Refining the model to design α-chymotrypsin superactivators: The role of the binding mode of quaternary ammonium salts

Article abstract of DOI:10.1039/d0nj04676k

A number of quaternary ammonium salts with bulky hydrophobic moieties are known to provoke the superactivation of α-chymotrypsin (α-CT) in aqueous solution. In particular, benzyl-substituted ammonium and dicationic ammonium-based salts have recently emerged as promising classes of compounds to induce α-CT superactivation and stabilization. Preliminary in silico modelling suggested the α-CT residue tryptophan 215 to be the major anchor point of these additives. In order to achieve a broader knowledge of the enzyme-additive interactions and to validate the modelling studies, new ammonium-based additives were designed and tested. The hydrophobic interaction resulted in being critical to improving superactivation, with [(2,3,5,6-tetramethyl-p-phenylene)dimethylene]bis[triethylammonium bromide] (bisEDuEAB) resulting as the most effective quaternary ammonium superactivating agent studied so far. Finally, a general agreement between in silico outcomes and kinetic parameters was observed, and data interpretation is discussed based on the proposed α-CT/additive binding modes. This journal is

Full text of DOI:10.1039/d0nj04676k

NJC
View Article Online
View Journal | View Issue
PAPER
Refining the model to design a-chymotrypsin
superactivators: the role of the binding mode of
Cite this: New J. Chem., 2020,
quaternary ammonium salts†
44, 20823
a
b
b
Francesco Gabriele, Laura Goracci, * Raimondo Germani
and
a
Nicoletta Spreti
*
A number of quaternary ammonium salts with bulky hydrophobic moieties are known to provoke the
superactivation of a-chymotrypsin (a-CT) in aqueous solution. In particular, benzyl-substituted
ammonium and dicationic ammonium-based salts have recently emerged as promising classes of
compounds to induce a-CT superactivation and stabilization. Preliminary in silico modelling suggested the
a-CT residue tryptophan 215 to be the major anchor point of these additives. In order to achieve a broader
knowledge of the enzyme–additive interactions and to validate the modelling studies, new ammonium-
based additives were designed and tested. The hydrophobic interaction resulted in being critical to
improving superactivation, with [(2,3,5,6-tetramethyl-p-phenylene)dimethylene]bis[triethylammonium
bromide] (bisEDuEAB) resulting as the most effective quaternary ammonium superactivating agent studied
so far. Finally, a general agreement between in silico outcomes and kinetic parameters was observed, and
data interpretation is discussed based on the proposed a-CT/additive binding modes.
Received 21st September 2020,
Accepted 9th November 2020
DOI: 10.1039/d0nj04676k
rsc.li/njc
the tertiary structure of the enzyme and to increase hydrophobic
interactions between the enzyme and the substrate.
1
. Introduction
a-Chymotrypsin (a-CT) (EC 3.4.21.1), a hydrophilic globular
Recently, the effect of co-solvents, stabilizers, pressure and
serine protease, represents one of the most studied model temperature on the a-CT hydrolysis of the peptide bond was
1
3–15
enzymes, and its structure and mechanism of action are well- also investigated.
1
–3
known since the 1970s.
For this reason, the effect of addi-
For many years our research group has been studying the
tives and new reaction media on the enzyme catalytic properties effect of hydrophilic additives on the activity and stability of
has been extensively studied and reviewed. Many studies in the a-CT; the enzyme was proved to be superactivated by a number
literature have been devoted to the superactivating effect of of cationic surfactants and a bulky hydrophobic head group
4
–6
16,17
hydrophilic additives on a-CT, such as cationic single-chain
seemed to play a decisive role.
In fact, in the presence of
7
–9
10
and gemini surfactants and polyamines. The enhancement cetyltributylammonium bromide (CTBABr), the instantaneous
of enzyme catalytic activity induced by these additives has been activity of the enzyme increased by 8 times compared to that of
attributed to hydrophobic interactions between the additives the buffer. However, further studies indicated that the amphi-
and the protein and this effect was mainly due to an increase of philic nature of the additives was not essential, since ammo-
the enzyme catalytic activity.
niums salts, like tetrabutylammonium bromide (TBABr), which
Polyelectrolytes are also able to induce a-CT super- possesses the same head group of CTBABr but lacks the long
1
1,12
activation,
modifying the electrostatic fields around the hydrocarbon chain, led to the same significant superactivation
enzyme. The favorable interactions between enzyme and sub- (i.e. 8-fold increase of instantaneous activity) and, in addition,
1
8
M
strate led to an increase in both affinity (K ) and catalytic enzyme superactivation is maintained for over two months.
constant (kcat). In particular, kosmotropes were able to stabilize
The effect produced by the ‘‘big head’’ additives on the
activity of a-CT with natural substrates was also assessed
1
9
using electron spray ionization mass spectrometry (ESI-MS)
a
Department of Physical and Chemical Sciences, University of L’Aquila, Via Vetoio,
and with model substrates containing more amino acid resi-
I-67100 Coppito, L’Aquila, Italy
dues than the model substrate, N-glutaryl-L-phenylalanine
b
Department of Chemistry, Biology and Biotechnology, University of Perugia,
2
0
p-nitroanilide (GPNA) used in previous studies.
via Elce di Sotto 8, I-06123, Perugia, Italy. E-mail: laura.goracci@unipg.it
The activation effect of the ‘‘big head’’ additives was ratio-
nalized by assuming that the proximity of the additive to the
†
Electronic supplementary information (ESI) available. See DOI: 10.1039/
d0nj04676k
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020 New J. Chem., 2020, 44, 20823--20833 | 20823

View Article Online
Paper
NJC
active site led to an increase in its hydrophobicity and therefore 2. Experimental
to a greater nucleophilicity of the serine 195 (S195) hydroxyl
2.1 Materials
group, which was mainly responsible for the enzymatic activity.
More recently, cationic ammonium-based additives bearing
Crystalline bovine pancreatic a-chymotrypsin (EC 3.4.21.1)
(a-CT, 24.8 kDa, type II: 3 times crystallized, dialyzed and
lyophilized), N-glutaryl-L-phenylalanine-p-nitroanilide (GPNA),
Tris used for buffer preparation and p-nitroaniline used for
molar absorption coefficient determination were purchased
from Sigma-Aldrich (St. Louis MO, USA). Buffered solutions of
enzyme and substrate were freshly prepared immediately
before use. All chemicals and solvents, all of analytical grade,
were purchased from Merck, and were used as received without
further treatment.
2
1
a benzylic group were synthesized and tested. In particular,
the effect of benzyltrimethylammonium bromide (BzTMABr),
benzyltributylammonium bromide (BzTBABr) and benzyldode-
cyldimethylammonium bromide (BzDDABr) was investigated. A
significant increase in the GPNA hydrolysis instantaneous rate
was observed, with a greater effect of BzTBABr. However, this
effect was accompanied by a faster deactivation of the enzyme
compared to pure buffer.
In addition to the benzyl-substituted ammonium salt, two
novel dicationic salts, (1,8-bis(tributylammonium)octane dibro-
mide (bisBOAB) and 1,4-bis(tributylammonium)xylene dibro-
2.2 Synthesis of additives
2
1
mide (bisBAB)) were also synthesized and tested. The major Mono- and dicationic quaternary ammonium salts were synthe-
difference between the two dicationic salts was in the linker sized by the quaternization of the respective tertiary amines
moiety, which was flexible and rigid, respectively. bisBOAB, with the appropriate bromo derivatives. Melting points were
which can be considered as a dimer of TBABr, induced similar determined on the Barloworld Scientific Stewart SMP3 apparatus
1
superactivation and stabilization effects when compared to the and are uncorrected. H-NMR spectra were registered on a Bruker
monomer, but at a lower concentration. In the case of the more AVANCE DRX 400 instrument using CDCl , CD OD or D O as
3
3
2
rigid bisBAB, the superactivation effect was reduced compared solvents at 25.0 1C. Chemical shifts are given in ppm relative to
1
to the more flexible bisBOAB and enzyme deactivation was the residual H solvent signal.
faster.
Trimethyl(3-phenylpropyl)ammonium bromide (PhPrTMABr)
The collected data were used to perform molecular model- was obtained by reaction of (3-bromopropyl)benzene (50 mmol)
ling studies aiming at rationalizing the observed kinetic effects. with trimethylamine (molar ratio 1 : 1.5) in EtOH, under mag-
2
1
The proposed in silico model suggested that the residue netic stirring for 12 h at room temperature. After the removal of
tryptophan 215 (W215), which is located adjacent to the cata- the solvent, the solid was purified by crystallization. Yield 12.1 g,
lytic site, may represent the anchor point for the quaternary 94%, m.p. 150–152 1C (ethyl acetate-methanol). dH (400 MHz,
ammonium salts possessing a superactivation effect. More CD
precisely, small size quaternary ammonium-based additives N (CH
with reduced flexibility (e.g. BzTBABr) were proposed to interact
3
OD) 2.10–2.18 (2H, m, CH
2
), 2.73 (2H, t, CH
), 3.33–3.38 (2H, m, Ph-CH ); 7.33–7.20 (5H, m, Ph).
Tributyl(3-phenylpropyl)ammonium bromide (PhPrTBABr)
with W215. Due to this binding mode, an increase in the was obtained starting from (3-bromopropyl)benzene (50 mmol)
hydrophobicity of the catalytic site occurs, with a consequent via reaction with tributylamine (molar ratio 1 : 1.5) in CH CN;
2
), 3.12 (9H, s,
+
22
)
3 3
2
3
increase in the kcat value. For dicationic ammonium salts, the the mixture was refluxed for 3 days. The reaction raw product
greater flexibility of the spacer in bisBOAB induced alternative after elimination of the solvent was taken up with ethyl ether
binding modes for the alkyl chains, while the rigid structure for until a solid was obtained; the latter was purified by double
bisBAB seemed to be not optimized to simultaneously interact crystallization from a mixture of ethyl acetate/ethyl ether.
with W215 and with the hydrophobic region generated by Yield 13.8, 72%, m.p. 79–81 1C (ethyl acetate–ethyl ether). dH
histidine 57 (H57). Therefore, the binding poses for dicationic (400 MHz, CD
ammonium salts were in agreement with the lower superactiva- 1.53 (6H, m, 3CH
tion effect. These modelling studies and experimental evi- (2H, t, Ph-CH ), 3.21 (8H, m, 4CH
dences suggested that enzyme catalytic properties could be Hexamethylene bis(triethylammonium)dibromide (bisEHAB)
3
OD) 0.99 (9H, t, 3CH
3
), 1.36 (6H, m, 3CH
2
),
2
), 2.10 (2H, m, Ph-CH
2
-CMx0048; -), 2.73
À
2
À
2
2
), 7.39–7.17 (5H, m, Ph).
strongly influenced not only by a number of structural features was synthesized by quaternization of 1,6-dibromohexane
of the additive (i.e. charge, charge density, size, and hydrophobic/ (41 mmol) with triethyl amine (molar ratio 1 : 2.1) in CH CN;
3
hydrophilic balance), but also by specific additive/enzyme the mixture was refluxed under magnetic stirring for 8 h. After
interactions.
removal of the solvent, the solid was purified by crystallization
This work aimed at validating the previously generated from an acetone/ethyl ether mixture. Yield 14.1 g, 75%, m.p.
hypotheses and the proposed in silico model for the interaction 267–268 1C decompn (acetone–ethy lether).
between ammonium-based additives and a-CT. To this aim,
3 3
dH (400 MHz, CD OD) 1.36 (18H, t, 6CH ), 1.55 (4H, m,
novel ammonium-based additives were designed, synthesized 2CH ), 1.84 (4H, m, 2CH ), 3.24 (4H, m, 2CH ). 3.40 (m, 12H,
2
2
2
2
3
and tested to validate the hypothesized binding mode. The 6CH2).
effect of additives on a-CT catalyzed hydrolysis of GPNA was
Hexamethylene bis(tributylammonium)dibromide (bisBHAB)
studied and the determination of kinetic parameters allowed us was prepared by quaternization of 1,6-dibromohexane (41 mmol)
to better understand the reasons for enzyme superactivation with tributylamine (molar ratio 1: 2.2) in acetonitrile; the mixture
and to estimate the predictive capacity of the model used.
was refluxed under magnetic stirring for 6 days. After removal of
2
0824 | New J. Chem., 2020, 44, 20823--20833 This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020

View Article Online
NJC
Paper
the solvent, the solid was purified by several times crystallization and K
m
in pure buffer and in the presence of additives were
from an acetone/ethyl ether mixture. Yield 8.8 g, 50%, m.p. derived by linear regression analysis of the double reciprocal
1
68–170 1C decompn. (acetone–ethyl ether).
Lineweaver–Burk plots in a range of substrate concentration
À3
À3
dH (400 MHz, D O) 0.95 (18H, t, 6CH ), 1.37 (16H, m, 8CH ), between 0.1 Â 10 M and 2.5 Â 10 M. Experiments were
2
3
2
2
4
1.66 (16H, m, 8CH ), 3.21 (16H, m, 8 CH ).
reproduced at least three times and the differences between
2
2
BisEDuEAB and bisEOMeEAB were obtained from durene duplicates in each experiment were always below 5%.
and 1,4-dimethoxybenzene respectively. These compounds
were initially converted into dibromo derivatives with formal- 2.3 Molecular modelling studies
dehyde and hydrobromic acid by reflux into acetic acid, according
The possible binding poses of the tested additives in the
surroundings of the a-CT catalytic site were explored using
25
to the procedure described in the literature.
,4-Bis(bromomethyl)-2,3,5,6-tetramethylbenzene, purified
by crystallization, m.p. 216–218 1C (ethyl acetate), yield 96%.
dH (400 MHz, CDCl ) 2.33 (12H, s, 4CH ), 4.60 (4H, s, 2CH ).
1
the FLAP (Fingerprints for Ligands and Proteins) software
27
(
Molecular Discovery Ltd, UK). The docking procedure to
21
3
3
2
study the a-CT/additive interaction is described elsewhere.
1
,4-Bis(bromomethyl)-2,5-dimethoxybenzene, purified by
crystallization, m.p. 202–203 1C (methanol), yield 86%. dH
400 MHz, CDCl ) 3.87 (6H, s, 2OCH ), 4.53 (4H, s, 2CH ),
.86 (s, 2H, Ph).
(2,3,5,6-Tetramethyl-p-phenylene)dimethylene] bis[triethyl-
Briefly, the X-ray a-CT structure (pdb code: 4CHA) was processed in
FLAP to describe the catalytic cavity in terms of the GRID Mole-
(
3
3
2
28,29
cular Interaction Fields (MIFs).
Probes used to generate the
6
MIFs were H (shape), DRY (hydrophobic interactions), N1 (H-bond
donor) and O (H-bond acceptor) interactions. Thus, the binding
[
ammonium bromide] (bisEDuEAB) was synthesized by reaction
of 1,4-bis(bromomethyl)-2,3,5,6-tetramethylbenzene (32 mmol)
poses of the additives in the a-CT cavity were generated using FLAP
30–32
in the structure-based mode,
with 50 conformations for each
with triethylamine (molar ratio 1 : 2.1) in CH CN; the mixture
3
additive considered. The 10 top-ranked poses according to the
Glob-Prod descriptor for each additive were visually inspected. The
same approach was used in this study to predict the most probable
binding mode of GPNA into the a-CT cavity.
was refluxed for 6 h. Purified by double crystallization from an
acetone-methanol mixture. Yield 11.5 g, 88%, m.p. 206–208 1C
decompn.
dH (400 MHz, D
.34 (12H, m, 6CH
2
O) 1.23 (18H, s, 6CH
3 3
), 2.44 (12H, s, 4CH ),
3
2
); 4.89 (4H, s, 2CH ).
2
[
(2,5-Dimethoxy-p-phenylene)dimethylene]bis[triethylammonium 3. Results and discussion
bromide] bisEOMeEAB was synthesized by reaction of 1,4-bis-
bromomethyl)-2,5-dimethoxybenzene (31 mmol) with triethylamine
molar ratio 1 : 2.1) in acetonitrile; the mixture was refluxed under
magnetic stirring for 10 h. Purified by double crystallization. Yield
3.1 General design approach
(
(
The design of novel ammonium-based additives was performed
based on a molecular modelling approach previously described.
21
Briefly, in that study the FLAP software was used to evaluate the
most probable binding pose of four ammonium based additives,
aiming at identifying the molecular interactions potentially respon-
sible for the catalytic effect. The four tested additives represented
three scaffolds (Fig. 1): (a) aromatic substituted quaternary
ammonium additives (BzTMABr, BzTBABr); (b) diammonium-
based additives with a flexible linker (bisBOAB); and (c)
diammonium-based additives with a rigid linker (bisBAB). Here,
the previously generated models for each additive were used as a
starting point for the rational design of new analogues, aiming
À1 not only at optimizing the catalytic effect, but also at validating
9
.8g, 75%, m.p. 218–220 1C (acetone–methanol).
dH (400 MHz, CD OD) 1.46 (18H, t, 6CH
), 3.98 (6H, s, 2OCH ), 4.57 (4H, s, Ph–CH
3
3
), 3.48–3.29
–N),
(12H, m, 6CH
2
3
2
2
6
7
.30 (2H, s, Ph).
2
.2 a-Chymotrypsin activity assay
2
1
The a-CT activity assay was already described elsewhere.
Briefly, a-CT activity was measured spectrophotometrically at
2
related to p-nitroaniline (pNA), the hydrolysis product of GPNA.
The molar absorption coefficient (e410) of pNA is 8800 M cm
5.0 Æ 0.1 1C, monitoring the increase in absorbance at 410 nm
À1
the hypothesis previously generated. Using the FLAP editing
tool, two new additives for each scaffold were designed, in the
attempt to optimize their interaction with the a-CT cavity
in pure buffer. Absorbance variation in the presence of the
additives at the tested concentrations was evaluated and used
in rate value estimations. A Shimadzu UV-160A UV-VIS spectro-
photometer equipped with a thermostated cell was used in this
study. The a-CT activity assay mixture was prepared in 0.1 M
(pdb code: 4CHA) in terms of the maximum overlap of GRID
molecular interaction fields (see the Methods section for
details). The structure of the new additives is also shown in
Fig. 1. The design and the observed catalytic effect for each
scaffold is separately discussed in the next paragraphs.
Tris–HCl buffer at pH 7.75, with an enzyme concentration of
À1
À3
0
.2 mg ml (8 mM) and a concentration of the substrate GPNA
of 2.5 Â 10 M. The pH of the mixture remained constant
during analysis. The linear increase of absorbance at 410 nm
due to pNA formation was recorded as a function of time for
3.2 Design of phenyl-based ammonium additives and their
effect on a-CT activity
300 seconds. Concerning the parameters discussed in this
study, the reaction rate of a-CT (i.e. moles of pNA formed per Concerning the phenyl-based ammonium additives (Fig. 1-a),
unit of time) was calculated from the slope of the initial linear previous studies demonstrated a superactivating effect of
curve of pNA concentration vs. time. The kinetic parameters kcat benzyl-substituted ammonium salts, especially BzTBABr, and
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020 New J. Chem., 2020, 44, 20823--20833 | 20825

View Article Online
Paper
NJC
Fig. 1 Design of new ammonium additives. Previously tested compounds are shown on the left, while designed structures are illustrated on the right. (a) aromatic
substituted quaternary ammonium additives; (b) diammonium-based additives with a flexible linker; and (c) diammonium-based additives with a rigid linker.
modeling suggested the interaction of the phenyl ring with the additive solution and in pure buffer (radd/r
W215 and the orientation of the tributylammonium moiety the additive concentration for PhPrTMABr and PhPrTBABr.
towards the catalytic triad as key factors in superactivation. As predicted by the in silico design, both additives produced
b
) as a function of
Therefore, we used the interaction model to evaluate whether a an activating effect on the enzyme, and a bell-shaped trend was
longer linker between the ammonium and the phenyl moiety observed as the concentration of the additive increased; the
could be favorable or detrimental for superactivation to occur. maximum of activity was recorded at 0.1 M for both additives.
In particular, due to synthetic accessibility, the new additives Comparing these results with those previously obtained with
PhPrTMABr and PhPrTBABr were designed and their most the benzyl analogues, however, it was observed that the
probable binding poses and similarity scores were compared presence of a propane linker produced a notable decrease in
to corresponding BzTMABr and BzTBABr (Fig. 2).
superactivity, almost 40% (from 4.7 to 2.9) for the trimethyl
Our model suggested that the presence of a propane linker derivative and more than 50% (from 12.4 to 5.4) for the tributyl
still allows the ammonium moiety of the additive to interact with one. Moreover, the positive effect of a bulky and hydrophobic
the hydrophobic moiety corresponding to H57 close to the ammonium group was less evident compared to benzyl addi-
catalytic triad. In the case of the trimethylammonium head tives. Kinetic parameters were determined to deeply under-
groups (Fig. 2a and b) the propane linker seems to facilitate stand the effect of these novel additives on enzyme activity
such interaction, increasing the percentage of poses oriented and the cause of their lower superactivation with respect to
towards the triad. However, compared to the benzyl analogues, BzTMABr and BzTBABr. Table 1 reports enzyme-substrate affi-
PhPrTMABr and PhPrTBABr are located much closer to the nity (Michaelis constant, K
catalytic site. Therefore, we decided to synthesize PhPrTMABr between kcat values in the presence of additive and in pure
and PhPrTBABr, which were predicted to be a-CT superactivators, buffer (kcat(add)/kcat(b)) and kcat/K values. Experiments were
M
), turnover number (kcat), ratio
M
but the extent of superactivation was uncertain, due to the performed at the additive concentration that produced the
location of the ammonium group that resulted very close to the maximum superactivation effect.
catalytic site.
Enzyme-substrate affinity decreased significantly in the
Thus, PhPrTMABr and PhPrTBABr were synthesized and presence of the two additives but, at the same time, an increase
their effect on a-CT activity was studied to compare the new of turnover number than in pure buffer was also found. These
results with those obtained for BzTMABr and BzTBABr. Fig. 3 trends were very similar to those observed with the benzyl
illustrates the behavior of the ratio between the reaction rate in analogues. As regards kcat, trimethylammonium-based derivatives
2
0826 | New J. Chem., 2020, 44, 20823--20833 This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020

View Article Online
NJC
Paper
Fig. 2 Most probable binding poses for BzTMABr (a), PhPrTMABr (b), BzTBABr (c) and PhPrTBABr (d). For each additive, the ten top-ranked binding poses
were analyzed and clustered in the two most different poses, associated with a percentage of occurrence. The similarity score S calculated according to
the Glob-Prod descriptor of FLAP is provided (S195: serine 195; H57: histidine 57; W215: tryptophan 215; D102: aspartate 102).
M
showed similar values, while a decrease of only 20% was observed increase in the K values for PhPrTMABr and PhPrTBABr is that,
from BzTBABr to PhPrTBABr. In summary, the lower superactiva- when a propane linker is used, the access of the substrate to the
tion produced by the new additives was mainly due to the catalytic cavity is partially hampered reducing the enzyme–sub-
enhancement in K , which was much more pronounced. In strate affinity. Alternatively, the additive may interact with GPNA
M
particular, the increase in K
BzTBABr was of about 3- and 4.5-fold with respect to the buffer, efficient.
respectively, while for PhPrTMABr and PhPrTBABr it was 8.4- and A crystal structure of the GPNA/a-CT complex is not available
1-fold higher than in the buffer. A possible explanation for this so far. Thus, the FLAP software was used to predict the most
M
in the presence of BzTMABr and making the interaction between enzyme and substrate less
1
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020 New J. Chem., 2020, 44, 20823--20833 | 20827

View Article Online
Paper
NJC
Fig. 3 Effect of PhPrTMABr (black) and PhPrTBABr (red) concentration on
the activity of a-chymotrypsin in 0.1 M Tris–HCl buffer, pH 7.75 at 25.0 1C.
Fig. 4 Simultaneous visualization of GPNA and the additives docked into
the a-CT cavity. GPNA is shown in purple. The most probable binding pose
for BzTBABr (a), BzTMABr (b), PhPrTBABr (c), and PhPrTMABr (d).
Table 1 a-Chymotrypsin kinetic parameters in the presence and absence
of additives in 0.1 M Tris–HCl buffer (pH 7.75) at 25.0 1C
2
À1
À1 À1
Additive
K
M
, mM 10 kcat, s
kcat(add)/kcat(b)
k
cat/K
M
, M
s
suggesting a potential electrostatic interaction that could per-
turb the orientation of the substrate into the catalytic cavity
(Fig. 4-d and Fig. S3, ESI†).
—
0.44
1.37
1.46
8.70
10.27
31.10
24.80
—
5.96
7.03
21.30
17.00
33.2
63.5
27.7
156.3
49.6
a
BzTMABr 0.4 M
PhPrTMABr 0.1 M 3.71
BzTBABr 0.15 M 1.99
PhPrTBABr 0.1 M 5.00
a
3.3 Diammonium additives with flexible linker and their
a
From ref. 21.
effect on a-CT activity
Concerning the diammonium additives with a flexible linker
probable binding mode of GPNA. Two most probable poses (Fig. 1-b), the superactivation effect of bisBOAB, was found to
were found (Fig. S1 in ESI†), in which only one was a potentially be lower than the one observed for BzTBABr, and we previously
reactive pose (Fig. S1-a, ESI†), with the corresponding mecha- hypothesized that this could be related to the greater flexibility.
nism of hydrolysis depicted in Fig. S2 (ESI†). Indeed, it is well- Thus, the design aimed at verifying the most probable binding
known that several binding modes are commonly associated modes for two analogues of bisBOAB, having a hexane linker
with a substrate, but only a few poses generally are compatible with a triethyl moiety (bisEHAB) or a tributyl moiety (bisBHAB),
with a reactive event catalyzed by the enzyme. Once the most to modulate hydrophobicity and/or flexibility.
probable reactive pose for GPNA is determined, this pose was
The most probable binding poses for bisBOAB and its
visualized in the a-CT cavity together with PhPrTMABr, analogues are shown in Fig. 5.
PhPrTBABr, BzTMABr and BzTBABr (Fig. 4).
The analysis of the binding poses indicated that the short-
Fig. 4 is not a real simulation of the interaction of the three ening of the linker still allows one tributyl-ammonium based
molecules (enzyme, additive and substrate), as the substrate derivative to anchor W215 and orients the other towards the
and the additive were docked into the protein cavity indepen- catalytic cavity where substrate is usually located (Fig. 5a and b). In
dently of each other. However, considering that the additive is particular, the most probable pose for bisBOAB (Fig. 5-a, right panel)
added to the enzyme buffered solution before the substrate, is actually very similar to the most probable one for bisBHAB (Fig. 5-
Fig. 4 suggests that, while BzTBABr and BzTMABr are posi- b, left panel), as bisBOAB makes a c-curve with the linker that
tioned well at the bottom of the region of interaction of the results spatially very similar to the C6-length. Concerning bisE-
substrate, simply defining its bottom edge (Fig. 4a and b, HAB, however, the reduced hindrance at the ammonium level
respectively), PhPrTBABr and PhPrTMABr are localized much and the shorter linker makes the binding poses much more
close to the substrate (Fig. 4c and d, respectively). In particular, variable (Fig. 5-c), suggesting a lower catalytic effect.
PhPrTBABr partially occupies the substrate region, and indeed
Fig. 6 shows the activity profile of a-CT in the presence of
this additive is associated with the highest KM value in the buffered solutions of newly synthesized bisEHAB and bisBHAB
series (Table 1). In the case of PhPrTMABr, the reduced size of at varying concentrations.
the head group does not induce an overlap of the poses for the
The two observed trends were rather different. Indeed, in
additive and the substrate; however, the N-trimethyl moiety of the presence of bisEHAB, the relative rate increased by just a
PhPrTMABr is very close to the carboxylic function of GPNA, factor of 2 and then remained almost constant in a wider
2
0828 | New J. Chem., 2020, 44, 20823--20833 This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020

View Article Online
NJC
Paper
Fig. 5 The most probable binding poses for bisBOAB (a), bisBHAB (b), bisEHAB (c). For each additive, the ten top-ranked binding poses were analyzed
and clustered in the two most different poses, associated with a percentage of occurrence. The similarity score S calculated according to the Glob-Prod
descriptor of FLAP is provided.
concentration range. On the other hand, a bell-shaped trend of
activity was obtained by varying bisBHAB concentration with a
maximum activity observed at 0.3 M, in which enzyme super-
activation was about 8-fold compared to that of pure buffer.
This behavior was similar to that previously obtained with
2
1
0
.4 M TBABr and 0.1 M bisBOAB. The presence of the three
additives, in fact, increased the rate of the hydrolysis reaction
by about 8-fold, confirming once again the essential role played
by the tributyl head group.
2
1
As previously described, a comparison of the kinetic para-
meters of TBABr with bisBOAB indicated that, despite a similar
superactivation, the addition of bisBOAB increased the enzyme–
substrate affinity, but also decreased the kcat value. Thus, the
effects of bisEHAB and bisBHAB on enzyme kinetic parameters
were also determined and results are shown in Table 2.
As would be expected, in bisEHAB solutions the two-fold
activation with respect to buffer was essentially due to an
increase in kcat value, with K
0%. The effect of bisBHAB on both enzyme–substrate affinity
and turnover number was very similar to that observed in
M
value that increased of about
7
Fig. 6 Effect of bisEHAB (black) and bisBHAB (red) concentration on the
activity of a-chymotrypsin in 0.1 M Tris–HCl buffer, pH 7.75 at 25.0 1C.
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020 New J. Chem., 2020, 44, 20823--20833 | 20829

View Article Online
Paper
NJC
Table 2 a-Chymotrypsin kinetic parameters in the presence and absence in the aromatic moiety. Therefore, having bisBAB as reference
of additives in 0.1 M Tris–HCl buffer (pH 7.75) at 25.0 1C
compounds, a first in silico study was performed to evaluate the
2
À1
À1 À1
binding poses for analogue additives bisEDuBAB and bisEO-
MeBAB, bearing a durene or a 1,4-dimethoxybenzene scaffold
instead of the dibenzyl moiety respectively (Fig. S4 in ESI†).
According to our model, the replacement of four aromatic
hydrogens with methyl groups in bisEDuBAB led to a similarity
score, which was identical to that of BisBAB (S = 1), and very
similar poses as well. Both for bisBAB and bisEDuBAB, a
tributylammonium moiety was involved in the interaction with
W215, with the additive displaying two different orientations in
the cavity. However, for bisEDuBAB the pose having the aro-
matic moiety located closer to the catalytic site resulted in
being more favored (Fig. S4-a and b in ESI†). When polar
substituents are added to the bisBAB scaffold, molecular mod-
elling studies indicated that a totally different binding mode is
preferred. In addition, the similarity score for bisEOMeBAB was
lower than the ones for bisBAB and bisEDuBAB, suggesting
Additive
—
bisEHAB 0.3 Ma 0.70
bisBOAB 0.1 M 1.86
bisBHAB 0.3 M 1.96
K
M
, mM 10 kcat, s
k
cat(add)/kcat(b)
k
cat/K
M
, M
s
0.44
1.46
3.28
19.29
18.44
—
33.2
46.9
2.25
13.21
12.63
103.8
94.1
a
From ref. 21.
bisBOAB, despite the increase in the concentration required to
achieve the same superactivation. This result seemed to confirm
the outcomes of modeling studies, according to which the linker
length did not significantly change the binding poses of the two
additives and therefore their effect on a-CT performance.
3
.4 Diammonium additives with a rigid linker and their effect
on a-CT activity
Concerning the diammonium additives with a rigid linker a less efficient interaction with the protein. The reduced
Fig. 1-c), we aimed at investigating the effect of substituents similarity score seems to be mainly related to the lack of
(
Fig. 7 The most probable binding poses for bisBAB (a), bisEDuEAB (b) and bisEOMeEAB (c). For each additive, the ten top-ranked binding poses were
analyzed and clustered in the two most different poses, associated with a percentage of occurrence. The similarity score S calculated according to the
Glob-Prod descriptor of FLAP is provided.
20830 | New J. Chem., 2020, 44, 20823--20833 This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020

View Article Online
NJC
Paper
interaction with W215, which induces bisEOMeBAB to move Table 3 a-Chymotrypsin kinetic parameters in the presence and absence
of additives in 0.1 M Tris–HCl buffer (pH 7.75) at 25.0 1C
towards the inner part of the cavity competing with GPNA (Fig.
S4-c and S5 in ESI†).
2
À1
Additive
K
M
, mM 10 kcat, s
k
cat(add)/kcat(b)
k
cat/K
M
Unfortunately, synthetic accessibility and solubility issues
hampered the investigation of bisEDuBAB and bisEOMeBAB;
thus, the triethyl analogues bisEDuEAB and bisEOMeEAB were bisEDuEAB 0.05 M
synthesized and tested instead. Fig. 7 shows the FLAP most
probable binding poses for the two tested compounds, and the
poses for bisBAB are also provided for reference.
—
0.44
1.49
0.96
1.46
12.00
24.80
6.59
—
8.20
16.99
4.51
33.2
80.5
258.3
68.6
a
bisBAB 0.1 M
bisEOMeEAB 0.1 M 0.96
a
From ref. 21.
For bisEDuEAB a similarity score of 1 was obtained, as for
bisBAB and bisEDuBAB (S = 1). Nevertheless, the hypothesized
binding modes were rather different, with bisEDuEAB being
not able to reach the oxyanion hole region. This effect seems to
be due, not to the substituents but to the triethyl ammonium
head group, as suggested by comparison with the bisEDuBAB
binding poses. Concerning bisEOMeEAB, the similarity score
was even lower than that for the tributyl analogue and again
lower than the ones for bisBAB and bisEDuAEB, suggesting a
less efficient interaction with the protein depending not only
on the substitution with polar groups but also on the head
group size. As for bisEOMeBAB, an interaction with W215 was
not observed.
bisBAB (6-fold), but it was the highest ever obtained with
cationic additives, both surfactants and salts, studied so far.
In this case, the hydrophobicity was not due to the bulky
tributyl head groups, but to the hydrophobic substituents
linked to the aromatic ring. Simultaneous visualization of
GPNA and bisEDuEAB docked into the a-CT cavity is shown
in the ESI† (Fig. S6). On the other hand, the presence of two
methoxy groups, capable of forming hydrogen bonds, probably
made the catalytic site much more polar, with a consequent
decrease in enzymatic superactivity.
Kinetic parameters were also determined and results are
provided in Table 3.
Enzyme-substrate affinity for both additives was lower than
M
buffer, K being double compared to the reference, but was less
Based on these considerations, a lower catalytic effect was
expected for bisEDuEAB and bisEOMeEAB compared to bis-
BAB. Fig. 8 reports the activity profile of a-CT in the presence of
bisEOMeEAB and bisEDuEAB as a function of additive
concentration.
than other superactivating additives, which for BzTBABr and
bisBOAB was about 2 mM and for cetyltributylammonium
1
6
bromide (CTBABr)
(
and tetrabutylammonium bromide
Results obtained with the two diammonium additives hav-
ing a rigid linker were quite different from each other and
compared with bisBAB, the reference additive previously
1
8
TBABr) were 3.7 mM and 6.1 mM, respectively.
An increase in kcat was observed for both additives, but their
effects were very different, i.e. 4.5-fold for bisEOMeEAB and 17-
fold for bisEDuEAB.
2
1
studied. Both curves of reaction rate versus surfactant concen-
tration were bell-shaped, with a maximum of activity at an
additive concentration of 0.1 M and 0.05 M for bisEOMeEAB
and bisEDuEAB, respectively. However, superactivation effect
produced by the most hydrophobic additive, bisEDuEAB (about
The value of the turnover number obtained with bisEDuEAB
was not the highest reached, since in the presence of BzTBABr
2
À1
and TBABr, it was equal to 31 Â 10 s but, given the greater
affinity of the enzyme for the substrate, the catalytic efficiency
was certainly the highest reached until now.
15-fold), not only was higher than bisEOMeEAB (4.5-fold) and
These results seemed to confirm once again the key role of
hydrophobic moieties on the structure of the additive in the
superactivation of a-CT, which can increase the nucleophilicity
of the catalytic serine. On the other hand, however, one or more
bulky head groups hinder the access of the substrate into the
active site, slightly reducing the superactivity.
4. Conclusions
In this research work, an in silico model previously used to
rationalize kinetic behaviors and superactivation effects of a
few quaternary ammonium salts toward a-CT was applied to
design new ammonium-based additives of three different
chemical series, with the double aim of testing the predicting
capabilities of the model and to get new insights into the key
interaction in a-CT superactivation. The in silico prediction well
correlated with experimental findings not only based on the
similarity scores obtained, but especially when a binding mode
Fig. 8 Effect of bisEOMeBAB (’) and bisEDuEAB ( ) concentration on
the activity of a-chymotrypsin in 0.1 M Tris–HCl buffer, pH 7.75 at 25.0 1C. analysis is performed.
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020 New J. Chem., 2020, 44, 20823--20833 | 20831

View Article Online
Paper
Regarding phenyl-based additives, the significant reduction
NJC
8 Y. A. Valiullina, E. A. Ermakova, D. A. Faizullin, A. B.
Mirgorodskaya and Y. F. Zuev, Structure and catalytic activity
of a-chymotrypsin in solutions of gemini surfactants, Russ.
Chem. Bull., 2014, 63(1), 273–279.
9 G.-Y. Bai, J.-L. Liu, J.-X. Wang, Y.-J. Wang, Y.-N. Li, Y. Zhao
and M.-H. Yao, Enzymatic Superactivity and Conforma-
tional Change of a-CT Induced by Cationic Gemini Surfac-
tant, Acta Phys.-Chim. Sin., 2017, 33(5), 976–983.
in enzyme superactivity observed when a propane linker
replaced a methylene was due to an increase in K , and
M
molecular modelling studies suggested that these additives
partially occupy the substrate interaction region, hindering its
access to the active site. The hydrophobic interaction resulted
in being critical to improve superactivation, and this was
especially evident for the series of diammonium additives
bearing a rigid linker. Indeed, bisEDuEAB, which was designed, 10 T. Kurinomaru, T. Kameda and K. Shiraki, Effects of multi-
synthesized and tested in this study, is the most effective
ammonium additive studied so far. In addition, bisEDuEAB
contrasts the paradigm that bulky hydrophobic head groups are
valency and hydrophobicity of polyamines on enzyme
hyperactivation of a-chymotrypsin, J. Mol. Catal. B: Enzym.,
2015, 115, 135–139.
needed to increase the superactivation effect, since the overall 11 A. Endo, T. Kurinomaru and K. Shiraki, Hyperactivation of
hydrophobicity seems to play a role. The presence of hydro-
phobic substituents linked to the aromatic ring instead of to
a-chymotrypsin by the Hofmeister effect, J. Mol. Catal. B:
Enzym., 2016, 133, S432–S438.
the head group produced a more hydrophobic microenviron- 12 A. Endo, T. Kurinomaru and K. Shiraki, Hyperactivation of
ment with a consequent increase in the nucleophilicity of the
catalytic serine residue and also allowed an easier entry of the
serine proteases by the Hofmeister effect, Mol. Catal., 2018,
455, 32–37.
substrate into the active site (lower KM value than tributyl 13 A. Wangler, R. Canales, C. Held, T. Q. Luong, R. Winter,
additives). The amino acid W215 was confirmed to be a key
residue for interaction with superactivating additives. Finally,
kinetic parameters were interpreted for the first time also taking
into account the potential binding mode of the substrate.
D. H. Zaitsau, S. P. Verevkin and G. Sadowski, Co-solvent
effects on reaction rate and reaction equilibrium of an
enzymatic peptide hydrolysis, Phys. Chem. Chem. Phys.,
2018, 20(16), 11317–11326.
1
1
1
1
4 C. Held, T. Stolzke, M. Knierbein, M. W. Jaworek, T. Q.
Luong, R. Winter and G. Sadowski, Cosolvent and pressure
effects on enzyme-catalysed hydrolysis reactions, Biophys.
Chem., 2019, 252, 106209.
5 M. Knierbein, C. Held and G. Sadowski, The role of mole-
cular interactions on Michaelis constants of a-chymotrypsin
catalyzed peptide hydrolyses, J. Chem. Thermodyn., 2020,
Conflicts of interest
There are no conflicts of interest to declare.
Acknowledgements
148, 106142.
The authors acknowledge the financial support by the University
of Perugia (Fondo Ricerca di Base 2019).
6 N. Spreti, F. Alfani, M. Cantarella, F. D’Amico, R. Germani
and G. Savelli, a-Chymotrypsin superactivity in aqueous
solutions of cationic surfactants, J. Mol. Catal. B: Enzym.,
1999, 6(1-2), 99–110.
References
7 F. Alfani, M. Cantarella, N. Spreti, R. Germani and G. Savelli,
a-Chymotrypsin Superactivity in Cetyltrialkylammonium
Bromide-Rich Media, Appl. Biochem. Biotechnol., 2000,
88(1-3), 001–016.
1
D. G. Herries, Enzyme Structure and Mechanism (Second
Edition), Biochem. Educ., 1985, 13(3), 146.
2
3
D. M. Blow, The Structure of Chymotrypsin, 1971, 3, 185–2126.
G. P. Hess, in The Enzymes, ed. B. P. D., Academic Press, 18 N. Spreti, P. Di Profio, L. Marte, S. Bufali, L. Brinchi and
New York, 3rd edn, 1971, vol. 3, pp. 213–248.
S. K. Verma and K. K. Ghosh, Effect of cationic surfactants
G. Savelli, Activation and stabilization of a-chymotrypsin by
cationic additives, Eur. J. Biochem., 2001, 268(24), 6491–6497.
on the enzymatic activity of a-chymotrypsin, Kinet. Catal., 19 F. De Angelis, A. Di Tullio, P. Del Boccio, S. Reale, G. Savelli
4
5
2011, 52(1), 6–10.
and N. Spreti, ESI-MS in the study of the activity of
a-chymotrypsin in aqueous surfactant media, Org. Biomol.
Chem., 2003, 1(17), 3125–3130.
M. S. Celej, M. G. D’Andrea, P. T. Campana, G. D. Fidelio
and M. L. Bianconi, Superactivity and conformational
changes on alpha-chymotrypsin upon interfacial binding 20 N. Spreti, M. V. Mancini, R. Germani, P. Di Profio and
to cationic micelles, Biochem. J., 2004, 378(3), 1059–1066.
E. Abuin, E. Lissi and C. Calder o´ n, Kinetics of N-glutaryl-l-
phenylalanine p-nitroanilide hydrolysis catalyzed by a-chymo-
G. Savelli, Substrate effect on a-chymotrypsin activity in
aqueous solutions of ‘‘big-head’’ ammonium salts, J. Mol.
Catal. B: Enzym., 2008, 50(1), 1–6.
6
7
trypsin in aqueous solutions of alkyltrimethylammonium 21 L. De Matteis, F. Di Renzo, R. Germani, L. Goracci, N. Spreti
bromides, J. Colloid Interface Sci., 2007, 308(2), 573–576.
S. K. Verma, B. K. Ghritlahre, K. K. Ghosh, R. Verma,
S. Verma and X. Zhao, Influence of Amine-Based Cationic
and M. Tiecco, a-Chymotrypsin superactivity in quaternary
ammonium salt solution: kinetic and computational studies,
RSC Adv., 2016, 6(52), 46202–46211.
Gemini Surfactants on Catalytic Activity of a-Chymotrypsin, 22 J. Weinstock, Anomalous Hofmann Elimination Reactions,
Int. J. Chem. Kinet., 2016, 48(12), 779–784.
J. Org. Chem., 1956, 21, 540–542.
20832 | New J. Chem., 2020, 44, 20823--20833 This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020

View Article Online
NJC
Paper
23 H. J. Barber and K. Gaimster, Hexamethonium Bitartrate 29 E. Carosati, S. Sciabola and G. Cruciani, Hydrogen Bonding
and Other Hexa-Alkonium Salts, J. Pharm. Pharmacol., 1951,
(1), 663–669.
4 T. Iwaguchi, Supporting Electrolyte of Polarography, Yakugaku
Zasshi, 1961, 81(6), 925–927.
5 L. Brinchi, R. Germani, L. Goracci, G. Savelli and
C. A. Bunton, Decarboxylation and Dephosphorylation in
New Gemini Surfactants. Changes in Aggregate Structures,
Langmuir, 2002, 18(21), 7821–7825.
Interactions of Covalently Bonded Fluorine Atoms: From
Crystallographic Data to a New Angular Function in the
GRID Force Field, J. Med. Chem., 2004, 47(21), 5114–5125.
30 G. Muratore, L. Goracci, B. Mercorelli, A. Foeglein,
P. Digard, G. Cruciani, G. Palu and A. Loregian, Small
molecule inhibitors of influenza A and B viruses that act
by disrupting subunit interactions of the viral polymerase,
Proc. Natl. Acad. Sci. U. S. A., 2012, 109(16), 6247–6252.
3
2
2
2
2
6 N. N. P. Mel’nikov and M. V. Prilutskaya, Chloromethylation 31 S. Massari, G. Nannetti, L. Goracci, L. Sancineto, G. Muratore,
of ethers of bivalent phenols and some of their transforma-
tions, Zh. Obshch. Khim., 1959, 29, 3746–3752.
S. Sabatini, G. Manfroni, B. Mercorelli, V. Cecchetti, M. Facchini,
G. Pal u` , G. Cruciani, A. Loregian and O. Tabarrini, Structural
Investigation of Cycloheptathiophene-3-carboxamide Derivatives
Targeting Influenza Virus Polymerase Assembly, J. Med. Chem.,
2013, 56(24), 10118–10131.
7 M. Baroni, G. Cruciani, S. Sciabola, F. Perruccio and
J. S. Mason, A Common Reference Framework for Analyz-
ing/Comparing Proteins and Ligands. Fingerprints for
Ligands And Proteins (FLAP): Theory and Application, 32 C. G. Fortuna, C. Bonaccorso, A. Bulbarelli, G. Caltabiano, L. Rizzi,
J. Chem. Inf. Model., 2007, 47(2), 279–294.
L. Goracci, G. Musumarra, A. Pace, A. Palumbo Piccionello,
A. Guarcello, P. Pierro, C. E. A. Cocuzza and R. Musumeci, New
linezolid-like 1,2,4-oxadiazoles active against Gram-positive mul-
tiresistant pathogens, Eur. J, Med. Chem., 2013, 65, 533–545.
2
8 P. J. Goodford, A computational procedure for determining
energetically favorable binding sites on biologically impor-
tant macromolecules, J. Med. Chem., 1985, 28(7), 849–857.
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020 New J. Chem., 2020, 44, 20823--20833 | 20833