Site-specific analysis of protein S-acylation by resin-assisted capture.

Protein S-acylation is a major posttranslational modification whereby a cysteine thiol is converted to a thioester. A prototype is S-palmitoylation (fatty acylation), in which a protein undergoes acylation with a hydrophobic 16 carbon lipid chain. Although this modification is a well-recognized determinant of protein function and localization, current techniques to study cellular S-acylation are cumbersome and/or technically demanding. We recently described a simple and robust methodology to rapidly identify S-nitrosylation sites in proteins via resin-assisted capture (RAC) and provided an initial description of the applicability of the technique to S-acylated proteins (acyl-RAC). Here we expand on the acyl-RAC assay, coupled with mass spectrometry-based proteomics, to characterize both previously reported and novel sites of endogenous S-acylation. Acyl-RAC should therefore find general applicability in studies of both global and individual protein S-acylation in mammalian cells.


Detection of S -acylated proteins by acyl-RAC
Following the indicated treatments/transfections, cells were collected and washed in cold PBS. After undergoing a freezethaw cycle, cells were lysed in lysis buffer (25 mM HEPES, 25 mM NaCl, 1 mM EDTA, pH 7.5) containing protease inhibitor cocktail (Roche). Lysis was improved by repeated passaging through a 28 gauge needle. For enrichment of membranes, lysates were depleted of nuclei via centrifugation at 800 g for 5 min. The supernatant was then centrifuged at 20,000 g for 30 min, and the pellet was resuspended in lysis buffer containing 0.5% Triton X-100. Total protein was quantifi ed with a bicinchononic acid (BCA) assay (Pierce) using BSA as the standard. Methodology for acyl-RAC, including blocking of free thiols with methyl methanethiosulfonate (MMTS), cleavage of thioester linkages, and capture of nascent thiols on thiopropyl Sepharose, was carried out essentially as described previously ( 23 ). In particular, equal amounts of protein (0.5-2.0 mg for immunoblot experiments and 10-20 mg for mass spectrometry experiments) were diluted to a concentration of 2 mg/ml in blocking buffer (100 mM HEPES, 1.0 mM EDTA, 2.5% SDS, 0.1% MMTS, pH 7.5) and incubated at 40°C for 10 min with frequent vortexing. Three volumes of cold acetone were added, and proteins were allowed to precipitate at Ϫ 20°C for 20 min. Following centrifugation of the solution at 5,000 g for 10 min, the pellet was extensively washed with 70% acetone, resuspended in 300 l of binding buffer (100 mM HEPES, 1.0 mM EDTA, 1% SDS, pH 7.5) and added to ‫ف‬ 40 l of prewashed thiopropyl Sepharose (GE-Amersham). To this mixture was added 40 l of either 2 M NH 2 OH (freshly prepared in H 2 O from HCl salt and brought to pH 7.5 with concentrated NaOH) or 2 M NaCl. Binding reactions were carried out on a rotator at room temperature for 2-4 h. Approximately 20 l of each supernatant was saved as the "total input." Resins were washed at least fi ve times with binding buffer. For immunoblot analysis, elution was performed using 60 l of binding buffer containing 50 mM DTT at room temperature for 20 min. Supernatants were removed and mixed with Laemmli loading buffer, heated to 95°C for 5 min, and separated via SDS-PAGE on a Mini-Gel apparatus (Bio-Rad).

On-resin trypsinization and mass spectrometric analysis of S -acylated sites
This procedure was performed essentially as described previously ( 23 ) but is fully detailed in the supplementary information (available at http://www.jlr.org).

Application of the acyl-RAC technique using purifi ed bovine brain membranes
The acyl-RAC assay is chemically analogous to the ABE assay, although it replaces the biotinylation/avidin pull-down adapted to immunoblotting techniques and is also adaptable to mass spectrometric-based identifi cation of individual S -acylated proteins (19)(20)(21)(22). However, the detection of biotinylated proteins requires expensive reagents and complicated procedures (e.g., repeated protein precipitations, SDS neutralization, and avidin pull down). We recently provided an initial description of a simple and robust alternative to ABE that uses the detection of S -acylated species via resin-assisted capture (acyl-RAC) in lieu of biotinylation ( 23 ). The method is rapid (the entire procedure can be completed in several hours) and is readily adapted to mass spectrometry techniques for identifying sites of S -acylation. Here we provide a detailed validation and expansion of the acyl-RAC method and demonstrate its effi cacy in detecting S -acylated protein substrates and sites of modifi cation.

Materials and reagents
All materials were obtained from Sigma Chemicals (St. Louis, MO), unless otherwise indicated. Sources of antibodies were mouse MAb ␣ -HA (code 2367; Cell Signaling Technology); and rabbit polyclonal antibody ␣ -H-Ras (code sc-520; Santa Cruz Biotechnology). Bovine brain membranes were isolated as described previously ( 24 ).

Mammalian cell culture and transfection
All cells were cultured at 37°C in a 5% C O 2 atmosphere. Cell lines were obtained from the Duke Cell Culture Facility and grown in DMEM (HEK293 cells) or McCoy's 5A medium (T24 cells) supplemented with 10% FBS, 100 U/ml penicillin, and 100 g/ml streptomycin. Cells were transfected with Superfect (Qiagen) per the manufacturer's instructions. In general, HEK293 cells were grown in 10 cm dishes to 70%-80% confl uency and transfected with 12 g of the indicated DNA and 48 l of Superfect (Qiagen). Approximately 24 h later, cells were harvested with cold PBS and used immediately.

Application of acyl-RAC to analysis of H-Ras, a model S -palmitoylated protein
To further explore the utility of acyl-RAC to detect S -acylation in an intact mammalian cell culture system, HEK293 cells were transfected with vectors encoding H-Ras, which is known to undergo S -palmitoylation on Cys 181 and Cys 184 ( 27 ) and S -farnesylation on Cys 186 ( 3,4 ). The highly modifi ed C terminus of human H-Ras is shown in Fig. 2A . As shown in Fig. 2B , acyl-RAC readily detected S -palmitoylation of H-Ras in a hydroxylamine-dependent manner. Importantly, the C 181/184 S double mutant, which cannot undergo S -acylation, was not detected. Furthermore, because the C 181/184 S mutant continues to undergo S -farnesylation on Cys 186 ( 28 ), these results confi rm the expected result that acyl-RAC does not detect S -prenylated proteins (because the thioether linkage is not susceptible to hydroxylamine cleavage). Further confi rmation that the protein species identifi ed by acyl-RAC are indeed S -acylated was provided by the observation that the degree of H-Ras S -palmitoylation was attenuated by incubation with 2-bromopalmitate, a known inhibitor of S -palmitoylation ( Fig. 2C ). Endogenous S -palmitoylated H-Ras could also be readily detected in the T24 bladder carcinoma cell line (supplementary Fig. IIA), in which the oncogenic G 12 V variant of H-Ras is known to drive the tumorigenic phenotype ( 29 ).
step with the use of direct conjugation to resin containing thiol-reactive thiopyridinyl groups ( Fig. 1 ). This strategy is advantageous for examining cysteine-based modifi cations because it is rapid and economical, and it allows the resinimmobilized proteins to be processed conveniently with virtually any chemical or enzyme treatment, except reductants (which would drive elution). As shown in supplementary Fig. I, acyl-RAC was applied to examine S -acylated proteins in bovine brain membranes, which are known to be rich in S -palmitoylated proteins. A number of proteins were readily detected by acyl-RAC in a hydroxylaminedependent manner via Coomassie staining of eluted proteins resolved by SDS-PAGE. In addition , two S -palmitoylated proteins known to be present in brain, G ␣ z ( 25 ) and GAP-43 ( 26 ), were readily detected by immunoblot analysis of acyl-RAC proteins, and only if the samples had been treated with hydroxylamine to cleave endogenous thioesters. In contrast, synaptophysin, which is not a substrate for S -acylation, was not detected by acyl-RAC. Thus,    Table I for the complete listing of sites identifi ed). C: Validation of MS data by transfection of HEK293 cells with putative S -acylated proteins followed by acyl-RAC and immunoblotting for the specifi ed individual proteins. In each case, the identifi ed sites of S -acylation were mutated to serine as noted. For MGST3, the identifi ed peptide contained two Cys residues (Cys 150 and Cys 151 ); therefore, both single and double mutants (DM) were subjected to acyl-RAC.

Validation of novel S -acylated targets identifi ed via MS-coupled acyl-RAC
To examine the fi delity of MS-based identifi cation of S -acylation sites from acyl-RAC-identifi ed proteins, three candidate proteins that were not previously studied in the context of S -acylation were selected for further analysis: the ␤ -subunit of the protein translocating system (Sec61b), ribosomal protein S11 (Rps11), and microsomal glutathione-S -transferase 3 (MGST3). These three proteins, and mutations of each in which the identifi ed S -acylated Cys had been changed to a Ser residue, were expressed in HEK293 cells. Cells were transfected with the respective HA-tagged constructs and then analyzed by acyl-RAC with anti-HA immunoblotting. As shown in Fig. 3C , acyl-RAC Although H-Ras is a highly studied prototype of S -acylated proteins, a more complex system was desired to verify the general applicability of acyl-RAC. To that end, a membrane-enriched fraction from HEK293 cells was pretreated with either buffer or palmitoyl-CoA, followed by analysis via acyl-RAC and direct visualization of captured proteins via SDS-PAGE and Coomassie staining (supplementary Fig. IIB). Capture of cellular proteins was both augmented by palmitoyl-CoA pretreatment and dependent on NH 2 OH during the assay, demonstrating that acyl-RAC can detect a range of S -palmitoylated proteins.

Mass spectrometry-coupled acyl-RAC for identifi cation of S -acylation sites
We also assessed the utility of acyl-RAC in identifi cation of specifi c sites of S -acylation on captured proteins, by using isobaric labeling and LC-MS/MS. Samples of a membrane-enriched fraction from HEK293 cells were subjected to the acyl-RAC procedure in the presence and absence of NH 2 OH, followed by on-resin trypsinization of captured proteins and isobaric labeling with either iTRAQ-114 atomic mass units (amu) (control) or iTRAQ-117 amu (plus NH 2 OH) reporter tags. Resins containing the proteins captured from both conditions were combined, and the resulting eluants were analyzed by LC-MS/MS. From a search of the human Swiss-Prot database, 93 putative sites of S -acylation on 88 peptides were identifi ed (supplementary Table I), including a number of sites previously known to undergo S -palmitoylation (Fig. 3B). Of the 88 identifi ed peptides, 84 peptides contained at least one Cys residue (the database search was not restricted to Cys-containing peptides and therefore provided another internal control). As an example, data obtained from the ␣ -subunit of the heterotrimeric G-protein G s , which is palmitoylated on the N-terminal Cys 3 ( 30,31 ), are shown (Fig. 3A). This N-terminal peptide containing Cys 3 was identifi ed by MS-coupled acyl-RAC, whereas none of the 7 other potential Cys-containing peptides from G s were identifi ed in the analysis. These data further validate the utility of acyl-RAC for identifying sites of S -palmitoylation in intact cells.
Several other established sites of S -acylation that were identifi ed by using acyl-RAC are shown in Fig. 3B . These sites include Cys 9 and Cys 10 within the ␣ -subunit of the heterotrimeric G-protein, G 11 ( 32 ), Cys 181 and Cys 184 of H-Ras ( Fig. 2A, B, and see reference [ 27 ]), as well as 6 Cys residues within SNAP23, of which 4 are conserved in its betterstudied homolog SNAP25, a known S -palmitoylated protein ( 33 ). Also identifi ed was the active site cysteine, Cys 632 , of E1 ubiquitin activating enzyme 1 (UBA1). This protein is known to form a thioester with the C-terminal glycine of ubiquitin at Cys 632 ( 34 ), which is required for ubiquitin transfer to downstream E2 proteins. Although UBA1 contains 19 Cys residues, acyl-RAC detected only the Cys-containing peptide from the active site, demonstrating that acyl-RAC is capable of identifying diverse types of S -acylation. detected all three proteins in a hydroxylamine-dependent fashion, whereas mutations of the putative acylation sites abrogated their detection by acyl-RAC. These fi ndings confi rm the fi delity of acyl-RAC in detecting both known and novel sites of S -acylation in intact cells.

DISCUSSION
Fatty acylation of proteins is increasingly recognized as a regulator of protein localization and a facilitator of signaling from cellular membranes (8)(9)(10)(11)35 ). Given the broad range of proteins that are known to undergo S -acylation (e.g., heterotrimeric G ␣ isoforms, H-Ras, N-Ras, Src family members, and G-protein-coupled receptors [GPCR]), substantial effort has been focused on understanding the mechanisms and biological consequences of S -acylation. These efforts will be aided signifi cantly by the development of effi cient tools for detecting S -acylated proteins and sites of modifi cation. Here we have described in detail and validated the acyl-RAC methodology, which effi caciously detects endogenous S -acylation. Notably, the acyl-RAC procedure can be completed in an afternoon and is fully compatible with modern proteomic methodologies to unambiguously identify proteins and their sites of modifi cation.
The acyl-RAC methodology should help provide new insight into the dynamics of S -acylation. The major advantage of acyl-RAC, however, is the simplicity with which it can be performed. Compared with the ABE assay, acyl-RAC uses far fewer procedures, thus minimizing the number of steps at which mistakes might inadvertently occur. Insofar as acyl-RAC is analogous to the detection of S -nitrosylated proteins by resin-assisted capture (SNO-RAC) , one would expect acyl-RAC to be similar with regard to sensitivity to the ABE assay (except for high-molecularweight proteins, where resin-based approaches like acyl-RAC and SNO-RAC are likely superior) ( 23 ).
In combination with metabolic labeling approaches (e.g., [ 3 H]palmitate, palmitate-based click chemistry), acyl-RAC can provide a powerful approach to the analysis of dynamic fatty acylation. It should be emphasized that acyl-RAC (as in the case of ABE) detects all types of S -acylation, (i.e., the presence of thioesterifi ed Cys residues) and cannot characterize the nature of the endogenous acyl group. For example, the active site, Cys 632 , of UBA1 undergoes S -acylation with a glycyl ubiquitin moiety, but UBA1 is apparently not S -palmitoylated. When questions regarding the acyl group arise, metabolic labeling with radiolabeled palmitate or chemically derivatized palmitate analogs (14)(15)(16)(17), as well as inhibitors of palmitoylation (e.g., 2-bromopalmitate), can help distinguish the nature of the acyl moiety. The acyl-RAC method should facilitate analysis of cellular protein S -acylation under physiological and pathophysiological conditions.