The effect of receptor clustering on vesicle-vesicle adhesion

Article abstract of DOI:10.1021/ja0657612

As part of our studies into how the localization of cell adhesion molecules into lipid rafts may affect cell adhesion, we developed Cu(1), a synthetic copper(iminodiacetate)-capped receptor able to phase separate from fluid phospholipid bilayers. The extent to which Cu(1) clustered into adhesive patches on the surface of vesicles could be controlled by changing vesicle composition. Extensive receptor phase separation significantly enhanced vesicle-vesicle adhesion; only vesicles with adhesive patches (blue fluorescence) adhered to their conjugate histidine-coated vesicles (red fluorescence) to form large vesicle aggregates (shown). Copyright

Full text of DOI:10.1021/ja0657612

Published on Web 10/25/2006
The Effect of Receptor Clustering on Vesicle-Vesicle Adhesion
Robert J. Mart, Kwan Ping Liem, Xi Wang, and Simon J. Webb*
Manchester Interdisciplinary Biocentre and the School of Chemistry, UniVersity of Manchester, 131 Princess St,
Manchester M1 7DN, U.K.
Received August 8, 2006; E-mail: s.webb@manchester.ac.uk
Chart 1. Fluorescent Lipids Cu(1) and Cu(2) with Conjugate
Histidine-Capped Lipid 3
The lateral phase separation of cell membrane constituents into
distinct regions called “lipid rafts” is central to important cellular
functions like signal transduction, cell adhesion, and endocytosis.1
These lipid rafts, phase separated domains composed of protein,
glycosphingolipids, and cholesterol, are liquid ordered “islands”
moving within a liquid disordered phospholipid bilayer matrix.2
The localization of cell adhesion molecules (CAMs) within lipid
rafts has recently been shown3 to play a key role in cell adhesion
and motility. Indeed the modification of adhesive properties caused
by concentrating CAMs within lipid rafts has important medical
consequences, affecting viral attachment to cells,1 lymphocyte and
axon chemotaxis,4,5 and cancer cell metastasis.3a
We have developed a perfluoroalkyl-pyrene membrane anchor
that enables synthetic lipids to phase separate within phospholipid
bilayers in the solid ordered (so) or liquid ordered (lo) phases, even
at membrane concentrations as low as 1% mol/mol.6a This motif
offers a unique opportunity to create lipid raft mimics; phase
separated domains of synthetic receptors floating within a fluid lo
matrix. Given the key role of lipid rafts in cell adhesion, we aimed
to create rafts of synthetic CAMs in the membranes of vesicles,
then study how they affect the extent of vesicle adhesion. These
simple mimics of cells would be crosslinked through relatively weak
Cu(iminodiacetate)-histidine (Cu(IDA)-His) bonds.7,8 This inter-
action, with K around 103 M-1, is similar in strength to natural
adhesive interactions, like selectin-sialyl Lewis X, which have
individual strengths less than 104 M-1.9 Thus, we hope that receptor
clustering will enhance these individually weak links as it does in
nature.10
The creation of vesicle assemblies is of great current interest,11
with both homogeneous vesicle assemblies12 and assemblies of
different types of vesicles13 being created. To create the latter, which
have more potential as tissue mimics and functional biomaterials,
we synthesized a complementary pair of CAM mimics: a Cu(IDA)-
capped lipid incorporating our perfluoroalkyl-pyrene motif, Cu(1),
and its conjugate receptor, an L-histidine-capped lipid 3 (Chart 1).
Cu(2) is a control compound that does not phase separate in vesicles
of any composition, yet is known to mediate vesicle aggregation
by poly L-histidine.7 The pyrene groups in Cu(1) and Cu(2) have
dual roles; the ratio of excimer to monomer emission intensity (E/M
ratio) reflects the extent of lipid phase separation, and they also
allow visualization of Cu(1) and Cu(2) containing vesicles by
fluorescence microscopy.
Given that previous studies had established that lipids containing
perfluoroalkyl groups formed domains in ordered bilayers,6 we
anticipated that Cu(1) would phase separate in DMPC/chol vesicles,
which are in the lo phase at 25 °C. However Cu(1) should not phase
separate from vesicles composed solely of DMPC; since these are
in the liquid disordered (ld) phase at 25 °C, Cu(1) will be distributed
evenly over the surface of the vesicles. As expected, fluorescence
spectroscopy showed significant phase separation for [Cu(1)-
DMPC/chol] vesicles, with E/M ) 0.5, while only a small amount
of excimer was observed in [Cu(1)-DMPC] vesicles, E/M ) 0.2.
Receptor Cu(2), at the same 5% mol/mol loading in DMPC/chol
vesicles, did not form any measurable excimer.
The formation of vesicle aggregates can be inferred from
increases in solution turbidity and directly observed using fluores-
cence microscopy. Job plots are particularly useful for detecting
vesicle aggregation as the absorbance of a noninteracting mixture
of vesicles can be predicted from the absorbances of the unmixed
suspensions; positive deviations from the predicted absorbance
indicate the formation of vesicle aggregates. Mixing our control
vesicles without any phase separation, [Cu(2)-DMPC/chol], with
[3-DMPC/chol] vesicles gave noninteracting mixtures, with no
significant deviation from the expected absorbance and no ag-
gregates observable by fluorescence microscopy. Given that previ-
ous work had shown that poly L-histidine can cross-link vesicles
containing Cu(2),7 this suggests that multivalent interactions with
polymeric ligands are stronger than analogous multivalent interac-
tions between vesicles. In contrast, a positive deviation in absorb-
Vesicles (0.8 µm diameter) containing 5% mol/mol of lipids H21,
H22, or 3 (at 1 mM) in dimyristoyl phosphatidylcholine (DMPC)
or DMPC mixed 50 % w/w with cholesterol (DMPC/chol) were
formed by extrusion. Lipids H21, H22, or 3 were admixed into the
appropriate phospholipids prior to vesicle formation; the affinity
of the chelating headgroups in H21 and H22 for copper(II) is around
4 × 107 M-1,7 so later addition of 1 equiv of copper(II) afforded
synthetic receptors Cu(1) and Cu(2) quantitatively.
9
14462
J. AM. CHEM. SOC. 2006, 128, 14462-14463
10.1021/ja0657612 CCC: $33.50 © 2006 American Chemical Society

C O M M U N I C A T I O N S
phatidylcholine (DSPC) (so phase at 25 °C), which have extensive
phase separation of Cu(1) (E/M ratio of 0.8), then even larger
aggregates were observed by fluorescence microscopy.14
To the best of our knowledge, this is the first time a biomimetic
system has been used to show that the lateral distribution of adhesive
agents within a membrane can directly affect the membrane’s
adhesive properties. It is revealing that partitioning receptors into
adhesive clusters, acknowledged as a major factor in controlling
the adhesive properties of cells, can be replicated and shown to
enhance the adhesive properties of vesicles. This binding enhance-
ment may stem from receptor preorganization, where part of the
entropic penalty inherent in receptor clustering at a binding interface
is prepaid though lipid phase separation.15 Alternatively, the chelate
effect may enhance vesicle-vesicle binding; after formation of the
first crosslinking bond to phase-separated Cu(1), this tethering link
between vesicles will facilitate additional intervesicular bonds to
other Cu(1) receptors in the same phase separated patch.16 We now
hope to exploit this discovery to develop biocompatible tissue
mimics that are structured on a submicrometer scale.
Figure 1. Job plots showing changes in turbidity observed for mixtures of
5% mol/mol 3 in DMPC/chol vesicles with vesicles containing 5% mol/
mol Cu(1) in DMPC (O) and DMPC/chol (b).
Acknowledgment. This work was supported by BBSRC Grant
B19722. We would like to thank Dr. Jason Micklefield for use of
a UV-vis spectrophotometer and the EPSRC National Mass
Spectrometry Service Centre, Swansea, U.K. for HRMS.
Figure 2. Fluorescence micrographs of vesicles containing 5% mol/mol
lipid 3 in DMPC/chol (red) mixed with vesicles containing 5% mol/mol
Cu(1) (blue) in (a) DMPC and (b) DMPC/chol.
Supporting Information Available: Syntheses and spectroscopic
data for H21, H22, and 3. Details of turbidimetric, fluorimetric, and
microscopic analyses. Emission spectra of [H21-DMPC/chol] and
[H21-DMPC] vesicles. Fluorescence micrographs of [3-DMPC/chol]
vesicles with [Cu(2)-DMPC/chol] or [Cu(1)-DSPC] vesicles.
ance of 11% was found when [Cu(1)-DMPC/chol] vesicles were
mixed with [3-DMPC/chol] vesicles, implying the phase separation
of Cu(1) was enabling the formation of vesicle aggregates (Figure
1). In line with this explanation, mixing [3-DMPC/chol] vesicles
with [Cu(1)-DMPC] vesicles, which contain only weakly phase
separated Cu(1), gave a much smaller (<2%) deviation in the
absorbance at 700 nm.
To confirm the presence of vesicle aggregates, these three
mixtures were analyzed by fluorescence microscopy. Since lipid 3
lacks a fluorophore, vesicles containing 3 were doped with 0.1%
mol/mol of the red fluorescent membrane dye, rhodamine B DPPE,
which contrasts with the blue fluorescence of Cu(1) and Cu(2).
Fluorescence microscopy revealed that mixtures with little deviation
in turbidity contained only isolated vesicles ([Cu(2)-DMPC/chol]
+ [3-DMPC/chol]) or small clusters of vesicles ([Cu(1)-DMPC]
+ [3-DMPC/chol]). In contrast, mixtures of [Cu(1)-DMPC/chol]
and [3-DMPC/chol] vesicles showed large aggregates containing
both types of vesicle (Figure 2).
These aggregates, with diameters between 20 and 80 µm, were
fully formed within 2 min of mixing the two vesicle populations.
The aggregates seemed stable for several days, and we observed
no movement of the fluorescent synthetic lipids between the
different vesicle populations.
The formation of these distinctive aggregates seems to depend
upon the degree of phase separation of the Cu(IDA) lipid. If the
structure of the membrane anchor in Cu(1) was changed to give a
lipid, Cu(2), that no longer phase separated in the bilayer, then no
adhesion was observed. If membrane composition was changed to
diminish phase separation of Cu(1), then large vesicle aggregates
did not form. Furthermore, if [3-DMPC/chol] vesicles were mixed
with vesicles containing 5% mol/mol Cu(1) in distearoyl phos-
References
(1) Binder, W. H.; Barragan, V.; Menger, F. M. Angew. Chem., Int. Ed. 2003,
42, 5802-5827.
(2) (a) Pike, L. J. Biochem. J. 2004, 378, 281-292. (b) Vereb, G.; Szo¨llo¨si,
J.; Matko´, J.; Nagy, P.; Farkas, T.; V´ıgh, L.; Ma´tyus, L.; Waldmann, T.
A.; Damjanovich, S. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 8053-8058.
(3) (a) Huang, Q.; Shen, H.-M.; Shui, G.; Wenk, M. R.; Ong, C.-N. Cancer
Res. 2006, 66, 5807-5815. (b) Causeret, M.; Taulet, N.; Comunale, F.;
Favard, C.; Gauthier-Rouviere, C. Mol. Biol. Cell 2005, 16, 2168-2180.
(4) Pizzo, P.; Viola, A. Microbes Infect. 2004, 6, 686-692.
(5) Ibanez, C. F. Neuron 2004, 42, 3-5.
(6) (a) Webb, S. J.; Greenaway, K.; Bayati, M.; Trembleau, L. Org. Biomol.
Chem. 2006, 4, 2399-2407. (b) Elbert, R.; Folda, T.; Ringsdorf, H. J.
Am. Chem. Soc. 1984, 106, 7687-7692.
(7) Webb, S. J.; Trembleau, L.; Mart, R. J.; Wang, X. Org. Biomol. Chem.
2005, 3, 3615-3617.
(8) Kent, M. S.; Yim, H.; Sasaki, D. Y.; Satija, S.; Seo, Y.-S.; Majewski, J.
Langmuir 2005, 21, 6815-6824.
(9) Jacob, G. S.; Kirmaier, C.; Abbas, S. Z.; Howard, S. C.; Steininger, C.
N.; Welply, J. K.; Scudder, P. Biochemistry 1995, 34, 1210-1217.
(10) (a) Vitte, J.; Benoliel, A.-M.; Eymeric, P.; Bongrand, P.; Pierres, A.
Biophys. J. 2004, 86, 4059-4074. (b) Chen, A.; Moy, V. T. Biophys. J.
2000, 78, 2814-2820. (c) Cohen, M.; Joester, D.; Geiger, B.; Addadi, L.
ChemBioChem 2004, 5, 1393-1399.
(11) Paleos, C. M.; Tsiourvas, D. Top. Curr. Chem. 2003, 227, 1-29.
(12) (a) Tsogas, I.; Tsiourvas, D.; Nounesis, G.; Paleos, C. M. Langmuir 2005,
21, 5997-6001. (b) Menger, F. M.; Zhang, H. J. Am. Chem. Soc. 2006,
128, 1414-1415. (c) Kisak, E. T.; Kennedy, M. T.; Trommeshauser, D.;
Zasadzinski, J. A. Langmuir 2000, 16, 2825-2831.
(13) (a) Marchi-Artzner, V.; Gulik-Krzywicki, T.; Guedeau-Boudeville, M.-
A.; Gosse, C.; Sanderson, J. M.; Dedieu, J.-C.; Lehn, J.-M. ChemPhys-
Chem 2001, 2, 367-376. (b) Menger, F. M.; Seredyuk, V. A. J. Am.
Chem. Soc. 2003, 125, 11800-11801.
(14) However there was evidence of vesicle fusion in this case; vesicles up to
5 µm in diameter were observed. See the Supporting Information.
(15) (a) Marchi-Artzner, V.; Lorz, B.; Gosse, C.; Jullien, L.; Merkel, R.; Kessler,
H.; Sackmann, E. Langmuir 2003, 19, 835-841.
(16) Doyle, E. L.; Hunter, C. A.; Phillips, H. C.; Webb, S. J.; Williams, N. H.
J. Am. Chem. Soc. 2003, 125, 4593-4599.
JA0657612
9
J. AM. CHEM. SOC. VOL. 128, NO. 45, 2006 14463