Showcasing research from Dr Martin Marek's Structural
Biology group, Loschmidt Laboratories, Masaryk
University, Czech Republic.
D e c o d i n g t h e i n t r i c a t e n e t w o r k o f m o l e c u l a r i n t e r a c t i o n s
o f a h y p e r s t a b l e e n g i n e e r e d b i o c a t a l y s t
Computational design of protein catalysts with enhanced
stabilities is a challenging task. We report X-ray structures
of a hyperstable engineered haloalkane dehalogenase
(Tm = 73.5 °C), which highlight key thermostabilization
effects. Unexpectedly, mutations toward bulky aromatic
amino acids at the protein surface triggered long-distance
(-27 A) backbone changes due to cooperative effects.
These cooperative interactions produced an unprecedented
double-lock system that dramatically reduced the volumes
of enzyme access tunnels. Despite these spatial restrictions,
experimental tracing of the access tunnels using krypton
derivative crystals demonstrates that the transport of
ligands is still effective.
A s f e a t u r e d i n :
S e e Jiri D a m b o r s k y ,
M a r t i n M a r e k e r a / . ,
Chem. Sci., 2020,11,11162.
^ £ 2 ^ ™ S I I w rsc.li/chemical-scienceOF CHEMISTRY
Registered charity number: 20789C
Chemical
Science
ROYAL SOCIETY
OF CHEMISTRY
EDGE ARTICLE View Article Online
V i e w Journal I View Issue
11) Check for updates
Cite this: Chem. ScL, 2020,11,11162
All publication charges for this article
have been paid for by the Royal Society
of Chemistry
Received 17th June 2020
Accepted 10th September 2020
DOI: 10.1039/d0sc03367g
rsc.li/chemical-science
Decoding the intricate network of molecular
interactions of a hyperstable engineered
biocatalystf
Klara Markova4a b
Klaudia Chmelova4a D
Sergio M. Marques,a D
Philippe Carpentier,
.,.ab ab . cd
David Bednar,a b
Jiri Damborsky
* a b
and Martin Marek
* a b
C o m p u t a t i o n a l design of protein catalysts w i t h e n h a n c e d stabilities for use in research a n d e n z y m e
t e c h n o l o g i e s is a c h a l l e n g i n g task. Using f o r c e - f i e l d calculations a n d p h y l o g e n e t i c analysis, w e previously
d e s i g n e d t h e haloalkane d e h a l o g e n a s e DhaA115 w h i c h c o n t a i n s 11 m u t a t i o n s that c o n f e r u p o n it
o u t s t a n d i n g t h e r m o s t a b i l i t y ( T m = 73.5 °C; A T m > 23 °C). A n u n d e r s t a n d i n g o f t h e structural basis o f this
hyperstabilization is required in o r d e r t o d e v e l o p c o m p u t e r a l g o r i t h m s a n d predictive tools. Here, w e
r e p o r t X-ray structures of DhaA115 at 1.55 A a n d 1.6 A resolutions a n d their m o l e c u l a r d y n a m i c s
trajectories, w h i c h unravel t h e intricate n e t w o r k o f interactions that r e i n f o r c e t h e a f t a - s a n d w i c h
architecture. Unexpectedly, m u t a t i o n s t o w a r d bulky a r o m a t i c a m i n o acids at t h e p r o t e i n surface
triggered l o n g - d i s t a n c e (~27 A) b a c k b o n e c h a n g e s d u e t o c o o p e r a t i v e effects. T h e s e c o o p e r a t i v e
interactions p r o d u c e d an u n p r e c e d e n t e d d o u b l e - l o c k system that: (i) i n d u c e d b a c k b o n e c h a n g e s , (ii)
closed t h e m o l e c u l a r gates t o t h e active site, (iii) r e d u c e d t h e v o l u m e s o f t h e m a i n a n d slot access
tunnels, a n d (iv) o c c l u d e d t h e active site. Despite these spatial restrictions, e x p e r i m e n t a l tracing o f t h e
access t u n n e l s using k r y p t o n derivative crystals d e m o n s t r a t e s that t r a n s p o r t of ligands is still effective.
O u r findings highlight key t h e r m o s t a b i l i z a t i o n effects a n d provide a structural basis for designing n e w
t h e r m o s t a b l e protein catalysts.
Introduction
Enzymes have evolved for b i l l i o n s of years, a n d w i l l c o n t i n u e to
do so as l o n g as life o n earth exists.1
T h e y catalyze a l m o s t a l l
c h e m i c a l reactions that occur i n l i v i n g organisms, a n d m a n y of
t h e m have been successfully incorporated into diverse industrial,
e n v i r o n m e n t a l a n d b i o m e d i c a l technologies.2
Often, w i l d
type enzymes do n o t fully m e e t the d e m a n d s of these h a r s h
technological processes, a n d p u n c t u a l m u t a t i o n s are engineered
into t h e m to improve their physico-chemical properties
for technological applications. T h e key parameter for a l l
enzymes to be employed i n i n d u s t r i a l catalysis is
"Loschmidt Laboratories, Department of Experimental Biology and RECETOX, Faculty
of Science, Masaryk University, Kamenice 5, 625 00 Brno, Czech Republic. E-mail:
jiri@chemi. muni, cz; martin. marek@recetox. muni, cz
international Clinical Research Center, St. Anne's University Hospital Brno, Pekařská
53, 656 91 Brno, Czech Republic
'Universitě Grenoble Alpes, CNRS, CEA, Interdisciplinary Research Institute of
Grenoble (IRIG), Laboratoire Chimie et Biologie des Métaux (LCBM), 17 Avenue des
Martyrs, 38054 Grenoble, France
d
European Synchrotron Radiation Facility (ESRF), 71 Avenue des Martyrs, 38043
Grenoble, France
f Electronic supplementary information (ESI) available. See DOI:
10.1039/d0sc03367g
% These authors contributed equally.
thermostability, w h i c h allows t h e m to w i t h s t a n d elevated
temperatures d u r i n g biocatalytic processes.3
E n h a n c i n g protein thermostability involves changes that
shift the f o l d i n g - u n f o l d i n g balance toward the folded f o r m .
Stabilizing substitutions c a n either stabilize the folded conform
a t i o n or destabilize the u n f o l d e d one. T h e m o s t direct way to
stabilize proteins is to create or strengthen attractive interactions
between a m i n o acids i n the folded c o n f o r m a t i o n .
A l t h o u g h proteins w i l l c o n t i n u e to u n f o l d anyway, these
stronger interactions w i l l either slow d o w n u n f o l d i n g or speed
up r e f o l d i n g processes." A structured f o r m c a n be stabilized
t h r o u g h non-covalent interactions i n c l u d i n g h y d r o p h o b i c
interactions, hydrogen b o n d s , salt bridges a n d v a n der W a a l s
forces.5
Increasing the n u m b e r of s t a b i l i z i n g electrostatic
interactions between residues of opposite charge reinforces
proteins' t h e r m a l stability.6
H y d r o p h o b i c interactions have
been s h o w n to contribute proportionally m o r e effectively to
protein stability t h a n hydrogen b o n d s . 7
T h e h y d r o p h o b i c effect
is i n d e e d the d o m i n a n t d r i v i n g force i n protein folding, a n d
d e s i g n i n g a well-packed h y d r o p h o b i c core is therefore usually
a n efficient strategy for e n g i n e e r i n g stable proteins.4
Haloalkane dehalogenases (HLDs; E C 3.8.1.5) are a/p-hydrolases
that catalyze the hydrolytic cleavage of the carbon-halogen
b o n d i n diverse halogenated aliphatic hydrocarbons via S N 2
nucleophilic substitution. The reaction requires the a d d i t i o n of
11162 I Chem. Sei., 2 0 2 0 , 11, 11162-11178 This journal is © The Royal Society of Chemistry 2 0 2 0
V i e w Article Online
Edge Article
a water molecule a n d releases a halide i o n together w i t h a proton,
finally p r o d u c i n g the corresponding alcohol.8
Structurally, H L D s
consist o f a canonical a/p-hydrolase fold, w h i c h is composed of
a central eight-stranded P-sheet d o m a i n surrounded by several ahelices
{i.e. a P a sandwich architecture). A n additional versatile
helical cap d o m a i n is observed to be specific to each H L D
enzyme.9
T h e active site contains a catalytic pentad, w h i c h
consists of a nucleophile, a base, a catalytic acid, a n d two halidestabilizing
residues.9
I n a l l H L D s , the active site is positioned i n
a hydrophobic pocket buried between the a/p-fold core a n d the
cap d o m a i n , a n d this catalytic center is connected w i t h the b u l k
solvent via a m a i n t u n n e l a n d a slot t u n n e l . 1 0
B o t h o f these
tunnels are crucial determinants of the specific catalytic activity
and the substrate selectivity o f each H L D enzyme.1 1
'1 2
Recently, we developed FireProt,1 3
'1 4
a fully automated a n d
robust computational pipeline c o m b i n i n g energy- a n d evolutionbased
approaches to design highly stable multi-point mutant
proteins. W e employed FireProt to enhance the thermostability of
DhaA, a n H L D enzyme f r o m Rhodococcus rhodochrous [Tm =
50.5 °C; To p t = 45 °C). After several iteration cycles, we obtained
an 11-point D h a A mutant, hereafter referred to as DhaA115, with
outstanding thermostability (Tm = 73.5 °C) a n d thermophilicity,
as demonstrated by a substantial shift i n the o p t i m a l catalytic
temperature (To p t = 65 °C).1 3
C o m p u t a t i o n a l m o d e l i n g showed
that 3 of the 11 stabilizing residues line the m a i n access tunnel, 3
other residues are buried w i t h i n the protein core a n d the last 5
residues are exposed to solvent o n the protein surface.1 3
W e
inferred that 8 of these mutations (C128F, T148L, A172I, C176F,
D198W, V219W, C262L a n d D266F), w h i c h were identified by the
energy-based approach, potentially enhance the stability of the
enzyme by i m p r o v i n g the p a c k i n g of atoms w i t h i n the protein
interior and/or by strengthening hydrophobic interactions.1 3
However, the stabilizing effects of the 3 r e m a i n i n g mutations
(E20S, F80R a n d A155P), proposed by the evolution-based
approach, cannot be reproduced by force-field calculations.1 5
Experimental data are l a c k i n g to explain the structural basis for
the engineered hyperstability of DhaA115.
To fill this gap, we crystallized a n d solved high-resolution
structures of the hyperstable enzyme D h a A 1 1 5 . Analyses of
these crystal structures h i g h l i g h t specific a m i n o acid constellations
that p r i m a r i l y reinforce the a P a - s a n d w i c h architecture
and the helical cap d o m a i n via m u l t i p l e newly-established
interactions of the non-polar, h y d r o p h o b i c a n d aromatic 7 t - u
stacking types. Surprisingly, we f o u n d that placement of bulky
aromatic a m i n o acids o n the protein surface triggered some
unexpected long-distance changes i n the protein backbone.
Essentially, these changes cause the gates a n d the internal
volumes of b o t h the m a i n a n d the slot access tunnels to be
restricted, a n d consequently the enzyme active site appears
somewhat occluded. Interestingly, despite the active site
occlusion, experimental m a p p i n g of the enzyme tunnels by
krypton derivatization o f the DhaA115 crystals, supported by
protein dynamics simulations, showed that l i g a n d molecules
can still be transported t h r o u g h the enzyme tunnels. Collectively,
o u r findings demonstrate that the hyperstabilization
engineered i n D h a A led to massive r e d u c t i o n i n the v o l u m e of
its access tunnels, a n d that the enzymes are still capable of
This journal is © The Royal Society of Chemistry 2 0 2 0
Chemical Science
operating since they are permeable to substrates, products a n d
water molecules. T h i s permeability is t h e n increased at elevated
temperature, as previously demonstrated by the shifted o p t i m a l
catalytic temperature (Topt = 65 °C) of the DhaA115 enzyme.1 3
Results
Crystal structure of the hyperstable enzyme DhaA115
To o b t a i n precise structural i n f o r m a t i o n about h o w the D h a A
enzyme is thermostabilized, we focused o u r efforts o n crystallization
of the m o s t stabilized enzyme variant, DhaA115. W e
obtained crystals that belonged to the space group P12tl a n d
diffracted at 1.6 A resolution (Table 1). T h e final m o d e l contains
Table 1 Crystallographic data c o l l e c t i o n a n d r e f i n e m e n t statistics
Krypton-soaked
Data collection" Native DhaA115 DhaA115
Wavelength (A) 0.861 0.861
Space group P12il P21 21 21
Cell dimensions
a, b, c (A) 70.19, 68.12, 83.92 67.98, 82.04, 144.18
c, ß, J (°) 90, 104.82, 90 90, 90, 90
Resolution (A) 48.07-1.6 (1.657-1.6) 49.2-1.55
(1.605-1.55)
Total reflections 673 251 (64 239) 1 562 749 (155 793)
Unique reflections 99 075 (9734) 116 063 (11 342)
^merge 6.85 (61.9) 6.3 (83.5)
Hal 15.55 (2.55) 24.89 (3.45)
Completeness (%) 98.16 (97.14) 98.76 (98.12)
Multiplicity 6.8 (6.6) 13.5 (13.7)
CC(l/2) 99.9 (85.6) 100 (94.7)
Wilson B-factor 17.23 18.26
Refinement
Resolution (A) 48.07-1.6 49.2-1.55
(1.657-1.6) (1.605-1.55)
No. reflections 99 075 (9733) 116 063 (11 335)
^work/^free 0.158/0.178 0.160/0.181
Number of atoms
Protein 4837 4827
Ligand 62 125
Water 567 622
B-factors
Protein 20.72 19.20
Ligand 31.74 31.56
Water 32.49 33.33
RMS deviations
Bond lengths (A) 0.006 0.006
Bond angles (°) 0.84 0.90
Ramachandran 96.02 96.37
favored (%)
Ramachandran 3.98 3.63
allowed (%)
Ramachandran 0 0
outliers (%)
PDB ID code 6SP5 6SP8
" Values i n parentheses are for the highest-resolution shell.
Chem. Sei., 2 0 2 0 , 11, 11162-11178 | 11163
Chemical Science
V i e w Article Online
Edge Article
two enzyme molecules per asymmetric u n i t a n d has good values
for deviation f r o m the ideal (root mean-square deviation o n the
C a atoms of 0.3 A; F i g . S l f ) , w i t h .R-factor a n d R-free values of
0.16 a n d 0.18 respectively (Table 1). A l m o s t a l l of the residues
were built i n density, except for a few residues at the disordered
a m i n o - a n d carboxy-terminal ends.
DhaA115 adopts a c a n o n i c a l H L D fold similar to that o f the
wild-type D h a A ( R M S D o n the C a atoms of 0.6 A ; F i g . S2f),
f o r m i n g a single a|3a s a n d w i c h architecture (a/p-hydrolase core]
w i t h a characteristic h e l i c a l cap d o m a i n (Fig. 1). T h e a/p core is
c o m p o s e d of a central eight-stranded P-sheet, w i t h a P2 strand
i n a n anti-parallel orientation. T h i s a/P core is s a n d w i c h e d by
helices ( a l - a 3 ) o n one side a n d ( a 8 - a l l ) o n the other. The
helical (a4-a7) cap d o m a i n , w h i c h is positioned between the P6
strand a n d the a8 helix, shields the a/p-hydrolase core to w h i c h
it is a n c h o r e d via L 9 a n d L 1 4 loops. T h e enzyme active site is
located i n a p r e d o m i n a n t l y h y d r o p h o b i c cavity f o r m e d at the
interface between the a/p-hydrolase core a n d the cap d o m a i n .
The overall topology o f the secondary structure elements is very
similar to that of the wild-type D h a A . However, specific backbone
re-arrangements are observed, w h i c h encompass the L 9 ,
L10 a n d L 1 4 loops a n d the a4 a n d a9 helices (Fig. 1 a n d S2f).
A
B
Stabilizing rK
mutations v>
N - h y d r o l a s e c o r e
j i _
c a p d o m a i n
4F
i_ J L
h y d r o l a s e core - C
DhaA115
L5 u2
- 6 E
DhaA115 1 M S E I G T G F P F D P HY V E V LGMRMHY V D V G P R D G T PV L F L H G N P T S S Y LWRN I I P H V A P S H R C I A P D L I G M G K S D K P D L D Y R F D 82
DhaA
a2 L6 ß5 L7 a 3 L8 L9 a4 L10 a 5 '
DhaA115 83 D H V R Y L D A F I E A L G L E E V V L V I HDWG S A LG F HWA KR N P ERV KG I A FjM E F I RP I P T W D E W P E F A R E
DhaA D H V R Y L D A F I E A L G L E E V V L V I H D W G S A L G F H W A K R N P E R V K G I A I M E F I R P I P T W D E W P E
L11 a 5 L12 a6 L13 a7 L14
L F Q A F R T P i
_^L15 _ß7
I D V G R E L I I D 164
E L I I D 164
L16
DhaA115 165 Q N A F I EG|
DhaA
J L P K j V VW,V V R P L T E V E M D H Y R E P F L K P VWR E P LWR F P N E L P I A G E P A N I WA L V E A Y M NWL HQ S P V P K L L FWG T PG V L 246
5 E P A N I H A L V
L18 «10 L19
DhaA115 247 I P P A
DhaA
E A A R L A E S L P N B K T V B I G P G L H Y L Q E D N P D L I G S E I A R W L P A L 293
E A A R L A E S L P N | K T V | I G P G L H Y L Q E D N P D L I G S E I A R W L P A L 293
L148 P155
Side view Top view
Fig. 1 Overall s t r u c t u r e of DhaA115. (A) S c h e m a t i c r e p r e s e n t a t i o n of t h e p r o t e i n s e q u e n c e s h o w i n g t h e d o m a i n t o p o l o g y of DhaA115 a n d t h e
positions of t h e stabilizing m u t a t i o n s . (B) S t r u c t u r e - b a s e d s e q u e n c e a l i g n m e n t of DhaA115 a n d DhaA. T h e stabilizing m u t a t i o n s are s h o w n in
violet frames. S e c o n d a r y s t r u c t u r e e l e m e n t s f o u n d in DhaA115 are s h o w n a b o v e t h e a l i g n m e n t . Catalytically essential residues are p o i n t e d o u t
w i t h red dots. (C) C a r t o o n r e p r e s e n t a t i o n of DhaA115 s t r u c t u r e w i t h t h e c e n t r a l e i g h t - s t r a n d e d (3-sheet (yellow), t h e a/(3-hydrolase helices (blue)
and t h e helical c a p d o m a i n ( b r o w n ) . T h e stabilizing m u t a t i o n s are s h o w n as p u r p l e spheres. T h e i s o t h i o c y a n a t e (SCN) m o l e c u l e b o u n d in t h e
e n z y m e active site is s h o w n as spheres. (D) C l o s e - u p v i e w of t h e e n z y m e ' s active site w i t h key catalytically essential residues (grey sticks) a n d
s i m u l a t e d a n n e a l i n g o m i t e l e c t r o n density m a p c o n t o u r e d at 2a a r o u n d t h e i s o t h i o c y a n a t e (SCN) m o l e c u l e . M o l e c u l a r c o n t a c t s b e t w e e n p r o t e i n
residues a n d SCN are s h o w n as y e l l o w dashed lines.
11164 | Chem. ScL, 2 0 2 0 , 11, 11162-11178 This journal is © The Royal Society of Chemistry 2 0 2 0
V i e w Article Online
Edge Article
Additionally, we u n a m b i g u o u s l y identified i n the electron
density m a p the presence of bis-tris propane (B3P), glycerol
(GOL) a n d isothiocyanate (SCN) molecules, w h i c h were b o u n d
to the DhaA115 enzyme. Consistent w i t h this, bis-tris propane
a n d isothiocyanate were required i n the crystallization solution,
while the glycerol was used for cryo-protection. T h e bis-tris
propane a n d glycerol molecules are b o u n d o n the protein
surface, the former b e i n g also involved i n crystal-packing
contacts. There are three S C N - b i n d i n g sites per enzyme molecule;
two of t h e m are also located o n the enzyme surface while
the last one is deeply b u r i e d i n the active site cavity (Fig. S l f ) . A s
s h o w n i n F i g . I D , the latter S C N a n i o n interacts w i t h three
catalytic residues: the n u c l e o p h i l i c aspartate D 1 0 6 (2.6 A) a n d
the two halide-stabilizing residues, W 1 0 7 (3.5 A) a n d N 4 1 (3.6
A). It is also i n close contact w i t h the non-catalytic proline P206
(3.3 A). T h i s S C N - b i n d i n g site thus overlaps w i t h the halideb
i n d i n g site, where the halide a n i o n product is usually
captured d u r i n g the dehalogenation reaction.
Solution structure of the hyperstable DhaA115
W h i l s t the wild-type D h a A is a m o n o m e r i c enzyme, we previously
noted that the stabilized DhaA115 variant forms
a m i n o r i t y of dimers a n d high-molecular-weight oligomeric
states.1 5
W e therefore speculated as to whether the d i m e r
observed i n the asymmetric u n i t of the crystal (Fig. S l f ) m i g h t
also exist i n solution. T o test this hypothesis, we employed
small-angle X-ray light scattering (SAXS) analysis to probe the
DhaA115 structure i n solution. T h e SAXS profile of the DhaA115
solution closely fits the scattering profile calculated u s i n g
a single DhaA115 m o n o m e r of the crystal structure ( x 2
= 1.25),
but consistently does n o t correspond at a l l to the scattering
curve calculated u s i n g the d i m e r of the crystal asymmetric u n i t
(x2
= 50.3; F i g . 2). Furthermore, the radius of gyration (Rg )
determined for the merged data has a value o f 18.34 A. T h e
representative pair distance d i s t r i b u t i o n function, P(r), evaluated
by the indirect Fourier transform w i t h the G N O M
Chemical Science
package,1 6
is s h o w n i n F i g . 2. T h e profile has a bell-like shape
w i t h a m a i n peak at 23.4 A , a n d trails off to a m a x i m u m
d i m e n s i o n (Dm a x) o f ~ 5 7 A . Finally, the ab initio m o d e l reconstructed
f r o m the experimental SAXS data perfectly a c c o m m o dates
a m o n o m e r of the DhaA115 crystal structure (Fig. 2).
O u r SAXS results demonstrate that the purified DhaA115 is
indeed a m o n o m e r i c enzyme. C o m p l e m e n t a r y PISA calculat
i o n s 1 7
showed that the b u r i e d solvent-accessible area i n the
crystal contact DhaA115 d i m e r is ~ 2 4 1 A 2
, w h i c h represents
only 2 . 1 % o f the total solvent-accessible surface area of the
m o n o m e r (~11 298 A 2
) . T a k e n together, the SAXS experiments
a n d the PISA calculations provide evidence that the crystal
contact d i m e r observed i n the asymmetric u n i t (Fig. S l f ) is n o t
biologically relevant a n d does n o t exist i n solution. O u r data
suggest that the DhaA115 dimers observed by Beerens a n d cow
o r k e r s 1 5
m u s t employ a d i m e r i z a t i o n interface different f r o m
that observed i n o u r crystal p a c k i n g (Fig. S l f ) .
Localization of the stabilizing mutations
Computer-aided design predicted eleven a m i n o acid substitutions,
whose simultaneous i n t r o d u c t i o n into the D h a A enzyme
resulted i n a highly thermostable enzyme variant, D h a A 1 1 5 ,
w i t h Tm = 73.5 °C a n d shifted o p t i m a l catalytic temperature
[Topt = 65 °C).1 3
A s s h o w n i n Fig. 1A, the designed mutations are
evenly distributed a l o n g the protein sequence, w i t h 6 of t h e m
localized i n the a/p-hydrolase core (E20S + F80R + C128F +
V 2 1 9 W + C 2 6 2 L + D266F) a n d the r e m a i n i n g 5 m u t a t i o n s i n the
cap d o m a i n (T148L + A155P + A172I + C 1 7 6 F + D198W). Careful
inspection o f the DhaA115 structure revealed that n i n e o u t of
the eleven mutations are located i n the secondary structure
elements (Fig. IB), the other two (F80R + A155P) i n the
secondary structure/loop transitions.
Structural implications of evolution-based mutations
There are three m u t a t i o n s that were designed by a protein
evolution-based a p p r o a c h , 1 3
n a m e l y E20S, F80R a n d A155P. A l l
Scattering vector q ( A ) r (A) view
Fig. 2 Solution s t r u c t u r e o f DhaA115 d e t e r m i n e d by SAXS. (A) E x p e r i m e n t a l SAXS scattering c u r v e for DhaA115 (black dots) is s h o w n against t h e
calculated scattering curves for t h e DhaA115 m o n o m e r (blue line) a n d DhaA115 crystal d i m e r (green line). (B) Distance distribution f u n c t i o n of
DhaA115 c o m p u t e d f r o m t h e X - r a y scattering pattern using t h e G N O M p r o g r a m . (C) Ab initio m o l e c u l a r e n v e l o p e g e n e r a t e d f r o m SAXS data
analysis. T h e m o l e c u l a r SAXS e n v e l o p e of t h e DhaA115 m o n o m e r is s h o w n in a s e m i - t r a n s p a r e n t grey c o l o r s u p e r p o s e d o n t h e DhaA115
m o n o m e r of t h e crystal s t r u c t u r e r e p r e s e n t e d as a blue c a r t o o n .
This journal is © The Royal Society of Chemistry 2 0 2 0 Chem. Sei., 2 0 2 0 , 11, 11162-11178 | 11165
Chemical Science
V i e w Article Online
Edge Article
these residues are located o n the enzyme surface (Fig. 1C),
where they were f o u n d to either participate i n the surface charge
network i m p o r t a n t for protein-solvent interactions (Fig. 3A) or
rigidify the solvent-exposed flexible loop (Fig. 3B). Specifically,
the replacement of a surface phenylalanine w i t h a n arginine
(F80R) disrupted the cation-7t interaction between F80 a n d
R204 (4.0 A) present i n the wild-type D h a A a n d established new
ionic interactions w i t h D78, D82 a n d D 8 3 . Moreover, there is
a newly established water-mediated hydrogen-bonded network
involving R80, D82, D83 a n d R86 (Fig. 3A).
Similarly, the serine residue (E20S) participates i n the
formation of a n extensive solvent-mediated interaction network,
i n w h i c h the residues L18, S20, D73 a n d Y87 are involved.
Strikingly, the water-mediated interactions between the L18, S20
a n d Y87 residues apparently rigidify the L I loop connecting the
p i a n d P2 strands, a n d help to protect the central P-sheet
(Fig. 3A). B o t h stabilizing S20 a n d R80 residues, w h i c h are
located ~ 1 6 . 5 A apart f r o m one another, participate extensively i n
local protein-water interactions, w h i c h contribute to the global
solvent hydrogen-bonded network (Fig. 3A).
The last of the m u t a t i o n s deduced by the evolution-based
a p p r o a c h is the substitution of a n alanine by p r o l i n e (A155P]
i n the L 1 0 loop that connects the a4 a n d a 5 ' helices w i t h i n the
cap d o m a i n . T h i s s u b s t i t u t i o n forces the L10 loop to adopt
a c o n f o r m a t i o n different f r o m that observed i n D h a A . Specifically,
the i n t r o d u c e d p r o l i n e residue (P155) is present i n transc
o n f o r m a t i o n , w h i c h brings its carbonyl oxygen into a p o s i t i o n
where it c a n interact w i t h the m a i n - c h a i n nitrogen of V157 (2.6
A) (Fig. 3B). I n a d d i t i o n , the new c o n f o r m a t i o n of the L10 loop
enables the molecule to establish two new m a i n - c h a i n to m a i n c
h a i n hydrogen b o n d s , n a m e l y between the carbonyl oxygen of
T154 a n d the nitrogen of G158 (2.9 A), a n d between the carbonyl
a t o m of A151 a n d the nitrogen of T154 (3.1 A). T h e L 1 0 loop
interacts extensively w i t h a n u n d e r n e a t h a7 helix t h r o u g h
m u l t i p l e water-mediated hydrogen b o n d s i n D h a A 1 1 5 , b u t not
i n D h a A (Fig. 3B), w h i c h again m a y have a positive effect o n
protein-solvent interactions.
Structural implications of energy-based mutations
The r e m a i n i n g 8 m u t a t i o n s i n D h a A 1 1 5 (C128F, T148L, A172I,
C176F, D 1 9 8 W , V219W, C 2 6 2 L a n d D266F) were inferred by
force-field c a l c u l a t i o n s . 1 3
Interestingly, all these a m i n o acids
were m u t a t i o n s toward residues of the h y d r o p h o b i c or aromatic
type a n d always w i t h a sterically b u l k i e r side c h a i n . Prior to our
E v o l u t i o n - i n f e r r e d m u t a t i o n s E n e r g y c a l c u l a t i o n - d e r i v e d m u t a t i o n s
S20 + R80 L148 + I172 + F176 W219 + F266
B
P155 F128 + L262 D198W
Fig. 3 Stabilizing a m i n o acid interactions o b s e r v e d in DhaA115. C l o s e - u p views of stabilizing residues a n d their interacting n e i g h b o u r s : S20 and
R80 (A), P155 (B), L148,1172 a n d F176 (C), F128 and L262 (D), W219 and F266 (E) and W 1 9 8 (F). T h e stabilizing residues are s h o w n as p u r p l e sticks
and s e m i - t r a n s p a r e n t spheres, t h e s u r r o u n d i n g p r o t e i n residues are s h o w n as grey sticks, and w a t e r m o l e c u l e s as red spheres. H y d r o g e n b o n d s
are s h o w n as y e l l o w dashed lines. T h e p r o t e i n p h y l o g e n y - i n f e r r e d m u t a t i o n s are in panels A and B, w h i l e t h e e n e r g y c a l c u l a t i o n - p r e d i c t e d
m u t a t i o n s are in panels C - F .
11166 | Chem. ScL, 2 0 2 0 , 11, 11162-11178 This journal is © The Royal Society of Chemistry 2 0 2 0
Edge Article
V i e w Article Online
Chemical Science
current work, these mutations were a s s u m e d to reinforce
h y d r o p h o b i c interactions a n d improve the protein p a c k i n g . 1 3
O u r DhaA115 structure strongly confirms these a s s u m p t i o n s ,
but we also see n e w structural effects that were n o t previously
predicted by the c o m p u t a t i o n a l design.
Crucially, we observe that the majority of the energy-based
mutations (7 o u t of the 8, D 1 9 8 W b e i n g the exception) cooperate
w i t h each other a n d jointly contribute to the stabilization
of the protein fold. Firstly, a triplet of mutations (T148L, A172I
a n d C176F) localized i n the cap d o m a i n interact w i t h each
other, b u t also strongly reinforce the h y d r o p h o b i c a n d aromatic
Tt-Tt interactions w i t h the n e i g h b o r i n g residues. These three
stabilizing m u t a t i o n s interlock the a.4, <x5' a n d a5 helices a n d
adjacent L14 loop, thus rigidifying the top of the cap d o m a i n
[Fig. 3C). Specifically, F176 forms a parallel-displaced stacking
interaction w i t h F149 (5.9 A) a n d a T-shaped edge-to-face
stacking interaction w i t h F144 (5.8 A), w h i l e L 1 4 8 a n d 1172
are involved i n m u l t i p l e h y d r o p h o b i c a n d non-polar contacts
w i t h s u r r o u n d i n g residues. C o m p l e m e n t a r y calculations
carried o u t by the Protein Interaction Calculator (PIC)1 8
identified
15 n e w h y d r o p h o b i c interactions i n DhaA115 d u e to this
triple substitution. These interactions are n o t present i n w i l d type
D h a A enzyme (Fig. 4A).
Similar cooperation takes place i n the core o f the enzyme
w i t h the substitutions C128F a n d C 2 6 2 L . T h e replacement of
these two polar cysteine residues by the b u l k i e r aromatic
DhaA
Thermostabilisation
DhaA115
T148L + A172I + C176F — 1 1 5 new hydrophobic interactions
— 4 2 hydrophobic interactions
— T 12 new hydrophobic interactions
4 1 aromatic-sulphur interaction
Protein backbone 14 n e w
hydrogen bonds
— T 1 new hydrophobic interactions
^ . — 4 1 hydrophobic interaction
— 4 1 hydrogen bond
E140) IL246)
D198W — T 3 new hydrophobic interactions
— 1 1 new cation-7t interaction
— 4 2 ioinic interactions
Fig. 4 Selected r e s i d u e - r e s i d u e interaction n e t w o r k s in D h a A a n d DhaA115. ( A - D ) Panels d e p i c t i n g a m i n o acid interactions n e t w o r k s for
stabilizing m u t a t i o n s T 1 4 8 L / A 1 7 2 I / C 7 6 F (A), C 1 2 8 F / C 2 6 2 L (B), V 2 1 9 W / D 2 6 6 F (C) a n d D 1 9 8 W (D). T h e Protein Interactions Calculator (PIC) server
was used t o m a p r e s i d u e - r e s i d u e interactions. A m i n o acid residues that w e r e m u t a t e d are s h o w n in violet circles a n d t h e o t h e r s in w h i t e . N e w l y
established r e s i d u e - r e s i d u e interactions are d e p i c t e d by violet a n d o r a n g e lines.
This journal is © The Royal Society of Chemistry 2 0 2 0 Chem. Sci., 2 0 2 0 , 11, 11162-11178 | 11167
V i e w Article Online
Chemical Science
phenylalanine (F128) a n d the h y d r o p h o b i c leucine (L262]
enabled the f o r m a t i o n of m u l t i p l e new van der W a a l s contacts
a n d non-polar interactions w i t h i n the b u r i e d h y d r o p h o b i c core
of the a/|3-hydrolase d o m a i n . T h e side c h a i n of F128 interacts
via a T-shaped edge-to-face stacking w i t h F113 (5.2 A), a n d via
a Y-shaped stacking interaction w i t h F113 (5.3 A), f o r m i n g the
aromatic cage. Next to this, the m u t a t e d L262 residue complements
a leucine-rich region that already includes L229, L238,
L255 a n d L259 i n w i l d type D h a A (Fig. 3D). Overall, our structure
shows that F128 a n d L262 cooperatively contribute to tighte
n i n g of the p a c k i n g between the central P-sheet a n d the adjacent
a-helical (a8 a n d a9) shell. T h e PIC calculations detected
12 newly-established h y d r o p h o b i c contacts as a result of these
interactions m e d i a t e d by F128 a n d L262 (Fig. 4B).
Besides the short-distance cooperativities a m o n g the introd
u c e d m u t a t i o n s described above, we also observed longdistance
cooperativity effects between m u t a t i o n s V 2 1 9 W a n d
D266F. Interestingly, b o t h substitutions are located o n the
protein surface, where the p l a c i n g of the b u l k i e r aromatic side
chains of W 2 1 9 a n d F266 triggered unexpected changes i n the
protein b a c k b o n e (Fig. 3E). Specifically, these h y d r o p h o b i c
residues tend to m i n i m i z e solvent exposure, a n d to c o m p a c t the
protein fold via interactions w i t h a m i n o acids b u r i e d i n their
surroundings. First, the side c h a i n of W 2 1 9 adopts a flipped-in
c o n f o r m a t i o n , w h i c h enables its indole nitrogen to be
hydrogen-bonded w i t h the carboxyl group of E223. M o r e
importantly, the flipped-in c o n f o r m a t i o n of W 2 1 9 has triggered
a major structural re-arrangement of the L 9 loop (Fig. 3E). These
changes are a c c o m p a n i e d by the f o r m a t i o n of a n e w network of
interaction between residues R133, E140, E251 a n d L246. T h i s
new re-arrangement is further favored by a slight t i l t i n g (~7°) of
a9 helix indirectly i n d u c e d by F266 a n d to a lesser extent by
L262 (Fig. 3E). T h e b u l k i e r m u t a t e d F266 (P9) pushes the side
c h a i n of W 2 4 0 (P7) toward the a9 helix w h i c h tilts, t h e n this
slight reorientation allows R133, E140, E251 a n d L246 to
interact w i t h each other (Fig. 3E). O u r analysis has unraveled a n
epistatic interaction network by w h i c h the simultaneous
substitutions of two remote residues (~27 A) o n the protein
surface (V219W a n d D266F) c a n trigger long-distance structural
re-arrangements (Fig. 3E a n d 4C).
Finally, the last energy calculation-derived m u t a t i o n ,
D 1 9 8 W , substitutes a n aspartate that forms two ionic interactions
w i t h K 7 4 a n d K195 i n D h a A . However, its replacement
w i t h the b u l k y aromatic tryptophan (D198W) established
a strong cation-7t interaction w i t h K 7 4 (4.7 A) a n d three additional
n e w h y d r o p h o b i c contacts w i t h F193 (4.1 A), L194 (4.8 A]
a n d V197 (3.8 A) (Fig. 3F a n d 4D).
Thermostabilization induced unexpected changes in the
protein backbone
W e s h o w that the computer-aided t h e r m o s t a b i l i z a t i o n of D h a A
via 11 p o i n t m u t a t i o n s (DhaA115) not only affected the corres
p o n d i n g side-chain to side-chain and/or side-chain to m a i n c
h a i n interactions, b u t also i n d u c e d major p r o t e i n backbone
changes (Fig. 5 a n d S2f). There is a l m o s t perfect agreement
between the c o m p u t a t i o n a l design a n d the crystallographic
11168 | Chem. Sci., 2 0 2 0 , 11, 11162-11178
Edge Article
structure of D h a A 1 1 5 i n the re-arrangement of the L 1 0 loop. The
DhaA115 crystallographic structure revealed that the introd
u c e d p r o l i n e (P155) adopts a fraras-conformation l e a d i n g to
structural re-arrangement of the L 1 0 loop, a n d c o n f i r m e d the
p r e d i c t i o n of the protein design (Fig. 3B a n d 5E).1 3
The major discrepancies between the designed a n d X-ray
structures of D h a A 1 1 5 consist i n different structural organizations
e n c o m p a s s i n g the L 9 a n d L14 loops, a n d the a4 a n d a9
helices. Careful i n s p e c t i o n of a l l structural m o d e l s - X-ray D h a A
template, D h a l l 5 design a n d X-ray D h a l l 5 - provided a n
explanation for these dissimilarities (Fig. 5). Firstly, the
substitution of a relatively s m a l l valine w i t h a b u l k y a n d
aromatic tryptophan (V219W) triggers the major structural
change i n the L 9 loop. T h i s L 9 re-arrangement is m o s t
p r o n o u n c e d i n the D h a A 1 1 5 X-ray structure, where the W 2 1 9 is
observed to adopt a flipped-in c o n f o r m a t i o n , w h i c h t h e n forces
the L 9 loop to re-arrange substantially (Fig. 5B). However,
a different c o n f o r m a t i o n was observed i n the predicted
DhaA115 design structure, where the c o r r e s p o n d i n g W 2 1 9
adopts a flipped-out orientation, w h i c h is not likely to exert
analogous steric pressure o n the L 9 loop to re-arrange to the
same extent as that observed i n the D h a A 1 1 5 X-ray structure
(Fig. 5C). Moreover, the structural re-arrangement of the L 9 loop
occurred c o n c o m i t a n t l y w i t h a slight tilt of the a4 helix (~6.3°),
w h i c h tightly presses against the o p p o s i n g a5 helix i n the
DhaA115 X-ray structure (Fig. 5D a n d S3f). Secondly, our
DhaA115 X-ray structure shows that the aspartate-tophenylalanine
s u b s t i t u t i o n (D266F) triggers structural changes
that are m o r e severe t h a n those predicted. As s h o w n i n F i g . 5,
the presence of the b u l k y side c h a i n of F266 indirectly, t h r o u g h
interaction w i t h W 2 4 0 , displaced the a9 helix toward the slot
t u n n e l entry. Here, it is i m p o r t a n t to note that another introd
u c e d m u t a t i o n , L262, is also likely responsible for the
displacement of the a9 helix (Fig. 5). As a result, several residues
l i n i n g the slot t u n n e l entry, especially R133, E140, E251 a n d
L246, are dramatically re-arranged, a n d f o r m m u l t i p l e new
hydrogen b o n d s , creating a so-called slot t u n n e l lock i n the
crystal structure of D h a A 1 1 5 (Fig. 5).
O u r observations p o i n t to the fact that the stabilizing
m u t a t i o n s m a y substantially affect the enzyme access tunnels
that connect the active site w i t h the b u l k solvent. These tunnels
are f u n c t i o n a l a n d ensure proper transport of substrate a n d
product molecules to a n d f r o m the active site, a n d are k n o w n
determinants of catalytic properties for this enzyme family.9
Thermostabilization reduced the volume of the enzyme access
tunnels
The b u r i e d active site of the wild-type D h a A enzyme is connected
w i t h the b u l k solvent via two tunnels, n a m e l y the m a i n
p i access t u n n e l a n d the slot p2 access t u n n e l (Fig. 6). It has
been previously s h o w n that engineering of one of these tunnels
may yield enzyme variants w i t h m o d i f i e d substrate preference,
enantioselectivity a n d thermostability.1 2
'1 9
'2 0
Analysis of the enzyme access tunnels a n d the active site cavity
i n DhaA115 revealed that the volumes of b o t h the m a i n a n d the
slot access tunnels are greatly reduced, a n d that the enzyme
This journal is © The Royal Society of Chemistry 2 0 2 0
V i e w Article Online
Edge Article Chemical Science
DhaA115 X-ray DhaA115 design
L10
Slot tunnel p l 5 5 - ^
semi-lock, V 1 5 7 d o m a l n
L 9 /
Superposition
DhaA X-ray + DhaA115 X-ray
L10
/
s* Cap
L14 I Q v 7/ Mm^.aA domain
Superposition
DhaA115 X-ray + DhaA115 design
L10
Cap
a 4 domain
Fig. 5 C o o p e r a t i v e effects of distant surface bulky side chains in slot t u n n e l closure. ( A - C ) C l o s e - u p c a r t o o n representations of X-ray structure
of DhaA (A), DhaA115 c o m p u t a t i o n a l design (B) a n d DhaA115 X - r a y s t r u c t u r e (C). N o t e that in DhaA (A) t h e slot t u n n e l entry is n o t b l o c k e d by a n y
residues, w h i l e in b o t h t h e DhaA115 X-ray structure (B) a n d t h e DhaA115 design (C) it is m a r k e d l y impaired. This is m o s t p r o n o u n c e d in t h e X-ray
structure, w h e r e t h e residue W 2 1 9 a d o p t s a f l i p p e d - i n c o n f o r m a t i o n leading t o a m a j o r structural r e - a r r a n g e m e n t of t h e L9 l o o p , w h i c h , t o g e t h e r
w i t h t h e « 9 helix a n d L16 l o o p , f o r m s t h e s o - c a l l e d slot t u n n e l lock (C). T h e stabilizing residues are s h o w n as purple sticks a n d s e m i - t r a n s p a r e n t
spheres. H y d r o g e n b o n d s are s h o w n as y e l l o w d a s h e d lines. (D a n d E) Structural superposition o f DhaA X-ray o n DhaA115 X-ray (D) a n d DhaA115
X-ray o n DhaA115 design (E). T h e stabilizing residues are s h o w n as purple sticks. Red a r r o w s depict m a j o r structural c h a n g e s in t h e b a c k b o n e .
N o t e that t h e stabilizing m u t a t i o n s V219W a n d D 2 6 6 F i n d u c e d structural r e - a r r a n g e m e n t s of b o t h L9 a n d t h e « 9 helix, leading t o t h e c l o s i n g o f
the slot t u n n e l .
active site is occluded (Fig. 6). U n l i k e i n DhaA, C A V E R calculations
o n the static X-ray DhaA115 structure d i d not detect any
tunnels w i t h m i n i m u m radius above 0.9 A (Fig. 6 C a n d D).
Careful inspection of the DhaA115 crystal structure revealed that
the m a i n access t u n n e l is blocked by the triplet of stabilizing
residues, namely L148,1172 a n d F176. T h e tight p a c k i n g of these
residues w i t h a few neighboring residues (F144, F149 a n d K175]
is the major h a l l m a r k o f a m a i n t u n n e l lock (Fig. 6C).
W h i l e partial closure of the m a i n access t u n n e l was previously
observed i n several engineered D h a A v a r i a n t s , 1 2
' 1 9
the
simultaneous closure of b o t h the m a i n a n d the slot tunnels is
a u n i q u e feature seen for the first t i m e i n D h a A 1 1 5 . T h i s doublelock
is enabled by: (i) the triplet of residues (L148, 1172 a n d
F176), coupled w i t h the re-positioning of the a4 helix, that lock
the m a i n access t u n n e l a n d (ii) the structural re-arrangement of
the L 9 l o o p coupled w i t h the re-positioning o f the a9 helix,
w h i c h lock the slot access t u n n e l . A s described above, we show
that residues W 2 1 9 , L262 a n d F266 are key drivers of these latter
structural re-arrangements, w h i c h b r i n g the residues R 1 3 3 ,
This journal is © The Royal Society of Chemistry 2 0 2 0 Chem. Sei., 2 0 2 0 , 11, 11162-11178 | 11169
V i e w Article Online
Chemical Science Edge Article
C u t a w a y e n z y m e surface Main tunnel entrance
T148
Slot tunnel entrance
E140
Open
active site
. ^ N41
W107 Halide-binding
site /
Slot tunnel
entry
tunnel
E251
P249
DhaA S
Main
tunnel entry
B C u t a w a y e n z y m e surface
N41
4 -
DhaA115
W 1 0 7
Halide-binding ^
site
O f
)
Close-up view of main tunnel lock
Main tunnel
Close-up view of slot tunnel lock
a4
Slot
tunnel
DhaA DhaA115 DhaA115
Fig. 6 Visualization o f access tunnels. C u t a w a y surface representations o f e n z y m e active site a n d access t u n n e l s f o r t h e w i l d t y p e DhaA (A) a n d
the hyperstable DhaA115 (B) e n z y m e s . N o t e that t h e D h a A e n z y m e has a s o - c a l l e d o p e n active site c o n n e c t e d w i t h t h e o u t s i d e bulk solvent via
b o t h t h e m a i n a n d slot access tunnels. In contrast, in DhaA115 t h e main a n d slot access t u n n e l s are t o a large e x t e n t b l o c k e d because o f t h e direct
or indirect effects o f stabilizing m u t a t i o n s . Visualization o f t h e C A V E R - c a l c u l a t e d main (C) a n d slot (D) access tunnels. T h e main t u n n e l w a s
calculated (cyan) for w i l d t y p e D h a A (left panel). T h e t u n n e l is absent in DhaA115 because o f t h r e e stabilizing residues (L148,1172 a n d F176) w h i c h ,
t o g e t h e r w i t h o t h e r h y d r o p h o b i c a n d a r o m a t i c residues (F144, F149 a n d L209) f o r m t h e s o - c a l l e d m a i n t u n n e l lock (right panel). Similarly, t h e slot
t u n n e l c a l c u l a t e d (blue) f o r w i l d t y p e DhaA (left panel) is absent in DhaA115 because o f a c o n f o r m a t i o n a l c h a n g e o f t h e L9 l o o p t r i g g e r e d by t h e
p r e s e n c e o f stabilizing residue W 2 1 9 a n d a positional shift o f t h e « 9 helix. This n e w c o n s t e l l a t i o n allows t h e residues R133 a n d E140 t o f o r m a s o called
slot t u n n e l lock (right panel). T h e stabilizing residues are s h o w n as p u r p l e spheres, a n d t h e o t h e r p r o t e i n residues are s h o w n as grey sticks.
E140, L246 a n d E251 close to each other to create the slot t u n n e l
lock (Fig. 5B a n d 6D).
Experimental tracking of the access tunnels using krypton
Despite the fact that the crystal structure o f DhaA115 showed
a greatly reduced v o l u m e for b o t h enzyme access tunnels, it has
been previously s h o w n that this hyperstable enzyme (Tm = 73.5
°C) still possesses dehalogenase activity, w i t h a shifted o p t i m a l
catalytic temperature ( T o p t = 65 °C). W e therefore a i m e d at
experimental m a p p i n g o f the enzyme access tunnels to test
whether s m a l l molecules c a n still be transported to the active
site o f D h a A 1 1 5 . W e employed a "soak-and-freeze" m e t h o d that
allows crystals to be processed i n a pressurized krypton atmosphere.2
2
T h e DhaA115 crystals were soaked i n krypton at
11170 I Chem. Sei., 2 0 2 0 , 11, 11162-11178 This journal is © The Royal Society of Chemistry 2 0 2 0
V i e w Article Online
Edge Article
a pressure of 150 bar, thereafter flash-frozen w h i l e still u n d e r
h i g h pressure, a n d t h e n recovered i n l i q u i d nitrogen, i n w h i c h
the derivatives are stable for X-ray data collections. A n o m a l o u s
diffraction data were collected at the k r y p t o n X-ray a b s o r p t i o n
edge (a wavelength o f 0.861 A).
Interestingly, s o a k i n g D h a A 1 1 5 crystals i n a high-pressure
krypton a t m o s p h e r e yielded crystals w i t h a h i g h e r symmetry
space g r o u p , P212121, a n d they diffracted to 1.55 A resolution
(Table 1). T h e final structural m o d e l of the k r y p t o n derivative of
a D h a A 1 1 5 crystal a g a i n revealed two enzyme m o l e c u l e s i n the
asymmetric u n i t ( R M S D o n C a ' s of 0.2 A ; F i g . S4f), a n d h a d
good values for deviations f r o m the ideal geometry, w i t h Rfactor
a n d R-free values o f 0.16 a n d 0.18 respectively (Table 1).
Importantly, k r y p t o n derivatization d i d not i n d u c e any structural
changes i n the p r o t e i n b a c k b o n e , a n d s u p e r p o s i t i o n w i t h
the native D h a A 1 1 5 s h o w e d a n R M S D for the C a a t o m s o f only
0.2 A (Fig. S5t).
T h e derivatized D h a A 1 1 5 structure shows 12 b i n d i n g
krypton sites ( K ^ - K r ^ ) per enzyme m o l e c u l e (Fig. 7 a n d S4f).
Two krypton a t o m s ( K r 6 a n d K r 7 ) are f o u n d i n the p r e d o m i nantly
h y d r o p h o b i c cavity of the enzyme active site, a n d i n close
proximity to the isothiocyanate (SCN) b o u n d i n the halideb
i n d i n g site. T w o a d d i t i o n a l krypton atoms ( K r 2 a n d K r 3 ]
occupy the slot t u n n e l , a n d one k r y p t o n a t o m (Kr5 ) sits i n the
entrance of the m a i n t u n n e l entry (Fig. 7B). Three krypton
atoms (Kr8 , K r 9 a n d K r u ) are b o u n d i n separate internal
h y d r o p h o b i c cavities, far away f r o m the active site. T h e
r e m a i n i n g four k r y p t o n a t o m s {Kiu K r 4 , K r 1 0 a n d K r 1 2 ) are
f o u n d at the surface of the protein, o c c u p y i n g excavations of
moderate h y d r o p h o b i c i t y or m e d i a t i n g crystal p a c k i n g contacts.
Finally, t u n n e l calculations o n the krypton-soaked D h a A 1 1 5
structure detected partial restoration of b o t h m a i n a n d slot
Chemical Science
tunnels (Fig. 7C), a l t h o u g h these tunnels are still m u c h r e d u c e d
as c o m p a r e d to those observed i n the wild-type D h a A enzyme
(Fig. 6). B e l o w we c o m p a r e these results w i t h a d y n a m i c a l
overview of the tunnels. O u r structural data demonstrate that
the engineered hyperstabilization l e d to massive r e d u c t i o n s i n
the v o l u m e s of b o t h enzyme access tunnels, b u t also that these
latter are still p e r m e a b l e for s m a l l m o l e c u l e s d u r i n g catalysis.
These observations suggest that p e r m e a b i l i t y is likely increased
w i t h elevated temperature, as previously s h o w n by the shift i n
the o p t i m a l catalytic temperature ( X o p t = 65 °C) o f D h a A 1 1 5 . 1 3
Analysis of structure flexibility by molecular dynamic
simulations
T h e structure o f D h a A 1 1 5 was subjected to m o l e c u l a r d y n a m i c s
(MD) analysis to assess its intrinsic flexibility a n d to c o m p a r e it
w i t h that of the w i l d type D h a A , a study w h i c h was reported
previously.1 5
T h e physical c o n d i t i o n s at w h i c h the M D s were
performed, n a m e l y 310 K (37 °C), p H 7.5, a n d salt concentration
of 0.1 M , c o r r e s p o n d to the c o n d i t i o n s at w h i c h H L D s are
typically characterized i n terms of activity a n d specificity.1 1
'1 2
These s i m u l a t i o n s , r u n i n duplicate for a total t i m e o f 200 n s ,
were properly equilibrated a n d converged (Fig. S6Af). T h e M D s
showed that the b a c k b o n e of D h a A 1 1 5 was similarly rigid to, or
slightly m o r e r i g i d t h a n , that of D h a A (Fig. S7Af). T h e m a i n
difference was a r o u n d residues 31 a n d 155, w h i c h presented
considerably h i g h e r B-factors i n the wild-type D h a A t h a n i n the
stabilized m u t a n t .
Accelerated M D s (aMDs) were also carried out. T h e a M D
technique is a n e n h a n c e d - s a m p l i n g m e t h o d that applies
a boost of potential energy that raises the energy of local
m i n i m a a n d thus decreases the energy barriers, resulting i n
A B C
Krypton-soaked DhaA115 Kryptons in enzyme active site Visualisation of access tunnels
Fig. 7 Experimental tracking of t h e e n z y m e access tunnels. (A) C a r t o o n representation of DhaA115 k r y p t o n derivative crystal s t r u c t u r e w i t h t h e
central e i g h t - s t r a n d e d (3-sheet (yellow), oc/p-hydrolase helices (blue) a n d helical c a p d o m a i n (brown). T h e locations of t h e k r y p t o n a t o m sites (Krj
t o Kru) in pressurized DhaA115 are s h o w n as blue spheres w i t h a n o m a l o u s Fourier m a p s c o n t o u r e d at 4<r (green mesh). (B) C l o s e - u p v i e w of t h e
active site a n d access t u n n e l s in t h e DhaA115 k r y p t o n derivative crystal structure. T w o k r y p t o n a t o m s (Kr6 a n d Kr7 ) are f o u n d in t h e e n z y m e active
sites, t w o (Kr2 a n d Kr3 ) o c c u p y t h e slot tunnels, a n d o n e k r y p t o n (Kr5 ) sits o n t o p o f t h e c a p d o m a i n . T h e stabilizing residues are s h o w n as purple
sticks a n d t h e o t h e r protein residues are s h o w n as grey sticks. (C) Visualization o f t h e access t u n n e l s in t h e DhaA115 k r y p t o n derivative crystal
s t r u c t u r e c a l c u l a t i o n by CAVER. T h e m a i n (cyan) a n d slot (blue) t u n n e l s are partially restored u p o n pressurization. T h e main t u n n e l is d e p i c t e d in
cyan a n d t h e slot t u n n e l in light blue. T h e stabilizing residues are s h o w n as purple spheres, r e m a i n i n g residues are s h o w n as grey sticks, a n d
k r y p t o n a t o m s as blue spheres.
This journal is © The Royal Society of Chemistry 2 0 2 0 Chem. Sei., 2 0 2 0 , 11, 11162-11178 | 11171
V i e w Article Online
Chemical Science
higher conformational transition rates. T h i s m e t h o d c a n be
useful for better s a m p l i n g the d y n a m i c behavior a n d the
conformational space o f biomolecular systems over longer timescales.2
3
'2 4
These a M D s were well equilibrated a n d stable
(Fig. S6Bf). T h e B-factors were higher i n the a M D s t h a n i n the
classical M D s , a n d this w a s expected f r o m simulations that
promote higher conformational diversity (Fig. S7Bf). Moreover,
i n the l o n g r u n , only the region a r o u n d residue 31 r e m a i n e d
m o r e rigid i n DhaA115 t h a n i n the wild-type D h a A , a n d the rest
of the protein showed s i m i l a r flexibility. T h e region a r o u n d
residue 77 seems to be slightly, b u t n o t significantly, m o r e
flexible i n the m u t a n t t h a n i n the wild-type enzyme.
W e also a i m e d to assess whether the structural changes
i n d u c e d by the mutations that were seen i n the crystal structures
are also m a i n t a i n e d d u r i n g M D s i m u l a t i o n s . A s expected,
b o t h DhaA115 a n d D h a A relaxed f r o m their respective crystallographic
conformations after the e q u i l i b r a t i o n steps. T h e
differences were m i n i m a l , w i t h R M S D slightly higher for the
wild-type D h a A t h a n for the DhaA115 m u t a n t (Fig. S6 a n d Table
S l f ) . One o f the m a i n differences previously highlighted was the
flipped-in c o n f o r m a t i o n o f the m u t a t e d residue W 2 1 9 . D u r i n g
all s i m u l a t i o n s this residue m a i n t a i n e d s u c h a n orientation a n d
showed very l o w deviations f r o m the DhaA115 crystal structure
( R M S D i n a M D s = 0.88 ± 0.28 A ; Table S l f ) , w i t h s i m i l a r values
as observed for V219 i n D h a A ( R M S D i n a M D s = 0.82 ± 0.27 A).
W h e n we analyzed the potential effect o f W 2 1 9 o n loop L9, we
observed that the initial distances between W 2 1 9 a n d P134 or
W 2 1 9 a n d P136 (both prolines b e i n g localized i n the L 9 loop]
were slightly increased d u r i n g the s i m u l a t i o n s . O u r results thus
suggest that the crystal p a c k i n g h a d some influence o n shorte
n i n g these distances as c o m p a r e d to the average structures i n
solution (rfi-rf3 distances i n Table S l f ) . However, the distances
f r o m W 2 1 9 to the L 9 loop were significantly longer i n DhaA115
t h a n i n the case o f V219 i n D h a A , i n d i c a t i n g that the effects o f
W 2 1 9 i n levering u p loop L 9 were preserved d u r i n g M D
simulations.
Another unexpected backbone change i n a9 helix positioning
observed i n the DhaA115 crystal structures is d u e to the D266F
mutation. T h e distances f r o m D266/F266 (located i n the (38
strand) to the backbone o f W240 (in the P7 sheet) are very stable
a n d close to those i n the crystal structures, a n d the conformation
of this region is not especially different f r o m the wild-type D h a A
(distance d4 i n Table S l f ) . However, the average distances
between the a9 helix a n d P7 sheet (d5) a n d between the a9 helix
a n d P8 sheet (rf6) remained greater i n the simulations o f DhaA115
as compared to D h a A (Table S l f ) . These M D observations
support the crystallographic findings that the protein backbone
was unpredictably re-arranged d u e to the thermostabilization
process. W e also verified the effects o n the structural rearrangement
at the m o u t h o f the slot tunnel, presumably locked through
an extended hydrogen b o n d network (Fig. 5B). T h e residues
R133, E140, E251 a n d R254 were m a k i n g H-bonds a n d were
intermittently i n contact w i t h one another. I n spite o f the initial
difference i n the crystal structure o f D h a A for E140, d u r i n g the
simulations the average distances between these four residues
were very similar for DhaA115 a n d DhaA, a n d consistent with
a true hydrogen-bond network (distances d7-d9 i n Table S l f ) .
11172 | Chem. Sci., 2 0 2 0 , 11, 11162-11178
Edge Article
Tunnel properties i n molecular dynamic simulations
The access tunnels were calculated d u r i n g the M D simulations
u s i n g C A V E R 3.02.2 1
T h i s allowed u s to assess h o w their
geometry a n d potential relevance changed i n c o m p a r i s o n w i t h
the static crystal structures. W e f o u n d that the tunnels i n
DhaA115 were still very narrow d u r i n g the M D simulations
(Table S2 a n d Fig. S8 a n d S9f), w i t h n o tunnels detected for the
selected probe size (0.7 A) i n a large fraction o f the trajectories.
W h e n tunnels were detected, they still h a d very l o w bottleneck
radii (BR-value) m o s t o f the time. For example, the m o s t defined
tunnel, p2b, showed a n average BR-value o f 0.84 ± 0.11 A . Very
importantly, however, these tunnels were able to occasionally
o p e n u p to significant values, e.g. tunnels p 2 b a n d p i showed
m a x i m u m BR-values o f 1.34 a n d 1.41 A .
The krypton-derivatized DhaA115 crystal structure was also
simulated ( R M S D plots i n Fig. S6f), a n d its tunnels showed
behavior very s i m i l a r to those i n the native DhaA115 structure,
i n terms o f the preferred tunnels a n d their m a i n properties. As
expected, this analysis showed that krypton-derivatization d i d
not alter the natural d y n a m i c a l properties o f DhaA115. W h e n
we looked at the a M D s , the access tunnels o f DhaA115 fluctuated
m o r e i n terms o f radius, length a n d topology (larger
standard deviations f r o m the m e a n values). S u c h behavior
could be expected f r o m a s i m u l a t i o n that promotes conformational
transitions, s u c h as a M D . T h e tunnels detected here also
h a d larger average BR-values t h a n i n the M D s , b u t m o r e
importantly, they displayed considerably larger BR-value
m a x i m a across the s i m u l a t i o n s (tunnels p4 a n d p i h a d
m a x i m a l BR-values o f 1.69 a n d 1.61 A respectively). W i t h
bottleneck radii o f this magnitude, these access tunnels c a n be
considered to be o p e n to the transit o f solvent a n d s m a l l ligands
(1.4 A is the m i n i m u m radius required for a water molecule to
pass through). These values are significantly larger t h a n the size
of the access tunnels f o u n d i n the crystal structures a n d
demonstrate the importance o f statistical analysis o f tunnels i n
dynamics rather t h a n d r a w i n g conclusions based solely o n
a single static crystallographic structure. O u r results show that
a l t h o u g h DhaA115 has a well-packed structure w i t h a n occluded
active site pocket, it is still able to o p e n occasionally a n d allow
the transport o f substrates a n d products. These findings explain
the ability o f DhaA115 to catalyze the dehalogenase reaction.
Finally, f r o m a c o m p a r i s o n o f DhaA115 w i t h the wild-type
D h a A a n d D h a A 3 1 , another active D h a A variant bearing m u t a tions
that significantly narrowed its t u n n e l s , 1 2
DhaA115 has the
narrowest tunnels, b o t h i n the respective crystal structures a n d
i n the M D s i m u l a t i o n s . A n o t h e r major difference a m o n g the
variants is i n the topology o f the relevant tunnels. W h i l e for
D h a A a n d D h a A 3 1 , the m o s t i m p o r t a n t was the m a i n p i t u n n e l ,
the slot p 2 t u n n e l became m o r e structurally relevant i n
DhaA115.
Discussion
I m p r o v i n g enzyme stability is o n e o f the major tasks i n
contemporary protein engineering. M a n y c o m p u t a t i o n a l tools
have been developed to m a k e rational predictions o f the effects
This journal is © The Royal Society of Chemistry 2 0 2 0
V i e w Article Online
Edge Article
of m u t a t i o n s o n protein stability.1 3
'2 5
"2 7
In-depth structural
u n d e r s t a n d i n g of these effects c a n help improve the accuracy of
computer algorithms.
Several distinct strategies have been employed to stabilize
proteins of the a/p-hydrolase fold family, namely: (i) structurebased
c o m p u t a t i o n a l approaches a n d i n f o r m e d mutagenesis
of flexible regions, (ii) sequence-based phylogenetic
approaches, a n d (iii) r a n d o m i z e d mutagenesis coupled w i t h
extensive library screening.2 8
Generally, stabilizing mutations
have been f o u n d to occur i n b o t h the cap a n d the catalytic
d o m a i n s , i n buried regions a n d surface-exposed areas.2 8
For
instance, de novo engineering of a disulfide b o n d physically
a n c h o r i n g the cap d o m a i n to the catalytic a/p-hydrolase d o m a i n
was successfully used for stabilization of l i p a s e s , 2 9 - 3 1
acetylcholine
esterase3 2
a n d the haloalkane dehalogenase D h l A . 3 3
Moreover, mutations l e a d i n g to enhanced interior p a c k i n g have
been r e p o r t e d , 3 4 - 3 6
as well as m u t a t i o n s introduced to decrease
flexibility a n d increase stability of the a/p-fold p r o t e i n s . 3 7 - 3 9
Plant esterase was stabilized to resist heat inactivation by
i n t r o d u c i n g proline residues into solvent-exposed loops.4 0
Several groups reported achieving increased stability of lipases
t h r o u g h single p o i n t substitutions e n a b l i n g the f o r m a t i o n of
new ionic interactions a n d salt b r i d g e s . 4 1 - 4 4
It was previously
s h o w n that n a r r o w i n g or b l o c k i n g access tunnels helps to
stabilize enzymes w i t h b u r i e d active sites a n d to increase their
resistance to organic co-solvents.1 9
'4 5
Re-engineering of the
access tunnels t h r o u g h five p o i n t m u t a t i o n s increased b o t h
catalytic activity toward 1,2,3-trichloropropane a n d t h e r m a l
stability i n D h a A 3 1 . 1 2
A l m o s t complete closure of the m a i n
t u n n e l while preserving the slot t u n n e l was observed i n the
stabilized m u t a n t D h a A 8 0 ( T m = 64.5 °C).1 9
The in-house FireProt server is a n automated c o m p u t a t i o n a l
tool c o m b i n i n g energy- a n d evolution-based approaches to
design highly heat-stable m u t a n t s . 1 3
T h e 11-point m u t a n t haloalkane
dehalogenase DhaA115, designed by FireProt, has the
highest thermostability of a l l the D h a A variants ever engineered
DhaA. However, the structural basis of this hyperstability was
poorly understood. In this w o r k we solved the high-resolution
structures of DhaA115 a n d c o m p a r e d t h e m w i t h those of the
w i l d type D h a A . Careful inspection of the DhaA115 structure
revealed that 9 out of the 11 stabilizing m u t a t i o n s are located i n
the secondary structure elements. T h e m u t a t i o n s designed by
a n evolution-based approach (E20S, F 8 0 R a n d A155P) participate
extensively i n the surface charge network, protein-solvent
interactions and/or rigidifying a solvent-exposed loop. W e
therefore conclude that newly-established protein-solvent
interactions o n the protein surface m i g h t be i m p o r t a n t factors
i n protecting the a/p-hydrolase core to stabilize the overall
protein fold. O u r structural data are thus i n agreement w i t h
previous observations by Beerens a n d co-workers,1 5
w h o have
s h o w n experimentally that stabilization by evolution-based
mutations is driven by b o t h entropy a n d enthalpy, the former
being difficult to predict f r o m force-field calculations.1 3
C o m p u t a t i o n a l p r e d i c t i o n tools s u c h as F o l d X 2 6
a n d Rosetta2 7
do not evaluate entropic contributions correctly due to underestimating
key factors s u c h as alternative protein
Chemical Science
conformations a n d specific interactions between a protein a n d
solvent molecules."
The r e m a i n i n g 8 mutations (C128F, T148L, A172I, C176F,
D198W, V219W, C 2 6 2 L a n d D266F) were inferred by a n energybased
a p p r o a c h . 1 3
It has been proposed that these mutations
s h o u l d reinforce h y d r o p h o b i c a n d aromatic interactions a n d
improve protein p a c k i n g . In general, our experimentally determ
i n e d DhaA115 structures c o n f i r m this proposal. W e observe
that the vast majority (7 out of 8) of the mutations cooperate
w i t h one another, s h o w i n g effects o n residue-to-residue p a c k i n g
a n d o n stabilization of the protein fold. A l l energy-deduced
mutations are h y d r o p h o b i c or aromatic (always sterically
bulkier t h a n the original residues); however this led to unexpected
structural effects. W e reveal that the replacement of
smaller residues w i t h a m i n o acids w i t h b u l k i e r side chains, e.g.,
V219W, C 2 6 2 L a n d D266F, leads to long-distance rearrangements
of the protein backbone, w h i c h were not predicted
i n the original c o m p u t a t i o n a l design. T h e backbone
rearrangements r e m a i n e d stable d u r i n g the M D simulations.
These unexpected structural effects led to the p r o d u c t i o n of the
so-called double-lock system: (i) they closed the active site
access gateways, (ii) the volumes of b o t h m a i n a n d slot enzyme
tunnels were reduced, a n d (iii) the active site was occluded. W e
t h i n k that the restricted tunnels are likely the major determinant
of the lower activity of DhaA115 at a temperature o p t i m a l
for D h a A . 1 3
E x p e r i m e n t a l tracking of the tunnels by kryptonderivatization
of DhaA115 crystals, supported by protein s i m u lations,
revealed that ligands c a n still be transported t h r o u g h
the tunnels as they c a n o p e n to a significant extent. W e expect
that this t u n n e l o p e n i n g w i l l be even m o r e p r o n o u n c e d at
higher temperatures, w h i c h w o u l d explain the shift i n the
temperature o p t i m u m to a higher range.
T a k e n together, our experimental a n d theoretical results
provide molecular insights into the engineered stability of
DhaA115 a n d the i m p a c t of introduced m u t a t i o n s o n functionally
i m p o r t a n t structural features of this hyperstable
enzyme. O u r data pave the way for s i m i l a r engineering efforts to
be applied to various protein catalysts f r o m the a/p-hydrolase
family, but also to other structurally unrelated protein folds.
Importantly, u n d e r s t a n d i n g of the structural basis of t h e r m a l
stability i n a protein designed by force-field calculations a n d
phylogenetic analysis provides valuable i n f o r m a t i o n for further
i m p r o v e m e n t of algorithms a n d c o m p u t a t i o n a l workflows for
achieving protein stabilization by rational protein d e s i g n . 4 6
One
of the lessons learned f r o m the structural analysis reported i n
this study is that the a c c u m u l a t i o n of experimentally verified
single-point m u t a t i o n s w i l l not lead to the structural basis of
stabilization observed w i t h D h a A 1 1 5 . M u l t i p l e substitutions
m u s t be introduced simultaneously to achieve cooperative
effects, like backbone changes, sealing of auxiliary access
tunnels, a n d f o r m a t i o n of occluded active sites. C o m p u t a t i o n a l
tools predicting the m u l t i p l e substitutions, s u c h as F i r e P r o t 1 3 1 4
a n d P R O S S 4 7
are already available for this type of design.
However, there is a space for further i m p r o v e m e n t of these
hybrid protein stabilization platforms. C o m p u t a t i o n a l design of
protein tunnels is underexplored strategy,4 8
'4 9
w h i c h c a n be
supported by the tools for calculation of access tunnels {e.g.,
This journal is © The Royal Society of Chemistry 2 0 2 0 Chem. Sei.. 2 0 2 0 , 11, 11162-11178 | 11173
V i e w Article Online
Chemical Science
C A V E R 2 1
) a n d ligands' passage {e.g., CaverWeb5 0
). The developm
e n t of novel algorithms a n d software tools for rational engineering
of protein loops is highly desirable, b u t still
challenging. N e w experimental d a t a 5 1
a n d better u n d e r s t a n d i n g
of structure-stability relationships are also essential premises
for developing m o r e reliable predictive m o d e l s by m a c h i n e
l e a r n i n g . 5 2
Experimental methods
Protein expression and purification
The DhaA115 enzyme was overproduced i n Escherichia coli, as
previously described. Briefly, DhaA115 was overexpressed i n E.
coli BL21(DE3) w i t h i n d u c t i o n by 0.5 m M IPTG at 20 °C for 16
hours. T h e cells were harvested by centrifugation at 11 806g at
4 °C for 10 m i n . T h e pellet was re-suspended i n purification
buffer A (500 m M N a C l , 10 m M imidazole, 20 m M p o t a s s i u m
phosphate buffer p H 7.5) a n d sonicated u s i n g a Sonic Dism
e m b r a t o r M o d e l 705 (Fisher Scientific, USA) i n 3 cycles, each
of 2 m i n (5 s pulse/5 s pause) w i t h a m p l i t u d e 5 0 % . D i s r u p t e d
cells were centrifuged at 21 OOOg at 4 °C for 1 h . His-tagged
DhaA115 protein was purified o n a Ni-chelating c o l u m n (NiN
T A Superflow cartridge) equilibrated i n purification buffer A.
The affinity-purified enzyme was eluted by a purification buffer
A supplemented w i t h 300 m M imidazole. T h e eluted protein
was further purified by size-exclusion chromatography o n
a H i L o a d 16/600 Superdex 200 gel filtration c o l u m n (GE
Healthcare) equilibrated i n G F buffer (50 m M N a C l , 10 m M Tris
p H 8.0). Peak fractions were pooled a n d concentrated w i t h an
A m i c o n Ultra centrifugal filter u n i t (Merck M i l l i p o r e Ltd) to
a final concentration of 11.5 m g m l - 1
. Protein concentration
was measured o n a DeNovixR® DS-11 Spectrophotometer
(DeNovix Inc., USA).
Crystallization, krypton-soaking and diffraction analysis
Diffraction-quality crystals of the DhaA115 enzyme were obtained
at 20 °C by m i x i n g equal volumes of DhaA115 (11.5 m g
m P 1
) w i t h reservoir s o l u t i o n c o m p o s e d of 1 8 - 2 4 % P E G 3350,
0.2 M K S C N a n d 0.1 M bis-tris propane ( p H 6.5), a n d crystallized
u s i n g the hanging-drop vapor diffusion technique. After 3-6
days, the crystals so grown were briefly transferred into reservoir
solution supplemented w i t h 2 2 % glycerol a n d flash-frozen i n
l i q u i d nitrogen.
Krypton derivatives were p r o d u c e d u s i n g 'soak a n d freeze'
methodology.2 2
T h e m e t h o d is a i m e d at d e c i p h e r i n g functional
tunnels i n proteins.5 3
'5 4
In practice, a crystal o b t a i n e d f r o m 1 : 1
protein (13.8 m g mP1
)/reservoir s o l u t i o n was fished out into
a capillary filled w i t h cryo-protective s o l u t i o n (24% P E G 3350,
0.2 M K S C N , 2 2 % glycerol a n d 0.1 M bis-tris propane p H 6.5).
The crystal was initially loaded into a pressure cell at a m b i e n t
pressure a n d temperature (294 K a n d 1 a t m respectively), i n
w h i c h the sample was t h e n pressurized i n a pure krypton gas
m e d i u m at 140 bar for 5 m i n u t e s . T h e n , still u n d e r pressure, the
crystal was directly flash-frozen i n the cell into the cold dense
krypton fluid phase w h i c h acts as a coolant. Finally, the pressure
was released, a n d the crystal was extracted f r o m the cell
11174 | Chem. Sci., 2 0 2 0 , 11, 11162-11178
Edge Article
a n d recovered i n l i q u i d nitrogen w i t h o u t b r e a k i n g the cryogenic
temperature c h a i n . A l l data were collected at the E S R F ID23-1
beamline (Grenoble, France)5 5
at a wavelength of 0.861 A (the
krypton X-ray absorption edge).
Structure determination, model building and refinement
The crystallographic data were processed u s i n g X D S 5 6
for
i n d e x i n g a n d integration a n d A i m l e s s 5 7
for merging. Initial
phases of DhaA115 were solved by molecular replacement u s i n g
P h a s e r 5 8
i m p l e m e n t e d i n the P h e n i x package.5 9
T h e structure of
D h a A (PDB: 4 H Z G ) was employed as a search m o d e l for
replacement i n DhaA115 m o n o m e r i c structures. T h e refinem
e n t was carried out w i t h several automated cycles i n the
phenix.refine p r o g r a m 6 0
a n d m a n u a l m o d e l b u i l d i n g was performed
i n C o o t . 6 1
Crystal structures of native a n d krypton
soaked DhaA115 were solved to resolutions of 1.6 A a n d 1.55 A
i n a m o n o c l i n i c P 1 2 ] l space group a n d P 2 1 2 1 2 1 respectively. The
final models were validated u s i n g tools provided by C o o t 6 1
a n d
M o l p r o b i t y . 6 2
V i s u a l i z a t i o n of structural data was done w i t h
P y M O L . 6 3
A t o m i c coordinates a n d structure factors of the native
DhaA115 a n d krypton-derivatized DhaA115 enzymes were
deposited i n the Protein Data B a n k u n d e r the P D B codes 6SP5
a n d 6SP8.
Small-angle X-ray scattering (SAXS)
The SAXS data sets were collected u s i n g the BioSAXS-1000,
Rigaku at C E I T E C (Brno, C z e c h Republic). Data were collected
at 293 K w i t h a focused (confocal OptiSAXS optic, Rigaku) C u K a
X-ray (1.54 A). T h e sample to detector (PILATUS 100K, Dectris)
distance was 0.48 m covering a scattering vector (q = 47tsin((9)/A)
range f r o m 0.008 to 0.6 A - 1
. Size exclusion buffer (41 m M
K 2 H P 0 4 , 9 m M K H 2 P 0 4 , p H 7.5) was used for the b l a n k
measurement. DhaA115 protein samples were measured at
concentrations of 8.5, 6.3, 4.3 a n d 2.2 m g m l - 1
. Six separate
images were collected for buffer a n d sample (5 m i n exposure
per image, 30 m i n total exposure). R a d i a l averaging, data
reduction a n d buffer subtractions were performed u s i n g SAXSLab3.0.0rl,
R i g a k u . Six i n d i v i d u a l scattering curves (5 m i n
exposures) were c o m p a r e d to check radiation damage a n d
averaged. Integral structural parameters were determined u s i n g
PRIMUS/qt ATSAS v.2.8.4.6 4
Data points before the G u i n i e r
region were truncated. I n d i v i d u a l scattering curves f r o m the
concentration series were m a n u a l l y merged for further analysis.
The ah initio m o d e l i n g for s u p e r i m p o s i t i o n w i t h the atomic
m o d e l was performed by D A M M I N ATSAS v.2.8.4, w i t h the
c o m p u t a t i o n m o d e set to "slow" a n d a l l other parameters kept
as default. E v a l u a t i o n of s o l u t i o n scattering a n d fitting to
experimental scattering curves was performed u s i n g C R Y S O L
ATSAS v.2.8.4; automatic constant subtraction was allowed a n d
other parameters were kept as default. S u p e r i m p o s i t i o n of the
atomic a n d ah initio models was performed by S U P C O M B ATSAS
v.2.8.4. Small-angle X-ray scattering datasets, experiment
details, the atomic m o d e l a n d fits have been deposited i n the
S m a l l Angle Scattering Biological Data B a n k (www.sasbdb.org)6 5
as entry S A S D H P 7 .
This journal is © The Royal Society of Chemistry 2 0 2 0
V i e w Article Online
Edge Article
Structural bioinformatics tools
R M S D values were calculated u s i n g pairwise structural alignm
e n t o n the D A L I server.6 6
Structural superposition was performed
u s i n g t h e secondary structure m a t c h i n g (SSM)
superimpose tool i n Coot.6 7
D i m e r interface a n d b u r i e d surface
areas were calculated by the PISA t o o l . 1 7
Analysis o f residue-toresidue
interactions i n t h e crystal structure was done u s i n g
Protein Interactions Calculator (PIC)1 8
w i t h default parameters.
Molecular dynamics simulations and analysis
The three-dimensional structures o f DhaA115 were used as
obtained f r o m the X-ray diffraction analysis, for b o t h native a n d
krypton-soaked crystals. T h e solvent, crystallization molecules
a n d krypton atoms were removed, a n d the double side chains
were corrected to retain only the m o s t populated c o n f o r m a t i o n
u s i n g the pdb4amber m o d u l e o f A m b e r T o o l s 1 4 . 6 8
H y d r o g e n
atoms were predicted u s i n g t h e H++ server,6 9
calculated i n
i m p l i c i t solvent at p H 7.5, 0.1 M salinity, a n d internal a n d
external dielectric constants o f 10 a n d 80 respectively. T h e
original crystallization solvent was added a n d t h e t L E a P
p r o g r a m i n A m b e r T o o l s 14 was used to prepare the topology
a n d coordinates files. F o r this, t h e force field f f l 4 S B 7 0
w a s
defined, N a +
a n d C P ions were added i n order to neutralize the
system a n d achieve a 0.10 M concentration o f N a C l , a n d a n
octagonal box o f T I P 3 P 7 1
water molecules w i t h the edges at least
10 A away f r o m the protein atoms was added.
The m o l e c u l a r dynamics ( M D ) s i m u l a t i o n s were carried o u t
w i t h the p m e m d . c u d a 7 2
' 7 3
m o d u l e o f A M B E R 14.6 8
I n total, five
m i n i m i z a t i o n steps a n d twelve steps o f e q u i l i b r a t i o n dynamics
were performed prior to t h e p r o d u c t i o n M D . T h e first four
m i n i m i z a t i o n steps, c o m p o s e d o f 2500 cycles o f steepest
descent followed b y 7500 cycles o f conjugate gradient, were
performed as follows: (i) i n the first step, a l l the atoms o f the
protein a n d ligand were restrained w i t h 500 kcal m o P 1
A - 2
h a r m o n i c force constant; (ii) i n the following three, only the
backbone atoms o f the protein a n d heavy atoms o f the ligand
were restrained w i t h , respectively, 500, 125, a n d 25 kcal m o P 1
A ~ 2
force constant. A fifth m i n i m i z a t i o n step, c o m p o s e d o f 5000
cycles o f steepest descent a n d 15 000 cycles o f conjugate
gradient, w a s performed w i t h o u t a n y restraints. T h e subsequent
M D s i m u l a t i o n s employed periodic boundary conditions,
the particle m e s h E w a l d m e t h o d for treatment o f the l o n g range
interactions beyond the 10 A cutoff,7 4
the S H A K E a l g o r i t h m 7 5
to
constrain the b o n d s involving the hydrogen atoms, the Langev
i n thermostat w i t h c o l l i s i o n frequency 1.0 p s - 1
, a n d a time step
of 2 fs. E q u i l i b r a t i o n dynamics were performed i n twelve steps:
(i) 20 p s o f gradual heating f r o m 0 to 310 K, u n d e r constant
volume, restraining t h e protein atoms a n d ligand w i t h
200 kcal m o P 1
A ~ 2
h a r m o n i c force constant; (ii) ten M D s o f 400
ps each, at constant pressure (1 bar) a n d constant temperature
(310 K), w i t h gradually decreasing restraints o n the backbone
atoms o f t h e protein a n d heavy atoms o f t h e l i g a n d w i t h
h a r m o n i c force constants o f 150,100, 75, 50, 2 5 , 1 5 , 1 0 , 5 , 1 , a n d
0.5 kcal m o P 1
A ~ 2
; (iii) 400 ps o f unrestrained M D at the same
conditions as t h e previous restrained M D s . T h e energy a n d
coordinates were saved every 10 ps. T h e p r o d u c t i o n M D s were
This journal is © T h e Royal Society o f Chemistry 2 0 2 0
Chemical Science
r u n for 100 n s u s i n g the same settings employed i n the last
equilibration step a n d performed i n duplicate for each system.
Accelerated M D (aMD) s i m u l a t i o n s were performed for each
system u s i n g the p m e m d . c u d a 7 2
' 7 3
m o d u l e o f A M B E R 14.6 8
T h e
systems were prepared a n d m i n i m i z e d as previously described
for the classical M D s , u s i n g the f f l 4 S B 7 0
force field. D u a l energy
boosts were applied to the torsional (V11
*1
) a n d total potential
(V1 0 1
) energy. T h e average d i h e d r a l ( V 0
d l h
) a n d total potential
[V0
tot
] energies o f each system were extracted f r o m the first 10 n s
of p r o d u c t i o n M D , a n d were used to calculate t h e respective
energy thresholds (E) a n d acceleration factors (a), as previously
described.7 6
i ^ 1 1 1
was set as 3.5 kcal m o P 1
p e r protein residue
above V o d l h
, a n d the c o r r e s p o n d i n g acceleration factor, a d l h
, was
set as 1/5 o f that difference; the total potential energy threshold,
Etot
, was defined as 0.2 kcal m o P 1
per a t o m o f the system above
V0
tot
, a n d the respective acceleration factor, a t o t
, w a s set as the
difference between those two energies. C a l c u l a t i n g the parameters
i n this way always yielded values o f E41
*1
ca. 2 7 % above the
respective V o d l h
. T h e a M D s were r u n w i t h o u t any restraints, w i t h
calculation steps o f 2 fs, saving the energy a n d coordinate every
10 ps. These s i m u l a t i o n s were r u n i n duplicate for 100 ns. T h e
a M D s were performed as a complementary m e t h o d to sample
the c o n f o r m a t i o n a l space equivalent to longer t i m e scales,
estimated at several orders o f m a g n i t u d e greater t h a n those o f
the M D s (between the us a n d m s time scales).2 3
'7 7
The trajectories were analyzed u s i n g the cpptraj7 8
m o d u l e o f
A m b e r T o o l s 14, a n d visualized u s i n g P y M O L 6 3
a n d V M D I.9.I.7 9
The s i m u l a t i o n s o f each type were c o m b i n e d to create a single
one u s i n g cpptraj,7 8
a n d aligned to the respective crystal structures
by m i n i m i z i n g the root-mean-square deviation (RMSD) o f
the backbone atoms, e x c l u d i n g the very flexible t e r m i n a l residues
o f each c h a i n (4-6 t e r m i n a l residues).
Access tunnel calculations
C A V E R 3.02,2 1
was used to calculate a n d cluster the tunnels i n
the crystal structures, aggregated M D a n d a M D s i m u l a t i o n s o f
DhaA115, a n d the previously reported analogous s i m u l a t i o n s o f
D h a A 1 5
a n d D h a A 3 1 . 7 6
D u r i n g the s i m u l a t i o n s the tunnels were
calculated for every 10 ps-spaced snapshot u s i n g a probe radius
of 0.7 A (0.5 A for the crystal structures o f DhaA115), a shell
radius o f 3 A a n d a shell d e p t h o f 4 A . T h e starting p o i n t for the
t u n n e l calculation was defined b y the geometric center o f the
carboxylic oxygen atoms o f the catalytic D106. T h e clustering
was performed b y the average-link hierarchical M u r t a g h algor
i t h m , w i t h a weighting coefficient o f 1 a n d clustering threshold
of 3.5 A. Approximate clustering was allowed only w h e n the total
n u m b e r o f tunnels was greater t h a n 20 000, a n d i t was performed
u s i n g 20 t r a i n i n g clusters.
Author contributions
K. C . a n d K. M . prepared the protein samples for crystallization,
performed initial crystallization screenings, a n d o p t i m i z e d
crystallization hits. K. M . prepared t h e protein samples for
SAXS. K. C , P. C. a n d M . M . carried out krypton-derivatization o f
crystals a n d collected diffraction data. K . C , P. C . a n d M . M .
Chem. Sci., 2 0 2 0 , 11, 11162-11178 | 11175
Chemical Science
solved the protein crystal structures. S. M . M . a n d D . B . performed
M D analyses. M . M . a n d J . D. designed the project,
supervised research a n d interpreted data. K. M . a n d M . M . wrote
the m a n u s c r i p t w i t h contributions f r o m a l l authors. A l l authors
have given approval to the final version of the manuscript.
Data and code availability
A t o m i c coordinates a n d structural factors have been deposited
i n the Protein Data B a n k (www.wwpdb.org) u n d e r P D B accession
codes: 6SP5 a n d 6SP8. SAXS datasets, experiment details,
atomic m o d e l a n d fits have been deposited i n the S m a l l Angle
Scattering Biological Data B a n k (www.sasbdb.org)6 5
as entry
S A S D H P 7 . A u t h o r s w i l l release the atomic coordinates a n d
experimental data u p o n article p u b l i c a t i o n .
Conflicts of interest
The authors declare n o c o m p e t i n g financial interest.
Acknowledgements
W e are grateful to Zbynek Prokop a n d A n t o n i n K u n k a (Masaryk
University, Brno) for their careful r e a d i n g of the m a n u s c r i p t .
The authors w o u l d like to express their thanks to the C z e c h
M i n i s t r y of E d u c a t i o n (grants 02.1.01/0.0/0.0/18_046/0015975,
L M 2 0 1 8 1 2 1 , LM2015047) a n d to the E u r o p e a n U n i o n (857560,
720776 a n d 814418). T h i s project has received f u n d i n g f r o m the
E u r o p e a n U n i o n ' s H o r i z o n 2020 research a n d i n n o v a t i o n prog
r a m m e u n d e r the M a r i e Sklodowska-Curie grant agreement
No. 792772. M . M . acknowledges financial support f r o m G A M U
of the M a s a r y k University (MUNI/H/1561/2018). T h e
c o m p u t a t i o n a l resources were s u p p l i e d by the project "eInfrastruktura
C Z " (e-INFRA LM2018140) provided w i t h i n the
p r o g r a m Projects of Large Research, Development a n d
Innovations Infrastructures. CIISB research infrastructure
project (LM2018127) is acknowledged for financial support of
the measurements at the X-ray Diffraction a n d Bio-SAXS Core
Facility. W e t h a n k T o m a s K l u m p l e r ( C E I T E C - M U , B r n o , C z e c h
Republic) for h i s assistance d u r i n g SAXS data collection a n d
processing. T h e crystallographic experiments were performed
o n b e a m l i n e ID23-1 at the E u r o p e a n Synchrotron Facility
(ESRF), Grenoble, France. W e are grateful to the m e m b e r s o f the
ESRF synchrotron for the use of their b e a m l i n e facilities a n d for
help d u r i n g data collection.
References
1 P. K. R o b i n s o n , Essays Biochem., 2015, 59, 1-41.
2 R. Singh, M . K u m a r , A. M i t t a l a n d P. K. M e h t a , 3 Biotech,
2016, 6, 174.
3 C . J . Y e o m a n , Y. H a n , D . D o d d , C . M . Schroeder, R. I. M a c k i e
a n d I. K. O. C a n n , Advances in applied microbiology, Elsevier,
2010, vol. 70, p p . 1-55.
4 R. Kazlauskas, Chem. Soc. Rev., 2018, 47, 9026-9045.
5 C . N . Pace, J . M . Scholtz a n d G . R. Grimsley, FEBS Lett., 2014,
588, 2177-2184.
V i e w Article Online
Edge Article
6 S. S. Strickler, A. V. G r i b e n k o , A. V. G r i b e n k o , T. R. Keiffer,
J . T o m l i n s o n , T. Reihle, V . V . Loladze a n d
G. I. M a k h a t a d z e , Biochemistry, 2006, 45, 2761-2766.
7 C . N . Pace, H . F u , K. L. Fryar, J . L a n d u a , S. R. Trevino,
B. A. Shirley, M . M . H e n d r i c k s , S. I i m u r a , K. Gajiwala,
J. M . Scholtz a n d G . R. G r i m s l e y , / . Mol. Biol, 2011, 408,
514-528.
8 J . D a m b o r s k y , E . Rorije, A . Jesenská, Y. Nagata, G . K l o p m a n
a n d W . J . G . M . Peijnenburg, Environ. Toxicol. Chem., 2001,
20, 2681-2689.
9 J . Damborsky, R. Chaloupková, M . Pavlova, E . Chovancova
a n d J . Brezovsky, i n Handbook of Hydrocarbon and Lipid
Microbiology, Springer, Berlin, Heidelberg, 2010, p p . 1 0 8 1 -
1098.
10 M . Petrek, M . Otyepka, P. Banas, P. Kosinová, J . Koča a n d
J. Damborsky, BMC Bioinf, 2006, 7, 316.
11 R. Chaloupková, J . Sýkorova, Z. Prokop, A. Jesenská,
M . M o n i n c o v a , M . Pavlova, M . T s u d a , Y. Nagata a n d
J. D a m b o r s k a , / . Biol. Chem., 2003, 278, 52622-52628.
12 M . Pavlova, M . Klvana, Z. Prokop, R. Chaloupková, P. Banas,
M . Otyepka, R. C. W a d e , M . T s u d a , Y. Nagata a n d
J. Damborsky, Nat. Chem. Biol, 2009, 5, 727-733.
13 D. Bednař, K. Beerens, E . Šebestová, J . B e n d i , S. Khare,
R. Chaloupková, Z. Prokop, J . Brezovsky, D. Baker a n d
J. Damborsky, PLoS Comput. Biol, 2015, 11, el004556.
14 M . M u s i l , J . Stourac, J . B e n d i , J . Brezovsky, Z. Prokop,
J. Z e n d u l k a , T. M a r t i n e k , D . B e d n a r a n d J . D a m b o r s k y ,
Nucleic Acids Res., 2017, 45, W 3 9 3 - W 3 9 9 .
15 K. Beerens, S. M a z u r e n k o , A. K u n k a , S. M . M a r q u e s ,
N . H a n s e n , M . M u s i l , R. Chaloupková, J . W a t e r m a n ,
J. Brezovsky, D . Bednar, Z. Prokop a n d J . Damborsky, ACS
Catal, 2018, 8, 9420-9428.
16 D. I. Svergun,/. Appl. Crystallogr., 1992, 25, 495-503.
17 E . K r i s s i n e l a n d K. Henrick,/. Mol. Biol, 2007, 372, 774-797.
18 K. G . T i n a , R. B h a d r a a n d N . Srinivasan, Nucleic Acids Res.,
2007, 35, W 4 7 3 - W 4 7 6 .
19 T. Koudelakova, R. Chaloupková, J . Brezovsky, Z. Prokop,
E. Šebestová, M . Hesseler, M . K h a b i r i , M . Plevaka,
D. K u l i k , I. K u t a Smatanova, P. Řezačova, R. Ettrich,
U. T. Bornscheuer a n d J . D a m b o r s k y , Angew. Chem., Int.
Ed., 2013, 52, 1959-1963.
20 V. Liškova, D. Bednar, T. P r u d n i k o v a , P. Řezačova,
T. Koudelakova, E . Šebestová, I. K. Smatanova,
J. Brezovsky, R. Chaloupková a n d J . D a m b o r s k y ,
ChemCatChem, 2015, 7, 648-659.
21 E . Chovancova, A . Pavelka, P. Beneš, O. Strnad, J . Brezovsky,
B. Kozlíkova, A. G o r a , V. Sustr, M . Klvana, P. Medek,
L. B i e d e r m a n n o v a , J . Sochor a n d J . Damborsky, PLoS
Comput. Biol, 2012, 8, el002708.
22 B. Lafumat, C. M u e l l e r - D i e c k m a n n , G . Leonard, N . C o l l o c ' h ,
T. Prangé, T. G i r a u d , F. Dobias, A. Royant, P. v a n der L i n d e n
a n d P. Carpentier,/. Appl. Crystallogr., 2016, 49, 1478-1487.
23 P. R. L. M a r k w i c k a n d J . A. M c C a m m o n , Phys. Chem. Chem.
Phys., 2011, 13, 20053.
24 D. H a m e l b e r g , J . M o n g a n a n d J . A . M c C a m m o n , /. Chem.
Phys., 2004, 120, 11919-11929.
11176 I Chem. Sci., 2 0 2 0 , 11, 11162-11178 This journal is © The Royal Society of Chemistry 2 0 2 0
V i e w Article Online
Edge Article
25 Y. D e h o u c k , J . M . Kwasigroch, D. Gilis a n d M . R o o m a n , BMC
Bioinfi, 2011, 12, 151.
26 J . Schymkowitz, J. Borg, F. Stricher, R. Nys, F. Rousseau a n d
L. Serrano, Nucleic Acids Res., 2005, 33, W 3 8 2 - W 3 8 8 .
27 C. A. Röhl, C. E. M . Strauss, K. M . S. M i s u r a a n d D. Baker,
Methods in Enzymology, Elsevier, 2004, vol. 383, pp. 66-93.
28 B. J . Jones, H . Y. L i m , J . H u a n g a n d R. J . Kazlauskas,
Biochemistry, 2017, 56, 6521-6532.
29 Z. H a n , S. H a n , S. Z h e n g a n d Y. L i n , Appl. Microbiol.
Biotechnol., 2009, 85, 117-126.
30 Q. A. T. Le, J . C. Joo, Y. J . Yoo a n d Y. H . K i m , Biotechnol.
Bioeng., 2012, 109, 867-876.
31 X.-W. Y u , N.-J. T a n , R. X i a o a n d Y. X u , PLoS One, 2012, 7,
e46388.
32 O. Siadat, A. Lougarre, L. L a m o u r o u x , C. Ladurantie a n d
D. Fournier, BMC Biochem., 2006, 7, 12.
33 M . G . Pikkemaat, A. B. M . L i n s s e n , H . J . C. Berendsen a n d
D. B. Janssen, Protein Eng., Des. Sei, 2002, 15, 185-192.
34 H . S. Y u n , H . J . Park, J . C. Joo a n d Y. J . Yoo,/. Ind. Microbiol.
Biotechnol, 2013, 40, 1223-1229.
35 R. K u m a r , R. S i n g h a n d J. Kaur,/. Mol. Catal. B: Enzym., 2013,
97, 243-251.
36 S.-B. Z h a n g a n d Z.-L. W u , Bioresour. Technol, 2011, 102,
2093-2096.
37 Z.-J. L u a n , H.-L. Y u , B.-D. M a , Y.-K. Q i , Q. C h e n a n d J . - H . X u ,
Ind. Eng. Chem. Res., 2016, 55, 12167-12172.
38 G . Y a n , S. C h e n g , G . Zhao, S. W u , Y. L i u a n d W . S u n ,
Biotechnol. Lett, 2003, 25, 1041-1047.
39 J . Z h a n g , Y. L i n , Y. S u n , Y. Ye, S. Z h e n g a n d S. H a n , Enzyme
Microb. Technol, 2012, 50, 325-330.
40 J . H u a n g , B. J . Jones a n d R. J . Kazlauskas, Biochemistry, 2015,
54, 4330-4341.
41 H . J . Park, K. Park, Y. H . K i m a n d Y. J . Yoo,/. Biotechnol,
2014, 192, 66-70.
42 R. R u s l a n , R. N . Z. R. A. R a h m a n , T. C. Leow, M . S. M . A l i ,
M . Basri a n d A. B. Salleh, Int. J. Mol. Sei, 2012, 13, 943-960.
43 P. K. Sharma, R. K u m a r , R. K u m a r , O. M o h a m m a d , R. Singh
a n d J . K a u r , Gene, 2012, 491, 264-271.
44 J.-P. W u , M . L i , Y. Z h o u , L.-R. Y a n g a n d G . X u , Biotechnol.
Lett, 2015, 37, 403-407.
45 S. G i h a z , M . Kanteev, Y. Pazy a n d A. F i s h m a n , Appl. Environ.
Microbiol, 2018, 84(23), e02143-18.
46 M . M u s i l , H . Konegger, J . H o n , D. B e d n a r a n d J . D a m b o r s k y ,
ACS Catal, 2019, 9, 1033-1054.
47 A. Goldenzweig, M . G o l d s m i t h , S. E. H i l l , O. G e r t m a n ,
P. L a u r i n o , Y. A s h a n i , O. D y m , T. Unger, S. Albeck,
J . Prilusky, R. L. L i e b e r m a n , A. A h a r o n i , I. S i l m a n ,
J . L. S u s s m a n , D. S. Tawfik a n d S. J . F l e i s h m a n , Mol. Cell,
2016, 63, 337-346.
48 N . Kreß, J . M . Halder, L. R. R a p p a n d B. Hauer, Curr. Opin.
Chem. Biol, 2018, 47, 109-116.
49 P. K o k k o n e n , D. Bednar, G . Pinto, Z. Prokop a n d
J . Damborsky, Biotechnol. Adv., 2019, 37, 107386.
50 J . Stourac, O. Vavra, P. K o k k o n e n , J . Filipovic, G . Pinto,
J . Brezovsky, J . D a m b o r s k y a n d D. Bednar, Nucleic Acids
Res., 2019, 47, W 4 1 4 - W 4 2 2 .
Chemical Science
51 S. M a z u r e n k o , ChemCatChem, 2020, 12, D O I : 10.1002/
cctc.202000933.
52 S. M a z u r e n k o , Z. Prokop a n d J . Damborsky, ACS Catal, 2020,
10, 1210-1223.
53 N . C o l l o c ' h , P. Carpentier, L. C. M o n t e m i g l i o , B. Vallone a n d
T. Prange, Biophys. /., 2017, 113, 2199-2206.
54 J . K a l m s , A. S c h m i d t , S. Frielingsdorf, P. v a n der L i n d e n ,
D. von Stetten, O. Lenz, P. Carpentier a n d P. Scheerer,
Angew. Chem., Int. Ed., 2016, 55, 5586-5590.
55 D. Nurizzo, T. M a i r s , M . Guijarro, V. Rey, J . Meyer,
P. Fajardo, J . Chavanne, J.-C. Biasci, S. McSweeney a n d
E. M i t c h e l l , / Synchrotron Radial, 2006, 13, 227-238.
56 W . K a b s c h , Acta Crystallogr., Sect. D: Biol. Crystallogr., 2010,
66, 125-132.
57 P. R. Evans a n d G . N . M u r s h u d o v , Acta Crystallogr., Sect. D:
Biol. Crystallogr., 2013, 69, 1204-1214.
58 A. J . M c C o y , R. W . Grosse-Kunstleve, P. D. A d a m s ,
M . D. W i n n , L. C. Storoni a n d R. J . Read, /. Appl.
Crystallogr., 2007, 40, 658-674.
59 P. D. A d a m s , P. V. Afonine, G . Bunköczi, V. B. C h e n ,
I. W . Davis, N . Echols, J . J . H e a d d , L.-W. H u n g ,
G. J . K a p r a l , R. W . Grosse-Kunstleve, A. J . M c C o y ,
N . W . Moriarty, R. Oeffner, R. J . Read, D. C. R i c h a r d s o n ,
J. S. R i c h a r d s o n , T. C. Terwilliger a n d P. H . Zwart, Acta
Crystallogr., Sect. D: Biol. Crystallogr., 2010, 66, 213-221.
60 P. V. Afonine, R. W . Grosse-Kunstleve, N . Echols, J . J . H e a d d ,
N . W . Moriarty, M . M u s t y a k i m o v , T. C. Terwilliger,
A. Urzhumtsev, P. H . Zwart a n d P. D. A d a m s , Acta
Crystallogr., Sect. D: Biol. Crystallogr., 2012, 68, 352-367.
61 P. E m s l e y a n d K. Cowtan, Acta Crystallogr., Sect. D: Biol.
Crystallogr., 2004, 60, 2126-2132.
62 C. J . W i l l i a m s , J . J . H e a d d , N . W . Moriarty, M . G . Prisant,
L. L. V i d e a u , L. N . Deis, V. V e r m a , D. A. Keedy,
B. J . H i n t z e , V. B. C h e n , S. J a i n , S. M . Lewis,
W . B. A r e n d a l l , J . Snoeyink, P. D. A d a m s , S. C. Lovell,
J. S. R i c h a r d s o n a n d D. C. R i c h a r d s o n , Protein Sei., 2018,
27, 293-315.
63 The PyMOL Molecular Graphics System Version 2.0,
Schrödinger, L L C , 2014.
64 D. Franke, M . V. Petoukhov, P. V. Konarev, A. Panjkovich,
A. T u u k k a n e n , H . D. T. Mertens, A. G . K i k h n e y ,
N . R. Hajizadeh, J . M . F r a n k l i n , C. M . Jeffries a n d
D. I. S v e r g u n , / Appl. Crystallogr., 2017, 50, 1212-1225.
65 E. V a l e n t i n i , A. G . K i k h n e y , G . Previtali, C. M . Jeffries a n d
D. I. Svergun, Nucleic Acids Res., 2015, 43, D 3 5 7 - D 3 6 3 .
66 L. H o l m a n d P. Rosenström, Nucleic Acids Res., 2010, 38,
W 5 4 5 - W 5 4 9 .
67 E. K r i s s i n e l a n d K. H e n r i c k , Acta Crystallogr., Sect. D: Biol.
Crystallogr., 2004, 60, 2256-2268.
68 D. A. Case, V. B a b i n , J . T. B e r r y m a n , R. M . Betz, Q. C a i ,
S. Cerutti, T. E. C h e a t h a m III, T. A. D a r d e n , R. E. D u k e ,
H . G o h l k e , A. W . Goetz, S. Gusarov, N . Homeyer,
P. J a n o w s k i , J . Kaus, I. Kolossväry, A. Kovalenko, T. S. Lee,
S. L e G r a n d , T. L u c h k o , R. L u o , B. M a d e j , K. M . M e r z ,
F. Paesani, D. R. Roe, A. Roitberg, C. Sagui, R. SalomonFerrer,
G . Seabra, C. L. S i m m e r l i n g , W . S m i t h , J . Swails,
R. C. Walker, J . W a n g , R. M . Wolf, X . W u a n d
This journal is © The Royal Society of Chemistry 2 0 2 0 Chem. Sei.. 2 0 2 0 , 11, 11162-11178 | 11177
Chemical Science
V i e w Article Online
Edge Article
P. A. K o l l m a n , AMBER 14, University of California, San
Francisco, 2014.
69 J . C. G o r d o n , J . B . Myers, T. Folta, V . Shoja, L. S. H e a t h a n d
A. Onufriev, Nucleic Acids Res., 2005, 33, W 3 6 8 - W 3 7 1 .
70 J . A. M a i e r , C . M a r t i n e z , K. Kasavajhala, L. W i c k s t r o m ,
K. E . H a u s e r a n d C . S i m m e r l i n g , / . Chem. Theory Comput,
2015, 11, 3696-3713.
71 W . L. Jorgensen, J . C h a n d r a s e k h a r , J . D . M a d u r a ,
R. W . Impey a n d M . L. K l e i n , / . Chem. Phys., 1983, 79, 9 2 6 -
935.
72 A . W . Götz, M . J . W i l l i a m s o n , D . X u , D . Poole, S. Le G r a n d
a n d R. C. Walker,/. Chem. Theory Comput, 2012, 8, 1 5 4 2 -
1555.
73 S. Le G r a n d , A. W . Götz a n d R. C . Walker, Comput. Phys.
Commun., 2013, 184, 374-380.
74 T. D a r d e n , D. Y o r k a n d L. P e d e r s e n , / Chem. Phys., 1993, 98,
10089-10092.
75 J.-P. Ryckaert, G . Ciccotti a n d H . J . C. B e r e n d s e n , / Comput.
Phys., 1977, 23, 327-341.
76 S. M . Marques, Z. Dunajova, Z. Prokop, R. C h a l o u p k o v a ,
J. Brezovsky a n d J . D a m b o r s k y , / . Chem. Inf. Model, 2017,
57, 1970-1989.
77 L. C . T. Pierce, R. Salomon-Ferrer, C. A. F. de Oliveira,
J. A. M c C a m m o n a n d R. C. Walker, /. Chem. Theory
Comput, 2012, 8, 2997-3002.
78 D. R. Roe a n d T. E . C h e a t h a m , /. Chem. Theory Comput.,
2013, 9, 3084-3095.
79 W . H u m p h r e y , A. Dalke a n d K. Schulten, /. Mol. Graphics,
1996, 14, 33-38.
11178 I Chem. Sei.. 2 0 2 0 , 11, 11162-11178 This journal is © The Royal Society of Chemistry 2 0 2 0