Synthesis of a hemifluorinated amphiphile designed for self-assembly and two-dimensional crystallization of membrane protein

Article abstract of DOI:10.1016/j.tetlet.2008.02.043

The work reported herein deals with the synthesis and the preliminary physical-chemical analysis of new hemifluorinated surfactant made up of one fluorinated chain linked to a tricarboxylic acid polar head which is able to complex a Ni atom and should fav

Full text of DOI:10.1016/j.tetlet.2008.02.043

Available online at www.sciencedirect.com
Tetrahedron Letters 49 (2008) 2247–2250
Synthesis of a hemifluorinated amphiphile designed for self-assembly
and two-dimensional crystallization of membrane protein
a
a,
b
a,
*
*
´
Julien Dauvergne , Ange Polidori , Catherine Venien-Bryan , Bernard Pucci
a
`
´
Universite´ d’Avignon et des pays du vaucluse, laboratoire de chimie bioorganique et des systemes moleculaires vectoriels,
33 rue Louis Pasteur, F-84000 Avignon, France
b Laboratory of Molecular Biophysics, University of Oxford, South Park Road, OX1 3QU Oxford, UK
Received 3 January 2008; revised 4 February 2008; accepted 6 February 2008
Available online 13 February 2008
This work is dedicated to the memory of our colleague Dr. Charles Mioskowski
Abstract
The work reported herein deals with the synthesis and the preliminary physical–chemical analysis of new hemifluorinated surfactant
made up of one fluorinated chain linked to a tricarboxylic acid polar head which is able to complex a Ni atom and should favor the two-
dimensional crystallization of membrane proteins. Such a compound forms a Langmuir film which is a fluid at 20 °C and not perturbed
by the presence of hydrocarbon detergent in aqueous solution.
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords: Hemifluorinated surfactants; Langmuir film; Membrane proteins; Nickel complexation; 2D proteins crystallization
Knowledge of the three-dimensional structure of pro-
teins helps the understanding of biological mechanisms at
the molecular level. Atomic resolution structures can cur-
rently be determined by NMR of proteins in solution, or
by X-ray diffraction of three-dimensional (3D) crystals,
or by electron diffraction of two-dimensional (2D) crystals,
the latter using electron microscopy. Some proteins may
form 2D crystals on a lipid monolayer at an air–water
interface.1 This method may be improved by anchoring
the protein onto a Langmuir film using specific ligands
attached to the lipids or surfactants.2–5 The polar head of
a lipid monolayer can be loaded with Nickel atoms which
can bind to a protein expressed with histidines tag. After
a complex is formed between the Ni atom and polyhistidine
tag, translational and rotational diffusion within the film
can result in the protein becoming organized into 2D
arrays. The quality and the rate of protein crystallization
are strongly dependent on the intrinsic film properties,
notably its fluidity.6–8,3,9,10 However, this technique cannot
be applied to membrane proteins, as the surfactants or
detergents commonly used to handle these proteins in solu-
tion would solubilize the lipid monolayer.11 To overcome
this problem, Mioskowski et al. have designed lipids with
a bi-tailed fluorinated hydrophobic segment.12,11 Perfluori-
nated compounds do not mix with either hydrocarbons or
hydrophilic compounds, as they are lipophobic and hydro-
phobic. Therefore, fluorinated surfactant monolayers are
not destabilized by the presence of hydrocarbon detergents
in the aqueous solution.12–14
Furthermore, adding a hydrocarbon segment to the end
of the fluorinated chains minimizes the association of the
fluorinated chains and consequential formation of a
pseudo-crystal phase, which is detrimental to the formation
of a fluid-phase lipid layer and hence the crystallization of
the protein.
*
Corresponding authors. Tel.: +33 4 90 14 44 45; fax: +33 4 90 14 44 49
(A.P.); tel.: +33 4 90 14 44 42; fax: +33 4 90 14 44 49 (B.P.).
E-mail addresses: ange.polidori@univ-avignon.fr (A. Polidori),
bernard.pucci@univ-avignon.fr (B. Pucci).
Thorough investigations remain to be done in this field
since until now only a few hemifluorocarbon surfactants
0040-4039/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2008.02.043

2248
J. Dauvergne et al. / Tetrahedron Letters 49 (2008) 2247–2250
have been designed and prepared for specific biological
application.15,16 In this context, we have previously shown
that non-ionic perfluorinated surfactants (such as
C6F13SOTHAM) can lead to stable Newton Black Films
(NBFs). The possibility of organizing detergent-solubilized
membrane proteins in a plane within the core of these
NBFs was also demonstrated.17 From these preliminary
results, we report herein the synthesis and some physical–
chemical properties of a new hemifluorocarbon surfactant
called PhenylHFNTANi made up of one fluorinated chain
linked to a tricarboxylic acid polar head which is able to
complex a Ni atom. To provide the necessary fluidity of
the lipid monolayer made of this fluorinated surfactant,
the perfluorinated tail is end-capped with a phenyl propyl
group (Fig. 1).
tion of Ni2+ was carried out in a chloroform solution:
triacid derivative 10 was stirred with a phosphate buffer
solution at pH 8 containing 1 equiv of NiCl2ꢀ6H2O. After
1 h we observed a blue coloration of the organic phase
showing the complexation of nickel cation by the NTA
group. The organic phase was decanted and concentrated
under reduced pressure to obtain the nickel-chelating com-
pound PhenylHFNTANi, which was purified by size exclu-
sion chromatography. PhenylHFNTANi, was obtained as
a blue-green amorphous solid. This compound is weakly
soluble in water. The structure and purity of Phen-
ylHFNTA was assessed and confirmed by proton, 19F
and 13C NMR spectroscopy, and mass spectroscopy.19
We have investigated some physical properties of
PhenylHFNTANi and C6F13SOTHAM using tensiometry
measurements20 and surface pressure measurements.21
The surface tensions of C6F13SOTHAM and Phen-
ylHFNTANi have been determined when the surface was
saturated with surfactant molecules (at a concentration of
surfactant above the critical micelle concentration—cmc-
value). Three hundred microliters of a solution of
C6F13SOTHAM (50 mM in water) was spread on a 20 ml
water solution (final concentration of 0.75 mM, about
twice the cmc value, which is 0.37 mM). We measured a
value of 21.5 mN mꢁ1 for the surface tension. Successive
additions of the detergent DDM (n-dodecyl-b-D-maltopyr-
anoside from Anatrace, USA) to the subphase leading to
final concentrations in the range from 0.09 mM (0.5 ꢂ cmc)
to 1.08 mM (6 ꢂ cmc) did not modify the surface tension
value of C6F13SOTHAM, indicating that the langmuir film
was not disturbed by the presence of detergent in solution.
The surface tension value of a saturated monolayer of
PhenylHFNTANi was 33.8 mN mꢁ1 ( 200 ll of a solution
of 18 mM PhenylHFNTANi in chloroform/hexane 1/1
was spread on a 20 ml buffer subphase21 giving a final con-
centration of 0.2 mM, about twice the cmc which is
0.1 mM). This value being comparable to the surface ten-
sion value of the DDM alone (34.9 mN mꢁ1), we chose
another hydrocarbon surfactant (H12-TAC) to perform
the test of stability of the PhenylHFNTANi monolayer
upon the addition of detergent in the subphase. H12-
TAC, a telomer surfactant derived from trishydroxymethy-
laminomethane with a cmc of 0.5 mM, has been extensively
tested for its detergent properties.22 The surface tension
value of H12-TAC is 39.3 mN mꢁ1 at a concentration four
times higher than its cmc. Successive additions of H12-TAC
(leading to final concentrations in the range from 0.25 mM
to 1.5 mM) to the subphase of a monolayer of Phen-
ylHFNTANi did not modify its surface tension value,
which kept constant at 33.8 mN mꢁ1. This demonstrates
that H12-TAC is not miscible with a monolayer of Phen-
ylHFNTANi. These data agree with the results of previous
work where a Langmuir film made from C18F17NTANi (a
comparable compound to Phenyl NTANi) was not per-
turbed by the presence of detergent in the subphase.17
These results show that both surfactants C6F13SO-
THAM and PhenylHFNTANi are not miscible in the
O
O
F
F F F F F
H
N
H
N
OH2
OH2
N
Ni
O
F F F F F F
PhenylHFNTANi
F F F F
O
O
O
O
OH
OH
OH
O
F
F
F
S
O
N
H
F
F F F F F
C6F13SOTHAM
Fig. 1. Structures of the partially fluorinated PhenylHFNTANi and
fluorinated sulfoxide surfactant C6F13SOTHAM.
PhenylHFNTANi surfactant was synthesized following
a convergent synthetic pathway through the preparation
of the hemifluorinated hydrophobic part and the lysine-
derived nickel-chelating nitrilotriacetic group. The hemi-
fluorinated segment was first synthesized from 1,6-diiodo-
perfluorohexane as starting material (Scheme 1). Phenyl-
3-propene was grafted onto the perfluorocarbon chain by
using AIBN as the radical initiator. The monofunctional-
ization of the perfluorinated chain afforded compound 1
in 51% yield. Radical addition of Boc-allylamine 2 was
carried out with sodium dithionite as the radical initiator
following the method developed by Huang.18 Compound
3 so obtained was dehalogenated by using tributyltin
hydride in the presence of AIBN to give compound 4 in
70% yield. Thus, treatment with TFA in methylene chlo-
ride led to the deprotection of amino group to provide
compound 5. Second time around, the tricarboxylic polar
head was obtained from Ne-benzyloxycarbonyl L-lysine
carboxymethyl ester. Substitution of ethyl bromoacetate
by lysine amino group provided triester 6 in 89% yields.
Thus, the hydrogenolysis of benzyloxycarbonyl group of
lysine derivative led to an amino group on the lateral chain
which was coupled with DCI to gave the imidazoyl com-
pound 8. The latter was coupled to amino group of com-
pound 5 through an urea link to provide compound 9.
The saponification of carboxylate gave the nitrilotriacetic
(NTA) compound 10 in 76% yield. The complexation reac-

J. Dauvergne et al. / Tetrahedron Letters 49 (2008) 2247–2250
2249
F
F F F F F
CO2Me
NH2.HCl
(d)
I
ZNH
F
F F F F F
I
1
(e)
(f)
(a)
Br
CO2Et
NH2
NHBoc
2
I
F
F F F F
F
CO2Me
N
NHBoc
ZNH
CO2Et
F
F F F F
F
I
CO2Et
6
3
(g)
(b)
CO2Me
N
F F F F F
F
H2N
CO2Et
NHBoc
7
CO2Et
F
F F F F
(h)
F
4
(c)
O
CO2Me
N CO2Et
F F F F F
F F F F
F
NH3
CF3COO
N
NH
N
F
F
CO2Et
8
(i)
5
CO2Et
F
F F F F
F
H
H
N
N
N
CO2Et
F
F F F F
F
O
CO2Me
9
(j)
CO2H
F
F F F F
F
H
N
H
N
N
CO2H
F
F F F F
F
O
O
CO2H
10
(k)
O
O
Ni
O
F
F F F F
F
H
N
H
N
OH2
OH2
N
F F F F F F
O
O
O
PhenylHFNTANi
Scheme 1. Synthesis of hemifluorinated PhenylHFNTANi surfactant. Reagents and conditions: (a) MeONa (7 equiv), CH3CN, D, 24 h, 89%; (b) H2,
Pd–C 10%, MeOH, 7 bars, 3 h, 100%; (c) CDI (4 equiv), DMAP cat, CH2Cl2, 2 h, 57%; (d) I(CF2)6I, AIBN cat, CH3CN, D, 72 h, 51%; (e) Boc2O, Et3N,
CH2Cl2, 4 h, 48%; (f) NaHCO3, Na2S2O4, CH3CN/H2O: 3/1, 5 h, 55%; (g) Bu3SnH, AIBN, CH3CN, D, 24 h, 70%; (h) CH2Cl2/TFA: 4/1, 1 h, 100%; (i)
Et3N, DMAP cat, CH2Cl2, 20 h, 69%; (j) KOH 0.5 M, MeOH, 4 days, 76%; (k) NiCl2ꢀ6H2O, phosphate buffer pH 8, CHCl3, 48 h, 100%.
monolayer with two detergents, respectively, DDM and
H12-TAC which are generally used for the solubilization
of membrane proteins. This property makes these (hemi)-
fluorinated compounds very good candidates and suitable
for the 2D crystallization of membrane proteins at an
air–water interface.
PhenylHFNTANi is fluid at this temperature. Same behav-
ior was observed when the buffer subphase was replaced by
water (data not shown). The p–A isotherm of a monolayer
made from pure C6F13SOTHAM spread on a water sub-
phase was performed (data not shown), too. No phase
transition pressure or collapse was noticed. The fluidity
and high solubility of this compound in water might
explain this behavior.
In conclusion, we have shown that a langmuir film made
of C6F13SOTHAM or PhenylHFNTANi is not perturbed
by the presence of detergent in the subphase. Moreover,
PhenylHFNTANi and C6F13SOTHAM films are fluid at
The p–A isotherm of a monolayer made from pure
PhenylHFNTANi spread on a buffer solution (NaCl
250 mM; Tris 20 mM, pH 8.0) is shown in Figure 2. The
PhenylHFNTANi isotherm was stable up to 38 mN mꢁ1
The collapse area was 43 A . PhenylHFNTANi did not
.
2
˚
display a phase transition pressure at 20 °C, indicating that

2250
J. Dauvergne et al. / Tetrahedron Letters 49 (2008) 2247–2250
´
´
4. Lebeau, L.; Schultz, P.; Celia, H.; Mesini, P.; Nuss, S.; Klinger, C.;
Olland, S.; Oudet, P.; Mioskowski, C. In Handbook of Non-medical
Applications of Liposomes, Barenholz, Y.; Lasic, D. D. E. Eds.; 1996;
pp 153–186.
45
40
35
30
25
20
15
10
5
Π (mN.m-1)
5. Uzgiris, E. E.; Kornberg, R. D. Nature 1983, 301, 125–129.
6. Courty, S.; Lebeau, L.; Martel, L.; Lenne, P. F.; Balavoine, F.;
Dischert, W.; Konovalov, O.; Mioskowski, C.; Legrand, J. F.;
Venien-Bryan, C. Langmuir 2002, 18, 9502–9512.
7. Darst, S. A.; Ahlers, M.; Meller, P. H.; Kubalek, E. W.; Blankenburg,
R.; Ribi, H. O.; Ringsdorf, H.; Kornberg, R. D. Biophys. J. 1991, 59,
387–396.
8. Ellis, M. J.; Hebert, H. Micron 2001, 32, 541–550.
9. Mosser, G.; Brisson, A. J. Struct. Biol. 1991, 106, 191–198.
10. Nuss, S.; Mioskowski, C.; Lebeau, L. Chem. Phys. Lipids 1999, 103,
21–35.
11. Lebeau, L.; Lach, F.; Venien-Bryan, C.; Renault, A.; Dietrich, J.;
Jahn, T.; Palmgren, M. G.; Kuhlbrandt, W.; Mioskowski, C. J. Mol.
Biol. 2001, 308, 639–647.
0
0
50
100
150
200
Area per molecule (Å2)
Fig. 2. Surface pressure (p)–area isotherm of PhenylHFNTANi mono-
layer spread on buffer subphase at 20 °C.
20 °C. Both conditions are key features in the 2D crystalli-
zation of membrane proteins on a lipid monolayer experi-
ments. The membrane proteins are usually solubilized in
detergent. The monolayer matrix made from these two
compounds is not perturbed by the presence of detergent
in the subphase and therefore provide a suitable anchor
for the solubilized membrane proteins. Then there is a
direct correlation between the fluidity of the monolayer
and the crystallization process. Once the proteins are
adsorbed and concentrated at the lipid or surfactant mono-
layer, the organization of proteins within a 2D crystal
results from the in-plane diffusion of the protein–surfactant
complexes both in rotation and in translation. This diffu-
sion process is partly controlled by the fluidity property
of the Langmuir film. We have shown that C6F13SO-
THAM and PhenylHFNTANi films are fluid at 20 °C.
We think that these two newly designed components are
good candidates for the crystallization of membrane
proteins
12. Held, P.; Lach, F.; Lebeau, L.; Mioskowski, C. Tetrahedron Lett.
1997, 38, 1937–1940.
13. Mukerjee, P. Colloids Surf., A 1994, 84, 1–10.
14. Mukerjee, P.; Yang, A. Y. S. J. Phys. Chem. 1976, 80, 1388m–
1390m.
15. Lebaupain, F.; Salvay, A. G.; Olivier, B.; Durand, G.; Fabiano, A.-S.;
Michel, N.; Popot, J-. L.; Ebel, C.; Breyton, C.; Pucci, B. Langmuir
2006, 22, 8881–8890.
16. Polidori, A.; Presset, M.; Lebaupain, F.; Ameduri, B.; Popot, J-. L.;
Breyton, C.; Pucci, B. Bioorg. Med. Chem. Lett. 2006, 16, 5827–5831.
17. Petkova, V.; Benattar, J.-J.; Zoonens, M.; Zito, F.; Popot, J.-L.;
Polidori, A.; Jasseron, S.; Pucci, B. Langmuir 2007, 23, 4303–
4309.
18. Huang, W. Y. J. Fluorine Chem. 1992, 58, 1–8.
19. Selected spectra for PhenylHFNTA: 1H NMR (DMSO-d6) 7.28 (m,
5H, phenyl); 3.50 (m, 4H, NCH2COOH); 3.36 (t, 1H, CHLys); 3.08 (t,
2H, CH2NH); 2.96 (t, 2H, CH2NH); 2.71 (t, 2H, CH2Ph); 2.22 (m,
4H, CH2CF2); 1.85 (m, 2H, CH2bPh); 1.61 (m, 4H, CH2bNHCONH);
1.35 (m, 4H, CH2 Lys). 19F NMR (DMSO-d6) ꢁ113.23 (4F, s,
CH2CF2); ꢁ121.69 (4F, s, CF2bCH2); ꢁ123.18 (4F, s, (CF2)2). 13C
NMR (DMSO-d6) 174.27 (CO2H Lys); 173.5 (2CH2CO2H); 158.55
(CO urea); 141.37 (C phenyl); 126.6–128.9 (CH phenyl); 64.7 (CH
Lys); 53.7 (CH2CO2H); 39.7 (NHCONHCH2); 39.1 (CH2NHCO);
34.4 (CH2Ph); 28.1ꢁ30.2 (CH2); 21.5–23.6 (CH2).
Acknowledgments
Mass spectrum for PhenylNTANi: (ESI-MS in
a
preseþnce of
ammonium acetate, TOF analyser) m/z: 836 [MꢁNi+2NHþ4 ꢃ , m/z:
We are grateful to Dr R. Thomas and E. Latter for mak-
ing facilities, including the tensiometer and the Langmuir
822 [Mꢁ2H2O+2H]+.
20. Surface tension measurements were made on a Kruss (Hamburg,
¨
´
balance available to us. C. Venien-Bryan thanks the Wel-
Germany) K10T maximum pull digital tensiometer with a Pt/Ir ring.
Solution of PhenylHFNTANi (18 mM) was prepared in chloroform/
hexane (1:1 v/v). C6F13SOTHAM (50 mM) was prepared in water.
21. The surface pressure p, was measured using an automated Wilhelmy
film balance (Nima technology Ltd Coventry, UK). The pressure-
measuring systems was equipped with a filter paper (Whatman 541,
periphery 4 cm). The trough was made from a 535 cm2 Teflon-coated
brass. Solution of PhenylHFNTANi (18 mM) was prepared in
chloroform/hexane (1:1 v/v). C6F13SOTHAM (19 mM) was prepared
in water. Three microliters of the surfactants were spread on the
subphase. The spreading solvent was allowed to evaporate for 15 min
prior to compression. The monolayer was compressed at a speed
of 50 cm2/min. The substrate solution was either distilled water
(surface tension, 73 mN mꢁ1) or a buffer (NaCl 250 mM, Tris 20 mM,
pH 8.0).
come Trust for support (Project Grant 064775). This work
was supported by E.U. Specific Targeted Research Project
IMPS (Innovative tools for membrane protein structural
`
´
proteomics) and the ‘Ministere de l’Enseignement Superieur
et de la Recherche’ (France).
References and notes
1. Thompson, D. H.; Zhou, M.; Grey, J.; Kim, H.-K. Chem. Lett. 2007,
36, 956–975.
2. Dietrich, J.; Venien-Bryan, C. Strategies for Two-dimensional Crys-
tallization of Proteins Using Lipid Monolayers; Imperial College Press,
2005.
22. Pucci, B.; Maurizis, J.-C.; Pavia, A. A. Eur. Polym. J. 1991, 27, 1101–
1106.
3. Jap, B. K.; Zulauf, M.; Scheybani, T.; Hefti, A.; Baumeister, W.;
Aebi, U.; Engel, A. Ultramicroscopy 1992, 46, 45–84.