formation of both SpE and the productive ternary
complex SpES (cf Scheme 2) Under these conditions,
the formation of SpE and SpES is governed by a high
Kss (Kss= 0.6 mm), which is about 10 times higher
than Km The Ki value of 1.03 lm reported by these
authors is 20 times higher than our Ki value for the
inhibition of the BuChE-catalysed hydrolysis of
ATMA (at low concentrations of ATMA) This
differ-ence supports our interpretation that these authors
were observing effects related to the substrate
activa-tion poractiva-tion of the mechanism and not to the primary
hydrolytic steps
Interaction of propanil with BuChE
Propanil (3¢,4¢-dichloroacetanilide) was not hydrolysed
by wild-type BuChE under our experimental
condi-tions, i.e [E] > [S] Yet, propanil binds to BuChE and
linearly inhibits the hydrolysis of o-NTFNAC and
ASCh over a large substrate concentration range Inhi-bition constants were determined from Dixon plots and Cornish–Bowden plots (data not shown) Propanil
is a pure competitive inhibitor (a = 0) of the BuChE-catalysed hydrolysis of both substrates: Ki= 0.49 ± 0.05 mm with ASCh, and Ki= 0.74 ± 0.58 mm with o-NTFNAC Thus, propanil interferes with the forma-tion of ES, but not with SpE or SpES
In the BuChE–ASCh complex, the choline head group strongly interacts with W82 [29] The fact that propanil is a competitive inhibitor suggests that it also binds to the p-cation binding site W82 These results imply that other acetyl anilide substrates (i.e o-NAC, o-NTFNAC, ATMA) may bind to W82 in the active centre This would place the substrate in the BuChE– acetanilide substrate complexes into a favourable posi-tion for the use of the catalytic triad Ser198⁄ H438⁄ E325 to make products The resistance of propa-nil to hydrolysis by BuChE probably results from electronic effects contributed by the polar chlorine atoms in the aromatic ring that could hamper the rota-tional flexibility of the amide bond [57] This could impair appropriate orientation of the carbonyl oxygen
in the oxyanion hole
Reaction of N-acetylanthranilic acid (NATAc) with BuChE
Hydrolysis of the negatively charged acetanilide NATAc by BuChE was attempted We chose this sub-strate because it is homologous to aspirin (N-acetyl-salicylic acid), a negatively charged acetyl ester that is
a BuChE substrate [58] We found that BuChE, even
at high concentration ([E] = 0.03 lm), does not hydro-lyse NATAc ([S]max= 0.5 mm) Moreover, NATAc
up to 1 mm did not inhibit the BuChE-catalysed hydrolysis of ASCh (0.035–1 mm) or o-NTFNAC (2–3 mm) Therefore, it does not appear to bind Com-petition of NATAc with the three selected ligands was not investigated
Active structure site responsible for AAA activity
It has recently been suggested that Ser224 is the nucle-ophile involved in the hydrolysis of aryl acylamides by BuChE [33] However, several lines of structural evi-dence clearly rule out this hypothesis First, kinetic analysis of organophosphate inhibition of the ester and AAA activities of BuChE indicates that there is a single nucleophilic serine, Ser198, for both activities [35] Second, studies on mutant forms, e.g silent allo-zyme and S198C⁄ D mutants of BuChE, support the kinetic findings with wild-type BuChE and rule out the
Trang 10hypothesis that a nucleophile other than Ser198 is
responsible for the AAA activity [35] Third, inspection
of the three-dimensional structure of human BuChE
shows that Ser224 is deeply buried inside the protein,
with Oc pointing away from the surface, about 6–7 A˚
from the bulk solvent [29] (Fig 6A) Therefore, no
access for substrate to Ser224 is possible
The latter problem was acknowledged by the
authors of the Ser224 proposal However, it was
argued that the binding of ligands such as
benzalkoni-um may induce a conformational change that activates
a Ser224⁄ His438 ⁄ E197 triad [33] Our present results
show that the effect of benzalkonium on the AAA activity of BuChE can be interpreted without postulat-ing the unmaskpostulat-ing of an alternative nucleophile A conformational change of the enzyme that would give accessibility to Ser224 is unlikely, because it would require a large movement of the main chain and subse-quent disorganization of the central b-sheet Moreover, catalysis relies on optimal angles and distances between the nucleophile and the base in order to allow the formation of short, strong hydrogen bonds The observed spatial position of Ser224 and His438 does not allow the formation of such a short, strong hydro-gen bond For example, the distance between Ne of His438 and Oc of Ser224 is 4.8 A˚ In addition, Ser224
is strongly locked in a dense hydrogen bond network that is essential for the integrity of the active site This dense hydrogen bond network prevents any conforma-tional change of this residue (Fig 6B) Ser224 notably makes a strong hydrogen bond with Glu325 (d 2.5 A˚) Moreover, Ser224 is also hydrogen bonded to two key water molecules that are strictly conserved in all crystal structures of BChE in com-plexes with charged or uncharged ligands that have been solved to date Ligands invariably fill the pocket near Trp82 without triggering any alteration of this hydrogen bond network Finally, His438 is totally restrained because of a stabilizing interaction with Phe398 In contrast with the observations reported for acetylcholinesterases [59–61], no mobility of the cata-lytic histidine of human BChE has been observed in crystal structures A change in the position of this cat-alytic histidine would be necessary for the formation
of a Ser224⁄ His438 ⁄ E197 triad
Conclusions
Despite the complexity of interactions between BuChE, tyramine, serotonin and benzalkonium, no fundamen-tal differences were found between the effects of these compounds on the AAA and esterase activities of human BuChE
The concentrations of tyramine and serotonin that activate or inhibit the AAA activity of BuChE (and also its esterase activity) are several orders of magni-tude higher than the concentrations of these com-pounds that can be encountered in vivo under physiological conditions or even during pathological processes The concentration of serotonin in human plasma of normal subjects is 9 nm [62]; it is increased several fold as a consequence of migraine headache, schizophrenia, hypertension or carcinoid syndrome The concentration of tyramine in the plasma of nor-mal subjects is about 7 nm [63]; it is increased as a
Ser198 Ser224
A
B
Ser224
Ser198
His438
Glu325
wat
wat
2.7 2.8
2.5 2.9 2.7
Fig 6 (A) Overall view of the three-dimensional structure of
human BuChE The solvent-accessible surface is represented by a
mesh Helices are represented as coils and b-sheets by arrows.
Ser224 is represented by cyan balls and Ser198 is represented by
green balls with their respective Oc in red (B) Hydrogen bond
net-work associated with Ser224 and Ser198 Participating residues are
represented as sticks and water molecules as balls Hydrogen bond
distances are given for the catalytic triad residues Ser198 ⁄ His438 ⁄
Glu325 (green) and Ser224 (cyan).