Journal of Structural and Functional Genomics 2: 29–35, 2001. © 2001 Kluwer Academic Publishers. Printed in the Netherlands.
29
Selenomethionine incorporation into a protein by cell-free synthesis Takanori Kigawa1,2,† , Emi Yamaguchi-Nunokawa1,†, Koichiro Kodama1,3, Takayoshi Matsuda1, Takashi Yabuki1 , Natsuko Matsuda1 , Ryuichiro Ishitani3, Osamu Nureki3 & Shigeyuki Yokoyama1,2,3,∗ 1 RIKEN Genomic Sciences Center,
1-7-22 Suehiro-cho, Tsurumi, Yokohama, 230-0045 Japan; 2 Cellular Signaling Laboratory, RIKEN Harima Institute at SPring-8, 1-1-1 Kouto, Mikazuki-cho, Sayo, Hyogo, 679-5148 Japan; 3 Department of Biophysics and Biochemistry, Graduate School of Science, The University of Tokyo, 7-3-1 Hongo, Bunkyo, Tokyo, 113-0033 Japan; ∗ Author for correspondence (e-mail:
[email protected]; fax: +81-45-503-9195) Received 12 June 2001; Accepted in revised form 12 September 2001
Key words: cell-free protein synthesis, multi-wavelength anomalous diffraction phasing, selenomethionine, structural genomics, X-ray crystallography
Abstract Multi-wavelength anomalous diffraction phasing is especially useful for high-throughput structure determinations. Selenomethionine substituted proteins are commonly used for this purpose. However, the cytotoxicity of selenomethionine drastically reduces the efficiency of its incorporation in in vivo expression systems. In the present study, an improved E. coli cell-free protein synthesis system was used to incorporate selenomethionine into a protein, so that highly efficient incorporation could be achieved. A milligram quantity of selenomethionine-containing Ras was obtained using the cell-free system with dialysis. The mass spectrometry analysis showed that more than 95% of the methionine residues were substituted with selenomethionine. The crystal of this protein grew under the same conditions and had the same unit cell constants as those of the native Ras protein. The three-dimensional structure of this protein, determined by multi-wavelength anomalous diffraction phasing, was almost the same as that of the Ras protein prepared by in vivo expression. Therefore, the cell-free synthesis system could become a powerful protein expression method for high-throughput structure determinations by X-ray crystallography. Abbreviations: LC-MS – liquid chromatography-electrospray mass spectrometry; MAD – multi-wavelength anomalous diffraction; MALDI-TOF MS – matrix assisted laser desorption ionization-time of flight mass spectrometry; NMR – nuclear magnetic resonance. Introduction Multi-wavelength anomalous diffraction (MAD) phasing [1] is a powerful approach for the phasing of macromolecular crystal structures. As it fundamentally requires only a single crystal containing the heavy atom that causes anomalous diffraction, such as selenium, it would free us from the time-consuming heavy atom searching steps. Selenomethionine incorporated proteins are useful for the MAD phasing [2], and this has greatly improved the speed and ac† The two authors contributed equally to this work.
curacy of protein structure determinations by X-ray crystallography. Therefore, this method would be especially suitable for high-throughput protein structure determinations in structural genomics. The methionine in the cell culture medium is replaced by selenomethionine for its incorporation into proteins when in vivo expression methods are used. In many cases, this results in the low production of the target protein, because the cytotoxicity of selenomethionine may strongly inhibit the cell growth [3]. In addition, the amino acid metabolism in an in vivo expression system drastically reduces the efficiency of
30 selenomethionine incorporation [3]. These problems of selenomethionine incorporation in the in vivo expression method can be solved by using a cell-free protein synthesis system, in which both amino acid metabolism and cytotoxicity are not concerns. We have already shown that highly efficient and selective stable-isotope labeling of proteins can be achieved by a cell-free system [4–6]. Low productivity problem of the cell-free synthesis has been solved by the modification of reaction conditions [5, 7, 8] and the development of the dialysis method [6, 9, 10]. Then, the productivity has become comparable to that of the in vivo expression method [6]. We have succeeded in producing about hundred of mouse cDNA products as glutathione-S-transferase fused forms [11], indicating that the system could be applicable to variety of proteins. Thus, the MAD phasing in combination with the selenomethionine incorporation by the cellfree system is expected to become a powerful highthroughput protein structure determination method in the structural genomics field. In the present study, an improved E. coli cell-free protein synthesis system with dialysis [6] was applied to the selenomethionine incorporation of the Ras protein. A high-resolution crystal structure of the Ras protein has already been determined [12]. A milligram quantity of fully selenomethionine substituted Ras protein was obtained, and then its three-dimensional structure was solved by the MAD phasing.
Materials and methods Cell-free synthesis of Ras protein by the dialysis method The E. coli S30 cell extract was prepared by modification of the procedure of Pratt [13]. As there is no enough space for describing the detail of our procedure, it will be published elsewhere. T7 RNA polymerase was prepared according to Zawadzki and Gross [14]. The human c-Ha-Ras protein used in this study consisted of 171 amino acid residues, and lacked the C-terminal 18 amino acid residues [15]. The cellfree synthesis reaction with the dialysis method was carried out according to our previous study [6] as follows. The internal solution (3 ml) consisted of 58 mM Hepes-KOH (pH 7.5), 1.8 mM DTT, 1.2 mM ATP, 0.8 mM each of CTP, GTP, and UTP, 80 mM creatine phosphate (Roche), 0.25 mg/ml creatine kinase (Roche), 4.0% PEG 8000 (Sigma), 0.64 mM
Figure 1. LC-MS spectrum of the selenomethionine-substituted Ras protein. Peaks A (m/z = 19, 681) and B (m/z = 19, 708) correspond to the deduced mass of the completely selenomethionine-substituted Ras protein, and its formylated form, respectively. Inset: SDS-PAGE analysis of the purified selenomethionine-substituted Ras protein.
3 , 5 -cyclic AMP, 68 µM L(−)-5-formyl-5,6,7,8tetrahydrofolic acid, 175 µg/ml E. coli total tRNA (Roche), 210 mM potassium glutamate, 27.5 mM ammonium acetate, 10.7 mM magnesium acetate, 1.0 mM each of the 20 amino acids, 0.5 U/µl RNase inhibitor (TOYOBO), 0.05% sodium azide, 6.7 µg/ml pK7-Ras plasmid [4], 0.27 mg/ml T7 RNA polymerase, and 0.9 ml S30 extract. The external solution (30 ml) contained the components of the internal solution except for the creatine kinase, the plasmid vector, the T7 RNA polymerase, the S30 extract, and the RNase inhibitor, and also contained an additional 4.2 mM magnesium acetate, corresponding to the magnesium carryover from the S30 extract. For the selenomethionine incorporation, the methionine in the internal and external solutions was replaced by selenomethionine (Sigma). The reaction unit was incubated at 37 ◦ C with shaking at 160 rpm for 4 hours. The Ras protein was purified from the internal solution using successive chromatographies on SOURCE Q 15 (Amersham Pharmacia Biotech) and Superdex 75 (Amersham Pharmacia Biotech) columns on an ÄKTAexplorer 10S (Amersham Pharmacia Biotech). In order to estimate the level of se-
31 was of sufficient quality to trace the backbone and to assign the side chains using the selenium sites as landmarks. An atomic model was fitted into the electron density map using the program O [23], and then the crystallographic positional and simulated annealing refinements were carried out with CNS [24].
Results and discussion Selenomethionine incorporation by cell-free synthesis
Figure 2. Crystal of the selenomethionine-substituted Ras protein.
lenomethionine incorporation, the selenomethioninecontaining Ras protein was subjected to liquid chromatography-electrospray mass spectrometry (LCMS). X-ray crystallography For comparison with the Ras protein prepared by the cell-free synthesis, the protein was prepared by the in vivo system as described [16]. Crystals of Ras proteins were grown by the hanging-drop method at 4 ◦ C using a precipitant solution containing 200 mM calcium acetate, 100 mM sodium cacodylate (pH 6.5), and 18% PEG 8000 (Crystal Screen Kit solution #46, Hampton Research). To identify the bound guanine nucleotide, a dissolved crystal of Ras protein was subjected to matrix assisted laser desorption ionizationtime of flight mass spectrometry (MALDI-TOF MS), and then was analyzed according to the method of Akashi and co-workers [17]. Data of crystals of Ras proteins were collected at 20 ◦ C on an Raxis-IV image plate detector mounted on a rotating anode X-ray source (Rigaku) . MAD data of the selenomethionine substituted Ras protein were measured on a MAR CCD detector at the BL44B2 beam line at SPring-8. All of the data were processed with the DENZO and SCALEPACK programs [18]. Selenium atom sites were identified from the anomalous difference Patterson map by the RSPS program [19]. Selenium parameters were refined and phases were calculated using SHARP [20]. Sites were confirmed by the direct method with the program Shakeand-Bake [21]. Phases were improved by density modification using SOLOMON [22]. The resulting map
First, we examined the inhibitory effect of selenomethionine on the cell-free protein synthesis. The productivity of the cell-free synthesis with selenomethionine was almost the same as that without selenomethionine, indicating that the selenomethionine hardly inhibited the protein synthesis in the cell-free system (data not shown). Then, the selenomethionine substituted Ras protein was produced on a large scale, using the cell-free synthesis system with dialysis (3 ml internal/30 ml external, see Materials and methods). About 2 mg of the selenomethionine-substituted Ras protein was purified to homogeneity (Figure 1). To determine the level of selenomethionine incorporation, the purified Ras protein was subjected to LC-MS and two peaks (m/z = 19, 681 and 19,708) were observed (Figure 1). Peak A (m/z = 19, 681) corresponds to the deduced mass, 19,682, of the Ras protein with its four methionine residues completely replaced by selenomethionine, and Peak B (m/z = 19, 708) corresponds to the formylated form (the deduced mass of 19,710) within the experimental error. On the other hand, we could not observe any peak that corresponds to the deduced mass of incompletely selenomethionine-substituted protein or its formylated form. Considering the signal-to-noise ratio of LC-MS analysis, we concluded that the selenomethionine incorporation efficiency is higher than 95%. If an in vivo expression method is used for selenomethionine incorporation into a protein, it often results in low incorporation and low productivity. This low incorporation efficiency causes the serious problem of protein sample heterogeneity, which is not good for crystallization. Our result indicates that this problem of in vivo expression must be solved, and then, selenomethionine substituted proteins suitable for MAD phasing could be obtained by using the cell-free system.
32 Table 1. Crystallographic, phasing, and refinement data Crystal statistics Native (in vivo)1 a= b = 82.7 c = 103.4 219,513 14,692 14.94 99.7 (99.4)3 8.9 9.1 (36.8)3
Unit cell dimensions (Å) Measurements Unique reflections Multiplicity Completeness (%)
/<σ> Rsym 4 (%)
Native (Cell-free)1 a = b = 82.7 c = 103.2 232,493 14,678 15.84 99.8 (100)3 9.0 7.0 (32.7)3
Selenomethionine-substituted2 a = b = 82.7 c = 102.7 271,633 14,558 18.66 99.5 (98.6)3 16.6 5.7 (16.0)3
MAD data collection2 λ (Å) Resolution range (Å) Measurements Unique reflections Completeness (%) >/<σ> Rsym 4 (%)
edge 0.97925 20-2.0 274,641 14,539 99.6 (99.8)3 19.6 5.6 (15.1)
peak 0.97882 20-2.0 219,328 14,566 99.5 (98.5)3 19.6 5.7 (14.8)
remote 1 0.97115 20-2.0 276,044 14,547 99.6(100)3 19 6.1 (16.7)
remote 2 0.98660 20-2.0 271,022 14,539 99.2 (84.7)3 19.3 5.3 (14.2)
Phasing statistics (20–2.0 Å) Mean figure of merit Rcullis 5 (dispersive) R5cullis (anomalous) Phasing power6
– 0.7 –
0.72 0.63 0.62 1.92
0.56 0.58 2.24
0.67 0.95 1.77
Refinement statistics Rwork 7 (Rfree 8 )(%) Rmsd bond lengths (Å) Rmsd bond angles (◦ ) Rmsd impropers (◦ )
23.8 (28.3) 0.006 1.110 0.701
1 These data were collected on an Raxis-IV (Rigaku). 2 These data were collected at the BL44B2 beam line at SPring-8. 3 Numbers in parentheses correspond to the values in the highest resolution shell. 4R sym = h j | Ihj − < Ih > |/h j Ihj , where Ihj is the intensity of observation j of reflection h. 5R cullis = [|FH | − (|FPH | − |FP |)]/(|FPH | − |FP |) (only for centric reflections), where |FH | represents the calculated
heavy atom structure factor. 6 Phasing power = [ |F |2 / (|F 2 1/2 where n is the number of derivative reflections and |FH| is n H n PH(obs) | − |FPH(calc) |) ]
the calculated intensity of the heavy atom structure. 7R work = h ||FO |−|FC ||/|FO | for all reflections above a I/σ cutoff of 2.0, where FO and FC are observed and calculated
structure factors, respectively. 8R free was calculated against 10% of the complete data set excluded from refinement.
Crystals of Ras proteins prepared by the cell-free system We made crystals of three kinds of Ras proteins: the native protein prepared in vivo in E. coli cells (crystal I), that prepared by the cell-free system (crystal II), and the selenomethionine-substituted protein prepared by the cell-free system (crystal III, Figure 2). All of these Ras proteins were crystallized under the same conditions (see Materials and methods). To compare
the characteristics of crystals I and II, diffraction data of these crystals were obtained with an Raxis-IV. The statistics of these crystals are summarized in Table 1. Both of them belong to space group P 65 22 and diffract to 2.0 Å. Their unit-cell dimensions are almost the same. The isomorphous difference between them is 6.7% on average, suggesting that these two crystals are almost identical. These results suggest that the Ras protein produced by the cell-free system forms almost
33
Figure 3. Stereodiagram of the electron density map around the Phe156 in helix 5 of the Ras protein. The map was calculated using the experimental MAD phase.
the same three-dimensional structure as that of the Ras protein produced by the in vivo expression method. To analyze the bound guanine nucleotide in crystal III, the dissolved crystal was subjected to MALDI-TOF MS [17]. The GDP molecule was found, and thus, the Ras protein in crystal III was revealed to be in the GDP-bound form (data not shown). MAD analysis of the selenomethionine substituted Ras protein The MAD data of crystal III were collected at BL44B2 beam line at SPring-8, and are summarized in Table 1. This crystal also belongs to space group P 65 22, and its unit-cell dimensions are almost the same as those of crystals I and II. A high resolution (2.0 Å) electron density map (Figure 3) was obtained from our data by MAD phasing, and then, the three-dimensional structure of the Ras protein was determined with only the obtained data. The coordinates (PDB accession code: 1IOZ) of the protein were obtained (Figure 4), except for residues 61-67 in loop 4, and residues 170 and 171 at the C-terminus, which have poor electron density. We also obtained the coordinates for the bound GDP molecule. Comparison with other Ras structures The three-dimensional structure of the GDP-bound Ras protein was already determined by X-ray crystallography [12] (PDB accession code: 1Q21) and nuclear magnetic resonance (NMR) spectroscopy [25, 26] (PDB accession code: 1CRQ and 1AA9, respectively). In the X-ray crystallography study, residues 60–69 in loop 4 are disordered [12], and in the NMR spectroscopy studies, the residues in loop 4 exhibit a significantly rapid internal motion on the subnanosecond time scale [25, 26], indicating that loop
Figure 4. Ribbon representation of the GDP-bound Ras protein. The bound GDP is colored cyan. (a) The structure determined by the MAD phasing in this study. (b) The structure reported by de Vos et al. [12] (PDB code; 1Q21). Figures were made by the program MOLMOL [32].
4 of the GDP-bound Ras protein is highly mobile in nature. Thus, it is reasonable that residues 61–67 in loop 4 are disordered in our structure. The rms deviations between the structure previously determined by X-ray crystallography (PDB accession code: 1Q21) and our structure are 0.32 Å for the backbone atoms and 0.87 Å for all non-hydrogen atoms, respectively, excluding residues 59–71 in loop 4 and residues 170 and 171 at the C-terminus. Thus, our structure of the protein prepared by the cell-free system is almost identical to the structure of the protein prepared by the in vivo expression method. This confirms that proteins produced by the cell-free system would be useful for structure determinations by X-ray crystallography. Future aspect of selenomethionine incorporation by cell-free synthesis Cell-free selenomethionine incorporation demonstrated in the present study would be applicable
34 to varieties of proteins. With respect to the protein molecular mass limit of the cell-free synthesis, it should be noted that we succeeded in producing a 110 kDa mouse protein [11], and other groups also reported that proteins larger than 100 kDa could be produced using the cell-free system [8] (http://biochem.roche.com/rts/results_main.htm). The modification of the reaction conditions still seems to raise the upper limit of the protein size. As well as glutathione S-transferase-fused proteins [11], histidine-tagged proteins could also be produced (unpublished results). Proteins of low solubility could be prepared in soluble form by adding chaperon molecules in the system [27]. Functional single-chain Fv fragment could be produced by the cell-free system in an oxidized condition, indicating that proper disulfide bridge formation and rearrangement will take place in the system [28]. The membrane integrated protein could be produced and then inserted into membrane by the cell-free synthesis with microsomal membrane fraction [29]. Signal sequence of the precursor protein could be properly processed in the system [30]. It is also possible to prepare the cytotoxic protein in an active form [31]. Therefore, it is expected that a wide range of proteins can be substituted with selenomethionine by cellfree protein synthesis. The cell-free system is also useful for the stable-isotope labeling of proteins for NMR spectroscopy [6]. Thus, the cell-free system will become a powerful protein expression method especially for high-throughput protein structure determinations, and will strongly advance structural genomics.
2. 3.
4. 5.
6. 7. 8. 9. 10. 11.
12.
13.
14. 15.
16.
17. 18.
Acknowledgements We thank Drs Koji Takio, Satoko Akashi, Naoshi Dohmae, and Hiroshi Nakayama (Biomolecular Characterization Division, Characterization Center, RIKEN) for measuring LS-MS and MALDI-TOF MS. This work was supported in part by a grant from RIKEN, Japan, a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology, Japan (12780502), and a grant for the ‘Research for the Future’ Program from the Japan Society for the Promotion of Science, Japan (JSPS-RFTF96I00306). References 1. Hendrickson, W.A. (1991) Science 254, 51–58.
19. 20. 21. 22. 23. 24.
25. 26.
27.
Hendrickson, W.A., Horton, J.R. and LeMaster, D.M. (1990) EMBO J. 9, 1665–1672. O’Gara, M., Adams, G.M., Gong, W., Kobayashi, R., Blumenthal, R.M. and Cheng, X. (1997) Eur. J. Biochem. 247, 1009–1018. Kigawa, T., Muto, Y. and Yokoyama, S. (1995) J. Biomol. NMR 6, 129–134. Yabuki, T., Kigawa, T., Dohmae, N., Takio, K., Terada, T., Ito, Y., Laue, E.D., Cooper, J.A., Kainosho, M. and Yokoyama, S. (1998) J. Biomol. NMR 11, 295–306. Kigawa, T., Yabuki, T., Yoshida, Y., Tsutsui, M., Ito, Y., Shibata, T. and Yokoyama, S. (1999) FEBS Lett. 442, 15–19. Kim, D.-M., Kigawa, T., Choi, C.-Y. and Yokoyama, S. (1996) Eur. J. Biochem. 239, 881–886. Madin, K., Sawasaki, T., Ogasawara, T. and Endo, Y. (2000) Proc. Natl. Acad. Sci. U.S.A. 97, 559–564. Davis, J., Thompson, D. and Beckler, G.S. (1996) Promega Notes Magazine 56, 14–18. Kim, D.M. and Choi, C.Y. (1996) Biotechnol. Prog. 12, 645– 649. Yokoyama, S., Hirota, H., Kigawa, T., Yabuki, T., Shirouzu, M., Terada, T., Ito, Y., Matsuo, Y., Kuroda, Y., Nishimura, Y., Kyogoku, Y., Miki, K., Masui, R. and Kuramitsu, S. (2000) Nat. Struct. Biol. 7 Suppl., 943–945. de Vos, A.M., Tong, L., Milburn, M.V., Matias, P.M., Jancarik, J., Noguchi, S., Nishimura, S., Miura, K., Ohtsuka, E. and Kim, S.H. (1988) Science 239, 888–893. Pratt, J.M. (1984) In Transcription and translation (Eds., Hames, B.D. and Higgins, S.J.), IRL Press, Oxford, Washington DC, pp. 179–209. Zawadzki, V. and Gross, H.J. (1991) Nucleic Acids Res. 19, 1948. Ha, J.-M., Ito, Y., Kawai, G., Miyazawa, T., Miura, K., Ohtsuka, E., Noguchi, S., Nishimura, S. and Yokoyama, S. (1989) Biochemistry 28, 8411–8416. Yamasaki, K., Kawai, G., Ito, Y., Muto, Y., Fujita, J., Miyazawa, T., Nishimura, S. and Yokoyama, S. (1989) Biochem. Biophys. Res. Commun. 162, 1054–1062. Akashi, S., Shirouzu, M., Yokoyama, S. and Takio, K. (1996) J. Mass Spectrom. Soc. Jpn. 44, 269–277. Otwinowski, Z. and Minor, W. (1997) Macromol. Crystallogr. A276, 307–326. Knight, S.D. (2000) Acta Crystallogr. D56, 42–47. de la Fortelle, E. and Bricogne, G. (1997) Methods Enzymol. 276, 472–494. Hauptman, H.A., Xu, H., Weeks, C.M. and Miller, R. (1999) Acta Crystallogr. 55, 891–900. Collaborative Computational Project Number 4 (1994) Acta Crystallogr. D50, 760–763. Jones, T.A., Zou, J.Y., Cowan, S.W. and Kjeldgaard, M. (1991) Acta Crystallogr. A47, 110–119. Brünger, A.T., Adams, P.D., Clore, G.M., DeLano, W.L., Gros, P., Grosse-Kunstleve, R.W., Jiang, J.S., Kuszewski, J., Nilges, M., Pannu, N.S., Read, R.J., Rice, L.M., Simonson, T. and Warren, G.L. (1998) Acta Crystallogr. D54, 905–921. Kraulis, P.J., Domaille, P.J., Campbell-Burk, S.L., Van Aken, T. and Laue, E.D. (1994) Biochemistry, 33, 3515–3531. Ito, Y., Yamasaki, K., Iwahara, J., Terada, T., Kamiya, A., Shirouzu, M., Muto, Y., Kawai, G., Yokoyama, S., Laue, E.D., Wälchli, M., Shibata, T., Nishimura, S. and Miyazawa, T. (1997) Biochemistry 36, 9109–9119. Tsalkova, T., Zardeneta, G., Kudlicki, W., Kramer, G., Horowitz, P.M. and Hardesty, B. (1993) Biochemistry 32, 3377–3380.
35 28.
Ryabova, L.A., Desplancq, D., Spirin, A.S. and Pluckthun, A. (1997) Nat. Biotechnol. 15, 79–84. 29. Cayanis, E., Bayley, H. and Edelman, I.S. (1990) J. Biol. Chem. 265, 10829–10835. 30. Markoff, L. (1989) J. Virol. 63, 3345–3352.
31. 32.
Martemyanov, K.A., Shirokov, V.A., Kurnasov, O.V., Gudkov, A.T. and Spirin, A.S. (2001) Protein Expr. Purif. 21, 456–461. Koradi, R., Billeter, M. and Wüthrich, K. (1996) J. Mol. Graph. 14, 51–55.