Chemical probes for the rapid detection of fatty-acylated proteins in mammalian cells

Article abstract of DOI:10.1021/ja0685001

The attachment of lipids onto proteins modulates the activity of proteins in many biological settings. The analysis of protein lipidation, however, is challenging due to the relatively few methods for the detection of lipid-modified proteins. Here we describe the synthesis of ω-azido-fatty acids as non-radioactive chemical probes for the rapid visualization of fatty-acylated proteins in mammalian cells. Following metabolic installation of the ω-azido-fatty acids onto target proteins by cellular enzymes, fatty-acylated proteins are selectively biotinylated with a phosphine-biotin reagent via the Staudinger ligation and visualized by streptavidin blotting. Depending on the chain length of the ω-azido-fatty acids, N-myristoylated and S-palmitoylated proteins can be visualized selectively in cell lysates and on specific proteins. These chemical probes provide new tools to analyze fatty acylation of proteins in living cells. Copyright

Full text of DOI:10.1021/ja0685001

Published on Web 02/17/2007
Chemical Probes for the Rapid Detection of Fatty-Acylated Proteins in
Mammalian Cells
†
†
†
†
Howard C. Hang, Ernst-Jan Geutjes, Gijsbert Grotenbreg, Annette M. Pollington,
‡
,†
Marie Jose Bijlmakers, and Hidde L. Ploegh*
Whitehead Institute for Biomedical Research and Department of Biology, Massachusetts Institute of Technology,
Cambridge, Massachusetts 02142, and Department of Immunobiology, Guy’s Hospital, King’s College London,
London SE1 9RT, United Kingdom
Received November 27, 2006; E-mail: ploegh@wi.mit.edu
The attachment of saturated fatty acids onto proteins modulates
their interactions with membranes and other proteins, which
ultimately regulates their signaling properties, subcellular localiza-
1
tion, trafficking, and enzymatic activity. In eukaryotes, the two
major forms of protein fatty acylation are N-myristoylation and
S-palmitoylation (Figure 1A), which are characterized by the
addition of myristic acid to the N-terminal glycine of nascent
polypeptides or the modification of cysteine residues with palmitic
acid, respectively. These two forms of fatty acylation can occur
separately on proteins or in concert to regulate protein function
and have been shown to play key roles in many biological processes,
such as T cell activation, apoptosis, tumorogenesis, neuronal
signaling, cellular growth, and differentiation.1
Figure 1. (A) Two major forms of protein fatty acylation in mammalian
cells. (B) Strategy for the rapid detection of protein fatty acylation in
The study of protein fatty acylation has been challenging due to
limited methods for biochemical analysis. Radiolabeled fatty acids
mammalian cells with ω-azido-fatty acids (1-4) followed by bioorthogonal
labeling with phosphine-biotin via Staudinger ligation. The chain length
of the ω-azido-fatty acids (1-4) corresponds to 12, 14, 15, and 16 carbons,
respectively. The chemical synthesis of ω-azido-fatty acids is described in
the Supporting Information Scheme 1.
2
have been utilized to visualize protein lipidation but suffer from
2
low specific activity and are cumbersome to handle. Alternatively,
mass spectrometry has also been used to detect fatty acylation of
2
purified proteins. The development of non-radioactive methods
to analyze protein fatty acylation in cell lysates or even in vivo
would facilitate the analysis of protein fatty acylation. For example,
the acyl-biotinyl exchange protocol provides a non-radioactive
3
method for the visualization of protein S-acylation and has enabled
the global analysis of protein S-palmitoylation in yeast cell lysates.4
Nonetheless, this method still requires metabolic labeling with
radioactive fatty acids to confirm protein fatty acylation in living
4
cells. Here we describe the development of ω-azido-fatty acids as
non-radioactive probes for the rapid detection of protein fatty
acylation in mammalian cells (Figure 1B).
To explore whether ω-azido-fatty acids could serve as chemical
probes for protein fatty acylation, we synthesized a series of
ω-azido-fatty acids 1-4 of varying chain length, which contain
Figure 2. Visualization of ω-azido-fatty acid protein labeling in mammalian
cell lysates. (A) Analysis of ω-azido-fatty acid protein labeling in RAW264.7
cell lysates. - or + hydroxylamine (NH2OH) treatment. Cells were
incubated with various ω-azido-fatty acids (1, 20 µM), (2-4, 100 µM), or
DMSO control (-) for 6 h and harvested; / marks endogenously biotinylated
proteins. (B) The ω-azido-fatty acid labeling of endogenous Lck immuno-
precipitated from Jurkat cells. (C) The ω-azido-fatty acid labeling of wild-
type Lck and G2A Lck mutant overexpressed in HEK 293T cells.
12, 14, 15, and 16 carbons, respectively (Figure 1B and Supporting
Information Scheme 1). Mammalian cells (RAW264.7 macro-
phages) were incubated with the panel of ω-azido-fatty acids 1-4
and harvested after 6 h. Cell lysates were prepared and subjected
to bioorthogonal labeling with phosphine-biotin via the Staudinger
5
ligation (Figure 1B). Biotinylated proteins were then separated by
(
Figure 2A), whereas the longer chain ω-azido-fatty acids 2-4
gel electrophoresis and analyzed by streptavidin blotting (Figure
afford similar profiles of protein labeling that are distinct and
partially overlapping with ω-azido-fatty acid 1 (Figure 2A).
Furthermore, protein labeling with the ω-azido-fatty acids 1-3 in
living cells was time- and dose-dependent (Supporting Information
Figure 1), demonstrating that active cellular metabolism is required
for ω-azido-fatty acid incorporation and that polypeptides visualized
by this approach are not the result of post-lysis enzymatic activity.
The metabolic labeling of mammalian proteins with ω-azido-fatty
acids 1-4 followed by bioorthogonal reaction with phosphine-
biotin affords a very sensitive non-radioactive method to detect
2
A). Treatment of mammalian cells with these chemical probes
afforded selective protein labeling that was dependent on the chain
length of the ω-azido-fatty acids 1-4 (Figure 2A). In the absence
of any ω-azido-fatty acid, only endogenously biotinylated proteins
are detected by streptavidin blotting, demonstrating the specificity
of our bioorthogonal labeling conditions (Figure 2A). The 12-carbon
ω-azido-fatty acid 1 yields the most robust profile of protein labeling
†
Massachusetts Institute of Technology.
‡
King’s College London.
2744
9
J. AM. CHEM. SOC. 2007, 129, 2744-2745
10.1021/ja0685001 CCC: $37.00 © 2007 American Chemical Society

C O M M U N I C A T I O N S
fatty-acylated proteins, as labeled proteins can be visualized within
minutes by streptavidin blotting.
acid analogue 1 and analyzed for Lck N-myristoylation by
streptavidin blotting after immunoprecipitation (Figure 2C). The
myristic acid analogue 1 only labeled wild-type Lck and not the
G2A mutant, demonstrating that compound 1 was attached specif-
ically to the N-terminal glycine of Lck. Collectively, these
experiments establish that ω-azido-fatty acids can be efficiently
metabolized by mammalian cells and selectively installed on sites
of fatty acylation on proteins, depending on the chain length of the
ω-azido-fatty acids.
The distinct profiles of protein labeling afforded by ω-azido-
fatty acid 1 compared to 2-4 suggest that these chemical probes
target different sets of fatty-acylated proteins in mammalian cells.
To determine whether the ω-azido-fatty acids selectively target
N-myristoylated or S-palmitoylated proteins, we subjected ω-azido-
2
fatty acid labeled proteins to hydroxylamine (NH OH) treatment,
which selectively removes fatty acids attached to proteins via a
3
thioester linkage. Treatment of ω-azido-fatty acid labeled proteins
New tools are needed to further our understanding of protein
fatty acylation. Here we demonstrate that ω-azido-fatty acids 1 and
3 can be efficiently metabolized by mammalian cells and serve as
selective probes to rapidly visualize N-myristoylation and S-
acylation, respectively. In addition to the more sensitive detection
of fatty-acylated proteins with these chemical probes, the ability
to biotinylate fatty-acylated proteins provides an opportunity for
enrichment and proteomic analysis of lipidated proteins. These
studies are currently underway and will be reported in due course.
The development ω-azido-fatty acids adds to the toolbox of
with NH
ω-azido-fatty acid 1 were resistant to NH
In contrast, proteins visualized with ω-azido-fatty acids 2-4 were
sensitive to NH OH treatment and suggest that the majority of
2
OH demonstrated that the polypeptides visualized with
2
OH cleavage (Figure 2A).
2
ω-azido-fatty acids 2-4 are attached to proteins through a thioester
bond (Figure 2A). Proteins labeled with the 14-carbon ω-azido-
fatty acid 2 were differentially sensitive to NH OH, which suggests
2
that compound 2 may serve as a substrate analogue for both myristic
and palmitic acid (Figure 2A). Co-incubation with myristic acid
selectively decreased protein labeling with ω-azido-fatty acid 1 in
a dose-dependent manner, whereas lauric acid and palmitic acid
had no effect at the same concentration (Supporting Information
Figure 2A and B). Alternatively, protein labeling with ω-azido-
fatty acid 3 was selectively reduced by co-incubation with longer
chain fatty acids, such as palmitic acid or stearic acid and not with
lauric or myristic acid (Supporting Information Figure 2C).
Furthermore, proteins labeled with ω-azido-fatty acid 1 were
sensitive to co-incubation with cycloheximide, whereas proteins
targeted by ω-azido-fatty acid 3 were not (Supporting Information
Figure 2D), consistent with the co-translational addition of myristic
acid to proteins by N-myristoyltransferases6 and the post-
translational modification of palmitoylated proteins by S-palmi-
12
chemical reporters to monitor protein glycosylation and farnesy-
lation.13
Acknowledgment. We thank Dr. Jannie Borst (Netherlands
Cancer Institute) for the gift of polyclonal anti-Lck sera. H.C.H.
acknowledges the Damon Runyon Cancer Research Foundation for
a postdoctoral fellowship. A.M.P. acknowledges the National
Science Foundation for graduate fellowship. M.J.B. acknowledges
the Medical Research Council for a Career Development Award.
This work was funded by NIH grant to H.L.P. (5RO1A1034893-
13).
Supporting Information Available: Procedures for chemical
synthesis, metabolic labeling, and streptavidin blots. This material is
available free of charge via the Internet at http://pubs.acs.org.
7
toyltransferases. Collectively, these results suggest that the ω-azido-
fatty acid 1 selectively targets N-myristoylated proteins, whereas
the longer chain ω-azido-fatty acids 2-4 primarily label S-acylated
proteins.
References
(
1) Resh, M. D. Nat. Chem. Biol. 2006, 2, 584-590.
To further establish fatty-acylated protein labeling with ω-azido-
fatty acids, we analyzed a well-characterized fatty-acylated protein,
Lck, a Src-family protein kinase that is essential for T cell
(2) Resh, M. D. Methods 2006, 40, 191-197.
(
(
3) Drisdel, R. C.; Green, W. N. Biotechniques 2004, 36, 276-285.
4) Roth, A. F.; Wan, J.; Bailey, A. O.; Sun, B.; Kuchar, J. A.; Green, W.
N.; Phinney, B. S.; Yates, J. R., III; Davis, N. G. Cell 2006, 125, 1003-
1013.
8
activation. Human Jurkat T cells were metabolically labeled with
(
5) Saxon, E.; Bertozzi, C. R. Science 2000, 287, 2007-2010.
ω-azido-fatty acid 1 or 3, cell lysates were prepared, labeled with
phosphine-biotin, immunoprecipitated for endogenous Lck with
polyclonal anti-Lck sera,9 and analyzed for fatty acylation by
streptavidin blotting (Figure 2B). Lck immunoprecipitated from
Jurkat cell lysates was efficiently labeled with the myristic acid
analogue 1 as well as by the palmitic acid analogue 3, albeit at
much lower levels (Figure 2B). The differences observed between
Lck labeling with compounds 1 and 3 are consistent with N-
myristoylation being a constitutive modification, whereas S-
palmitoylation is often dynamic and substoichiometric. These results
are similar to those observed using radiolabeled fatty acids.10 To
unequivocally demonstrate that the myristic acid analogue 1 labels
the N-terminal glycine residue of Lck and not the side chain of
lysine residues elsewhere in the protein, we analyzed wild-type Lck
and an N-terminal glycine to alanine Lck mutant (G2A), which
blocks N-myristoylation of proteins.11 Human embryonic kidney
(6) Farazi, T. A.; Waksman, G.; Gordon, J. I. J. Biol. Chem. 2001, 276,
3
9501-39504.
7) Smotrys, J. E.; Linder, M. E. Annu. ReV. Biochem. 2004, 73, 559-587.
(8) Kabouridis, P. S.; Magee, A. I.; Ley, S. C. EMBO J. 1997, 16, 4983-
(
4998.
(
9) Taher, T. E.; Smit, L.; Griffioen, A. W.; Schilder-Tol, E. J.; Borst, J.;
Pals, S. T. J. Biol. Chem. 1996, 271, 2863-2867.
(
(
(
10) Bijlmakers, M. J.; Isobe-Nakamura, M.; Ruddock, L. J.; Marsh, M. J.
Cell Biol. 1997, 137, 1029-1040.
11) Paige, L. A.; Nadler, M. J.; Harrison, M. L.; Cassady, J. M.; Geahlen, R.
L. J. Biol. Chem. 1993, 268, 8669-8674.
12) (a) Vocadlo, D. J.; Hang, H. C.; Kim, E. J.; Hanover, J. A.; Bertozzi, C.
R. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 9116-9121. (b) Hang, H. C.;
Yu, C.; Kato, D. L.; Bertozzi, C. R. Proc. Natl. Acad. Sci. U.S.A. 2003,
1
00, 14846-14851. (c) Prescher, J. A.; Dube, D. H.; Bertozzi, C. R.
Nature 2004, 430, 873-877. (d) Dube, D. H; Prescher, J. A.; Quang, C.
N.; Bertozzi, C. R. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 4819-4824.
(e) Rabuka, D.; Hubbard, S. C.; Laughlin, S. T.; Argade, S. P.; Bertozzi,
C. R. J. Am. Chem. Soc. 2006, 128, 12078-12079. (f) Sawa, M.; Hsu,
T.-L.; Itoh, T.; Sugiyama, M.; Hanson, S. R.; Vogt, P. K.; Wong, C.-H.
Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 12371-12376.
(
13) Kho, Y.; Kim, S. C.; Jiang, C.; Barma, D.; Kwon, S. W.; Cheng, J.;
Jaunbergs, J.; Weinbaum, C.; Tamanoi, F.; Falck, J.; Zhao, Y. Proc. Natl.
Acad. Sci. U.S.A. 2004, 101, 12479-12484.
(HEK) 293T cells were transfected with cDNA constructs encoding
wild-type Lck or the G2A mutant Lck, labeled with the myristic
JA0685001
J. AM. CHEM. SOC.
9
VOL. 129, NO. 10, 2007 2745