Synthetic mimics of small mammalian cell surface receptors

Article abstract of DOI:10.1021/ja046663o

Receptors on the surface of mammalian cells promote the uptake of cell-impermeable ligands by receptor-mediated endocytosis. To mimic this process, we synthesized small molecules designed to project anti-dinitrophenyl antibody-binding motifs from the surface of living Jurkat lymphocytes. These synthetic receptors comprise N-alkyl derivatives of 3β-cholesterylamine as the plasma membrane anchor linked to 2,4-dinitrophenyl (DNP) and structurally similar fluorescent 7-nitrobenz-2-oxa-1,3-diazole (NBD) headgroups. Insertion of two β-alanine subunits between a DNP derivative and 3β- cholesterylamine yielded a receptor that avidly associates with cell surfaces (cellular t_1/2 ～ 20 h). When added to Jurkat cells at 10 μM, this receptor enhanced uptake of an anti-DNP IgG ligand by ～200-fold in magnitude and ～400-fold in rate within 4 h (ligand internalization t _1/2 ～ 95 min at 37°C). This non-natural receptor mimics many natural receptors by dynamically cycling between plasma membranes and intracellular endosomes (recycling t_1/2 ～ 3 min), targeting of protein ligands to proposed cholesterol and sphingolipid-enriched lipid raft membrane microdomains, and delivery of protein ligands to late endosomes/lysosomes. Quantitative dithionite quenching of fluorescent extracellular NBD headgroups demonstrated that other 3β-cholesterylamine derivatives bearing fewer β-alanines in the linker region or N-acyl derivatives of 3β-cholesterylamine were less effective receptors due to more extensive trafficking to internal membranes. Synthetic cell surface receptors have potential applications as cellular probes, tools for drug delivery, and methods to deplete therapeutically important extracellular ligands.

Full text of DOI:10.1021/ja046663o

Published on Web 11/25/2004
Synthetic Mimics of Small Mammalian Cell Surface Receptors
Siwarutt Boonyarattanakalin, Scott E. Martin, Sheryl A. Dykstra, and
Blake R. Peterson*
Contribution from the Department of Chemistry, The PennsylVania State UniVersity,
104 Chemistry Building, UniVersity Park, PennsylVania 16802
Received June 7, 2004; E-mail: brpeters@chem.psu.edu
Abstract: Receptors on the surface of mammalian cells promote the uptake of cell-impermeable ligands
by receptor-mediated endocytosis. To mimic this process, we synthesized small molecules designed to
project anti-dinitrophenyl antibody-binding motifs from the surface of living Jurkat lymphocytes. These
synthetic receptors comprise N-alkyl derivatives of 3â-cholesterylamine as the plasma membrane anchor
linked to 2,4-dinitrophenyl (DNP) and structurally similar fluorescent 7-nitrobenz-2-oxa-1,3-diazole (NBD)
headgroups. Insertion of two â-alanine subunits between a DNP derivative and 3â-cholesterylamine yielded
a receptor that avidly associates with cell surfaces (cellular t1/2 ∼ 20 h). When added to Jurkat cells at 10
µΜ, this receptor enhanced uptake of an anti-DNP IgG ligand by ∼200-fold in magnitude and ∼400-fold
in rate within 4 h (ligand internalization t1/2 ∼ 95 min at 37 °C). This non-natural receptor mimics many
natural receptors by dynamically cycling between plasma membranes and intracellular endosomes (recycling
t1/2 ∼ 3 min), targeting of protein ligands to proposed cholesterol and sphingolipid-enriched lipid raft
membrane microdomains, and delivery of protein ligands to late endosomes/lysosomes. Quantitative
dithionite quenching of fluorescent extracellular NBD headgroups demonstrated that other 3â-cholesteryl-
amine derivatives bearing fewer â-alanines in the linker region or N-acyl derivatives of 3â-cholesterylamine
were less effective receptors due to more extensive trafficking to internal membranes. Synthetic cell surface
receptors have potential applications as cellular probes, tools for drug delivery, and methods to deplete
therapeutically important extracellular ligands.
Introduction
internalizing cell surface receptors generally results in clustering
of receptor/ligand complexes in plasma membrane subdomains
Receptors on the surface of mammalian cells function as
sensors and mediators of uptake of specific ligands in the
extracellular environment. These biomolecules reside on the
cellular plasma membrane, the protective lipid bilayer that
envelops the cell to guard the inner cellular machinery from
potentially toxic or opportunistic extracellular materials. Only
small hydrophobic compounds passively diffuse across this
hydrophobic barrier to rapidly penetrate into cells. More
hydrophilic macromolecules, such as proteins and DNA, gener-
ally require specific active transport mechanisms to access the
cell interior. Cell surface receptors are components of this active
transport machinery and deliver cargo, such as nutrients and
growth factors, into cells via the uptake process of receptor-
mediated endocytosis.1
that invaginate and pinch off as intracellular endocytic vesicles.
These internalized vesicles become increasingly acidified by
the activation of proton pumps and fuse in the cytoplasm to
form larger acidic (pH ∼ 6) endosomes.4 The altered chemical
environment within endosomes typically dissociates internalized
receptor/ligand complexes. The free receptor can traffic to other
endosomes, can cycle back to the cell surface via exocytosis,
and can often be reused for uptake of ligand up to several
hundred times.4 The free ligand either exits endosomes and
proceeds to other intracellular locations or more often becomes
trapped in lysosomes, more acidic (pH ∼ 5.0) organelles
containing degratory enzymes that are highly active at this low
pH.4
Cell surface receptors involved in receptor-mediated endocy-
tosis have been extensively targeted to enable cellular uptake
of poorly permeable molecules. Ligands of these receptors have
been used to deliver linked drugs, molecular probes, and
macromolecules into cells.5 Receptors targeted in this way
include the glycosylphosphatidylinositol (GPI) lipid-anchored
Mammalian cell surface receptors comprise at least three basic
components: a ligand recognition motif projecting from the cell
surface, a linker region, and a membrane-binding element that
anchors the receptor to the plasma membrane. As an example,
the small receptor ganglioside GM1 (1, Figure 1) projects a
pentasaccharide headgroup from the cell surface that binds the
bacterial protein cholera toxin as a ligand. This molecular
recognition event enables cholera toxin to penetrate into cells
by receptor-mediated endocytosis.2,3 Binding of ligands to
(2) Smith, D. C.; Lord, J. M.; Roberts, L. M.; Johannes, L. Semin. Cell. DeV.
Biol. 2004, 15, 397-408.
(3) Sandvig, K.; van Deurs, B. FEBS Lett. 2002, 529, 49-53.
(4) Maxfield, F. R.; McGraw, T. E. Nat. ReV. Mol. Cell. Biol. 2004, 5, 121-
132.
(5) Vyas, S. P.; Singh, A.; Sihorkar, V. Crit. ReV. Ther. Drug Carrier Syst.
(1) Conner, S. D.; Schmid, S. L. Nature 2003, 422, 37-44.
2001, 18, 1-76.
9
10.1021/ja046663o CCC: $27.50 © 2004 American Chemical Society
J. AM. CHEM. SOC. 2004, 126, 16379-16386
16379

A R T I C L E S
Boonyarattanakalin et al.
Figure 1. Structures of the small natural receptor ganglioside GM1, novel synthetic receptors, and related control compounds.
folate receptors,6-8 transferrin receptors,9 and low-density
lipoprotein (LDL) receptors10 that are often overexpressed on
rapidly proliferating cells.
As an alternative to natural receptors targeted for the delivery
of impermeable molecules, our laboratory is investigating
synthetic mimics of cell surface receptors. Synthetic receptors
derived from 3â-cholesterylamine linked to fluorescein, biotin,
and peptides as protein-binding motifs have been directly loaded
into plasma membranes of living mammalian cells to promote
the endocytosis of impermeable protein ligands.11-13 This
approach has been termed “synthetic receptor targeting” (Figure
2).12 Other elegant approaches for cell surface engineering have
also been reported14-18 and reviewed.19 A chemical approach
for rewiring related cellular adhesion processes has also been
described.20 The cellular uptake of ligands mediated by 3â-
cholesterylamine-derived receptors has been proposed12 to
mimic endocytic cellular penetration mechanisms employed by
Figure 2. Synthetic receptor targeting approach for enhancing the cellular
uptake of impermeable ligands. Synthetic mimics of cell surface receptors
are added to living mammalian cells. These cells internalize cognate ligands,
such as macromolecular antibodies (IgG), bound to bacterial protein A (PrA)
by synthetic receptor-mediated endocytosis.
certain protein toxins and viruses.21 These mechanisms are
highly complex at the molecular level and are thought to involve
plasma membrane subdomains termed lipid rafts22,23 and/or
endocytosis involving the protein clathrin.1,24,25
Lipid rafts are proposed to comprise liquid-ordered phases
of the cellular plasma membrane enriched in cholesterol and
sphingolipids, such as ganglioside GM1 (1).23,26,27 These
domains have recently been visualized by two-photon micro-
scopy studies of living cells,28 and aspects of lipid rafts have
been reconstituted in model systems.29 Many proteins covalently
or noncovalently associated with cholesterol, sphingolipids, or
saturated lipids of cellular membranes are thought to associate
(6) Lu, Y.; Sega, E.; Leamon, C. P.; Low, P. S. AdV. Drug DeliVery ReV. 2004,
56, 1161-1176.
(7) Rui, Y. J.; Wang, S.; Low, P. S.; Thompson, D. H. J. Am. Chem. Soc.
1998, 120, 11213-11218.
(8) Lee, R. J.; Low, P. S. J. Biol. Chem. 1994, 269, 3198-3204.
(9) Qian, Z. M.; Li, H.; Sun, H.; Ho, K. Pharmacol. ReV. 2002, 54, 561-587.
(10) Chung, N. S.; Wasan, K. M. AdV. Drug DeliVery ReV. 2004, 56, 1315-
1334.
(11) Hussey, S. L.; He, E.; Peterson, B. R. J. Am. Chem. Soc. 2001, 123, 12712-
12713.
(12) Martin, S. E.; Peterson, B. R. Bioconjugate Chem. 2003, 14, 67-74.
(13) Hussey, S. L.; Peterson, B. R. J. Am. Chem. Soc. 2002, 124, 6265-6273.
(14) de Kruif, J.; Tijmensen, M.; Goldsein, J.; Logtenberg, T. Nat. Med. 2000,
6, 223-227.
(21) Manes, S.; del Real, G.; Martinez, A. C. Nat. ReV. Immunol. 2003, 3, 557-
568.
(15) Saxon, E.; Bertozzi, C. R. Science 2000, 287, 2007-2010.
(16) Mahal, L. K.; Yarema, K. J.; Bertozzi, C. R. Science 1997, 276, 1125-
1128.
(22) Brown, D. A.; London, E. Annu. ReV. Cell DeV. Biol. 1998, 14, 111-136.
(23) Simons, K.; Ikonen, E. Science 2000, 290, 1721-1726.
(24) Kirchhausen, T. Annu. ReV. Biochem. 2000, 69, 699-727.
(25) Schmid, S. L. Annu. ReV. Biochem. 1997, 66, 511-548.
(26) Brown, D. A.; London, E. J. Biol. Chem. 2000, 275, 17221-17224.
(27) Maxfield, F. R. Curr. Opin. Cell Biol. 2002, 14, 483-487.
(28) Gaus, K.; Gratton, E.; Kable, E. P.; Jones, A. S.; Gelissen, I.; Kritharides,
L.; Jessup, W. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 15554-15559.
(29) Edidin, M. Annu. ReV. Biophys. Biomol. Struct. 2003, 32, 257-283.
(17) Prescher, J. A.; Dube, D. H.; Bertozzi, C. R. Nature 2004, 430, 873-877.
(18) George, N.; Pick, H.; Vogel, H.; Johnsson, N.; Johnsson, K. J. Am. Chem.
Soc. 2004, 126, 8896-8897.
(19) Kellam, B.; De Bank, P. A.; Shakesheff, K. M. Chem. Soc. ReV. 2003, 32,
327-337.
(20) Kato, M.; Mrksich, M. J. Am. Chem. Soc. 2004, 126, 6504-6505.
9
16380 J. AM. CHEM. SOC. VOL. 126, NO. 50, 2004

Synthetic Mimics of Small Mammalian Cell Surface Receptors
A R T I C L E S
Scheme 1 a
with lipid rafts30 that segregate and concentrate membrane
proteins, regulate the activation of specific signal transduction
pathways,31 and control the endocytosis of specific receptors.32
Binding of cholera toxin to the lipid raft-associated receptor
ganglioside GM1 (1) promotes uptake of the toxin by endocy-
tosis. In cell lines, such as lymphocytes, that lack the raft-
associated protein caveolin, this endocytosis is thought to be
mediated by the protein clathrin.33 However, the molecular
mechanisms linking 1 to clathrin are not yet well understood,
and small receptors, such as 1, are also thought to internalize
ligands via distinct lipid raft-dependent endocytic pathways.33,34
We report here the construction of novel synthetic mimics
of small cell surface receptors (2-5) and related green
fluorescent analogues (6-10) as mechanistic probes (Figure 1).
When added to living mammalian cells, these compounds were
designed to insert in the cellular plasma membrane, project
protein-binding headgroups from the cell surface, capture
cognate soluble antibodies (IgG), and internalize these macro-
molecular ligands. We found that the most efficient synthetic
receptor (2) mimics many natural receptors by avidly associating
with the cell surface, rapidly cycling between plasma membranes
and intracellular endosomes, targeting of protein ligands to
cholesterol and sphingolipid-enriched lipid rafts, and delivery
of protein ligands to late endosomes/lysosomes. Remarkably,
subtle molecular modifications substantially altered the traf-
ficking of these compounds to internal membranes, resulting
in modulation of the efficiency of synthetic receptors as delivery
systems for impermeable ligands.
a Reagents and conditions: (a) oxalyl chloride, DMSO, CH2Cl2, TEA,
-78 °C; (b) L-selectride, THF, -78 °C; (c) PPh3, HN3, DEAD, benzene;
(d) LiAlH4, Et2O, 0 °C; (e) 2-nitrobenzenesulfonyl chloride, DIEA, THF;
(f) boc-3-chloropropylamine, K2CO3, DMA, 120 °C; (g) TFA, CH2Cl2 (2:
25); (h) 6-(2,4-dinitrophenyl)aminohexanoic acid succinimidyl ester, DIEA,
CH2Cl2; (i) PhSH, K2CO3, THF/DMF (1:4); (j) 6-(N-(7-nitrobenz-2-oxa-
1,3-diazol-4-yl)amino)hexanoic acid succinimidyl ester, DIEA, CH2Cl2; (k)
boc-â-alanine NHS ester, DIEA, CH2Cl2; (l) 5-carboxyoregongreen NHS
ester, DIEA, DMF.
Results
Design and Synthesis of Receptors for Anti-DNP IgG
Ligands. To enable mammalian cells to capture and internalize
anti-dinitrophenyl IgG ligands, synthetic receptors were designed
that incorporate 2,4-dinitrophenyl (DNP) and structurally similar
green fluorescent 7-nitrobenz-2-oxa-1,3-diazole (NBD) head-
groups. These headgroups were linked to N-alkyl and N-acyl
derivatives of 3â-cholesterylamine via tethers containing 6-ami-
nohexanoic acid and â-alanine subunits (Figure 1). To evaluate
the affinity of these headgroups for rabbit polyclonal anti-DNP
IgG, related fluorescent derivatives were evaluated with fluo-
rescence polarization assays. These assays indicated that both
DNP and fluorescent NBD derivatives bind tightly to this IgG
with apparent Kd values of 23 ( 1.5 nM (DNP) and 820 ( 144
nM (NBD) (data shown in Figure S1, Supporting Information).
The weaker binding of NBD compared with that of DNP is
consistent with previous biophysical studies of similar receptor/
ligand systems.35 Detailed binding studies of anti-DNP IgG to
DNP-linked lipids embedded in model membranes on a solid
support were recently reported.36
of cholesterol (11) to ketone 12, provided the nosyl-protected
3â-cholesterylamine 15 in 40% yield over five steps. This
compound (15) was elaborated to receptors 2-10 by Fukuya-
ma’s amine synthesis methodology39 and/or deprotection and
sequential amide bond formation reactions. Synthetic procedures
and compound characterization data are provided in the Sup-
porting Information.
The Linker Region of Synthetic Receptors Controls the
Magnitude and Kinetics of Ligand Uptake. Synthetic recep-
tors (2-9) were evaluated as mediators of cellular uptake of
protein ligands in living Jurkat lymphocytes (a human T-cell
line). Confocal laser scanning microscopy and flow cytometry
were employed to examine cells that were treated with synthetic
receptors, washed to remove unincorporated compounds, and
subsequently treated with a polyclonal rabbit anti-DNP antibody
(IgG) complexed with fluorescent conjugates of the IgG-binding
protein A (PrA) from Staphylococcus aureus.40,41 Confocal
microscopy of cells treated with the green fluorescent NBD
derivative 6 (10 µM, 1 h) and anti-DNP IgG-labeled red
fluorescent with PrA-Alexa Fluor-594 (PrA-AF594, 4 h)
revealed receptor 6 localized both on the cell surface (embedded
in the plasma membrane) and in intracellular compartments
(Figure 3, panel A). The red fluorescent IgG ligand was taken
To construct synthetic receptors 2-10, a previously reported37
synthesis of 3â-cholesterylamine was modified, as shown in
Scheme 1. This approach, incorporating a Swern oxidation38
(30) Zacharias, D. A.; Violin, J. D.; Newton, A. C.; Tsien, R. Y. Science 2002,
296, 913-916.
(31) Simons, K.; Toomre, D. Nat. ReV. Mol. Cell. Biol. 2000, 1, 31-39.
(32) Ikonen, E. Curr. Opin. Cell Biol. 2001, 13, 470-477.
(33) Singh, R. D.; Puri, V.; Valiyaveettil, J. T.; Marks, D. L.; Bittman, R.;
Pagano, R. E. Mol. Biol. Cell 2003, 14, 3254-3265.
(38) Mancuso, A. J.; Swern, D. Synthesis 1981, 3, 165-185.
(39) Fukuyama, T.; Jow, C. K.; Cheung, M. Tetrahedron Lett. 1995, 36, 6373-
(34) Nichols, B. J. Curr. Biol. 2003, 13, 686-690.
6374.
(35) Lancet, D.; Pecht, I. Biochemistry 1977, 16, 5150-5157.
(36) Yang, T.; Baryshnikova, O. K.; Mao, H.; Holden, M. A.; Cremer, P. S. J.
Am. Chem. Soc. 2003, 125, 4779-4784.
(40) Cedergren, L.; Andersson, R.; Jansson, B.; Uhlen, M.; Nilsson, B. Protein
Eng. 1993, 6, 441-448.
(41) Tashiro, M.; Montelione, G. T. Curr. Opin. Struct. Biol. 1995, 5, 471-
(37) Kan, C. C.; Yan, J.; Bittman, R. Biochemistry 1992, 31, 1866-1874.
481.
9
J. AM. CHEM. SOC. VOL. 126, NO. 50, 2004 16381

A R T I C L E S
Boonyarattanakalin et al.
treated with receptor 6 followed by excess anti-DNP IgG/PrA-
AF594 for 4 h revealed that the protein ligand did not deplete
the receptor from the cell surface, despite extensive uptake of
this protein complex. These and other results presented here
are consistent with a delivery mechanism involving binding of
the IgG/PrA ligand to the synthetic receptor 6 at the cell surface,
internalization of this complex by endocytosis, dissociation of
IgG from 6 in acidic endosomes, and return of receptor 6 to
the cell surface by plasma membrane recycling.
To investigate the efficiency of uptake mediated by the higher
affinity DNP-based receptors 2-5, the magnitude and kinetics
of delivery of a green fluorescent anti-DNP/PrA-AF488 complex
were quantified by flow cytometry (Figure 3, panels B and C).
Since this technique does not distinguish cell surface-associated
fluorescence from intracellular fluorescence, cells were washed
with the hapten 6-(2,4-dinitrophenyl)aminohexanoic acid (DNP-
CO2H, 100 µM) prior to analysis to competitively displace any
noninternalized antibody. The longest of the synthetic receptors
(2) proved to be the most efficient in both magnitude and relative
rate of delivery of the IgG ligand into cells. After treatment of
cells with receptor 2 for 1 h at a final concentration of 10 µM,
this receptor enhanced the cellular uptake of added anti-DNP
IgG by 208-fold in magnitude and 391-fold in rate compared
with basal levels of endocytosis. Dose-dependent competition
experiments with DNP-CO2H added with the antibody further
confirmed the specificity of this uptake (Figure 3, panel B, right).
Remarkably, sequential removal of â-alanine subunits from the
linker of 2, affording receptors 3 and 4, substantially reduced
receptor efficiency. Additionally, conversion of the secondary
amine of receptor 4 to the secondary amide of 5 essentially
abolished internalization of IgG, reducing delivery compared
with that of 4 by over 50-fold at 10 µM (Figure 3, panel B).
The most efficient receptor (2, 1 µM) promoted internalization
of the fluorescent protein ligand with first-order kinetics (k )
0.44 h-1) corresponding to a ligand half-life on the cell surface
of 95 min (Figure 3, panel C, right). This value was quantified
by cooling receptor-treated cells to 4 °C to block endocytosis,
allowing the protein ligand to bind the receptor (2) on the cell
surface, warming cells to 37 °C to trigger uptake, and subsequent
competitive removal of cell surface protein prior to analysis.
Under the ligand uptake conditions shown in Figure 3, no
significant cellular cytotoxicity was observed after 24 h (see
the Supporting Information for details).
Linker Structure Controls the Subcellular Localization
of Synthetic Receptors. To investigate the substantial differ-
ences in biological activities of receptors 2-5, cells were treated
with structurally similar green fluorescent receptors (6-9).
Epifluorescence microscopy and fluorescence quenching assays
were employed to examine differences in cellular fluorescence
and receptor subcellular localization. As shown in Figure 4
(panel A), the overall cellular fluorescence resulting from
treatment with 10 µM of these compounds was within 2-fold
(Figure 4 and Figure S2 of the Supporting Information).
However, the subcellular localization of these compounds
differed significantly. Receptor 6, bearing two â-alanine linker
subunits, was nearly equally distributed between the cell surface
and internal membranes. However, the single â-alanine-contain-
ing receptor 7 exhibited lower cell surface localization, and the
absence of â-alanine in the linker of 8 rendered this compound
Figure 3. Synthetic receptor-mediated uptake of protein ligands by Jurkat
lymphocytes. Cells were pretreated with receptors for 1 h at 37 °C prior to
addition of an anti-DNP IgG/PrA complex. Panel A: confocal laser scanning
and differential interference contrast (DIC) microscopy of cells treated with
the green fluorescent receptor 6 (10 µM) followed by addition of red
fluorescent anti-DNP/PrA-AF594 (for 4 h). Yellow pixels indicate co-
localization of green and red fluorophores. Panel B: dose-dependent
magnitude of uptake of anti-DNP/PrA-AF488 (after 4 h) quantified by flow
cytometry. Right: competitive inhibition of this uptake (mediated by
receptor 2, 1 µM) by addition of 6-(2,4-dinitrophenyl)aminohexanoic acid
(DNP-CO2H). Panel C: kinetics of uptake of anti-DNP/PrA-AF488
quantified by flow cytometry. Left: relative differences in receptor uptake
rates quantified by accumulation of cellular fluorescence. Right: determi-
nation of the rate and half-life of ligand internalization mediated by receptor
2 at 1 µM.
up by cells under these conditions, but it only partially
co-localized with intracellular 6, suggesting dissociation of the
receptor/ligand complex within cells (Figure 3, panel A). This
intracellular red fluorescence was 59-fold lower in the absence
of receptor 6 (Figure S2, Supporting Information), consistent
with a synthetic receptor-mediated uptake mechanism. The IgG-
loaded intracellular compartments were identified as late en-
dosomes and lysosomes by co-localization studies in cells
transfected with EGFP-lgp120, a fluorescent protein marker of
these compartments (Figure S3, Supporting Information).42
Comparison of cells treated with receptor 6 alone with cells
(42) Patterson, G. H.; Lippincott-Schwartz, J. Science 2002, 297, 1873-1877.
9
16382 J. AM. CHEM. SOC. VOL. 126, NO. 50, 2004

Synthetic Mimics of Small Mammalian Cell Surface Receptors
A R T I C L E S
Figure 5. Cycling of receptor 6 between the cell surface and intracellular
endosomes in Jurkat lymphocytes. Panels A-E: DIC (left) and epifluo-
rescence (right) micrographs of cells captured after treatment with 6 (10
µM, 1 h) followed by quenching of cell surface fluorophores at 4 °C with
sodium dithionite (30 mM, 5 min). Cells were washed with ice-cold media
(t ) 0 min, panel A), split into two portions, and return of green fluorescence
to the cell surface was examined as a function of time and temperature
(panels A-C). Cells were again cooled to 4 °C prior to detection of NBD
headgroups at the cell surface with red fluorescent anti-DNP/PrA-AF633
by epifluorescence microscopy (panels D and E, t ) 30 min) and flow
cytometry (panel F, t ) 30 min). NBD fluorescence is shown in panels
A-C; PrA-AF633 fluorescence is shown in panels D-F.
the magnitude and kinetics of IgG uptake observed with
receptors 2-5. Thus, the relative distribution of synthetic
receptors between the cell surface and intracellular membranes
is a major factor controlling the efficiency of these compounds
as mediators of ligand uptake.
Figure 4. Subcellular localization of synthetic receptors in living Jurkat
lymphocytes. Panel A: DIC (left images) and epifluorescence micrographs
(right images) of cells treated for 1 h with green fluorescent receptors 6-9
(10 µM). Panel B: quantification of cell surface fluorescence by quenching
of exposed fluorophores with cell-impermeable sodium dithionite. Cells
treated, as shown in panel A, were washed with sodium dithionite (5 mM)
for 10 s at 22 °C, and residual fluorescence was quantified by flow
cytometry. Panel C: quantification of the cellular half-life of receptor 2.
Cells were treated with 2 (10 µM) for 1 h, resuspended in receptor-free
media for the times indicated, and 2 was detected on the cell surface with
anti-DNP/PrA-AF488.
To evaluate the cellular stability of receptor 2, the half-life
of this compound on living Jurkat lymphocytes was quantified
by addition of a soluble antibody. Cells were treated with this
receptor for 1 h (10 µM); the unincorporated receptor was
removed by washing cells with fresh media, and the abundance
of 2 on the cell surface as a function of time was detected with
a green fluorescent anti-DNP/PrA-AF488 complex. Curve fitting
of the flow cytometry data shown in Figure 4 (panel C) revealed
a cellular half-life of 20 h. This value is similar to the ∼24 h
half-life of natural folate receptor proteins that are anchored to
the cell surface by covalently attached GPI lipids.45
predominately intracellular. Remarkably, the amide analogue
9 was essentially completely associated with different internal
membranes, which explains the inability of this analogue to
mediate significant cellular uptake of the anti-DNP IgG ligand.
In contrast to the fluorescent N-alkyl derivatives of 3â-
cholesterylamine (6-8), this N-acyl analogue (9) nearly com-
pletely co-localized with a red fluorescent probe of the cellular
golgi apparatus and nuclear membrane, indicating a unique
intracellular destination of this ineffective receptor (data shown
in Figure S4 of the Supporting Information).
Synthetic Receptors Rapidly Cycle between the Plasma
Membrane and Intracellular Endosomes. Many natural cell
surface receptors undergo dynamic plasma membrane recycling.4
In this process, free receptors and receptor/ligand complexes
rapidly traffic between the cell surface and intracellular endo-
somes. Receptors typically dissociate from ligands in these
acidic compartments and cycle back to the cell surface up to
several hundred times, whereas free ligands are often sorted to
degratory lysosomes. To examine whether the green fluorescent
receptor 6 might undergo similar dynamic cycling, a fluorescence-
quenching assay with sodium dithionite was employed. As
shown in Figure 5, Jurkat cells were treated with receptor 6
(10 µM) and cooled to 4 °C to stop plasma membrane recycling.
The fraction of receptors 6-9 at the cell surface was
quantified by irreversible quenching of surface-exposed NBD
fluorophores. This was accomplished by brief treatment of cells
with the relatively cell-impermeable reducing agent sodium
dithionite (sodium hypodisulfite, NaO2S-SO2Na), which rapidly
reduces the nitro functionality of extracellular NBD headgroups
to the amine, enabling quantification of differences in cell
surface and intracellular fluorescence.43,44 Quantitative dithionite
quenching assays demonstrated substantial differences in recep-
tor subcellular localization (Figure 4, panel B) that paralleled
(43) Lentz, B. R.; Talbot, W.; Lee, J.; Zheng, L. X. Biochemistry 1997, 36,
2076-2083.
(44) Alakoskela, J. I.; Kinnunen, P. K. Biophys. J. 2001, 80, 294-304.
(45) Chung, K. N.; Roberts, S.; Kim, C. H.; Kirassova, M.; Trepel, J.; Elwood,
P. C. Arch. Biochem. Biophys. 1995, 322, 228-234.
9
J. AM. CHEM. SOC. VOL. 126, NO. 50, 2004 16383

A R T I C L E S
Boonyarattanakalin et al.
Fluorescence at the cell surface was quenched by treatment with
ice-cold sodium dithionite to reduce the nitro group of 6, and
cells were washed with cold media to remove this reducing agent
(Figure 5, panel A). This cell culture was split into two equal
portions; one portion was maintained at 4 °C (Figure 5, panel
B) for 30 min, and the other portion was warmed to 37 °C for
30 min to reactivate plasma membrane recycling (Figure 5, panel
C). Significant return of green fluorescence back to the plasma
membrane was observed only in the sample warmed to 37 °C
(compare panels B and C of Figure 5). Moreover, NBD
headgroups that returned to the cell surface were detected with
the NBD-binding red fluorescent anti-DNP/PrA-AF633 com-
plex. In the cells warmed to 37 °C, a 5-fold increase in receptor
6 at the plasma membrane was observed (Figure 5, panels D-F).
These results demonstrate that the intracellular fraction of
synthetic receptors is in dynamic exchange with the cellular
plasma membrane, consistent with the synthetic receptors
accessing an active plasma membrane recycling pathway.
To quantify the rate of plasma membrane recycling, the
oregon green-derived receptor 10 was employed. Unlike the
more strongly cell-associated DNP and NBD-based receptors,
this compound (10) binds proteins in cell culture media and
can be rapidly and efficiently depleted from cell surfaces by
washing cells with “back exchange” media containing bovine
serum albumin (BSA, 1%) and methyl-â-cyclodextrin (Me-â-
CD, 2 mM). This concentration of Me-â-CD has been shown
to affect neither cellular endocytosis nor plasma membrane
recycling.46 The oregon green fluorophore of 10 was also chosen
for these studies because, unlike the structurally similar car-
boxyfluorescein, the fluorescence of oregon green is not
appreciably quenched in the acidic environment of endosomes.47
Cells treated with 10 (1 µM) exhibited bright green fluorescence
both at the cell surface and in the intracellular (endosomal)
compartments; this distribution of cellular fluorescence was
similar to that of cells treated with the structurally related
receptor 7 (data not shown). Cells were treated with 10 and
cooled to 4 °C to stop dynamic plasma membrane recycling.
These cells were washed four times with ice-cold back exchange
media to extensively deplete 10 from the cell surface without
affecting the population of this compound in intracellular
endosomes. Back exchange media prewarmed to 37 °C was
subsequently added to reactivate plasma membrane recycling.
This temperature jump resulted in rapid return of 10 to the
plasma membrane by exocytosis, and the added back exchange
media efficiently removed the surface-exposed recycled recep-
tors. Analysis of cellular fluorescence as a function of time
provided the fluorescence decay data shown in Figure 6. Fitting
of these data with a one-site exponential decay model yielded
a recycling half-life of 2.8 ( 0.8 min, indicating rapid recycling
of synthetic cell surface receptors. The fluorescent lipid C6-
NBD-sphingomyelin is known to exhibit similar recycling
kinetics in mammalian cells.48
Figure 6. Quantification of efflux kinetics of receptor 10 from endosomes
to the plasma membrane of Jurkat lymphocytes by back exchange. Cells
treated with 10 (1 µM, 1 h) were cooled to 4 °C and cell surface fluorophores
removed (>90%) by repeated washing with ice-cold back exchange media
containing BSA (1%) and methyl-â-cyclodextrin (2 mM). Cells were
resuspended in back exchange media at 37 °C to resume recycling; cellular
fluorescence was quantified by flow cytometry, and loss of fluorescence
was evaluated with a one-site exponential decay model.
sphingolipid-enriched lipid raft subdomains of cellular plasma
membranes.49 These subdomains are generally biochemically
characterized as low-density fractions of the plasma membrane
insoluble in buffers containing nonionic detergents, such as
Triton X-100 at 4 °C.29 Ganglioside GM1 (1) is a prototypical
example of a small lipid raft-associated cell surface receptor,
and complexes of this glycolipid with fluorescent cholera toxin
protein are often used as markers for lipid rafts.50 In addition,
macromolecular receptors, including B-cell receptors, T-cell
receptors, and growth factor receptors, are thought to associate
with these membrane subdomains.49 In certain cases, binding
of ligands to receptors initiates formation of lipid rafts.51 This
ligand-mediated concentration of specific membrane lipids has
been proposed to recruit intracellular raft-associated kinases to
activate cellular signaling pathways.31
To determine whether synthetic cell surface receptors (2-5)
might cofractionate with ganglioside GM1 (1) in lipid rafts,
sucrose density gradient ultracentrifugation was employed to
fractionate cellular membranes containing receptor/ligand com-
plexes (Figure 7). As positive and negative control experiments,
Jurkat cells were treated with green fluorescent conjugates of
cholera toxin B subunit and transferrin proteins. Unlike binding
of cholera toxin to ganglioside GM1 (1), binding of transferrin
to the transferrin receptor on the cell surface does not result in
association with lipid rafts.50 As shown in Figure 7, a low-
density fluorescent membrane fraction (#5) was observed upon
treatment of cells with the cholera toxin B subunit, but this
fraction was not observed upon treatment with transferrin,
consistent with previous studies of lipid raft association of these
proteins.50 Application of this analysis to receptors 2-5 (1 µM)
revealed that only receptors 2 and 3 enabled the isolation of
this fluorescent low-density membrane fraction (#5). These
results paralleled the magnitude of protein uptake mediated by
synthetic receptors 2-5 at the receptor concentration studied
(Figure 3, panel B). Unlike the IgG-bound receptors, fraction-
ation of cells treated with the green fluorescent receptor 6 alone
did not reveal any significant lipid raft association of the free
receptor (data not shown). Free receptors may be associated
with less well-ordered “lipid shell” domains of cellular mem-
Synthetic Receptor/Ligand Complexes Cofractionate with
Lipid Raft Components of Cellular Plasma Membranes.
Another hallmark of many natural cell surface receptors is the
proposed association of these biomolecules with cholesterol and
(46) Rodal, S. K.; Skretting, G.; Garred, O.; Vilhardt, F.; van Deurs, B.; Sandvig,
K. Mol. Biol. Cell. 1999, 10, 961-974.
(49) Munro, S. Cell 2003, 115, 377-388.
(50) Harder, T.; Scheiffele, P.; Verkade, P.; Simons, K. J. Cell Biol. 1998, 141,
929-942.
(47) Sun, W. C.; Gee, K. R.; Klaubert, D. H.; Haugland, R. P. J. Org. Chem.
1997, 62, 6469-6475.
(51) Dietrich, C.; Volovyk, Z. N.; Levi, M.; Thompson, N. L.; Jacobson, K.
Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 10642-10647.
(48) Hao, M.; Maxfield, F. R. J. Biol. Chem. 2000, 275, 15279-15286.
9
16384 J. AM. CHEM. SOC. VOL. 126, NO. 50, 2004

Synthetic Mimics of Small Mammalian Cell Surface Receptors
A R T I C L E S
headgroup, differences in the extent of trafficking of receptors
to internal membranes appear to predominately control the
efficiency of these compounds as mediators of ligand uptake.
The impact of changes in receptor subcellular localization was
particularly evident when comparing the N-alkyl derivatives 4
and 8 with the analogous N-acyl derivatives 5 and 9. These
amide analogues (5 and 9) were much less effective receptors
due to their unique localization at the internal membranes of
the golgi apparatus and nuclear membrane of living cells.
Differences in trafficking between these N-alkyl and N-acyl
derivatives of 3â-cholesterylamine may relate to structural
similarities of N-acyl derivatives to natural cholesteryl esters.
These esters are known to localize at intracellular membranes
of the golgi apparatus.55
Synthetic receptor 2, bearing two â-alanine linker subunits
between the DNP headgroup and 3â-cholesterylamine, stably
associated with living Jurkat lymphocytes. After removal of the
unincorporated receptor from cell culture media, this receptor
was detected on the cell surface with a cellular half-life of 20
h. This stability is similar to that of natural folate receptors that
are attached to the cellular plasma membrane by GPI lipids (t1/2
∼ 24 h).45 Fluorescence quenching assays with the structurally
similar fluorescent receptor 6 and endosomal efflux assays with
receptor 10 revealed that these compounds undergo continuous
trafficking (recycling t1/2 ∼ 3 min) between the cell surface and
intracellular endosomes. To maintain membrane balance, these
synthetic receptors presumably make a round trip between the
cell surface and endosomes in approximately 10 min. Other
synthetic membrane probes, such as C6-NBD-sphingomyelin,
have been shown to similarly cycle between these membranes.48
Certain macromolecular receptors also rapidly cycle between
the cell surface and intracellular endosomes. For example,
transmembrane LDL receptors complete this round trip in 10
min, cycling several hundred times in their 20 h life span.56
IgG-bound receptors 2 and 3 also cofractionated with cholera
toxin bound to the natural receptor ganglioside GM1 (1) in
detergent-resistant lipid raft fractions of cellular plasma mem-
branes, suggesting mechanistic similarities between these syn-
thetic and natural systems. A simple model of synthetic receptor-
mediated endocytosis summarizing these results is shown in
Figure 8. The actual series of events in living cells is highly
complex at the molecular level and presumably involves the
interplay of multiple endosomal compartments.4
Figure 7. Association of receptor/ligand complexes with putative lipid raft
fractions of plasma membranes of Jurkat lymphocytes. Panel A: cells were
treated with the AF488-labeled protein ligand for 30 min at 4 °C. Panel B:
cells were treated with receptors 2-5 (1 µM) for 1 h at 37 °C followed by
an anti-DNP/AF488 conjugate for 30 min at 4 °C. Cells were fractionated
by ultracentrifugation for 20 h at 4 °C in a 2-40% sucrose step gradient
containing 0.2% Triton X-100. CT-B: cholera toxin B subunit.
branes.52 These results suggest additional mechanistic similarities
between uptake of IgG-bound synthetic receptors and uptake
of cholera toxin mediated by ganglioside GM1 (1).
Discussion
We report the synthesis of mimics of small cell surface
receptors. These synthetic receptors project DNP and NBD
antibody-binding motifs from the surface of mammalian cells,
capture anti-DNP IgG ligands, and trigger efficient endocytosis
of these macromolecules. We demonstrate that synthetic recep-
tors constructed from N-alkyl derivatives of 3â-cholesterylamine
can access membrane trafficking pathways that rapidly cycle
between the plasma membrane and intracellular endosomes.
Access to these pathways may relate to the ability of these
compounds to mimic cholesterol, a major component of both
mammalian plasma membranes and endosomes53 that plays key
roles in endocytic trafficking.54 Remarkably, the extent of
receptor localization on the cell surface was found to be critically
dependent on the structure of the linker between the protein-
binding headgroup and the membrane anchor. Insertion of
â-alanine subunits into this linker increased the population of
synthetic receptors on the cell surface, resulting in much higher
ligand uptake efficiency. The most effective receptor, 2, bearing
two â-alanine linker subunits, partitioned equally between the
cell surface and internal endosomal membranes, as evidenced
by fluorescence quenching assays with structurally similar 6.
Removal of these â-alanines resulted in substantially greater
association of synthetic receptors with internal endosomal
membranes. Although insertion of â-alanines may also favorably
impact the affinity of the IgG ligand for the receptor on the
cell surface by improving the accessibility of the protein-binding
Synthetic mimics of mammalian cell surface receptors may
have applications as cellular probes, as tools for drug delivery,
and as mediators of cellular uptake of therapeutically important
ligands. Because subtle molecular changes substantially affect
the subcellular localization of 3â-cholesterylamine derivatives,
these and related57 compounds may provide useful probes of
cellular mechanisms that regulate the segregation, localization,
and sorting of membrane-associated biomolecules. Synthetic cell
surface receptors may also enable novel approaches for cellular
targeting and drug delivery. Cellular targeting applications may
be facilitated by the structural similarity of these compounds
to cholesterol, which may allow packaging of synthetic receptors
(55) Reaven, E.; Tsai, L.; Azhar, S. J. Biol. Chem. 1996, 271, 16208-16217.
(56) Brown, M. S.; Goldstein, J. L. Angew. Chem., Int. Ed. Engl. 1986, 25,
583-602.
(57) Sato, S. B.; Ishii, K.; Makino, A.; Iwabuchi, K.; Yamaji-Hasegawa, A.;
Senoh, Y.; Nagaoka, I.; Sakuraba, H.; Kobayashi, T. J. Biol. Chem. 2004,
279, 23790-23796.
(52) Anderson, R. G.; Jacobson, K. Science 2002, 296, 1821-1825.
(53) Hao, M.; Lin, S. X.; Karylowski, O. J.; Wustner, D.; McGraw, T. E.;
Maxfield, F. R. J. Biol. Chem. 2002, 277, 609-617.
(54) Puri, V.; Watanabe, R.; Dominguez, M.; Sun, X.; Wheatley, C. L.; Marks,
D. L.; Pagano, R. E. Nat. Cell Biol. 1999, 1, 386-388.
9
J. AM. CHEM. SOC. VOL. 126, NO. 50, 2004 16385

A R T I C L E S
Boonyarattanakalin et al.
Figure 8. Simple model of synthetic receptor-mediated endocytosis. Synthetic receptors embedded in the cellular plasma membrane rapidly cycle between
the cell surface and intracellular endosomes. Binding of ligand (IgG) results in association with lipid rafts and uptake of the complex by endocytosis.
Dissociation in endosomes frees the receptor to return to the cell surface. The protein ligand is sorted to late endosomes and lysosomes.
in LDL particles for selective delivery to tumor cells.58 In
addition, because cells can become resistant to drugs by altering
the expression of transporters or receptors at the plasma
membrane, synthetic receptors that promote the internalization
of drugs might provide novel drug delivery strategies. Other
potential applications of synthetic receptor targeting might focus
on eliminating autoreactive antibodies, tumor-promoting growth
factors, or inflammatory cytokines involved in numerous
diseases. Treatment with appropriately designed synthetic recep-
tors might enable diverse cell types to remove these therapeuti-
cally important proteins from circulation. Synthetic receptors
that function on cell surfaces have significant potential as
biological modulators and probes.
Acknowledgment. We thank the National Institutes of Health
(R01-CA83831) for financial support. We thank Ms. Elaine
Kunze and Ms. Susan Magargee for assistance with flow
cytometry and confocal microscopy.
Supporting Information Available: Additional supporting
figures (S1-S4) and Experimental Section. This material is
available free of charge via the Internet at http://pubs.acs.org.
(58) Lenz, M.; Miehe, W. P.; Vahrenwald, F.; Bruchelt, G.; Schweizer, P.;
Girgert, R. Anticancer Res. 1997, 17, 1143-1146.
JA046663O
9
16386 J. AM. CHEM. SOC. VOL. 126, NO. 50, 2004