These authors contributed equally to this work.
Probing Protein–Protein Interactions with a Genetically Encoded Photo-crosslinking Amino Acid
Article first published online: 15 JUN 2011
Copyright © 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Volume 12, Issue 12, pages 1854–1857, August 16, 2011
How to Cite
Ai, H.-w., Shen, W., Sagi, A., Chen, P. R. and Schultz, P. G. (2011), Probing Protein–Protein Interactions with a Genetically Encoded Photo-crosslinking Amino Acid. ChemBioChem, 12: 1854–1857. doi: 10.1002/cbic.201100194
- Issue published online: 5 AUG 2011
- Article first published online: 15 JUN 2011
- Manuscript Received: 21 MAR 2011
- US Department of Energy, Division of Materials Sciences. Grant Number: DE-FG03-00ER46051
- Skaggs Institute for Chemical Biology
- amber suppression;
- protein–protein interactions;
- unnatural amino acids
Considerable effort has been devoted to mapping the complex interactions that make up the molecular circuitry of living cells.1 For weak or transient interactions that are not easily identified through affinity-based approaches,2 methods such as yeast two-hybrid screening,3 protein-fragment complementation,4 chemical crosslinking, and photo-crosslinking5 are particularly useful. Photo-crosslinkers can be directly incorporated into proteins,6, 7 nucleic acids,8, 9 and carbohydrates10, 11 by exogenously supplying cells with appropriate precursors. Alternatively, photo-crosslinking moieties can be site-specifically introduced into proteins chemically or enzymatically,12 or in response to nonsense codons by means of orthogonal amber suppressor tRNA/aminoacyl-tRNA synthetase pairs.13 For example, amino acids containing benzophenone, diazirine, and aryl azide side chains have all been genetically encoded in prokaryotic and eukaryotic organisms,14–18 and used to crosslink protein–protein and protein–DNA interactions in vitro and in living cells.19–23 However, the size and photochemical reactivity of these photo-crosslinkers might not be ideal for a given site in a target protein of interest. Recently, an aliphatic diazirine amino acid, 3′-azibutyl-N-carbamoyl-lysine (AbK, Figure 1 A) was introduced into proteins by using the wild-type M. barkeri pyrrolysyl tRNA/tRNA synthetase pair (wt-mbPylRS/tRNAPyl).24 Site-specific incorporation of AbK into glutathione S-transferase allowed covalent crosslinking of the two subunits of the dimeric protein in E. coli by using UV light.25 Inspired by this success, we sought to solve another limitation: in mammalian cells, photo-crosslinking amino acids typically have a low efficiency of incorporation in response to amber (TAG) codons.18, 25 Here we report that a new aminoacyl-tRNA synthetase engineered from wt-mbPylRS, when coupled with tRNAPyl, can significantly increase the coding efficiency of AbK in both E. coli and mammalian cells. Moreover, when Abk was substituted for Asp144 in cyclin-dependent kinase 5 (Cdk5), the diazirine moiety photo-crosslinked Cdk5 to its substrate, p21-activated kinase 1 (Pak1).
Previously, wild-type M. maize PylRS and mbPylRS have been used to incorporate a variety of carbamoyl lysine derivatives into proteins.26–29 We also confirmed that the wt-mbPylRS/tRNAPyl pair can indeed suppress the TAG codon in the presence of AbK. A plasmid encoding wt-mbPylRS (pBK-PylRS) was used to cotransform E. coli DH10B cells along with a second plasmid pMyo-Lys99TAG (harboring a His6-tagged myoglobin gene containing a Lys99TAG mutation, and one copy of the pyrrolysyl-tRNA gene).29 Protein expression was induced in the presence of 1 mM AbK. Ni-NTA affinity chromatography was used to purify full-length myoglobin, and eluted protein was analyzed by SDS-PAGE. Although full-length protein was observed when AbK was present in culture media, the yield was low (0.7±0.2 mg myoglobin per liter of culture; Figure 1 B).
Previous work has also made use of mutant PylRS/tRNAPyl pairs to encode lysine analogues.30–33 We postulated that the wt-mbPylRS/tRNAPyl pair could be subjected to directed evolution to optimize its amber suppression efficiency in the presence of AbK. We therefore used a mbPylRS library randomized at residues Leu270, Tyr271, Leu274, and Cys313 (and an additional Tyr349Phe mutation)31 to select synthetases that can efficiently charge AbK to tRNAPyl. A series of positive and negative selections were performed in E. coli strain DH10B as previously described.34 In brief, in the positive selection, PylRS variants were introduced into E. coli containing a plasmid encoding chloramphenicol acetyltransferase with the Asp112TAG mutation, and cells were selected for chloramphenicol resistance in the presence of 1 mM AbK; the negative selection was carried out with a toxic barnase gene with amber mutations at multiple sites (Gln2TAG, Asp44TAG, and Gly65TAG), and the selection was carried out in the absence of AbK. After three positive and two negative rounds of selection, single E. coli colonies containing PylRS mutants were obtained. The sequences of mutant synthetases converged to a unique clone (AbKRS; mbPylRS-Leu274Met/Cys313Ala/Tyr349Phe). Next, we tested amber suppression of pMyo-Lys99TAG by the AbKRS/tRNAPyl pair. Full-length myoglobin was produced in good yield in the presence of 1 mM AbK (12±1.1 mg myoglobin per liter of culture), while full-length myoglobin was not observed in the absence of AbK (Figure 1 B). The resulting protein was characterized by electrospray ionization mass spectrometry (ESI-MS): the observed molecular mass confirmed that AbK was site-specifically incorporated at residue 99 of myoglobin (Figure 2 A and the Supporting Information). When the protein was exposed to 360 nm light, mass spectrometry revealed complete photolysis in 10 min (Figure 2 B and the Supporting Information).35
To test whether AbKRS engineered in E. coli can be directly adapted to encode AbK in mammalian cells, a pCMV-AbK plasmid was constructed to express both AbKRS and tRNAPyl (Figure S2).30 HEK (human embryonic kidney) 293T cells were transfected with pCMV-AbK (in which tRNAPyl and AbKRS are expressed under the U6 and CMV promoters, respectively) and pEGFP-Tyr39TAG (His6-tagged enhanced green fluorescent protein gene with a TAG mutation at residue 39). Transiently transfected cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10 % fetal bovine serum (FBS) in the absence or presence of 1 mM AbK for 48 h. EGFP fluorescence was imaged with an inverted fluorescence microscope, and was only observed when AbK was added to the culture medium (Figure 3 A). A portion of EGFP-positive cells were subjected to 360 nm irradiation for 10 min. Cells (with or without UV treatment) were lysed, and EGFP proteins were purified by Ni-NTA affinity chromatography. The yield (∼0.4 mg EGFP from 1×107 cells) was notably higher than when other photo-crosslinking amino acids were used in similar experiments.18 ESI-MS analysis indicated that AbK was site-specifically incorporated into EGFP, and its photolysis can be efficiently induced in living mammalian cells by UV irradiation. (Figure 3 B and the Supporting Information).35
Kinase-catalyzed phosphorylation plays a central role in cellular regulation and signaling, and the identification of kinase interactions remains a challenge in cell biology. To address this, we tested the ability of AbK to crosslink the kinase Cdk5 to its cognate substrate Pak1 in mammalian cells. Cdk5 and its activator p35 are known to bind and phosphorylate the downstream kinase Pak1 in a RacGTP-dependent manner.36, 37 We selected a residue (Asp144) positioned within the Cdk5 kinase catalytic cleft, and generated the Asp144TAG mutant of His8-tagged Cdk5 (pCdk5-Asp144TAG). The plasmids pCdk5-Asp144TAG and pCMV-AbK were used to cotransfect HEK 293T cells. For comparison, the pCMV-PylRS plasmid encoding wt-mbPylRS and tRNAPyl was used in the control group. Cells were lysed, and Cdk5 proteins were probed with anti-His antibody (Figure 4 A). When 1 mM AbK was present in culture media, full-length Cdk5 was detected in both groups of cells. However, Western blot analysis showed that the mutant synthetase significantly improved the amber-suppression efficiency in mammalian cells, consistent with our previous results in E. coli. Next, a constitutively active Rac mutant (Rac-Gln61Leu), p35 and Pak1 were expressed in HEK 293T cells containing pCMV-AbK and pCdk5-Asp144TAG. Transfected cells were cultured in the presence of 1 mM AbK. After 48 h, cells were washed with phosphate-buffered saline (PBS), exposed to 360 nm light for 10 min, and then lysed. Cdk5 as well as any crosslinked proteins were purified on Ni-NTA affinity columns. SDS-PAGE and Western blot analysis with anti-Pak1 antibody showed that Pak1 was crosslinked to Cdk5 (Figure 4 B and C). The crosslinked band on the SDS-PAGE gel was subjected to in-gel trypsin digestion followed by LC-MS/MS-based proteomic analysis. A database search further confirmed the presence of Pak1 and Cdk5 proteins in the gel slice (Supporting Information).
In conclusion, wt-mbPylRS/tRNAPyl has been engineered to genetically encode the aliphatic photo-crosslinking amino acid AbK in high efficiency in both E. coli and mammalian cells. Due to its small size and flexible nature, AbK might cause less structural perturbation than other photo-crosslinking amino acids when incorporated into proteins.38 In addition, the improved efficiency of AbK incorporation could be useful when the target protein of interest is difficult to express.
Protein expression and purification fromE. colicells: Plasmids pMyo-Lys99TAG and pBK (harboring the mutant synthetase gene) have been described previously.29 Both were used to cotransform DH10B E. coli cells. A single colony was used to inoculate 2YT medium [25 mL, containing L-arabinose (0.2 %), tetracycline (20 μg mL−1), and kanamycin(50 μg mL−1)] in the presence or absence of AbK (1 mm) at 30 °C for 14 h. Cells were harvested by centrifugation and lysed with B-PER II protein extraction reagent (Pierce). His6-tagged myoglobin was purified with Ni-NTA agarose beads (Qiagen) under native conditions according to the manufacturer's instructions.
Protein expression and purification from HEK 293T cells: The mammalian expression vector pCMV has been described previously.30, 31 In brief, the plasmid contains a copy of the synthetase gene driven by a CMV promoter, and a copy of the tRNA gene under the control of a human U6 promoter (Figure S2). The new plasmid pCMV-AbK was created by replacing the synthetase in the previous pCMV-NBK plasmid.30, 31 HEK 293T cells were grown in DMEM supplemented with 10 % fetal bovine serum (FBS). Cells at 70 % confluency were transfected with mixtures of the corresponding plasmids by using Lipofectamine 2000 (Invitrogen). The culture medium was further supplemented with AbK (1 mm) as appropriate. When expressing EGFP in HEK 293T cells, pCMV (12 μg) and pEGFP-Tyr39TAG (12 μg) were mixed with Lipofectamine (60 μL) to transfect cells in 10 mm diameter cell culture dishes; when expressing Cdk5 mutants for anti-His blot, pCMV (2 μg) and pCdk5 (2 μg) were used with Lipofectamine (10 μL) to transfect cells in 6-well plates. Cells were harvested 48 h after transfection, washed with PBS (three times), and then collected and lysed with radio-immunoprecipitation assay (RIPA) buffer (Sigma) on ice for 30 min. Lysates were cleared with a bentchtop centrifuge at 5000 g for 2 min and used directly for Western blot or were purified by Ni-NTA agarose beads (Qiagen). In the Ni-NTA purification step, the manufacturer's instructions for native conditions were followed.
Protein electrospray mass spectrometry: Proteins were directly injected into an Agilent 1100 Series LC/MS instrument (quadrupole MSD equipped with a short C-8 column to remove salt impurities). Observed spectra were deconvoluted to derive protein masses by using the Agilent LC/MSD Deconvolution package provided along with the instrument. The instrument detects protein masses within the expected ±0.02 % mass error.
Photo-crosslinking: To photo-crosslink Cdk5 and Pak1, cells were first transfected with appropriate plasmid mixtures. For HEK 293T cells cultured in 10 mm diameter dishes, a mixture of Lipofectamine (40 μL), pCMV-AbK (4 μg), pCdk5 (4 μg), pCMV-Pak1 (3 μg, Addgene, Cambridge, MA, USA), pCMV-p35 (3 μg, Addgene), and pRac-Q61L gene (2 μg, Addgene) was used for transfection. After 48 h of protein expression in the corresponding medium, cells were washed in PBS (3×20 mL), and then irradiated under a 360 nm UV lamp for 10 min on ice (Black Ray Lamp, Model XX-20BLB, VWR, cat. no. 21 474–21 676). Next, cells were collected and disrupted with RIPA buffer. Ni-NTA agarose beads were again used to purify the protein complexes. The manufacturer's protocol was slightly modified so that imidazole in PBS (35 mm, ten bead column) was used to increase the wash stringency. After the proteins were eluted with imidazole (500 mM in PBS), SDS-PAGE and Western blot analysis were performed. The proteomic analysis of the gel slice was carried out by the Scripps Center for Mass Spectrometry (La Jolla, CA), and is further described in the Supporting Information.
This work was supported by the US Department of Energy, Division of Materials Sciences, under Award No. DE-FG03-00ER46051 and the Skaggs Institute for Chemical Biology. We thank Corey Dambacher, Dr. Jonathan Day, Frank Peters, and Virginia Seely for manuscript preparation.
- 2Curr. Protein Mol. Biol. 2001, 10.16.11–10.16.29., , ,
- 5Crosslinking Reagents Technical Handbook, Thermo Scientific Pierce, Rockford, 2009.
Detailed facts of importance to specialist readers are published as ”Supporting Information”. Such documents are peer-reviewed, but not copy-edited or typeset. They are made available as submitted by the authors.
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.