Developing highly active small molecule ice recrystallization inhibitors based upon C-linked antifreeze glycoprotein analogues

Article abstract of DOI:10.1039/c3ra43835j

Ice recrystallization during cryopreservation results in a significant amount of cellular damage making compounds that exhibit ice recrystallization inhibition (IRI) activity desirable as a novel class of cryoprotectants. Herein, we report a systematic structure-function study on a previously identified IRI active C-linked antifreeze glycoprotein (C-AFGP) analogue (1). These studies indicate that while C-AFGPs containing three tripeptide repeats exhibit weak IRI activity 5.5 μM, a minimum number of four tripeptide repeats is required for potent IRI activity at this concentration. In addition, the galactosyl-ornithine building block 5 exhibited only moderate activity at 22 mM, but IRI activity was significantly increased upon addition of two glycine units to the C-terminal end of the amino acid bearing the C-linked galactopyranose residue. Finally, we report that conjugation of long alkyl chains (octyl, nonyl and decyl) to the C-linked galactosyl moiety of 1 can furnish IRI active small molecules. The ideal hydrocarbon chain length was 10 carbons for potent activity in this series of compounds. Longer hydrocarbon chain lengths dramatically decreased solubilities. The results of this study emphasize the importance of hydrophobic moieties for IRI activity and while consistent with previously reported small molecule carbohydrate-based and lysine-based ice recrystallization inhibitors, is the first example where a large IRI active glycoconjugate has been successfully truncated to small molecule IRI active components.

Full text of DOI:10.1039/c3ra43835j

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Developing highly active small molecule ice
recrystallization inhibitors based upon C-linked
antifreeze glycoprotein analogues†
Cite this: RSC Adv., 2013, 3, 26005
*
John F. Trant, Robyn A. Biggs, Chantelle J. Capicciotti and Robert N. Ben
Ice recrystallization during cryopreservation results in a significant amount of cellular damage making
compounds that exhibit ice recrystallization inhibition (IRI) activity desirable as novel class of
a
cryoprotectants. Herein, we report a systematic structure–function study on a previously identified IRI
active C-linked antifreeze glycoprotein (C-AFGP) analogue (1). These studies indicate that while C-AFGPs
containing three tripeptide repeats exhibit weak IRI activity 5.5 mM, a minimum number of four
tripeptide repeats is required for potent IRI activity at this concentration. In addition, the galactosyl–
ornithine building block 5 exhibited only moderate activity at 22 mM, but IRI activity was significantly
increased upon addition of two glycine units to the C-terminal end of the amino acid bearing the C-
linked galactopyranose residue. Finally, we report that conjugation of long alkyl chains (octyl, nonyl and
decyl) to the C-linked galactosyl moiety of 1 can furnish IRI active small molecules. The “ideal”
hydrocarbon chain length was 10 carbons for potent activity in this series of compounds. Longer
hydrocarbon chain lengths dramatically decreased solubilities. The results of this study emphasize the
importance of hydrophobic moieties for IRI activity and while consistent with previously reported small
molecule carbohydrate-based and lysine-based ice recrystallization inhibitors, is the first example where a
large IRI active glycoconjugate has been successfully truncated to small molecule IRI active components.
Received 22nd July 2013
Accepted 10th October 2013
DOI: 10.1039/c3ra43835j
www.rsc.org/advances
property is referred to as thermal hysteresis (TH). AFGPs are
also potent inhibitors of ice recrystallization. Ice recrystalliza-
Introduction
Antifreeze glycoproteins (AFGPs, Fig. 1) protect Antarctic teleost
sh from cryo-injury associated with the sub-zero temperatures
in which they live. AFGPs are comprised of a (Thr-Ala-Ala)n tri-
peptide repeat where the secondary hydroxyl group of the
threonine residue is glycosylated with b-D-galactosyl-(1–3)-a-N-
acetyl-D-galactosamine.1–4 One of the most interesting proper-
ties of AFGPs is their ability to bind to the surface of ice crystals
resulting in a localized depression of the freezing point relative
to the melting point, ultimately inhibiting ice growth.5–7 This
tion is dened as large crystals growing larger at the expense of
small ones. The ability to inhibit this process is a very desirable
property for cryopreservation as it has been reported that a
majority of cell damage incurred during cryopreservation is due
to injury from ice recrystallization.8,9 AFGPs have attracted
signicant synthetic interest due to their potential applications
in cryogenics and cryosurgery, as well as various commercial
and industrial processes where ice growth is problematic.4,10–13
However, due to their ice binding ability and subsequent
change in ice crystal habit,2,5,14 biological antifreezes such as
AFGPs actually increase cellular damage at the temperatures
associated with cryopreservation.15 Consequently, it is not
surprising that the use of AFGPs as cryoprotectants for biolog-
ical materials has largely been unsuccessful.
Our laboratory has recently demonstrated that carbon-linked
(C-linked) glycopeptide analogues of AFGPs (C-AFGPs) with the
structure shown in Fig. 2 exhibit “custom-tailored” antifreeze
activity.16,17 These C-AFGPs have ice recrystallization inhibition
(IRI) activity that is equipotent to AFGP-8, but they do not exhibit
TH activity and thus function as novel cryoprotectants.17,18
However, despite the recent advances in the synthesis of
complex glycopeptides their synthesis remains complex and
costly, thus the rational design of small molecules having the
Fig. 1 Structure of native AFGPs.
Department of Chemistry, University of Ottawa, 10 Marie Curie, Ottawa, ON, K1N
6N5, Canada. E-mail: rben@uottawa.ca; Fax: +1-613-562-5170; Tel: +1-613-562-5800
† Electronic supplementary information (ESI) available: Experimental procedures
and spectroscopic data for new compounds. See DOI: 10.1039/c3ra43835j
This journal is ª The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 26005–26009 | 26005

View Article Online
RSC Advances
Paper
activity. In addition, the IRI activity of the galactosyl–ornithine
analogue 5 and derivatives bearing hydrocarbon chains of
varying lengths were also of interest.
Glycopeptides 2, 3, and 4 were prepared using standard
solid-phase peptide synthesis techniques as previously
described for glycoconjugate 1.17 TH activity was assessed at 10
mg mLÀ1 using nanolitre osmometry,24 however none of these
compounds exhibited TH activity or dynamic ice shaping sug-
gesting that there was no direct interaction with the ice lattice.
This is consistent with the lack of TH activity observed for 1.17
The IRI activity of each glycopeptide was assessed using a “splat-
cooling” assay.25 In this assay, the size of ice crystals in a frozen
ꢀ
wafer aer a 30 minutes annealing time at À6.4 C are quan-
tied and compared to a phosphate buffered saline (PBS)
solution that serves as a positive control for ice recrystalliza-
tion.26 Activity is then represented as the mean grain size ratio
between the compound tested and PBS (% MGS), thus smaller
bars on the graph represent more potent IRI activity. The IRI
results for glycopeptides 2, 3 and 4 and the galactosyl–ornithine
building block 5 are shown in Fig. 3. At 5.5 mM only the full
length C-AFGP 1 containing four tripeptide repeats retained
potent IRI activity. As the number of tripeptide repeats was
decreased IRI activity also decreased and the glycopeptide with
one tripeptide repeat (2) did not exhibit any ability to inhibit ice
recrystallization at this concentration. This result is consistent
with previous studies correlating antifreeze activity of synthetic
native AFGPs to the number of tripeptide repeats.11,12
Glycopeptides 2–4 and the galactosyl–ornithine derivative 5
were also assessed for IRI activity at 22 mM, the concentration
at which some carbohydrate and lysine derivatives were previ-
ously demonstrated to exhibit potent activity.20,21 At 22 mM
glycopeptides 2, 3, and 4 all exhibited IRI activity ranging from
moderate to high with 4, containing three tripeptide repeats,
exhibiting the highest degree of activity (Fig. 3). Galactosyl–
ornithine derivative 5 had activity that was similar to a 22 mM
Fig. 2 Structure–function studies on glycopeptide 1 to elucidate the key struc-
tural features necessary for its potent IRI activity.
capacity to inhibit ice recrystallization is very attractive. While
the structural features necessary for AFGPs to bind to ice and
exhibit TH activity have been extensively studied,7,11 the struc-
tural features necessary for potent IRI activity remain largely
unknown. Previous work by our laboratory has elucidated the
relationship between hydration and IRI activity.17,19 In addition,
we have recently reported the rst examples of potent IRI activity
with small carbohydrate-derived molecules.20 The results from
this study suggest that hydrophobic moieties were also impor-
tant for IRI activity. This hypothesis was further validated using
small molecule lysine-based IRIs where the presence of long
alkyl chains resulted in potent activity.21 The orientation of
hydrophobic domains has also been suggested to be crucial for
the IRI activity exhibited by polyvinyl alcohol (PVA).22
In the current manuscript, we report the results of a thor-
ough structure–function study of ice recrystallization inhibitor
C-AFGP 1 (Fig. 2) with the goal of identifying the smallest
structural sub-unit of 1 that can exhibit potent IRI activity. The
results from this study will help develop a better understanding
of how these molecules are inhibiting ice recrystallization and
facilitate the rational design of small molecule IRIs.
Results and discussion
The correlation between length (number of tripeptide repeats)
and thermal hysteresis (TH) and IRI activities in AFGPs has
indicated that these activities are greatly reduced when less
than four repeating tripeptide units are present.11,12 Our labo-
ratory has also reported that IRI activity is directly correlated to
the number of tripeptide repeats in C-AFGP analogues similar
to 1 where the glycopeptide length ranged from three to nine
repeating tripeptide units.23 However the minimum number of
tripeptide repeats necessary for IRI activity in C-AFGP analogue
1 is not known. Thus, a systematic structure–function investi-
gation of 1 was envisioned whereby glycoconjugates containing
one, two and three tripeptide repeats (Fig. 2, compounds 2, 3
Fig. 3 IRI activity of glycopeptides 1–4 (at 22 mM and 5.5 mM) and galactosyl–
ornithine derivative 5 (at 22 mM). All compounds are represented as a % MGS
and 4, respectively) would be prepared and assessed for IRI (mean grain size) of ice crystals relative to the PBS positive control.
26006 | RSC Adv., 2013, 3, 26005–26009
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
RSC Advances
solution of D-galactose. Glycoconjugate 2 is approximately twice containing a butyl chain. The zwitterionic ornithine derivative
as active as 5, indicating that the addition of glycine units to the 5, with four carbons in the hydrocarbon chain, had activity that
C-terminal end of the galactosyl–ornithine 5 is benecial for was also similar to that of 17, but was slightly more active than
activity. The IRI activity of 3 (containing two tripeptide repeats) the parent compound 12 and carboxylic acid analogue 18.
is equal to that of 2 (containing one tripeptide repeat) while 4 is Interestingly, derivative 19 possessing a ve carbon chain with a
much more active. This data establishes that while glyco- terminal OH group is noticeably more active than the parent
conjugates containing three tripeptide repeats exhibit weak IRI compound 13 that is lacking this primary hydroxyl group.
activity, a minimum number of four tripeptide repeats is
Analogues containing various structural isomers of three,
necessary for potent IRI activity at 5.5 mM.
four and ve carbon alkyl chains were also prepared utilizing the
The effect of conjugating hydrocarbon chains of varying general synthetic scheme outlined in Fig. 4A. The structures and
lengths at the amide-linkage of galactosyl derivative 5 on IRI IRI activity of these derivatives are presented in Fig. 5. Isopropyl
activity was also investigated. Many of these derivatives were derivative 20 had similar IRI activity as propyl derivative 11, but
synthesized by coupling carboxylic acid derivative 6 (ref. 17) with activity decreased slightly as the size of the alkyl substituent was
the appropriate primary amine, according to the general scheme increased. The isobutyl (21), sec-butyl (22) and 1-ethylpropyl (23)
shown in Fig. 4A (full synthetic details for all derivatives can be derivatives were all less active than 11. Cyclopropane and
found in ESI†). Carboxylic acid 7 and primary amide 8 both cyclopentane derivatives 24 and 25 had activity that was similar
exhibited weak IRI activity (Fig. 4C). With hydrocarbon chains of to propyl derivative 11. This was surprising as 25 is a cyclic
1–6 carbons (9–14) activity ranged from weak to moderate. analogue of 23, but 23 was inactive. Unsaturated analogues such
Moderately active analogues are equipotent to a 22 mM D-galac- as allyl and propargyl derivatives 26 and 27 showed a signicant
tose solution. Derivatives with seven and eight carbon atoms in increase in IRI activity compared to propyl analogue 11.
the hydrocarbon chain (15 and 16, respectively) exhibited very However, when the double bond was in conjugation with the
high IRI activity (approximately 10% of PBS) at 22 mM.
amide nitrogen (1-propenamide analogues 28 and 29) or a
The introduction of ionizable functional groups at the end of terminal methyl group was introduced (crotyl amide derivative
the hydrocarbon chains appears to have little effect upon IRI 30) IRI activity decreased. All of the unsaturated derivatives still
activity as the ammonium salt derivative 17 and carboxylic acid exhibited activity that was similar to or better than D-galactose.
derivative 18 had similar activity to the parent compound 12 Thus, for the propyl, butyl and pentyl analogues assessed in this
study, it appears that introducing unsaturation rather than steric
bulk is more benecial for IRI activity.
Fig. 4 Synthesis and IRI activity of alkylated derivatives. (A) General scheme for
the preparation of alkyl derivatives from carboxylic acid 6. (B) Structure of
carboxylic acid 7, primary amide 8, and alkylated derivatives 9–19. (C) IRI activity
of derivatives 5 and 7–19 at 22 mM. All compounds are represented as a % MGS
(mean grain size) of ice crystals relative to the PBS positive control.
Fig. 5 Structure and IRI activity of derivatives containing various structural
isomers of propyl, butyl and pentyl hydrocarbon chains. (A) Structure of deriva-
tives 20–30. (B) IRI activity of n-propyl derivative 11 and derivatives 20–30 at
22 mM. All compounds are represented as a % MGS (mean grain size) of ice
crystals relative to the PBS positive control.
This journal is ª The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 26005–26009 | 26007

View Article Online
RSC Advances
Paper
Fig. 6 Structure and IRI activity of long alkyl chain derivatives. (A) Structure of long alkyl chain derivatives 15, 16, and 31–35. (B) IRI activity of long alkyl chain
derivatives 15, 16, and 31–35. Each analogue was first examined at its highest soluble concentration, then at subsequent lower dilutions. All compounds are rep-
resented as a % MGS (mean grain size) of ice crystals relative to the PBS positive control.
The initial investigation of the alkylated derivatives at 22 mM nonspecic toxic effects in cellular systems due to surfactant-like
showed a dramatic increase in IRI activity when the length of characteristics makes this particular class of compounds with
the hydrocarbon chain was seven carbon atoms or greater longer alkyl chains not suitable as small molecule ice recrystal-
(Fig. 4C). In an attempt to further optimize and increase the lization inhibitors.
activity of these small molecules a series of analogues were
The exact mechanism by which the alkylated-C-galactosyl
synthesized (Fig. 4A) that contained hydrocarbon tails greater derivatives examined in this study exhibit IRI activity is
than eight carbons (derivatives 31–35, Fig. 6A). The IRI activity currently not known. However, a noteworthy observation about
of these derivatives is presented in Fig. 6B. Unfortunately, many the solution behaviour of some of these derivatives occurred
of the alkylated derivatives with longer hydrocarbon-tails were while assessing their activity. Upon standing for a day, several of
insoluble in PBS at 22 mM and were subsequently tested in a the solutions appeared to form a hydrogel. Further investigation
range of concentrations from 22 mM to 5.5 mM, with each into this property found that the nonyl analogue (31) formed a
analogue assessed at the highest concentration it was soluble. hydrogel at 6 mM, the octyl analogue (16) formed a hydrogel at
The IRI activities of the derivatives with heptyl and octyl 36 mM and the heptyl analogue (15) formed a hydrogel at a high
hydrocarbon chains (15 and 16) were also assessed to allow for a concentration of 98 mM. In contrast, the hexyl and pentyl
direct comparison of how activity varies with concentration and analogues did not form hydrogels. Our laboratory has previ-
hydrocarbon-tail length.
ously investigated the correlation between hydrogelation ability
As evident in Fig. 6B, many of the long-chain alkyl analogues and IRI activity with various amphiphilic molecules. These
have signicant IRI activity even at low concentrations. The studies have indicated that the ability to form a hydrogel is not a
nonyl (31) and decyl (32) analogues are soluble in PBS at 5.5 mM prerequisite for potent IRI activity.20,21 Thus, while the heptyl,
and have very potent activity at this concentration. At 5.5 mM octyl and nonyl derivatives (15, 16 and 31) formed hydrogels at
octyl derivative 16 is more active than heptyl derivative 15, and high concentrations and exhibited potent IRI activity, this
extending the alkyl chain by one or two more carbons further activity is likely due to the increase in the hydrocarbon chain
increases activity (31 and 32). A ten-fold dilution to 0.55 mM length rather than due to hydrogelation ability.21
curtails much of the activity of these derivatives (15, 16, 31 and
32). Interestingly, these derivatives (31 and 32) and the dodecyl
derivative 33 have similar moderate IRI activity at 0.55 mM
Conclusions
despite the increase in hydrocarbon chain length. At this The initial goal of this work was to investigate the relationship
concentration, the longer dodecyl hydrocarbon chain (derivative between structure and IRI activity for glycopeptide 1. It is clear
33) resulted in only a slight increase in activity compared to the that activity decreases rapidly as the number of tripeptide
other derivatives. While longer chain analogues such as tetra- repeats is decreased (glycopeptides 2, 3, and 4). A minimum of
decyl derivative 34 and hexadecyl derivative 35 were soluble at three tripeptide repeats is required for potent IRI activity at
5.5 mM, they exhibited only moderate activity at this concentra- 22 mM and while this analogue is weakly active at 5.5 mM, four
tion. Thus, even though activity increases when the hydrocarbon tripeptide repeats are required to obtain potent IRI activity at
chain length is increased, the poor solubility, and moderate IRI this concentration. Galactosyl–ornithine building block 5 also
activity at low concentrations combined with potential exhibited moderate IRI activity at 22 mM and the level of activity
26008 | RSC Adv., 2013, 3, 26005–26009
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
RSC Advances
was similar to a 22 mM D-galactose solution. These results
suggest that the peptide backbone and the number of tripeptide
9 A. Fowler and M. Toner, Ann. N. Y. Acad. Sci., 2005, 1066,
119.
repeats are crucial for the potent IRI activity of C-linked AFGP 10 (a) J. Garner and M. M. Harding, ChemBioChem, 2010, 11,
analogue 1. However, conjugation of long alkyl chains to the
same galactosyl moiety in 1 showed promising results for IRI
active small molecules at low concentrations as derivatives with
octyl, nonyl and decyl chains (16, 31 and 32) retained potent
activity at 5.5 mM. These results suggest that the addition of
hydrophobic moieties (alkyl chains) is benecial for IRI activity,
2489; (b) R. Peltier, M. A. Brimble, J. M. Wojnar,
D. E. Williams, C. W. Evans and A. L. DeVries, Chem. Sci.,
2010, 1, 538; (c) M. I. Gibson, Polym. Chem., 2010, 1, 1141;
(d) C. J. Capicciotti, M. Doshi and R. N. Ben, in Recent
Developments in the Study of Recrystallization, ed. P. Wilson,
InTech, New York, 2013, pp. 177–224.
and are consistent with previous studies using carbohydrate- 11 Y. Tachibana, G. L. Fletcher, N. Fujitani, S. Tsuda, K. Monde
based and lysine-based surfactants/gelators as potent ice and S.-I. Nishimura, Angew. Chem., Int. Ed., 2004, 43, 856.
recrystallization inhibitors. However, while an increase in the 12 B. L. Wilkinson, R. S. Stone, C. J. Capicciotti, M. Thaysen-
length of the hydrocarbon chain resulted in increased IRI
Andersen, J. M. Matthews, N. H. Packer, R. N. Ben and
activity, longer alkyl chain derivatives 14 and 16 carbons in
R. J. Payne, Angew. Chem., Int. Ed., 2012, 51, 3606.
length were only soluble at 5.5 mM and exhibited moderate 13 R. Izumi, T. Matsushita, N. Fujitani, K. Naruchi, H. Shimizu,
activity. This work has demonstrated that it is possible to
develop small molecules that are IRI active at low concentra-
S. Tsuda, H. Hinou and S.-I. Nishimura, Chem.–Eur. J., 2013,
19, 3913.
tions and such molecules are targets for the development of 14 (a) J. A. Raymond, P. Wilson and A. L. DeVries, Proc. Natl.
other small molecule ice recrystallization inhibitors.
Acad. Sci. U. S. A., 1989, 86, 881; (b) A. L. DeVries, Comp.
Biochem. Physiol., Part B: Biochem. Mol. Biol., 1988, 90, 661.
15 (a) J. F. Carpenter and T. N. Hansen, Proc. Natl. Acad. Sci. U.
S. A., 1992, 89, 8953; (b) H. Chao, P. L. Davies and
J. F. Carpenter, J. Exp. Biol., 1996, 199, 2071; (c)
B. Rubinsky, A. Arav and A. L. DeVries, Cryobiology, 1992,
29, 69; (d) T. Wang, Q. Zhu, X. Yang, J. R. J. Layne and
A. L. DeVries, Cryobiology, 1994, 31, 185.
Acknowledgements
The authors acknowledge the Natural Sciences and Engineering
Research Council of Canada (NSERC), Canadian Blood Services
(CBS) and Canadian Foundation for Innovation (CFI) for
nancial support. The views expressed herein do not necessarily
represent the view of the federal government. J.F.T. was sup- 16 S. Liu and R. N. Ben, Org. Lett., 2005, 7, 2385.
ported by a NSERC PGS-D grant. C.J.C. thanks CBS for a GFP 17 P. Czechura, R. Y. Tam, E. Dimitrijevic, A. V. Murphy and
award.
R. N. Ben, J. Am. Chem. Soc., 2008, 130, 2928.
`
18 M. Leclere, B. K. Kwok, L. K. Wu, D. S. Allan and R. N. Ben,
Bioconjugate Chem., 2011, 22, 1804.
19 R. Y. Tam, S. S. Ferreira, P. Czechura, J. L. Chaytor and
R. N. Ben, J. Am. Chem. Soc., 2008, 130, 17494.
Notes and references
1 (a) A. L. DeVries and D. E. Wohlschlag, Science, 1969, 163,
`
1073; (b) A. L. DeVries, S. K. Komatsu and R. E. Feeney, 20 C. J. Capicciotti, M. Leclere, F. A. Perras, D. L. Bryce,
J. Biol. Chem., 1970, 245, 2901; (c) P. F. Scholander and
J. E. Maggert, Cryobiology, 1971, 8, 371.
H. Paulin, J. Harden, Y. Liu and R. N. Ben, Chem. Sci.,
2012, 3, 1408.
2 R. E. Feeney, T. S. Burcham and Y. Yeh, Annu. Rev. Biophys. 21 A. K. Balcerzak, M. Febbraro and R. N. Ben, RSC Adv., 2013, 3,
Biophys. Chem., 1986, 15, 59.
3232.
3 Y. Yeh and R. E. Feeney, Chem. Rev., 1996, 96, 601.
4 M. M. Harding, P. I. Anderberg and A. D. J. Haymet, Eur.
J. Biochem., 2003, 270, 1381.
5 J. A. Raymond and A. L. DeVries, Proc. Natl. Acad. Sci. U. S. A.,
1977, 74, 2589.
22 (a) R. C. Deller, T. Congdon, M. A. Sahid, M. Morgan,
M. Vatish, D. A. Mitchell, R. Notman and M. I. Gibson,
Biomater. Sci., 2013, 1, 478; (b) T. Congdon, R. Notman and
M. I. Gibson, Biomacromolecules, 2013, 14, 1578.
23 A. Eniade, M. Purushotham, R. N. Ben, J. B. Wang and
K. Horwath, Cell Biochem. Biophys., 2003, 38, 115.
6 (a) C. A. Knight, C. C. Cheng and A. L. DeVries, Biophys. J.,
1991, 59, 409; (b) C. A. Knight and A. L. DeVries, J. Cryst. 24 A. Chakrabartty and C. L. Hew, Eur. J. Biochem., 1991, 202,
Growth, 1994, 143, 301; (c) P. W. Wilson, CryoLetters, 1993,
1057.
14, 31.
25 C. A. Knight, J. Hallett and A. L. DeVries, Cryobiology, 1988,
25, 55.
26 J. Jackman, M. Noestheden, D. Moffat, J. P. Pezacki,
S. Findlay and R. N. Ben, Biochem. Biophys. Res. Commun.,
2007, 354, 340.
7 C. A. Knight, E. Driggers and A. L. DeVries, Biophys. J., 1993,
64, 252.
8 (a) P. C. Mazur, Science, 1970, 168, 939; (b) P. Mazur, Am.
J. Physiol., 1984, 247, C125.
This journal is ª The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 26005–26009 | 26009