Tyrosine-selective protein alkylation using π-allylpalladium complexes

Article abstract of DOI:10.1021/ja057106k

A new protein modification reaction has been developed based on a palladium-catalyzed allylic alkylation of tyrosine residues. This technique employs electrophilic π-allyl intermediates derived from allylic acetate and carbamate precursors and can be used to modify proteins in aqueous solution at room temperature. To facilitate the detection of modified proteins using SDS-PAGE analysis, a fluorescent allyl acetate was synthesized and coupled to chymotrypsinogen A and bacteriophage MS2. The tyrosine selectivity of the reaction was confirmed through trypsin digest analysis. The utility of the reaction was demonstrated by using taurine-derived carbamates as water solubilizing groups that are cleaved upon protein functionalization. This solubility switching technique was used to install hydrophobic farnesyl and C₁₇ chains on chymotrypsinogen A in water using little or no cosolvent. Following this, the C₁₇ alkylated proteins were found to associate with lipid vesicles. In addition to providing a new protein modification strategy targeting an under-utilized amino acid side chain, this method provides convenient access to synthetic lipoproteins. Copyright

Full text of DOI:10.1021/ja057106k

Published on Web 01/04/2006
Tyrosine-Selective Protein Alkylation Using π-Allylpalladium Complexes
S. David Tilley and Matthew B. Francis*
Department of Chemistry, UniVersity of California, Berkeley, California 94720-1460, and
Material Science DiVision, Lawrence Berkeley National Labs, Berkeley, California 94720-1460
Received October 19, 2005; E-mail: francis@cchem.berkeley.edu
Transition metal mediated reactions are emerging as a promising
set of strategies for protein modification, as organometallic
intermediates can functionalize amino acid side chains that are
difficult to target using traditional bioconjugation techniques.
Several studies have now demonstrated that these reactions can
modify proteins in aqueous solution with high chemoselectivity and
virtually complete functional group tolerance.1 Due to their versatil-
ity in organic synthesis, palladium-catalyzed reactions are a
welcome addition to this list, and pioneering studies have demon-
strated that palladium-catalyzed cross-coupling reactions can indeed
modify aryl halides2a,b and boronic acids2c introduced into synthetic
peptides.
For the direct modification of native proteins, however, we were
drawn to electrophilic π-allyl complexes because these species are
more appropriately matched to the nucleophilic functional groups
of the natural amino acids. π-Allyl species have enjoyed widespread
success in organic synthesis3 and offer the advantage that the allylic
acetate, carbonate, and carbamate precursors are completely inert
until activated by the palladium catalyst. Perhaps the most relevant
precedent for protein modification is the allylation of phenolic
groups,4 suggesting that tyrosine residues could be functionalized
if biomolecule-compatible reaction conditions could be realized.
We have found that this can be achieved, resulting in a highly
selective alkylation reaction that can install allylic functionality on
proteins under mild reaction conditions (Scheme 1). We report
herein the conditions and selectivity of this new bioconjugation
reaction, as well as a demonstration of its use for the installation
of hydrophobic groups through a solubility-switching technique.
These studies began with the development of aqueous reaction
conditions5 that would be expected to preserve the tertiary structure
of protein substrates. Chymotrypsinogen A (4a) was chosen as a
model protein for these studies, as previous Mannich-type alkylation
reactions have revealed the presence of a solvent-accessible tyrosine
residue on the protein surface.6 Upon exposure of a 200 µM solution
of this protein to rhodamine-labeled allylic acetate 1 (1 mM),
Pd(OAc)2 (40 µM, 4.3 ppm Pd), and triphenylphosphine tris-
(sulfonate) (TPPTS, 0.48 mM) as a water-soluble phosphine ligand,
efficient protein labeling was observed within 45 min at room
temperature. Analysis of the reaction mixture by ESI-MS and RP-
HPLC consistently indicated 50-65% conversion of the starting
protein to a singly alkylated product, corresponding to the addition
of 1 after loss of the acetate group (Figure 1a).7 A small amount of
Figure 1. Alkylation of tyrosine residues with fluorescent allyl acetate 1.
(a) ESI-MS analysis of modified chymotrypsinogen A (4a, 200 µM).
Reaction conditions: 1 mM 1, 40 µM Pd(OAc)2, 0.48 mM TPPTS, 100
mM phosphate buffer, pH 8.5, 45 min, room temperature. (b) The tyrosine
selectivity of the reaction was confirmed using a trypsin digest. The modified
T14-T15 fragment (containing Y171) can be identified at m/z 1516. (c)
Structure of chymotrypsinogen A, indicating the location of Y171. Y146
is presumed to be the second modification site. (d) Detection of protein
modification using SDS-PAGE and Coomassie staining (left) and fluo-
rescence visualization (right). Proteins: (5) horse heart myoglobin, (6)
H-Ras, (7) bacteriophage MS2, (8) R-chymotrypsin A. The modification
of 7 was carried out using 200 µM Pd(OAc)2; otherwise, all reactions were
run as described in (a) using the indicated protein concentrations. The
reaction analyzed in lane 6* was run with Oregon Green maleimide to verify
the presence of a reactive cysteine residue on 6.
doubly alkylated product was also observed. Consistent with a
mechanism involving the phenolate anion, the reaction was found
to proceed best at pH 8.5-9.0.8 Through HPLC analysis and
comparison to authentic material, it was determined that excess 1
was converted to diene 3,9 which could easily be removed from
the protein solution by gel filtration or extraction.
The site selectivity of the reaction was determined through
proteolytic digestion of the modified protein with trypsin. The T14-
15 fragment, which contains Y171, was identified as the predomi-
nant site of modification (Figure 1b,c). Fragments corresponding
to the alkylation of other nucleophilic side chains have not been
identified to date. The second modification site is presumed to be
Y146, a second surface-accessible residue. The two other tyrosine
residues in the protein sequence are significantly less accessible to
the reaction solvent and are presumed not to participate.
Due to the fluorescent nature of 1, the alkylated products could
also be detected using a gel-based assay. As shown in Figure 1d,
similar modification levels were obtained for reactions carried out
using only 5 µM 4a (compare lanes 1 and 2). As would be expected,
no labeling was detected in the absence of Pd(OAc)2 (lane 3). In
the presence of active catalyst, no modification was observed for
horse heart myoglobin (5, lane 4), a protein with 19 lysines but
with no surface-accessible tyrosine residues. Similarly, the reactive
cysteine thiolate of H-Ras (6) was not modified under the reaction
conditions (lane 5). The ability of this site to participate in alkylation
Scheme 1
9
1080
J. AM. CHEM. SOC. 2006, 128, 1080-1081
10.1021/ja057106k CCC: $33.50 © 2006 American Chemical Society

C O M M U N I C A T I O N S
Figure 2. Preparation of synthetic lipoproteins. (a) Allylic taurine car-
bamates 9 and 10 serve as convenient water-soluble lipid precursors. (b)
ESI-MS analysis of farnesylated 4a. Conditions: 200 µM 4a, 1 mM 9,
400 µM Pd(OAc)2, 2 mM TPPTS, phosphate buffer, pH 9, 100 min, room
temperature. (c) To confirm the product structure, cresol adduct 11 was
prepared under analogous reaction conditions. (d) Modification of 4a using
10. Identical reactions conditions were used, with the addition of 5% DMSO.
Figure 3. Incorporation of synthetic lipoproteins into lipid bilayers. A
mixture of 4a, 4c (singly modified), and 4d (doubly modified) was incubated
with 100 nm small unilamellar vesicles (SUVs) for 20 min in the presence
of 4 mM â-octyl-D-glucopyranoside. The unbound protein was removed
by passing the vesicles through Sephacryl S-1000 gel filtration resin. (b)
SDS-PAGE analysis indicated that only lipid-modified proteins were
associated with the vesicles.
reactions was verified by exposing this protein to Oregon Green
maleimide, a well-known reagent for the modification of cysteine
residues (lane 6*). In contrast, the protein coat of bacteriophage
MS2 (7, lane 7, demonstrated previously to possess reactive tyrosine
residues10) was successfully modified. Finally, the reaction also
proceeded to high levels of conversion when carried out on activated
R-chymotrypsin A (8). After disulfide reduction with DTT (which
cleaves the protein into three separate peptide chains), SDS-PAGE
analysis indicated that only the fragment at 10 kD (corresponding
to Y171) had been modified (lane 8).
effect highly selective transformations through the activation of
otherwise inert substrates. In addition to targeting an under-utilized
functional group, this reaction provides an important step forward
for the preparation of artificial lipoproteins. Current efforts are
focusing on the use of alternative phosphine ligands to direct
modifications to one of several tyrosine residues present on more
complex protein targets.
Acknowledgment. We gratefully acknowledge the DOE Nano-
scale Science, Engineering, and Technology Program for generous
financial support, as well as the NIH (GM072700-01) and the
University of California, Berkeley, Chemistry Department. We
would also like to thank Jodi Gureasko and John Kuriyan for
samples of H-Ras, and John Antos, Bill Galush, Bryan Jackson,
and Jay Groves for helpful discussions.
A particularly attractive feature of this reaction is its ability to
cleave a “disposable” group upon formation of the π-allyl complex.
This can allow the aqueous solubilization of otherwise prohibitively
hydrophobic molecules by coupling them to charged carrier groups,
such as taurine (Figure 2a). To demonstrate this possibility, water-
soluble farnesyl carbamate 9 (1 mM) was exposed to 200 µM 4a,
2 mM TPPTS, and 400 µM Pd(OAc)2 in phosphate buffer at pH 9.
ESI-MS analysis indicated that 30% overall conversion to farne-
sylated product 4b was observed. Although Pd-catalyzed rearrange-
ments of the double bonds can be envisioned during this reaction,
linear allyl ether 11 was the only product observed upon exposure
of p-cresol to comparable reaction conditions.11
Supporting Information Available: Full experimental procedures
and characterization data for all intermediates. This material is available
free of charge via the Internet at http://pubs.acs.org.
References
Similar success was achieved for the alkylation of 4a using
exceptionally hydrophobic C17 chains. Taurine carbamate 10
displays good solubility in 95:5 H2O:DMSO mixtures and affords
singly and doubly modified protein conjugates (Figure 2d). The
sudden increase in doubly modified product may arise from the
association of singly modified protein with additional molecules
of carbamate 10 or through relaxation of the protein structure by
the small amount of organic cosolvent.
This “solubility switching” strategy provides convenient access
to proteins that can be incorporated into lipid membranes. To
explore this possibility, a mixture of 4a, 4c, and 4d (resulting from
the modification of 4a with 10) was exposed to 100 nm small
unilamellar vesicles (SUVs) for 20 min (Figure 3a). A nonionic
surfactant, â-octyl-D-glucopyranoside, was also added to the solution
to discourage the aggregation of the lipidated proteins. Following
this step, the unbound protein was removed from the sample by
passing the vesicles through Sephacryl S-1000 gel filtration media.
The protein content of the resulting samples was then determined
using SDS-PAGE. As shown in Figure 3b, none of the free protein
samples were able to pass through the resin in the absence of SUVs,
and vesicles exposed to unmodified 4a carried no protein with them
through the media. In contrast, SUVs were able to carry the lipid-
modified proteins through the gel filtration column, suggesting that
they had been embedded in the membrane bilayers. More detailed
studies examining the nature of this association are in progress.
Through these studies, a new tyrosine modification reaction has
been developed, capitalizing on the ability of transition metals to
(1) (a) Wang, Q.; Chan, T. R.; Hilgraf, R.; Fokin, V. V.; Sharpless, K. B.;
Finn, M. G. J. Am. Chem. Soc. 2003, 125, 3192-3193. (b) Deiters, A.;
Cropp, T. A.; Mukherji, M.; Chin, J. W.; Anderson, J. C.; Schultz, P. G.
J. Am. Chem. Soc. 2003, 125, 11782-11783. (c) Link, A. J.; Tirrell, D.
A. J. Am. Chem. Soc. 2003, 125, 11164-11165. (d) Antos, J. M.; Francis,
M. B. J. Am. Chem. Soc. 2004, 126, 10256-10257. (e) McFarland, J.
M.; Francis, M. B. J. Am. Chem. Soc. 2005, 127, 13490-13491.
(2) (a) Dibowski, H.; Schmidtchen, F. P. Angew. Chem., Int. Ed. 1998, 37,
476-478. (b) Bong, D. T.; Ghadiri, M. R. Org. Lett. 2001, 3, 2509-
2511. (c) Ojida, A.; Tsutsumi, H.; Kasagi, N.; Hamachi, I. Tetrahedron
Lett. 2005, 46, 3301-3305.
(3) See: Trost, B. M.; Lee, C. Asymmetric Allylic Alkylation Reactions. In
Catalytic Asymmetric Synthesis, 2nd ed.; Ojima, I., Ed.; Wiley-VCH: New
York, 2000.
(4) (a) Muzart, J.; Genet, J. P.; Dennis, A. J. Organomet. Chem. 1987, 326,
C23-C28. (b) Trost, B. M.; Toste, F. D. J. Am. Chem. Soc. 1998, 120,
815-816. (c) Trost, B. M.; Toste, F. D. J. Am. Chem. Soc. 1998, 120,
9074-9075.
(5) For additional examples of palladium π-allyl chemistry in aqueous media,
see: (a) Genet, J. P.; Blart, E.; Savignac, M. Synlett 1992, 715-717. (b)
Lemaire-Audoire, S.; Savignac, M.; Pourcelot, G.; Genet, J. P.; Bernard,
J. M. J. Mol. Catal. A: Chem. 1997, 116, 247-258. (c) Genet, J. P.;
Savignac, M. J. Organomet. Chem. 1999, 576, 305-317. (d) Danjo, H.;
Tanaka, D.; Hayashi, T.; Uozumi, Y. Tetrahedron 1999, 55, 14341-14352.
(e) Uozumi, Y.; Shibatomi, K. J. Am. Chem. Soc. 2001, 123, 2919-2920.
(6) Joshi, N. S.; Whitaker, L. R.; Francis, M. B. J. Am. Chem. Soc. 2004,
126, 15942-15943.
(7) See Supporting Information for sample quantitation details.
(8) Both phosphate and carbonate buffers (each at 100 mM) have been used
in the reaction with comparable results.
(9) For a similar elimination of a π-allyl complex, see: Takacs, J. M.; Lawson,
E. C.; Clement, F. J. Am. Chem. Soc. 1997, 119, 5956-5957.
(10) Hooker, J. M.; Kovacs, E. W.; Francis, M. B. J. Am. Chem. Soc. 2004,
126, 3718-3719.
(11) Geranyl groups have also been transferred to phenols without olefin
rearrangement. See: Goux, C.; Lhoste, P.; Sinou, D. Synlett 1992,
725-727.
JA057106K
9
J. AM. CHEM. SOC. VOL. 128, NO. 4, 2006 1081