Evidence that a ' dynamic knockoutg' in Escherichia coli dihydrofolate reductase does not affect the chemical step of catalysis

Article abstract of DOI:10.1038/nchem.1296

The question of whether protein motions play a role in the chemical step of enzymatic catalysis has generated much controversy in recent years. Debate has recently reignited over possible dynamic contributions to catalysis in dihydrofolate reductase, foll

Full text of DOI:10.1038/nchem.1296

ARTICLES
PUBLISHED ONLINE: 11 MARCH 2012 | DOI: 10.1038/NCHEM.1296
Evidence that a ‘dynamic knockout’ in Escherichia
coli dihydrofolate reductase does not affect the
chemical step of catalysis
E. Joel Loveridge, Enas M. Behiry, Jiannan Guo and Rudolf K. Allemann
*
The question of whether protein motions play a role in the chemical step of enzymatic catalysis has generated much
controversy in recent years. Debate has recently reignited over possible dynamic contributions to catalysis in dihydrofolate
reductase, following conflicting conclusions from studies of the N23PP/S148A variant of the Escherichia coli enzyme. By
investigating the temperature dependence of kinetic isotope effects, we present evidence that the reduction in the hydride
transfer rate constants in this variant is not a direct result of impairment of conformational fluctuations. Instead, the
conformational state of the enzyme immediately before hydride transfer, which determines the electrostatic environment
of the active site, affects the rate constant for the reaction. Although protein motions are clearly important for binding and
release of substrates and products, there appears to be no detectable dynamic coupling of protein motions to the hydride
transfer step itself.
ne of the most contentious topics in modern enzymology is donor–acceptor distance and the angle of approach, and similar
the potential role of protein motions in the chemical step of to the more classical concept of orbital overlap)1,2,8,19,20. Hence, if
O
enzymatic catalysis. Although ligand association and dis- conformational fluctuations are directly linked to the chemical
sociation steps are known to be accompanied by sometimes very step of the reaction cycle and therefore affect either of these par-
large conformational changes, the coupling of protein dynamics ameters, then they will play a role in determining the KIE on
to the chemical coordinate of the reaction is less well defined and hydride transfer. Perturbation of these conformational fluctuations
is often inferred from indirect evidence such as the temperature will therefore have an effect on the KIEs. However, it is not sufficient
dependence of the primary hydrogen kinetic isotope effects to measure the KIE at a single temperature, as equilibrium changes
(KIEs) on the reaction1–3, or by computational methods4–7
.
to the enzyme structure may counteract any potential changes
Hydrogen (as hydride, hydrogen radical or proton) transfer brought about by alteration of the dynamics. The temperature
reactions are often used to investigate the potential involvement of dependence of the KIE is therefore a highly sensitive and much
protein motions in driving the chemistry of enzyme-catalysed used1–3,21–28 probe of the changes to the reaction, as it is unlikely
reactions. In these reactions, quantum-mechanical tunnelling that conflicting factors will exactly counterbalance over a wider
plays a role, and the width of the reaction barrier is as important temperature range.
in determining the reaction rate as its height. That hydrogen
The potential role of motions in promoting the chemical step of
tunnels in enzymatic reactions may be unsurprising given its the reaction has been extensively studied in dihydrofolate reductase
small mass, but recognition of hydrogen tunnelling has led to (DHFR). EcDHFR, the enzyme from Escherichia coli, adopts two
proposals that enzymes may have evolved to enhance the degree major conformations—closed and occluded (Fig. 1)—as it moves
of tunnelling so as to accelerate the reaction, and that enzyme through its catalytic cycle29. These conformations take their
dynamics may couple to the reaction coordinate to promote tunnel- names from the position of the M20 loop (residues 9–24), which
ling both by optimising the wavefunction overlap between the forms part of the active site. Before the chemical step of the cycle,
reactant and product states and by decreasing the donor–acceptor this loop closes over the active site, shielding the reactants from
distance1,2,8,9. These protein motions may range from microsecond solvent and providing an optimal environment for protonation of
to millisecond conformational fluctuations to localized subnanose- N5 of the dihydrofolate and hydride transfer to occur. This closed
cond ‘promoting vibrations’1,10–13. Although both experimental14,15 conformation is stabilized by hydrogen bonding to the neighbour-
and computational studies16 suggest that the degree of tunnelling ing FG loop (residues 116–132). In the product complexes,
is not increased in enzymatic reactions relative to the equivalent however, the FG loop occludes the active site, preventing
reactions in solution, and therefore that the contribution from the nicotinamide ring of the cofactor NADP(H) from entering.
tunnelling to catalysis is negligible, whether or not protein This occluded conformation is stabilized by a pair of hydrogen
motions promote hydrogen transfer is less clear and remains an bonds between residue N23 on the M20 loop and S148 on
issue of debate1,2,10,17
.
the GH loop (residues 142–149; Fig. 1). The chemical step of the
Although current models that relate a potential role of protein reaction, hydride transfer from NADPH to dihydrofolate with
motions in promoting hydrogen transfer to the temperature concomitant protonation of dihydrofolate, occurs within the
dependence of the KIE are not consistent with all available closed conformation29.
data1,2,16–18, it is clear that the KIE is highly sensitive to changes in
A ‘dynamic knockout’ of EcDHFR has recently been reported, in
the donor–acceptor distance and the degree of wavefunction which mutations were made that both prevent formation of the
overlap between the reactant and product states (affected by both occluded conformation through loss of hydrogen bonding
School of Chemistry, Cardiff University, Park Place, Cardiff, CF10 3AT, UK. e-mail: allemannrk@cf.ac.uk
*
292
NATURE CHEMISTRY | VOL 4 | APRIL 2012 | www.nature.com/naturechemistry
© 2012 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1296
ARTICLES
constant for hydride transfer and it was suggested that this was a
consequence of the loss of conformational flexibility30. However, a
subsequent computational study suggested instead that the
reduced rate constant is a consequence of changes to the reorganiz-
ation free energy of the reaction brought about by changes to the
electrostatic preorganization within the active site31. Here, we
investigate the effect of the N23PP/S148A mutations on EcDHFR
catalysis, and provide experimental evidence that the nature of the
chemistry in this ‘dynamic knockout’ of EcDHFR is essentially
the same as in the wild-type enzyme.
a
NADP+
Results and discussion
Creation of EcDHFR-N23PP/S148A and structural studies.
EcDHFR-N23PP/S148A was created using standard site-directed
mutagenesis techniques. The purified protein was first characterized
by circular dichroism spectroscopy, which showed only minor
differences to the wild-type enzyme (Supplementary Fig. S1). It
has previously been shown that the NADPþ:folate complex of
EcDHFR-N23PP/S148A has a similar structure to that of the wild-
type enzyme30. The melting temperature of EcDHFR-N23PP/S148A
was found to be 56.8+0.2 8C, which is slightly higher than the
value of 51.6+0.7 8C obtained for the wild-type enzyme32,
suggesting that the N23PP and S148A mutations have a stabilizing
effect on the enzyme.
Folate
FG
loop
N23PP
Steady-state kinetics. Michaelis–Menten curves for both substrate
and cofactor were measured at 20 8C and pH 7. The Michaelis
constants were found to be 2.4+0.3 mM for NADPH and 0.7+
0.2 mM for dihydrofolate. These are in good agreement with
the values of 4.8+1.0 mM and 0.7+0.2 mM obtained at 25 8C
for wild-type EcDHFR33, demonstrating that the affinity of the
enzyme for its reactants is not significantly affected by the
N23PP/S148A mutations. The value of kcat was found to be
1.94+0.09 s21 at 20 8C, in good agreement with the reported
value30. The temperature dependence of kcat was determined at
pH 7 (Supplementary Table S1). The activation energy was found
to be 52.6+1.2 kJ mol21, which is only slightly higher than that
of the wild-type enzyme (47.7+0.8 kJ mol21)34, as expected from
the relatively small (ꢀsixfold) reduction in kcat. It should be
noted, however, that direct comparison of the activation energies
is mechanistically uninformative as a different step within the
catalytic cycle is rate limiting in EcDHFR-N23PP/S148A than in
the wild-type enzyme30. The primary hydrogen KIE on kcat was
only slightly higher than unity at pH 7 (1.49+0.07), but was
2.79+0.02 at pH 9. This is in agreement with the steady-state rate
constant being predominantly limited by a physical step of the
catalytic cycle at pH 7 (with some contribution from the chemical
step, as expected from the relatively low hydride transfer rates30),
but almost exclusively limited by the chemical step at elevated pH,
as had been seen for wild-type EcDHFR33.
M20 loop
S148A
GH loop
b
N23
GH
loop
M20 loop
FG
loop
S148
Figure 1 | Crystal structures of wild-type EcDHFR and EcDHFR-N23PP/
S148A. a, Cartoon representations of wild-type EcDHFR (green,
PDB 1RX229) and EcDHFR-N23PP/S148A (orange, PDB 3QL030) as
DHFR:NADPþ:folate complexes, which serve as a model for the
Single turnover kinetics. The kcat at pH 7 is dominated by physical
steps within the reaction cycle (vide supra), so its KIE cannot be
interpreted mechanistically. Therefore, to investigate the effect of
the ‘dynamic knockout’30 on the chemical step of EcDHFR
catalysis, the temperature dependence of the KIE on the hydride
transfer step of catalysis, isolated under single turnover stopped-
flow conditions, was determined at pH 7 and 9 (Fig. 2, Table 1
and Supplementary Tables S2–S5). It has been shown recently
that the KIE on hydride transfer in EcDHFR is temperature-
dependent below pH 8 and temperature-independent at or above
this pH, demonstrating that the temperature-dependent KIEs
observed at pH 7 are indeed physiologically relevant, and the
temperature-independent KIEs observed at elevated pH, where
DHFR:NADPH:dihydrofolate Michaelis complexes29. Note the high similarity
in the overall fold of the two enzymes. b, Detailed view of the catalytically
important M20 loop in the closed (green, PDB 1RX229) and occluded
(yellow, PDB 1RX429) conformations in EcDHFR. In the closed conformation,
the substrate dihydrofolate and the cofactor NADPH are positioned ready for
hydride transfer to occur and the M20 loop shields the active site from
solvent. In the occluded conformation, the M20 loop prevents access of the
nicotinamide ring of NADP(H) to the active site. Ligands and amino acids
replaced in the EcDHFR variant are shown as sticks. Structural elements
discussed in the text and hydrogen bonding between the M20 and GH loops
in the occluded conformation are indicated.
between the M20 and GH loops (Fig. 1) and eliminate millisecond hydride transfer is rate-limiting33, are not21.
to timescale motions of the M20 loop in the Michaelis complex30.
The rate constants for hydride (kH) and deuteride (kD) transfer
This EcDHFR-N23PP/S148A variant displayed a reduced rate were considerably lower than those of the wild-type enzyme
NATURE CHEMISTRY | VOL 4 | APRIL 2012 | www.nature.com/naturechemistry
© 2012 Macmillan Publishers Limited. All rights reserved.
293

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1296
ARTICLES
a
6
5
4
3
2
1
0
2
1
b
0
–1
–2
–3
–4
3.2
3.3
3.4
3.5
3.6
3.2
3.3
3.4
3.5
3.6
c
1.4
1.2
1
1.4
1.2
1
d
0.8
0.6
0.4
0.2
0
0.8
0.6
0.4
0.2
0
3.2
3.3
3.4
3.5
3.6
3.2
3.3
3.4
3.5
3.6
103/T (K–1
)
103/T (K–1
)
Figure 2 | Pre-steady-state kinetics of EcDHFR and EcDHFR-N23PP/S148A. a, Arrhenius plots for hydride (circles) and deuteride (triangles) transfer for
EcDHFR (blue) and EcDHFR-N23PP/S148A (red) at pH 7. b, Arrhenius plots for hydride (circles) and deuteride (triangles) transfer for wild-type EcDHFR
(blue) and EcDHFR-N23PP/S148A (red) at pH 9. c, Plots of KIE on a logarithmic abscissa against the inverse temperature for wild-type EcDHFR (blue) and
EcDHFR-N23PP/S148A (red) at pH 7. d, Plots of KIE on a logarithmic abscissa against inverse temperature for wild-type EcDHFR (blue) and EcDHFR-
N23PP/S148A (red) at pH 9. Error bars (standard errors of the mean at 1s) are small and hence obscured by the symbols in most cases. The data show
that, although there are large differences between the rate constants for hydride transfer in EcDHFR and EcDHFR-N23PP/S148A, the magnitude and
temperature dependence of the KIE are virtually identical.
(Table 1 and Fig. 2)30. Both the magnitude and temperature conformational motions other than the closed-to-occluded
dependence of the KIE on the reaction catalysed by EcDHFR- conformational change ‘probably play a role in hydride transfer’30.
N23PP/S148A, on the other hand, were almost indistinguishable The occluded conformation itself cannot be relevant to the chemical
from those of wild-type EcDHFR. The values of kH obtained here step of the catalytic cycle because the nicotinamide ring of
(25.9+0.3 s21 at pH 7 and 0.24+0.01 s21 at pH 9 at 25 8C) are NADP(H) is not within the active site and so cannot react29. In
slightly different to those published previously (14 and 3.3 s21
,
addition, for the M20 loop to move from the closed to the occluded
respectively30). Activation enthalpies were slightly reduced in conformation, the loop must first open to allow the nicotinamide
EcDHFR-N23PP/S148A, with the activation entropy accounting ring to leave the active site29,36. We have already provided evidence
for the higher activation free energy and corresponding lower that fluctuations between the excited-state closed conformation
rate constants. The activation entropy for hydride transfer by and the ground-state occluded conformation seen by NMR in
EcDHFR-N23PP/S148A is ꢀ30 J K21 mol21 more negative than the EcDHFR:NADPþ:tetrahydrofolate product ternary complex
that of the wild-type enzyme at pH 7 (Table 1). The potential are not relevant to the hydride transfer step itself21. Similarly,
energy surface is therefore probably flatter in the conformational fluctuations between the excited-state occluded conformation and
direction in the wild-type enzyme than in EcDHFR- the ground-state closed conformation seen by NMR in the
N23PP/S148A, as has been found recently in a computational EcDHFR:NADPþ:folate complex (a model for the Michaelis
study31. The values of DG‡ obtained here (Table 1) are in excellent complex29) are not relevant to hydride transfer but are thought to
agreement with recently calculated values31. We also note that the precede it37.
similarity of the magnitude of the KIE for EcDHFR-N23PP/S148A
Unlike the closed-to-occluded conformational change, other
and EcDHFR suggests that the two enzymes have indistinguishable millisecond motions within the active site are not necessarily
donor–acceptor distances in the Michaelis complexes, in contrast orthogonal to the reaction coordinate and may therefore affect the
to what was observed in the X-ray crystal structures of the reaction chemistry30. However, the correlation between the loss of
E:NADPþ:folate complexes, where this distance was reported to conformational fluctuations and the reduction in the rate constant
differ by 0.3 Å (ref. 30). We have previously shown that changes to for hydride transfer seen in EcDHFR-N23PP/S148A does not
the donor–acceptor distance as small as 0.05 Å are detectable using necessarily demonstrate causation. On the contrary, the similarity
our methodology35.
between the temperature dependence of the KIEs of the wild-type
It has been suggested that the lower hydride transfer rate and variant enzymes strongly suggests that the nature of the
constants in EcDHFR-N23PP/S148A relative to the wild-type chemistry is unaffected by the ‘dynamic knockout’30 induced by
enzyme are the result of the loss of conformational flexibility in the N23PP/S148A mutations and therefore that the loss of
the variant enzyme30. Although Bhabha et al. state that ‘the conformational fluctuations is not the cause of the lower hydride
decreased hydride transfer rate does not simply arise from impair- transfer rate constants. The KIE is highly sensitive to changes in
ment of the closed-to-occluded transition’30, they argue that as the donor–acceptor distance. Indeed, changes to the donor–accep-
EcDHFR-S148A, unlike EcDHFR-N23PP/S148A, retains milli- tor distance at a fixed temperature cause larger changes to the KIE
second conformational fluctuations in regions of the active site than temperature does at a fixed donor–acceptor distance19.
other than the M20 loop, and has only a slightly lower rate constant Increasing the donor–acceptor distance increases the temperature
for hydride transfer than wild-type EcDHFR, millisecond dependence of the KIE as well as altering its magnitude19, and the
294
NATURE CHEMISTRY | VOL 4 | APRIL 2012 | www.nature.com/naturechemistry
© 2012 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1296
ARTICLES
Table 1 | Kinetic parameters for EcDHFR and EcDHFR-N23PP/S148A at 25 8C.
Parameter
Wild-type EcDHFR21
EcDHFR-N23PP/S148A
pH 7
pH 9
pH 7
pH 9
kH (s21
)
203.7 + 7.4
2.71 + 0.08
29.9 + 0.6
2.52+ 0.02
3.23 + 0.05
32.0 + 0.6
25.9 + 0.3
2.58 + 0.03
25.7 + 0.9
0.24 + 0.01
2.82 + 0.06
31.1 + 1.2
KIE
EHa (kJ mol21
)
EDa (kJ mol21
)
37.8 + 0.6
32.0 + 0.5
33.4 + 1.2
7.7 + 1.5
31.4 + 1.2
0.4 + 1.7
DEa (kJ mol21
)
7.9 + 0.9
20.1 + 0.8
AH (s21
AD (s21
AH/AD
)
)
(3.4+ 0.5) × 107
(3.2 + 0.1) × 108
0.11+ 0.02
(1.0 + 0.3) × 106
(3.2 + 0.7) × 105
3.08 + 1.02
(8.2 + 0.2) × 105
(7.4 + 0.2) × 106
0.11+ 0.04
23.2 + 0.9
(6.8 + 0.3) × 104
(2.8 + 0.1) × 104
2.42 + 0.07
28.6 + 1.2
DH‡H (kJ mol21
DH‡D (kJ mol21
)
)
27.4+ 0.6
29.6 + 0.6
35.3 + 0.6
2109.0 + 2.5
290.3 + 1.8
59.9 + 0.6
29.5 + 0.5
2138.3 + 5.1
2147.7 + 4.8
70.8 + 0.6
30.9 + 1.2
2139.9 + 7.5
2121.6 + 6.4
64.9 + 0.9
29.0 + 1.2
2160.6 + 18.4
2167.9 + 22.8
76.5 + 1.2
DS‡H (J K21 mol21
DS‡D (J K21 mol21
DG‡H (kJ mol21
DG‡D (kJ mol21
)
)
)
)
62.2 + 0.6
73.6 + 0.5
67.2 + 1.2
79.1 + 1.2
effect on the KIE of a change as small as 0.05 Å would be detectable millisecond timescale appear to be unimportant, kinetic studies
using our methodology35. Therefore, if millisecond conformational cannot rule out a role for faster motions. New models and exper-
fluctuations are coupled directly to the chemical step, then it is iments are therefore required before the large amount of apparently
highly unlikely that changes to these fluctuations would not mani- contradictory evidence relating to the mechanistic underpinning of
fest themselves in either the magnitude or the temperature DHFR catalysis can be reconciled and fully explained.
dependence of the KIE. The only logical conclusion is therefore
Conclusions
that no such coupling exists.
The nature of the chemistry, as reported by the magnitude and
temperature dependence of the KIEs on hydride transfer, appears
to be unaffected in EcDHFR-N23PP/S148A. This demonstrates
that, contrary to a recent report30, conformational fluctuations do
not seem to be directly involved in the chemical step of the reaction.
The electrostatic preorganization within the active site during the
reaction is determined by the conformational state of the enzyme
at the point of reaction; hence, subtle changes to the (equilibrium)
conformational state of the enzyme as a consequence of structural
changes brought about by amino acid replacements or changes
to the conditions of the reaction will nevertheless affect the rate
constants for the reaction. As can be seen from the change in
rate-limiting step in the steady state30, the dynamic properties of
EcDHFR and its mutants clearly affect the conformational
motions that are important for the turnover of substrate to
product, but no correlation between these conformational
motions and the chemical motions that direct hydride transfer
have yet been identified for DHFR.
We have previously argued against a direct role for confor-
mational and dynamic effects in the chemical step of DHFR cataly-
sis21,35,38,39. The data reported here provide further evidence in
support of this view. However, the rate constant for the reaction is
most probably determined by the conformational state of the
enzyme at the point of reaction, because the specific conformation
will determine the electrostatics in and around the active site and
therefore control the reaction free energy. This conclusion is
supported by our observation that the rate constant for hydride
transfer (but not the KIE) for the DHFR-catalysed reaction is
proportional to the dielectric constant of the medium but not to
its viscosity35,39,40. Hence, the rate constants for the chemical step
appear to be dependent on electrostatic effects but not directly
influenced by large-scale dynamical contributions. It is therefore
more likely that the dampening of the millisecond motions
around the active site brings about more subtle conformational
changes. This is not
a semantic argument; although our
results show that the difference in catalytic properties of EcDHFR
and EcDHFR-N23PP/S148A are indeed a simple consequence of
electrostatic effects as has been argued previously31, these electro-
static effects are affected by the conformational state of the
enzyme and its effect on the energy landscape of the reaction.
However, this is an equilibrium property of the enzyme and no
dynamic contribution from protein motions to the reaction
is required.
As well as showing that millisecond dynamics do not directly
influence the chemical step of the EcDHFR-catalysed reaction, our
work demonstrates a clear need both for better theoretical frame-
works that can explain non-classical temperature dependences
of KIEs on enzymatic hydride transfer (particularly temperature-
Methods
Chemicals. Dihydrofolate was prepared by dithionite reduction of folate (Sigma)43
NADPH and 4-(R)-NADPD were prepared from NADPþ (Melford) as described
previously21. EcDHFR and EcDHFR-N23PP/S148A were produced in E. coli
BL21(DE3) cells containing a pET11c-based plasmid encoding the enzymes and
.
purified by anion exchange chromatography on Q-sepharose resin followed by size
exclusion chromatography on a prep-grade Superdex 75 column (GE Healthcare).
Oligonucleotide primers were purchased from Eurofins. The concentrations of
NADPH/NADPD and H2F were determined spectrophotometrically using
extinction coefficients of 6,200 cm21 M21 at 339 nm and 28,000 cm21 M21 at
282 nm, respectively42.
Site-directed mutagenesis. A gene encoding EcDHFR-N23PP/S148A was created
using the Finnzymes Phusion site directed mutagenesis kit and primers
independent KIEs18) and for direct experimental evidence for (or 5′-GAAAACGCCATGCCGTGGCCGCCGCTGCCTGCCGATCTCGCC-3′
(N23PP) and 5′-GCGCAGAACGCTCACAGCTATTGC-3′ (S148A). The altered
against) a link between protein motions, ligand motions and the
chemistry proper. All current evidence supporting a role for
and inserted nucleotides are underlined.
DHFR dynamics in the actual chemical step is indirect and largely
based on current models of KIEs that may be inadequate given
their numerous inconsistencies with both experimental and compu-
tational data1,19,22,30,41,42. However, no direct experimental evidence
that refutes a role for dynamics in the chemical step has been
Circular dichroism spectroscopy. Circular dichroism spectra between 190 and
400 nm were measured in a nitrogen atmosphere by an Applied Photophysics
Chirascan spectrophotometer using 12.5 mM protein in 10 mM potassium
phosphate buffer (pH 7.0). Mean residue ellipticities [Q]MRE were calculated using
the equation [Q]MRE ¼ Q/(10ncl), where Q is the measured ellipticity (in mdeg), n
the number of backbone amide bonds, c the protein concentration (in mol l21) and l
provided either. Even though dynamic contributions on a the path length (in cm). Thermal denaturation experiments were performed between
NATURE CHEMISTRY | VOL 4 | APRIL 2012 | www.nature.com/naturechemistry
© 2012 Macmillan Publishers Limited. All rights reserved.
295

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1296
ARTICLES
10 and 80 8C using temperature steps of 1 8C with 1 min equilibration at the desired
temperature before measurement. Melting temperatures were determined from the
change in signal at 222 nm against temperature.
16. Kamerlin, S. C. L. & Warshel, A. An analysis of all the relevant facts and
arguments indicates that enzyme catalysis does not involve large contributions
from nuclear tunneling. J. Phys. Org. Chem. 23, 677–684 (2010).
17. Pisliakov, A. V., Cao, J., Kamerlin, S. C. L. & Warshel, A. Enzyme millisecond
conformational dynamics do not catalyze the chemical step. Proc. Natl Acad. Sci.
USA 106, 17359–17364 (2009).
18. Romesberg, F. E. & Schowen, R. L. Isotope effects and quantum tunneling in
enzyme-catalyzed hydrogen transfer. Part I. The experimental basis. Adv. Phys.
Org. Chem. 39, 27–77 (2004).
19. Liu, H. & Warshel, A. Origin of the temperature dependence of isotope effects in
enzymatic reactions: the case of dihydrofolate reductase. J. Phys. Chem. B 111,
7852–7861 (2007).
20. Wu, Y. D. & Houk, K. Theoretical transition structures for hydride
transfer to methyleneiminium ion from methylamine and dihydropyridine.
On the nonlinearity of hydride transfers. J. Am. Chem. Soc. 109,
2226–2227 (1987).
Steady-state kinetic measurements. Steady-state turnover rates were determined
spectrophotometrically using a JASCO V-660 spectrophotometer by following the
decrease in absorbance at 340 nm (e340 (NADPH þ H2F) ¼ 11,800 M21 cm21)44.
Potassium phosphate (100 mM) containing 100 mM NaCl and 10 mM
b-mercaptoethanol was used for measurements at pH 7, and at pH 9, MTEN buffer
(50 mM morpholinoethanesulfonic acid, 25 mM Tris, 25 mM ethanolamine,
100 mM NaCl and 10 mM b-mercaptoethanol) was used. Phosphate buffer was
used at pH 7 as its pKa is relatively insensitive to changes in temperature45. The
enzyme (50 nM) was pre-incubated at the desired temperature with NADPH
(0.5–100 mM) for 1 min to avoid hysteresis11 before the addition of H2F
(0.5–100 mM). For the temperature dependence of kcat or for KIE measurements,
100 mM each of both substrate and cofactor were used. The change in initial
rate with concentration was fit to the Michaelis–Menten equation using SigmaPlot
10. Each data point is the result of three independent measurements.
21. Loveridge, E. J. & Allemann, R. K. Effect of pH on hydride transfer
by Escherichia coli dihydrofolate reductase. ChemBioChem 12,
1258–1262 (2011).
Pre-steady-state kinetic measurements. The chemical step was observed under
single turnover conditions from the loss of fluorescence resonance energy transfer
from the enzyme to NADPH during the reaction. The enzyme (20 mM final
concentration) was pre-incubated with NADPH (8 mM final concentration) for at
least 5 min in 100 mM potassium phosphate buffer (pH 7.0) containing 100 mM
NaCl and 10 mM b-mercaptoethanol or MTEN (vide supra) buffer (pH 9.0). The
reaction was then initiated by rapidly mixing in H2F (200 mM final concentration) in
the same buffer. Varying the concentrations of the reagents showed that the
measured rate constants were limiting values. Measurements were made on either an
Applied Photophysics stopped-flow spectrophotometer, exciting the sample at
292 nm and measuring the emission using a 400 nm cutoff output filter, or a
Hi-Tech Scientific stopped-flow spectrophotometer, exciting the sample at 297 nm
and measuring the emission above 440 nm. Identical results were obtained
with both set-ups. All experiments were repeated at least three times on
each spectrometer.
22. Sikorski, R. S. et al. Tunneling and coupled motion in the Escherichia coli
dihydrofolate reductase catalysis. J. Am. Chem. Soc. 126, 4778–4779 (2004).
23. Kohen, A., Cannio, R., Bartolucci, S. & Klinman, J. P. Enzyme dynamics and
hydrogen tunnelling in a thermophilic alcohol dehydrogenase. Nature 399,
496–499 (1999).
24. Chowdhury, S. & Banerjee, R. Evidence for quantum mechanical tunneling in
the coupled cobalt–carbon bond homolysis-substrate radical generation reaction
catalyzed by methylmalonyl-CoA mutase. J. Am. Chem. Soc. 122,
5417–5418 (2000).
25. Fan, F. & Gadda, G. Oxygen-and temperature-dependent kinetic isotope effects
in choline oxidase: correlating reversible hydride transfer with environmentally
enhanced tunneling. J. Am. Chem. Soc. 127, 17954–17961 (2005).
26. Anandarajah, K. & Schowen, K. Hydrogen tunneling in glucose oxidation
by the archaeon Thermoplasma acidophilum. Z. Phys. Chem. 222,
1333–1347 (2008).
27. Hay, S., Pudney, C. R. & Scrutton, N. S. Structural and mechanistic
aspects of flavoproteins: probes of hydrogen tunnelling. FEBS J. 276,
3930–3941 (2009).
Received 18 November 2011; accepted 6 February 2012;
published online 11 March 2012
28. Heyes, D. J., Sakuma, M., de Visser, S. P. & Scrutton, N. S. Nuclear quantum
tunneling in the light-activated enzyme protochlorophyllide oxidoreductase.
J. Biol. Chem. 284, 3762–3767 (2009).
29. Sawaya, M. R. & Kraut, J. Loop and subdomain movements in the mechanism of
Escherichia coli dihydrofolate reductase: crystallographic evidence. Biochemistry
36, 586–603 (1997).
30. Bhabha, G. et al. A dynamic knockout reveals that conformational
fluctuations influence the chemical step of enzyme catalysis. Science 332,
234–238 (2011).
31. Adamczyk, A. J., Cao, J., Kamerlin, S. C. L. & Warshel, A. Catalysis by
dihydrofolate reductase and other enzymes arises from electrostatic
preorganization, not conformational motions. Proc. Natl Acad. Sci. USA 108,
14115–14120 (2011).
32. Swanwick, R. S., Shrimpton, P. J. & Allemann, R. K. Pivotal role of Gly 121 in
dihydrofolate reductase from Escherichia coli: the altered structure of a mutant
enzyme may form the basis of its diminished catalytic performance.
Biochemistry 43, 4119–4127 (2004).
33. Fierke, C. A., Johnson, K. A. & Benkovic, S. J. Construction and evaluation of the
kinetic scheme associated with dihydrofolate reductase from Escherichia coli.
Biochemistry 26, 4085–4092 (1987).
34. Maglia, G., Javed, M. H. & Allemann, R. K. Hydride transfer during catalysis
by dihydrofolate reductase from Thermotoga maritima. Biochem. J. 374,
529–535 (2003).
35. Loveridge, E. J. et al. The role of large-scale motions in catalysis by dihydrofolate
reductase. J. Am. Chem. Soc. 133, 20561–20570 (2011).
References
1. Nagel, Z. D. & Klinman, J. P. A 21st century revisionist’s view at a turning point
in enzymology. Nature Chem. Biol. 5, 543–550 (2009).
2. Limbach, H-H., Schowen, K. B. & Schowen, R. L. Heavy atom motions and
tunneling in hydrogen transfer reactions: the importance of the pre-tunneling
state. J. Phys. Org. Chem. 23, 586–605 (2010).
3. Allemann, R. K., Evans, R. M. & Loveridge, E. J. Probing coupled motions
in enzymatic hydrogen tunnelling reactions. Biochem. Soc. Trans. 37,
349–353 (2009).
4. Masgrau, L. et al. Atomic description of an enzyme reaction dominated by
proton tunneling. Science 312, 237–241 (2006).
´
5. Antoniou, D., Basner, J., Nun˜ez, S. & Schwartz, S. D. Computational and
theoretical methods to explore the relation between enzyme dynamics and
catalysis. Chem. Rev. 106, 3170–3187 (2006).
6. Olsson, M. H. M., Parson, W. W. & Warshel, A. Dynamical contributions to
enzyme catalysis: critical tests of a popular hypothesis. Chem. Rev. 106,
1737–1756 (2006).
7. McGeagh, J. D., Ranaghan, K. E. & Mulholland, A. J. Protein dynamics and
enzyme catalysis: insights from simulations. Biochim. Biophys. Acta 1814,
1077–1092 (2011).
8. Kuznetsov, A. & Ulstrup, J. Proton and hydrogen atom tunnelling in hydrolytic
and redox enzyme catalysis. Can. J. Chem. 77, 1085–1096 (1999).
9. Sutcliffe, M. J. & Scrutton, N. S. Enzymology takes a quantum leap forward. Phil.
Trans. R. Soc. Lond. A 358, 367–386 (2000).
10. Pudney, C. R. et al. Evidence to support the hypothesis that promoting vibrations
enhance the rate of an enzyme catalyzed H-tunneling reaction. J. Am. Chem. Soc.
131, 17072–17073 (2009).
11. Schwartz, S. D. & Schramm, V. L. Enzymatic transition states and dynamic
motion in barrier crossing. Nature Chem. Biol. 5, 551–558 (2009).
12. Pineda, J. R. E. T., Antoniou, D. & Schwartz, S. D. Slow conformational motions
that favor sub-picosecond motions important for catalysis. J. Phys. Chem. B 114,
15985–15990 (2010).
13. Henzler-Wildman, K. & Kern, D. Dynamic personalities of proteins. Nature 450,
964–972 (2007).
14. Doll, K. M. & Finke, R. G. A compelling experimental test of the hypothesis that
enzymes have evolved to enhance quantum mechanical tunneling in hydrogen
transfer reactions: the beta-neopentylcobalamin system combined with prior
adocobalamin data. Inorg. Chem. 42, 4849–4856 (2003).
36. Arora, K. & Brooks, C. L. III Functionally important conformations of the Met20
loop in dihydrofolate reductase are populated by rapid thermal fluctuations.
J. Am. Chem. Soc. 131, 5642–5647 (2009).
37. Osborne, M. J., Schnell, J., Benkovic, S. J., Dyson, H. J. & Wright, P. E. Backbone
dynamics in dihydrofolate reductase complexes: role of loop flexibility in the
catalytic mechanism. Biochemistry 40, 9846–9859 (2001).
38. Loveridge, E. J. & Allemann, R. K. The temperature dependence of the
kinetic isotope effects of dihydrofolate reductase from Thermotoga maritima
is influenced by intersubunit interactions. Biochemistry 49, 5390–5396 (2010).
39. Loveridge, E. J., Tey, L-H. & Allemann, R. K. Solvent effects on catalysis
by Escherichia coli dihydrofolate reductase. J. Am. Chem. Soc. 132,
1137–1143 (2010).
40. Loveridge, E. J., Evans, R. M. & Allemann, R. K. Solvent effects on
environmentally coupled hydrogen tunnelling during catalysis by
dihydrofolate reductase from Thermotoga maritima. Chem. Eur. J. 14,
10782–10788 (2008).
15. Doll, K. M., Bender, B. R. & Finke, R. G. The first experimental test of the
hypothesis that enzymes have evolved to enhance hydrogen tunneling. J. Am.
Chem. Soc. 125, 10877–10884 (2003).
296
NATURE CHEMISTRY | VOL 4 | APRIL 2012 | www.nature.com/naturechemistry
© 2012 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1296
ARTICLES
41. Liu, H. & Warshel, A. The catalytic effect of dihydrofolate reductase and
its mutants is determined by reorganization energies. Biochemistry 46,
6011–6025 (2007).
Acknowledgements
This work was supported by the UK Biotechnology and Biological Sciences Research
Council (BBSRC) (grant no. BB/E008380/1) and by Cardiff University.
42. Swanwick, R. S., Maglia, G., Tey, L-H. & Allemann, R. K. Coupling of
protein motions and hydrogen transfer during catalysis by
Escherichia coli dihydrofolate reductase. Biochem. J. 394,
259–265 (2006).
43. Blakley, R. Crystalline dihydropteroylglutamic acid. Nature 188, 231–232 (1960).
44. Stone, S. R. & Morrison, J. F. Kinetic mechanism of the reaction catalyzed
by dihydrofolate reductase from Escherichia coli. Biochemistry 21,
3757–3765 (1982).
Author contributions
E.J.L. performed the bulk of the experimental work. E.M.B. and J.G. performed additional
experiments. E.J.L. and R.K.A. designed the experiments, analysed the data and wrote the
manuscript. All authors commented on the manuscript.
Additional information
The authors declare no competing financial interests. Supplementary information
accompanies this paper at www.nature.com/naturechemistry. Reprints and permission
information is available online at http://www.nature.com/reprints. Correspondence and
requests for materials should be addressed to R.K.A.
45. Reijenga, J. C., Gagliardi, L. G. & Kenndler, E. Temperature dependence of
acidity constants, a tool to affect separation selectivity in capillary
electrophoresis. J. Chromatogr. A 1155, 142–145 (2007).
NATURE CHEMISTRY | VOL 4 | APRIL 2012 | www.nature.com/naturechemistry
297
© 2012 Macmillan Publishers Limited. All rights reserved.