Solubilizing and Stabilizing Proteins in Anhydrous Ionic Liquids through Formation of Protein-Polymer Surfactant Nanoconstructs

Article abstract of DOI:10.1021/jacs.5b13425

Nonaqueous biocatalysis is rapidly becoming a desirable tool for chemical and fuel synthesis in both the laboratory and industry. Similarly, ionic liquids are increasingly popular anhydrous reaction media for a number of industrial processes. Consequently, the use of enzymes in ionic liquids as efficient, environment-friendly, commercial biocatalysts is highly attractive. However, issues surrounding the poor solubility and low stability of enzymes in truly anhydrous media remain a significant challenge. Here, we demonstrate for the first time that engineering the surface of a protein to yield protein-polymer surfactant nanoconstructs allows for dissolution of dry protein into dry ionic liquids. Using myoglobin as a model protein, we show that this method can deliver protein molecules with near native structure into both hydrophilic and hydrophobic anhydrous ionic liquids. Remarkably, using temperature-dependent synchrotron radiation circular dichroism spectroscopy to measure half-denaturation temperatures, our results show that protein stability increases by 55 °C in the ionic liquid as compared to aqueous solution, pushing the solution thermal denaturation beyond the boiling point of water. Therefore, the work presented herein could provide a platform for the realization of biocatalysis at high temperatures or in anhydrous solvent systems.

Full text of DOI:10.1021/jacs.5b13425

Subscriber access provided by University of Florida | Smathers Libraries
Article
Solubilizing and stabilizing proteins in anhydrous ionic liquids
through formation of protein-polymer surfactant nano-constructs
Alex P. S. Brogan, and Jason P. Hallett
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.5b13425 • Publication Date (Web): 14 Mar 2016
Downloaded from http://pubs.acs.org on March 16, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 10
Journal of the American Chemical Society
1
2
3
4
5
6
Solubilizing and stabilizing proteins in anhydrous ionic liq-
uids through formation of protein-polymer surfactant nano-
constructs.
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
†
†*
Alex. P. S. Brogan , Jason P. Hallett
† Department of Chemical Engineering, Imperial College, London, UK, SW7 2AZ
ABSTRACT: Nonaqueous biocatalysis is rapidly becoming a desirable tool for chemical and fuel synthesis in both the
laboratory and in industry. Similarly, ionic liquids are an increasingly popular anhydrous reaction media for a number of
industrial processes. Consequently, the use of enzymes in ionic liquids as efficient, environment-friendly, commercial
biocatalysts is highly attractive. However, issues surrounding the poor solubility and low stability of enzymes in truly an-
hydrous media remain a significant challenge. Here, we demonstrate for the first time, that engineering the surface of a
protein to yield protein-polymer surfactant nanoconstructs allows for dissolution of dry protein into dry ionic liquids.
Using myoglobin as a model protein, we show that this method can deliver protein molecules with near native structure
into both hydrophilic and hydrophobic anhydrous ionic liquids. Remarkably, using temperature dependent SRCD spec-
troscopy to measure half denaturation temperatures, our results show that protein stability increases by 55 °C in the ionic
liquid as compared to aqueous solution, pushing the solution thermal denaturation beyond the boiling point of water.
Therefore, the work presented herein could provide a platform for the realization of biocatalysis at high temperatures or
in
anhydrous
solvent
systems.
Introduction
There have been reported a number of studies concern-
7–19
ing proteins and enzymes in ionic liquids.
Despite this
Ionic liquids, organic salts with melting temperatures
typically below 100 °C, are becoming increasingly popular
as promising solvents for many industrial processes and
interest, enzymes are poorly soluble in ionic liquids and
tend to be highly unstable in the absence of water. A
number of strategies are often employed to counteract
1
,2
7,12
various other applications. The major draw for using
ionic liquids is their attractive and highly tuneable chemi-
cal and physical properties, including high thermal,
chemical, and electrochemical stabilities in conjunction
this, including solid support immobilization or dilution
with large amounts of water, which improves enzyme
7
solubility, but at a potential cost to substrate solubility.
Alternatively, novel ionic liquids have been developed
2
4
,6,15,16,18
with negligible vapour pressures. Recently, the applica-
that are specifically protein/enzyme friendly.
How-
tion of ionic liquids in solubilising biopolymers such as
cellulose and lignin has paved the way for the develop-
ment of new technologies to process lignocellulosic bio-
ever, the high costs associated with these bespoke ionic
liquids will likely limit commercial application. Despite
these advances, the water content and low stability still
pose problems with respect to industrial applications, as
larger use of expensive proteins is coupled with difficult
separation of large amounts of water. Furthermore, recent
3
mass.
While research into the various applications for ionic
liquids has progressed, the use of biocatalysts for a num-
4
–10
11
7
ber of purposes has also been gaining ground. Enzymes
can perform many industrially relevant reactions under
significantly milder conditions than their chemical coun-
terparts. However, the low stability of proteins in
nonaqueous environments has limited their widespread
applicability to a select few processes. This creates a con-
flict between low-temperature aqueous processing, where
conditions are optimal for enzyme stability and selectivi-
ty, and organic solvent or high temperature processing,
which favour high substrate solubility and vastly im-
proved reaction kinetics respectively. We aim to resolve
this conflict through engineering protein surfaces towards
solubility and stability in anhydrous ionic liquid media,
providing a platform for high-temperature and nonaque-
ous bioprocessing.
reviews by Naushad et al. and Gao et al. highlight the
frequent contradictions that arise when studying the in-
teractions between proteins and ionic liquids, particularly
concerning how the chemical nature of the ions affects
protein stability and the role of water in this process. As a
result, despite the plethora of enzymes available for pro-
cesses in ionic liquids, there are to the best of our
knowledge, no examples of dry biocatalysts solubilized in
truly dry ionic liquids. This creates a need to develop new
biotechnologies whereby proteins and enzymes can be
both solubilized and stabilized in both dry hydrophobic
and hydrophilic ionic liquids. This will broaden the scope
of biocatalysis in ionic liquids, particularly if water can be
excluded and low thermal stability could be overcome, to
include the possibility of efficient nonaqueous biopro-
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 10
cessing in these novel solvent media, key examples being bodiimide activation. The resultant C-Mb solution was
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the production of biodiesel or hydrophobic pharmaceuti-
cals.
dialysed against deionized water for 48 h and filtered us-
ing a 0.22 µm syringe filter. This solution was then added
to a neutralized solution of the surfactant ethylene glycol
Solvent-free liquid proteins are a new class of hybrid
material, where surface modification of proteins and en-
zymes, by introduction of a polymer-surfactant “organic
-
1
ethoxylate lauryl ether (M ~ 690 g.mol – S – Sigma, UK)
n
and stirred for 12 h to produce the [C-Mb][S] nanocon-
struct. This was dialysed against deionized water for 48 h
to remove unbound S, and then freezedried to yield a
light brown powder that formed the solvent-free biofluid
after thermal annealing.
corona”, can allow for a protein-rich biofluid, devoid of
20-25
any solvent.
These have recently been shown to pre-
serve the three dimensional architecture of the biomole-
20,21
22
cule,
maintain backbone dynamics as if in water, and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
23
retain biological function and enzymatic activity all in
the absence of water. Furthermore, these novel fluids
Synthesis of ionic liquids
24
For the ionic liquid synthesis, all reagents were bought
significantly increase the thermal stability of the pro-
20
from Sigma UK, dried over P O , and with purified further
tein/enzyme , and in the case of lipase, enzyme activity
2
5
24
as necessary. 1-Butyl-1-methylpyrrolidinium chloride was
synthesized as follows. Freshly distilled N-
methylpyrrolidine (311.65 g, 3.66 mol) was added dropwise
to a mixture of freshly distilled 1-chlorobutane (309.18 g,
was enhanced at temperatures up to 150 °C. Molecular
dynamics simulations suggested that the dielectric con-
stant of the organic corona surrounding the protein mol-
25
ecules was similar to polar organic solvents. One poten-
3
.33 mol) and ethyl acetate (300 mL). The reaction was
tial drawback limiting application of these fluids in their
23
heated for 3 days at 75 °C and upon completion the mix-
ture was placed in a freezer overnight to encourage pre-
cipitation and then the solvent was removed via Schlenk
techniques. The precipitate was washed with ethyl ace-
tate, recrystallised in acetonitrile, and dried in vacuo to
yield 1-butyl-1-methylpyrrolidinium chloride (356.66 g,
60.3 %) as a white crystalline solid.
own right is that their viscosity is in the kPa.s range.
Mixing these biofluids with ionic liquids would decrease
this viscosity, and the enhanced thermal stability of the
modified protein would be well complemented by the
high thermal stability and negligible vapour pressure of
ionic liquids. The low dielectric of the modified surface
should be ideal for solubilizing proteins/enzymes in ionic
liquids and ought to provide protection from the ions,
which can be adept at disrupting hydrogen-bonded struc-
1-Butyl-1-methylpyrrolidinium
bis(trifluoromethylsulfonyl)imide ([bmpy][NTf ]) was
2
26
tures.
synthesized as follows.
A
mixture of 1-butyl-1-
methylpyrrolidinium chloride (33.78 g, 0.19 mol) and di-
chloromethane (50 mL) was added to lithium bis-
Here, we demonstrate our surface modification tech-
nique with a model myoglobin system in the presence of
several ionic liquids. Specifically, we have synthesized
protein-polymer surfactant nanoconstructs, yielding a
protein rich biofluid that, to the best of our knowledge,
exhibits the first example of high solubility of a dry pro-
tein in both hydrophilic and hydrophobic ionic liquids.
Thermal stability of the modified myoglobin increases by
(
trifluoromethylsulfonyl)imide (60.29 g, 0.21 mol) and
stirred vigorously at room temperature for 24 h. The mix-
ture was filtered and washed repeatedly with deionized
water (10 x 100 mL) until no chloride traces remained
(
tested with silver nitrate solution). The ionic liquid in
dichloromethane was purified through a short alumina
column topped with a thin layer of activated charcoal (ca.
5
5 °C in ionic liquid solution as compared to aqueous so-
<
5 mm). The solvent was then removed in vacuo (24 h)
lutions, demonstrating a positive impact of the ionic liq-
uid on the modified protein’s stability. Our results
demonstrate that engineering the surface of a protein
favourably alters the manner in which it interacts with
ionic liquids. This could be a general pathway to realiza-
tion of ionic liquid soluble proteins and enzymes with
high thermal stability, leading to the development of bio-
catalysts for myriad non-aqueous bioprocessing opportu-
nities.
giving 1-butyl-1-methylpyrrolidinium
bis(trifluoromethylsulfonyl)imide (71.29 g, 89 %) as a col-
ourless viscous liquid.
Similarly,
1-butyl-1-methylpyrrolidinium
trifluoro-
methanesulfonate ([bmpy][OTf]) was synthesized as fol-
lows. A mixture of 1-butyl-1-methylpyrrolidinium chloride
(67.88 g, 0.38 mol) and acetone (50 mL) was added to
sodium trifluoromethanesulfonate (71.92 g 0.42 mol) and
the reaction was stirred for 48 h at room temperature.
The resulting organic phase was filtered and washed with
water until chloride tests proved negative (silver nitrate
test). The solvent was then removed via rotary evapora-
tion and the ionic liquid was passed through a short alu-
mina column topped with a layer of activated charcoal
(ca. < 5 mm). The remaining liquid was dried in vacuo (24
Materials and Methods
Preparation of protein-polymer surfactant biofluids
Protein-polymer surfactant nanoconstructs were pre-
23
pared using established methods. Briefly, N,N′-bis(2-
aminoethyl)-1,3-propanediamine (Sigma, UK) was cou-
pled to the aspartic and glutamic acid residues of equine
skeletal muscle myoglobin (Mb - Sigma, UK) via car-
h)
giving
1-butyl-1-methylpyrrolidinium
trifluoro-
methanesulfonate (72.15 g, 65.18 %) as a colourless viscous
liquid.
2
ACS Paragon Plus Environment

Page 3 of 10
Spectroscopic methods
Journal of the American Chemical Society
fluid with a melting temperature of 27 °C (amorphous-
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
solid to liquid transition, SI Fig. 2). The solvent-free liq-
uid was soluble in water and polar organic solvents such
as ethanol. Importantly, the protein-polymer surfactant
conjugates were highly miscible (solubility of >50 wt%)
with both hydrophilic ([bmpy][OTf]) and hydrophobic
UV/Vis spectroscopy was performed on a Shimadzu
UV2600 fitted with a Peltier controlled heating environ-
ment. Aqueous solutions of Mb, C-Mb, and [C-Mb][S]
were measured in 10 mm quartz cuvettes at 25 °C. Sol-
vent-free [C-Mb][S] and [C-Mb][S] solutions were meas-
ured in 0.01 mm two-part quartz cell at 25 °C. All spectra
were recorded using the “Fast” setting with automatic
sampling enabled.
(
[bmpy][NTf ]) ionic liquids (Fig. 1). The homogeneity of
2
modified myoglobin-ionic liquid mixtures was in stark
contrast to the heterogeneous mixture of unmodified Mb
and [bmpy][OTf] (maximum solubility of 5.6 wt%) and
the complete lack of mixing of unmodified Mb with
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Synchrotron radiation circular dichroism (SRCD) spec-
troscopy was performed at Diamond Light Source on
beam line B23 using module A, fitted with a custom
Linkam stage for temperature control. Spectra were col-
lected between 260 nm and 180 nm with an integration
time of 2 s and data interval of 1 nm. Aqueous solutions
[bmpy][NTf
titration indicated water contents of 0.2 wt% and 0.3 wt%
for the [bmpy][OTf] and [bmpy][NTf ] mixtures respec-
] (solubility of <0.3 wt%, Fig. 1). Karl-Fischer
2
2
tively. These values correspond to 12-19 water molecules
per [C-Mb][S] construct. This is significantly fewer than
the 526 water molecules to cover the accessible surface
area of Mb, and fewer than the 60 required as a “lower
-
1
(
0.1 – 0.3 mg.mL , 10 mM pH 6.8 phosphate buffer) of Mb,
C-Mb, and [C-Mb][S] were run in 1 mm quartz cuvettes.
Solvent-free liquid [C-Mb][S] samples were measured as
thin films cast between quartz plates. For [C-Mb][S] in
ionic liquid, samples were prepared as 50 % solutions to
minimize solvent absorption, and cast as films between
quartz plates. Subsequent spectra of both solvent-free,
and ionic liquid solutions of [C-Mb][S] were normalized
to intensity recorded at 222 nm of samples in 0.01 mm
quartz cells (protein absorbance was too high to measure
spectra across the complete range in 0.01 mm quartz
cells). This was to ensure that mean residue ellipticity
could be reasonably calculated. Spectra were deconvolut-
ed to estimate secondary structure content with the
DICHROWEB service using the SP175 dataset, and the
20
limit” for protein motion. As a result, although there is
residual water present, the systems can be affectively
thought as being water free.
27–29
CDSSTR analysis program.
Errors presented for the
secondary structure for the solvent-free and ionic liquid
solutions account for potential variation in film thickness
between samples. Thermal denaturation of [C-Mb][S]
samples (aqueous, solvent-free, and ionic liquid) were
performed using the above parameters. Solvent-free and
ionic liquid samples were measured in the temperature
range 0 – 210 °C with spectra recorded every 10 °C. Aque-
ous samples were measured in the temperature range 25 –
95 °C with spectra recorded every 5 °C. Heating rate was
-
1
set to 100 °C.min and samples were equilibrated for 2
minutes before spectra were recorded.
FTIR spectroscopy was measured using a Perkin Elmer
Spectrum 100 fitted with a Specac attenuated total reflec-
tion (ATR) accessory. Spectra were recorded as accumula-
-
1
tions of 8 scans with a bandwidth of 2 cm .
Figure 1. Images showing 50 wt% mixtures of lyophilized Mb
in [bmpy][OTf] (a), [C-Mb][S] in [bmpy][OTf] (b), Mb in
Results and Discussion
[
bmpy][NTf ] (c), and [C-Mb][S] in [bmpy][NTf ] (d). Sam-
2 2
Conjugation of myoglobin (Mb) cationized with N,N′-
bis(2-aminoethyl)-1,3-propanediamine (C-Mb) with the
surfactant glycolic acid ethoxylate lauryl ether (S) yielded
a protein-polymer surfactant construct [C-Mb][S] with a
stoichiometry of 84 surfactant molecules per protein (SI
Fig. 1). Upon dehydration and thermal annealing, this
formed a dark brown-red, protein rich, solvent-free bio-
ples were prepared by freezedrying aqueous solutions of pro-
tein and ionic liquid.
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 10
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. UV/Vis spectra showing the Soret band region (a), ATR-FTIR spectra showing amide I and II region (b), and SRCD
spectra showing far-UV region (c) for aqueous [C-Mb][S] (black), solvent-free [C-Mb][S] (red), [C-Mb][S] in [bmpy][OTf] (blue),
and [C-Mb][S] in [bmpy][NTf ] (green). Ionic liquid solutions were prepared at a [C-Mb][S] concentration of 50 wt%, and aque-
2
ous solutions were prepared with 10 mM phosphate buffer (pH 6.8). All spectra we recorded at 25 °C. ATR-FTIR was not record-
ed for aqueous solution due to water interference.
-
1
-1
This demonstrates that modifying the protein surface to
form the solvent-free biofluids increased the solubility of
the dry protein in anhydrous ionic liquids, yielding the
first example (to the best of our knowledge at time of
writing) of a soluble protein solution in the near total
absence of water. Overcoming these solubility issues min-
imized protein aggregation in the hydrophilic ionic liquid,
and allowed for the introduction of protein into the more
I and amide II bands at 1655 cm and 1545 cm respective-
ly for [C-Mb][S] in both ionic liquids and the solvent-free
state, indicative towards a highly conserved α-helical
structure (Fig. 2b). This is in broad agreement with the
UV/Vis spectra, demonstrating that the global confor-
mation of Mb is maintained even in the anhydrous ionic
liquids. The presence of highly conserved Mb structure
was confirmed using synchrotron radiation circular di-
chroism (SRCD), which showed negative peaks at 222 nm
and 208 nm with a positive peak at 195 nm for [C-Mb][S]
in the ionic liquid, solvent-free, and aqueous environ-
ments (Fig. 2c). These peak positions are highly indicative
of a predominantly α-helical structure of [C-Mb][S],
agreeing well with the ATR-FTIR spectra. The SRCD spec-
tra revealed quite significant changes in the structure
content of [C-Mb][S] between the aqueous and non-
aqueous environments. Dehydration of aqueous [C-
Mb][S] nanoconjugates to yield the solvent-free liquid
caused a reversal in the ratio of the negative features at
hydrophobic [NTf ] system, which has not previously
2
been demonstrated.
To assess the impact of the ionic liquids on the 3-
dimensional global conformation of the modified protein,
the states of the secondary and tertiary structures of [C-
Mb][S] were determined using a number of spectroscopic
measurements. UV/Vis spectroscopy showed Soret bands
at 412 nm and 413 nm, and Q bands at 529 nm and 535
nm, for [C-Mb][S] in [bmpy][OTf] and [bmpy][NTf ] re-
2
spectively (Fig. 2a). This corresponds well with the Soret
and Q absorptions of [C-Mb][S] in the solvent-free liquid
state (413 nm, 535 nm) and in aqueous solutions (409 nm,
2
08 nm and 222 nm, where the increase in the 222 nm
intensity suggested an increase in α-helical structure. This
was confirmed through deconvolution of the spectra us-
ing the DICHROWEB service,
there was indeed a moderate increase in α-helix content
from 51 % to 54 % upon dehydration (Table 1).
5
30 nm) [Fig. 2a]. This indicates that in aqueous and ionic
liquid solutions, and in the solvent-free liquid state, the
prosthetic heme group of Mb is still bound to the protein.
Furthermore, the topology of the binding pocket remains
unchanged in the different environments. Attenuated
total reflectance (ATR) FTIR spectroscopy showed amide
27–29
which showed that
27–29
Table 1. Estimated secondary structure content as determined from SRCD spectra in Fig. 2c using the DICHROWEB service.
α-helix / %
β-sheet / %
turns / %
unordered / %
NRMSD
[C-Mb][S]
aqueous
5
5
1
11
11
27
0.009
[
C-Mb][S]
4 ± 6
10 ± 1
12 ± 2
24 ± 5
0.004
solvent-free
[C-Mb][S]
[bmpy][OTf]
6
6
9 ± 6
1
6 ± 3
8
10 ± 2
11
15 ± 3
20
0.005
0.008
[
[
C-Mb][S]
bmpy][NTf2]
4
ACS Paragon Plus Environment

Page 5 of 10
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. (a,b) SRCD spectra showing far-UV region for aqueous solutions of Mb (a) and [C-Mb][S] (b) showing response to
[
2
bmpy][OTf] concentrations of 0 % (black), 20 % (red), 50 % (blue), and 80 % (green). (c) Plots of mean residue ellipticity at
22 nm against [bmpy][OTf] concentration for aqueous solutions of Mb (black triangles) and [C-Mb][S] (red squares).
Meanwhile, the SRCD spectra of [C-Mb][S] in ionic liq- lighting the limitations of the series at high ionic liquid
1
1
uids showed an increase in the absolute intensities of all
the spectral features, particularly the positive feature at
concentrations (and here in the anhydrous ionic liquid).
The interactions involved are therefore more complex,
and given the presence of the surfactant coronal layer, it
is likely that the relative molecular bulk and lower charge
1
95 nm (Fig. 1c), with the increase in secondary structure
most apparent in [bmpy][OTf]. Deconvolution suggested
that moving from the solvent-free liquid state to
-
-
density of NTf2 as compared to OTf , may also affect in-
teractions with the protein surface. It is often accepted
that anions with high β values and nucleophilicities inter-
[
bmpy][OTf] and [bmpy][NTf ] caused an increase in the
2
α-helix content of [C-Mb][S] to 69 % and 61 % (from 51%)
respectively (Table 1).
7
,11
act strongly with protein surfaces and destabilize them.
As a result, it is often assumed that proteins are stabilized
by an absence of these interactions. However, the results
presented here indicate a more subtle picture where fa-
vourable interactions between ionic liquid anions and
proteins can arise and promote structural integrity.
Increasing α-helix content was concomitant with a de-
crease in the unordered content, suggesting that the ionic
liquids were inducing α-helix formation and therefore
increasing structure, relative to both the aqueous and
solvent-free protein states. It must be noted that the
structure of aqueous [C-Mb][S] represents a destabilized
form of Mb (51 % α-helix content compared to 72 % α-
helix content for native Mb – SI Fig. 3, SI Table 1) due to
the cationization process causing a destabilization of the
native myoglobin state, such that a shift to a molten
It is interesting to note that protein structuration only
occurred under anhydrous conditions. SRCD spectra of
aqueous solutions of Mb showed structure decreasing
with increasing concentration of [bmpy][OTf] (Fig. 3a), in
line with observations of proteins in organic solvents,
20
34,35
globule state occurs as described previously.
where low levels of water can often be destabilizing.
Although there was no increase in structure with ionic
liquid concentration, modification of the protein did pro-
vide some protection against these effects, as the struc-
ture of [C-Mb][S] only decreased at high (80 %) ionic liq-
uid concentrations (Fig. 3b). This becomes more apparent
using a plot of the mean residue ellipticity at 222 nm as an
indication of secondary structure (Fig. 3c). Although [C-
Mb][S] is in the aforementioned destabilized molten
globule state, more structure remained at high ionic liq-
uid concentration than for the native protein, demon-
strating that the surfactant corona around the modified
protein provides protection from the otherwise denatur-
ing effects of the ionic liquid. This moderate level of
structure preservation observed for [C-Mb][S] in response
to ionic liquid concentration was only valid for non-acidic
anions- these experiments repeated with the highly acidic
Therefore, the significant increase in secondary struc-
ture observed in the anhydrous ionic liquid solution could
be justified as an ionic liquid-induced reformation of the
native Mb global fold, as opposed to forming a new, non-
native structure. This structuration of the sections of un-
ordered backbone of [C-Mb][S] suggests that the ionic
liquids act as helix enhancing solvents, restoring the
structure of the modified protein toward that exhibited in
the unmodified state, consistent with polar organic envi-
3
0–32
ronments such as those containing fluoroalcohols.
Furthermore, these observations suggest that the OTf
-
anion was able to interact with the biomolecule with
-
greater effect than NTf , as may be expected given the
2
- 33
greater hydrogen bond basicity (β) of OTf . Additionally,
from a Hofmeister perspective, OTf should provide
greater stability than NTf , and here that trend remains.
-
-
2
19
ionic liquid [N2220][HSO ] yielded no apparent benefit of
However, it is worth noting that according to the overall
Hofmeister series, both of these anions ought to be desta-
bilizing. In the anhydrous solutions however, we see sig-
nificant stabilization of the protein fold, potentially high-
4
protein modification with respect to secondary structure
protection (SI Fig. 4). In these experiments, low concen-
trations (<10 %) of the ionic liquid destabilized both Mb
and [C-Mb][S], showing that the organic corona could not
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 10
-
offer any protection against the highly interacting HSO4
may offset some of the stability gains conferred by the
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
anion, and low pH, in aqueous conditions.
surfactant corona. Alternatively, in agreement with the
increase in secondary structure seen with the OTf anion,
-
In light of the increase in secondary structure observed
for [C-Mb][S] in [bmpy][OTf] and [bmpy][NTf ], tempera-
the strength of interaction between the anion and protein
may be playing a role. Thermal stability in the ionic liq-
uids appears to be associated with a decrease in the coop-
erativity of the unfolding, as evidenced through a shal-
lowing of the slope of the unfolding curve (Fig. 4c), indi-
cating that the ionic liquid induced structuration of [C-
Mb][S] came at a cost to protein flexibility, exacerbating
the relatively noncooperative unfolding of the molten-
globule like structure. This suggests that the observed
increase in structure was not a return to the compact
globule associated with Mb, but rather that the ionic liq-
uid may cause a shift in the mechanism of unfolding to-
ward multi-state denaturation. Unfortunately, the resolu-
tion of CD spectroscopy is not sufficient to establish this
absolutely and therefore a more detailed atomistic ap-
proach is necessary to confirm this hypothesis.
2
ture dependent SRCD spectra were recorded to assess
whether this translated to an increase in thermal stability
of the modified protein when solubilized in the ionic liq-
uids (Fig. 4) relative to the solvent-free state. Increasing
the temperature from 0 °C to 210 °C saw progressive re-
ductions of the intensity of the spectral features at 222
nm, 208 nm, and 195 nm indicative of thermal denatura-
tion occurring over a large temperature range (Fig. 4a,b).
This showed that in pure ionic liquids, [C-Mb][S] was
dynamically free enough to sample the conformations
required to unfold, even at low (<30 °C) temperatures.
Furthermore, the presence of an isodichroic point at 204
nm suggests that the denaturation process occurred via a
two-state transition from a native state to a denatured
state. With this in mind, the SRCD data was used to con-
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
struct a plot of fraction denatured (f ) against tempera-
Using a two-state model of denaturation, justified by
the existence of an isodichroic point at 204 nm in the
D
ture (Fig. 4c). In neat [bmpy][OTf] and [bmpy][NTf ], [C-
Mb][S] had half denaturation temperatures (T ) of 113 °C
and 108 °C respectively (Table 2). This was an increase in
thermal stability of 27/22 °C from the solvent-free liquid
2
SRCD data, we extracted values for the entropy (ΔS ) and
m
m
enthalpy (ΔH ) of denaturation from plots of f (see Sup-
m
D
plementary Information for details). For [C-Mb][S] in
-
state (T = 86 °C, Fig. 4c, SI Fig. 5, Table 2) and 55/50 °C
[bmpy][OTf], ΔS and ΔH were calculated as 89 J.K
m
m
m
1
-1
-1
-
from aqueous [C-Mb][S] (Tm = 58 °C, Fig. 4c, SI Fig. 6,
.mol and 9.8 kJ.mol , and in [bmpy][NTf ] as 86 J.K
2
1
-1
-1
Table 2). The former results are in agreement with previ-
.mol and 9.0 kJ.mol (Table 2, SI Fig. 7). Values of ΔS
m
20,21,24
-
ous findings
that dehydration to form the solvent-
and ΔH calculated for [C-Mb][S] in water were 170 J.K
m
1
-1
-1
free liquid increased the thermal stability of the protein.
However, the latter results reveal that even generic ionic
liquids can further enhance this stabilization, despite an
increase in translational entropy offered by the lower pro-
tein concentration, mirroring the observed increases in
secondary structure (Fig. 2c, Table 1). This indicates that
the ionic liquids are interacting favourably with the pro-
tein, causing an increase in stability that is a direct re-
sponse to the increase in secondary structure. Concur-
.mol and 9.9 kJ.mol respectively (Table 2, SI Fig. 8).
Given the negligible change in enthalpy between the ionic
liquid and aqueous solutions, the observed stabilization
of the protein in the ionic liquids was entirely due to a
reduction in entropy. This is in broad agreement with the
decrease in cooperativity of the transition, and serves as
evidence that increased protein stability in the ionic liq-
uid was not result of any increase in favourable interac-
tions within the protein.
rently, stabilization was marginally lower when the anion
-
was NTf , signifying that the hydrophobicity of this ion
2
Figure 4. (a,b) Temperature dependant Far-UV SRCD spectra showing thermal denaturation of [C-Mb][S] (50 % solutions) in
[bmpy][OTf] (a) and [bmpy][NTf ] (b) from 0 °C (blue) to 210 °C (red) at intervals of 10 °C. (c) Plot of fraction denatured as cal-
2
culated from SRCD spectra for aqueous [C-Mb][S] (black triangles), solvent-free [C-Mb][S] (red squares), [C-Mb][S] in
[bmpy][OTf] (blue circles), and [C-Mb][S] in [bmpy][NTf ] (green diamonds). Data were fitted with Sigmoid functions (solid
lines) in accordance with a two-state model of denaturation.
2
6
ACS Paragon Plus Environment

Page 7 of 10
Journal of the American Chemical Society
-
1
-1
-1
Tm / °C
ΔSm / J.K .mol
ΔHm / kJ.mol
1
2
3
4
5
6
7
[
C-Mb][S] aqueous
58 ± 1
171 ± 17
304 ± 30
89 ± 6
9.9 ± 0.2
25.7 ± 0.3
9.8 ± 0.1
9.0 ± 0.1
[C-Mb][S] solvent-free
86.4 ± 0.6
113 ± 4
[
[
C-Mb][S] [bmpy][OTf]
C-Mb][S] [bmpy][NTf2]
108 ± 2
86 ± 4
Table 2. Thermodynamic parameters for the thermal denaturation of [C-Mb][S] in aqueous solution, the solvent-free state, and
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
in ionic liquid. Values calculated using a two-state model applied to temperature dependent SRCD spectra (See Supplementary
Information
for
further
details).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
-
-
2
The observed ΔSm for the OTf and NTf ionic liquids
95 °C the Soret band once again reduced in intensity and
underwent a blueshift from 413 nm to 408 nm, indicating
a change to a more denatured structure, albeit slightly
more structured than for [C-Mb][S] in [bmpy][OTf]. Up-
on cooling to 25 °C, there was again a >90 % recovery in
the Soret intensity, with the peak position returning to
413 nm, indicating structural recovery in the more hydro-
phobic ionic liquid. These results are in broad agreement
with the SRCD data (Fig. 4), demonstrating further that
[C-Mb][S] has a high thermal stability in ionic liquids,
and also that the protein is able to refold from a partially
denatured state, in both hydrophilic and hydrophobic
ionic liquids. This indicated that despite apparent con-
striction of the protein in the ionic liquids, as evidenced
through the small changes in entropy on denaturation,
the protein was still able to sample multiple confor-
mations. Therefore, introduction of a coronal layer pro-
vides an interface between the protein surface (both na-
tive and denatured) and the ionic liquid, allowing for re-
versible perturbations in global conformation.
were comparable despite differences in β values and ex-
pected specific ion effects, suggesting entropic stabiliza-
tion was key across both systems. However, ΔH was
marginally lower in the NTf2 ionic liquid, indicating that
the slightly lower stability of [C-Mb][S] in the more hy-
drophobic environment was predominately due to a small
change in enthalpy.
m
-
We rationalize this as the more hydrophobic nature of
bmpy][NTf ] stabilizing the unfolded state of the protein
more than [bmpy][OTf], reducing the enthalpic cost of
denaturation. By contrast, ΔSm and ΔHm for solvent-free
[
2
-1
-1
[
C-Mb][S] were calculated at 304 J.K .mol and 25.7
-1
kJ.mol respectively (Table 2, SI Fig. 9), showing that in
the absence of any solvent the situation is much more
complex. Any enthalpic gains from the increased strength
of stabilizing interactions within the protein in the ab-
sence of solvent were offset by the conformational free-
dom of the protein within the surfactant-coronal layer.
Since myoglobin has a propensity to occupy a molten
globule state, engineering the surface in the manner re-
20
ported herein causes such a change. It is therefore plau-
sible that increased stability upon transferring the protein
from an aqueous phase to an ionic liquid phase may be a
consequence of this unusual molten globule structure.
However, given that previous results indicate that modi-
fying more robust enzymes such as lysozyme or lipase
21,24
give greater stabilities than the myoglobin system , we
believe that the ionic liquid stability will be transferable.
In light of possible changes to the unfolding pathway of
[C-Mb][S] in the ionic liquid, reversibility of denaturation
was probed using UV/Vis spectroscopy (Fig. 5). Equili-
brating the [C-Mb][S]/[bmpy][OTf] mixture at 95 °C
caused a reduction and broadening in the Soret absorb-
ance concomitant with a blue shift in the peak position
from 412 nm to 393 nm (Fig. 5a). This indicates a change
in the global conformation of myoglobin consistent with
partial denaturation. On cooling to 25 °C, the Soret band
sharpened and returned to 412 nm with a >90 % recovery
in intensity. This demonstrates that in the ionic liquid
solution, the protein was able to recover the structure lost
at high temperatures, consistent with the high thermal
stability observed using SRCD (Fig. 4). Furthermore, once
in a (partially) denatured form, the protein was able to
refold to its previous structure. This was repeated with
Figure 5. UV/Vis spectroscopy showing [C-Mb][S] in
[bmpy][OTf] (a) and [bmpy][NTf ] (b) at 25 °C (black), after
2
incubation at 95 °C (red), and after cooling back down to
[
bmpy][NTf ] as the ionic liquid (Fig. 5b). On heating to
2
25 °C (blue).
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 10
Conclusions
Corresponding Author
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
*j.hallet@imperial.ac.uk.
Here, we have demonstrated that surface modification
of myoglobin, yielding the solvent-free liquid [C-Mb][S],
can be used as a platform to successfully solubilize the
protein in both hydrophilic and hydrophobic ionic liquids
in the absence of water. Spectroscopic methods (UV/Vis,
SRCD, FTIR) establish that there has been preservation of
the secondary and tertiary structures of the protein, with
significant structuration of the protein occurring, particu-
larly in the case of [bmpy][OTf]. This signifies that the
ionic liquids are an environment highly conducive to-
wards protein stability, and SRCD measurements con-
firmed that thermal denaturation occurs over a large
temperature range. With a denaturation temperature of
113 °C, some 55 °C higher than that in aqueous solutions,
we were able to show that modification of myoglobin not
only increased solubility in ionic liquids, but allowed for a
thermal stability surpassing what is possible in water.
Thermally induced unfolding and subsequent refolding
from 95 °C reinforced that the modified protein was ro-
bust towards both the ionic liquid medium and thermal
stresses. Using a two state model of denaturation, the
increase in thermal stability from the aqueous to the ionic
liquid phase appeared to be the result of decreasing en-
tropy of denaturation. We attribute this to subtle solvent-
induced changes in the unfolding mechanism, resulting
in lower cooperativity of the transition. This is possibly a
direct manifestation of a putative molten globule state of
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version
of the manuscript.
ACKNOWLEDGMENT
We thank EPSRC (Frontier Engineering Grant EP/K038648/1)
for financial support. We also thank Dr. Giuliano Siligardi
and Dr. Rohanah Hussain at Diamond Light Source for ac-
cess to the B23 beam line, Paul Joseph Corbett for ionic liq-
uid synthesis, and Patricia Carry for access to FTIR and DSC.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
REFERENCES
(1)
Smiglak, M.; Pringle, J. M.; Lu, X.; Han, L.; Zhang, S.;
Gao, H.; MacFarlane, D. R.; Rogers, R. D. Chem. Commun. 2014,
5
0, 9228–9250.
2)
(
Hallett, J. P.; Welton, T. Chem. Rev. 2011, 111, 3508–
3576.
(3)
Brandt, A.; Gräsvik, J.; Hallett, J. P.; Welton, T. Green
Chem. 2013, 15, 550–583.
(4)
Bi, Y.-H.; Duan, Z.-Q.; Li, X.-Q.; Wang, Z.-Y.; Zhao, X.-
R. J. Agric. Food Chem. 2015, 63, 1558–1561.
(5)
Bokinsky, G.; Peralta-Yahya, P. P.; George, A.; Holmes,
B. M.; Steen, E. J.; Dietrich, J.; Lee, T. S.; Tullman-Ercek, D.;
Voigt, C. A.; Simmons, B. A.; Keasling, J. D. Proc. Natl. Acad. Sci.
U. S. A. 2011, 108, 19949–19954.
(6)
Deive, F. J.; Ruivo, D.; Rodrigues, J. V.; Gomes, C. M.;
Sanromán, M. Á.; Rebelo, L. P. N.; Esperança, J. M. S. S.;
Rodríguez, A. RSC Adv. 2014, 5, 3386–3389.
[C-Mb][S], although a more complex set of interactions
between the ionic liquid and the protein is also possible.
The net result of high thermal stability, with reversible
unfolding indicates that the forces stabilizing protein
structure, as well dynamic flexibility, can be well main-
tained in, and even enhanced by, ionic liquids.
(7)
Biochem. Eng. J. 2015, 99, 67–84.
8) Huang, Z.-L.; Yang, T.-X.; Huang, J.-Z.; Yang, Z.
Bioenerg. Res. 2014, 7, 1519–1528.
9) Guncheva, M.; Paunova, K.; Yancheva, D.; Svinyarov, I.;
Bogdanov, M. J. Mol. Catal. B Enzym. 2015, 117, 62–68.
10) van Rantwijk, F.; Sheldon, R. A. Chem. Rev. 2007, 107,
2757–2785.
Gao, W.-W.; Zhang, F.-X.; Zhang, G.-X.; Zhou, C.-H.
(
(
We have shown that through modification of the sur-
face of myoglobin, it is possible to greatly increase solu-
bility of the dry protein in dry ionic liquids. The resulting
protein rich fluids have highly conserved protein architec-
ture, and high thermal stability. The synthetic procedure
has previously been shown to be very versatile, with sol-
(
(11)
Naushad, M.; ALOthman, Z. A.; Khan, A. B.; Ali, M.
Int. J. Biol. Macromol. 2012, 51, 555–560.
(12)
Lozano, P.; Bernal, J. M.; Garcia-Verdugo, E.; Sánchez,
G.; Vaultier, M.; Burguete, I.; Luis, S. V. Green Chem. 2015, 17,
3706–3717.
21,23
vent-free biofluids of
a
number of proteins,
(
13)
158–2169.
14) Reddy, P. M.; Umapathi, R.; Venkatesu, P. Phys. Chem.
Chem. Phys. 2015, 17, 184–190.
Kumar, A.; Venkatesu, P. Process Biochem. 2014, 49,
24,36
37
enzymes,
and viruses having been reported. In com-
2
bination with the results presented herein, we propose
that this biotechnology could provide a platform for the
realization of highly soluble ionic liquid and temperature
stable enzymes, providing pathways for improved bio-
catalysis in ionic liquids and generally homogeneous bio-
catalysis in anhydrous systems. As a result, we have be-
gun a research program investigating the stability and
function of industrially relevant enzymes such as lipase,
β-glucosidase, xylanase, and laccase in ionic liquids.
(
(15)
Fujita, K.; MacFarlane, D. R.; Forsyth, M. Chem.
Commun. 2005, 70, 4804–4806.
(
16)
Bioprocess Eng. 2010, 15, 40–53.
17) Moniruzzaman, M.; Nakashima, K.; Kamiya, N.; Goto,
M. Biochem. Eng. J. 2010, 48, 295–314.
18) Moniruzzaman, M.; Kamiya, N.; Goto, M. Org. Biomol.
Chem. 2010, 8, 2887–2899.
Gorke, J.; Srienc, F.; Kazlauskas, R. Biotechnol.
(
(
(19)
Weingärtner, H.; Cabrele, C.; Herrmann, C. Phys.
ASSOCIATED CONTENT
Chem. Chem. Phys. 2012, 14, 415–426.
(20) Brogan, A. P. S.; Siligardi, G.; Hussain, R.; Perriman, A.
W.; Mann, S. Chem. Sci. 2012, 3, 1839–1846.
Supporting Information. Thermal analysis, aqueous SRCD
spectra, thermal denaturation plots, calculated free energy
plots. This material is available free of charge via the Internet
at http://pubs.acs.org..
(
21)
S. J. Phys. Chem. B. 2013, 117, 8400–8407.
22) Gallat, F.-X.; Brogan, A. P. S.; Fichou, Y.; McGrath, N.;
Brogan, A. P. S.; Sharma, K. P.; Perriman, A. W.; Mann,
(
Moulin, M.; Härtlein, M.; Combet, J.; Wuttke, J.; Mann, S.;
AUTHOR INFORMATION
8
ACS Paragon Plus Environment

Page 9 of 10
Journal of the American Chemical Society
Zaccai, G.; Jackson, C. J.; Perriman, A. W.; Weik, M. J. Am. Chem.
Soc. 2012, 132, 13168–13171.
(23) Perriman, A. W.; Brogan, A. P. S.; Cölfen, H.; Tsoureas,
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
N.; Owen, G. R.; Mann, S. Nat. Chem. 2010, 2, 622–626.
(
24) Brogan, A. P. S.; Sharma, K. P.; Perriman, A. W.; Mann,
S. Nat. Commun. 2014, 5, 5058.
25) Brogan, A. P. S.; Sessions, R. B.; Perriman, A. W.;
Mann, S. J. Am. Chem. Soc. 2014, 136, 16824–16831.
26) Hunt, P. A.; Ashworth, C. R.; Matthews, R. P. Chem.
(
(
Soc. Rev. 2015, 44, 1257–1288.
(27) Lees, J. G.; Miles, A. J.; Wien, F.; Wallace, B. A.
Bioinformatics 2006, 22, 1955–1962.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(28) Whitmore, L.; Wallace, B. A. Biopolymers 2008, 89,
392–400.
(29) Whitmore, L.; Wallace, B. A. Nucleic Acids Res. 2004,
32, W668–W673.
(
30) Cort, J. R.; Andersen, N. H. Biochem. Biophys. Res.
Comm. 1997, 233, 687–691.
31) Sönnichsen, F. D.; Van Eyk, J. E.; Hodges, R. S.; Sykes,
(
B. D. Biochemistry 1992, 31, 8790–8798.
(32) D’Amico, M.; Raccosta, S.; Cannas, M.; Martorana, V.;
Manno, M. J. Phys. Chem. B. 2011, 115, 4078–4087.
(33) Ab Rani, M. A.; Brant, A.; Crowhurst, L.; Dolan, A.; Lui,
M.; Hassan, N. H.; Hallett, J. P.; Hunt, P. A.; Niedermeyer, H.;
Perez-Arlandis, J. M.; Schrems, M.; Welton, T.; Wilding, R. Phys.
Chem. Chem. Phys. 2011, 13, 16831–16840.
(
34) Cowan, D. A. Comp. Biochem. Physiol. 1997, 118, 429–
38.
(35) Griebenow, K.; Klibanov, A. M. J. Am. Chem. Soc. 1996,
118, 11695–11700.
36) Sharma, K. P.; Zhang, Y.; Thomas, M. R.; Brogan, A. P.
S.; Perriman, A. W.; Mann, S. J. Phys. Chem. B. 2014, 118, 11573–
1580.
37) Patil, A. J.; McGrath, N.; Barclay, J. E.; Evans, D. J.;
Cölfen, H.; Manners, I.; Perriman, A. W.; Mann, S. Adv. Mater.
012, 24, 4557–4563.
4
(
1
(
2
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 10
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
10
ACS Paragon Plus Environment