CRYSTALLIZATION NOTE Crystallization of Restriction Endonuclease BamHI with Nonspecific DNA

Size: px
Start display at page:

Download "CRYSTALLIZATION NOTE Crystallization of Restriction Endonuclease BamHI with Nonspecific DNA"

Transcription

1 Journal of Structural Biology 130, (2000) doi: /jsbi , available online at on CRYSTALLIZATION NOTE Crystallization of Restriction Endonuclease BamHI with Nonspecific DNA Hector Viadiu,* Rebecca Kucera, Ira Schildkraut, and Aneel K. Aggarwal *Department of Biochemistry and Molecular Biophysics, Columbia University, New York, New York 10032; New England Biolabs, 32 Tozer Road, Beverly, Massachusetts 01915; and Structural Biology Program, Department of Physiology and Biophysics, Box 1677, Mount Sinai School of Medicine, 1425 Madison Avenue, New York, New York Received November 22, 1999, and in revised form February 4, 2000 Restriction endonucleases show extraordinary specificity in distinguishing specific from nonspecific DNA sequences. A single basepair change within the recognition sequence results in over a millionfold loss in activity. To understand the basis of this sequence discrimination, it is just as important to study the nonspecific complex as the specific complex. We describe here the crystallization of restriction endonuclease BamHI with several nonspecific oligonucleotides. The 11-mer, 5 -ATGAATCCATA-3, yielded cocrystals with BamHI, in the presence of low salt, that diffracted to 1.9 Å with synchrotron radiation. The cocrystals belong to the space group P with unit cell dimensions of a Å, b 91.1 Å, c 66.4 Å, 90, 90, 90. This success in the cocrystallization of BamHI with a nonspecific DNA provides insights for future attempts at crystallization of other nonspecific DNA protein complexes Academic Press Key Words: restriction endonuclease; BamHI; nonspecific complex; protein DNA complex; crystallization. Protein DNA selectivity is an essential event in many biological processes, ranging from transcription and replication to restriction and modification. The central problem faced by the DNA binding proteins in controlling these processes is how to select the correct DNA sequence from the nonspecific sequences in a cell. Most of the structural studies to date, however, have concentrated on proteins bound to their specific DNA sequences (Phillips and Moras, 1999). This has increased our understanding of how specific sequences are recognized, but has left a gap in our knowledge of how specific and nonspecific sequences are distinguished in the cell. Among DNA binding proteins, restriction endonucleases are perhaps the most extreme in their DNA selectivity. Type II restriction endonucleases are among the best characterized DNA binding proteins, recognizing DNA sequences that vary between four and eight basepairs and requiring Mg 2 as a cofactor to catalyze the hydrolysis of DNA (Roberts and Halford, 1993). Their DNA selectivity is remarkable a single base pair change within the recognition sequence can result in over a million-fold decrease in activity (Lesser et al., 1990; Engler, 1998). To understand the basis of this extreme selectivity, it is important to see both the nonspecific and the specific states of the protein. Thus, to complement our earlier work on the endonuclease BamHI bound to a specific DNA site (Newman et al., 1995; Viadiu and Aggarwal, 1998), we undertook the cocrystallization of BamHI with a nonspecific DNA sequence. Overall, the crystallization of nonspecific DNA protein complexes presents unique obstacles, and we describe here the strategy that led to the nonspecific cocrystals diffracting to 1.9 Å resolution. A comparison between BamHI nonspecific and specific structures promises to reveal the basis of its extreme DNA selectivity. Also, our results will aid future attempts to crystallize other nonspecific DNA protein complexes. The overexpression and purification protocols for BamHI have been described (Jack et al., 1991). The oligonucleotides were synthesized on a solid support using phosphoramidite chemistry, without removing the trityl group in order to aid purification by reverse-phase HPLC (Aggarwal, 1990). All of the crystallization experiments were performed by the hanging drop method at 20 C. The protein DNA solution contained one part BamHI ( mg ml 1 in 0.5 M KCl, 20 mm potassium phosphate (ph 6.9), 1 mm dithiothreitol, 10% glycerol) to one part annealed DNA (10 mg ml 1 in water). To set up the hanging drops, 1 µl of the /00 $35.00 Copyright 2000 by Academic Press All rights of reproduction in any form reserved.

2 82 CRYSTALLIZATION NOTE TABLE I DNA Sequences Used in the Cocrystallization of Non-specific BamHI-DNA Complex Oligonucleotide Sequence Oligonucleotide Sequence Specific 5 -TATGGATCCATA-3 BA16-BA35 5 -TATCGGATTCCATA mer (1.8 Å) 3 -ATACCTAGGTAT-5 14-mer 3 - TAGCCTAAGGTATA-3 BA03 5 -TATGAATTCATA-3 BA36 BA37 5 -TATGGTTCCATA-3 12-mer 3 -ATACTTAAGTAT-5 12-mer ( Å) 3 -ATACCAAGGTAT-5 BA04 5 -TATCATATGATA-3 BA38 5 -TATGGTACCATA-3 12-mer 3 -ATAGTATACTAT-5 12-mer ( Å) 3 -ATACCATGGTAT-5 BA07 5 -TATAGAATTCTATA-3 BA39 BA40 5 -TATCGATCCATA-3 14-mer 3 -ATATCTTAAGATAT-5 12-mer 3 -ATAGCTAGGTAT-5 BA08 5 -TACATATGTA-3 BA41 5 -TATCGATCGATA-3 10-mer 3 -ATGTATACAT-5 12-mer ( Å) 3 -ATAGCTAGCTAT-5 BA09 5 -TTACATATGTA -3 BA42 5 -TATGGAUCCATA-3 11-mer 3 - ATGTATACATT-5 12-mer 3 -ATACCUAGGTAT-5 BA10 5 -TTATGAATTCATA-3 BA45 BA46 5 -TCGGAATCCGCA-3 13-mer 3 - ATACTTAAGTATT-5 12-mer 3 -AGCCTTAGGCGT-5 BA13 BA14 5 -TAGGAATCCGTA-3 BA47 BA48 5 -GCGGAATCCGCG-3 12-mer (4.0 Å) 3 -ATCCTTAGGCAT-5 12-mer ( Å) 3 -CGCCTTAGGCGC-5 BA15 BA16 5 -TATGGAATCCGATA-3 BA27 BA49 5 -ATGAATCCATA-3 14-mer 3 -ATACCTTAGGCTAT-5 11-mer (1.9 Å) 3 -TACTTAGGTAT-3 BA17 BA18 5 -TATGAGAATCCAGATA-3 BA50 BA33 5 -TATGAATCCATAT-3 16-mer 3 -ATACTCTTAGGTCTAT-5 13-mer 3 -ATACTTAGGTATA-5 BA19 BA20 5 -TATGATGAATCCTAGATA-3 BA51 BA52 5 -TTTTACATATGTA mer 3 -ATACTACTTAGGATCTAT-5 13-mer 3 - AATGTATACATAA-5 BA21 BA22 5 -TATGAGAATCCTAGATA-3 BA53 BA54 5 -TACATACGTAA-3 17-mer 3 -ATACTCTTAGGATCTAT-5 11-mer 3 -ATGTATGCATT-5 BA23 BA24 5 -TATGGAATCCAGATA-3 BA55 BA56 5 -TTTACATATGTA-3 15-mer 3 -ATACCTTAGGTCTAT-5 12-mer 3 -AAATGTATACAT-5 BA25 BA26 5 -TATGAATCCATA-3 BA57 BA58 5 -ATACATACGTAA-3 12-mer (5.0 Å) 3 -ATACTTAGGTAT-5 12-mer 3 -TATGTATGCATT-5 BA27 BA ATGAATCCATA-3 BA59 5 -TATGGCGCCATA-3 11-mer 3 -TTACTTAGGTA mer 3 -ATACCGCGGTAT-5 BA29 BA ATGAATCCATAA-3 BA64 BA65 5 -TACGAATCCGTA-3 12-mer 3 -TTTACTTAGGTA mer 3 -ATGCTTAGGCAT-5 BA25 BA TATGAATCCATA-3 BA66 BA67 5 -TAGGAATCCCTA-3 12-mer 3 -TATACTTAGGTA mer 3 -ATCCTTAGGGAT-5 BA32 BA33 5 -TTATGAATCCATA -3 BA82 BA83 5 -ATCATACGATA-3 13-mer 3 - ATACTTAGGTATA-5 11-mer 3 -TAGTATGCTAT-5 BA32 BA34 5 -TTATGAATCCATA -3 BA84 BA85 5 -ATGAATTCATA-3 13-mer 3 - TACTTAGGTATAA-5 11-mer 3 -TACTTAAGTAT-5 BA86 BA87 5 -ATAGATCCATA-3 BA96 BA97 5 -ATGGCTCCATA-3 11-mer 3 -TATCTAGGTAT-5 11-mer 3 -TACCGAGGTAT-5 BA88 BA89 5 -ATGGGTCCATA-3 BA98 BA99 5 -ATGTATACATA-3 11-mer (10.0 Å) 3 -TACCCAGGTAT-5 11-mer 3 -TACATATGTAT-5 BA90 BA91 5 -ATGTATCCATA-3 BA100 BA ATGCATGCATA-3 11-mer (2.8 Å) 3 -TACATAGGTAT-5 11-mer 3 -TACGTACGTAT-5 BA92 BA93 5 -ATGCATCCATA-3 BA102 BA ATCCTAGGATA-3 11-mer (3.5 Å) 3 -TACGTAGGTAT-5 11-mer 3 -TAGGATCCTAT-5 BA94 BA95 5 -ATGGTTCCATA-3 11-mer 3 -TACCAAGGTAT-5 BA68 BA69 5 -TATGGATCCATATATGAATCCATA-3 24-mer 3 -ATACCTAGGTATATACTTAGGTAT-5

3 CRYSTALLIZATION NOTE 83 TABLE I Continued Oligonucleotide Sequence Oligonucleotide Sequence BA70 BA71 22-mer BA72 BA73 21-mer BA74 BA75 20-mer BA76 BA77 23-mer BA78 BA79 20-mer BA80 BA81 19-mer 5 -TATGGATCCATATATGAATCCATA-3 3 -ATACCTAGGTATATACTTAGGTAT-5 5 -TATGGATCCATAGAATCCATA-3 3 -ATACCTAGGTATCTTAGGTAT-5 5 -TATGGATCCATGAATCCATA-3 3 -ATACCTAGGTACTTAGGTAT TATGGATCCATTATGAATCCATA-3 3 -ATACCTAGGTAATACTTAGGTAT-5 5 -TAGGATCCATATGAATCCTA-3 3 -ATCCTAGGTATACTTAGGAT TAGGATCCATAGAATCCTA-3 3 -ATCCTAGGTATCTTAGGAT- 5 protein DNA solution was mixed with 1 µl of the precipitant solution and placed above a sealed well containing 1 ml of the precipitant solution. Three days were sufficient to observe maximum crystal growth. To confirm the presence of DNA in the crystals, several crystals were washed three to five times, dissolved in loading buffer, and loaded into a 20% denaturing PAGE. A 10-µl aliquot from the last wash solution was used as a negative control. X-ray diffraction data were measured on (1) an R-axis IV imaging plate area detector mounted on a Rigaku RU 200 rotating anode (CuK ) X-ray generator (100 ma and 50 kv) and on (2) an ADSC charge coupled device detector at the Cornell High Energy Synchrotron Source (CHESS beamline A1, Å). Crystals were frozen under a stream of dry nitrogen gas at 110 K (Oxford Cryosystems). The crystals grown in polyethylene glycol (PEG) 8000 were transferred to a 20% glycerol solution prior to freezing while the crystals grown in 20% MPD did not require additional cryoprotection. The HKL suite was used to integrate the X-ray reflections (Otwinoski and Minor, 1997). Table I lists the various DNA oligomers used in the cocrystallization experiments. Oligomers BA03 and BA04 were the first to be attempted; both are a variation of the 12-mer used in the crystallization of the BamHI specific complex. BA03, containing the EcoRI site (5 -GAATTC-3 ), has 2 basepairs changed with respect to the specific site, while BA04 has all 6 central basepairs changed to create a completely random site (5 -CATATG-3 ). The ends were kept constant at 5 -TAT-3 because they were found to favor the crystallization of the specific complex. However, the only crystals we obtained were of the protein alone. An important variable in protein DNA crystallization is the length of the DNA. Thus, we synthesized an additional set of oligonucleotides in the range of 10 to 18 basepairs, most of which were blunt-ended, with the central 6 basepairs being half EcoRI and half BamHI (5 -GAATCC-3 ). By screening with this larger collection of nonspecific oligonucleotides, we obtained rod-like crystals (with oligonucleotides BA13-BA14, BA25-BA26, BA36-BA37, and BA38) using PEG as the precipitant in the presence of MnCl 2 or CaCl 2. However, none of these crystals diffracted beyond 7 8 Å. Although the addition of methanol, ethanol, and isopropanol resulted in better crystals, we did not observe an improvement in resolution. However, a breakthrough was made when we lowered the salt concentration. Because BamHI s ability to discriminate between specific and nonspecific sites decreases at lower salt concentration, we changed the protein buffer from 0.5 to 0.05 M KCl. At this lower salt concentration, the protein solution turned turbid, but protein aggregation could be prevented by mixing with DNA. Under these conditions, we identified small crystals in a crystallization mix that contained 30% MPD, 100 mm sodium acetate (ph 4.6), and 200 mm CaCl 2. These conditions were eventually refined to 16 20% MPD, 10 mm sodium acetate (ph 4.8), and 5 mm CaCl 2, using the different oligomers. The addition of low concentrations of sodium acetate and CaCl 2 was essential for crystallization, but, interestingly, at slightly higher concentrations it was detrimental for crystal growth. The best cocrystals, with dimensions of mm, were obtained with oligonucleotides BA27 BA49, which diffracted to 1.9-Å resolution at CHESS (Fig. 1). The cocrystals belong to space group P with unit cell dimensions of a Å, b 91.1 Å, c 66.4 Å, 90, 90, 90. The data measured at CHESS merge with an overall R sym of 5.7% (28.4% in the last resolution shell); they have an I/ (I) of 20.0 (3.8) and a completeness of 96.1% (85.2%). They are being

4 84 CRYSTALLIZATION NOTE FIG. 1. The cocrystals of BamHI with oligonucleotide BA27 BA49 had a tendency to grow in bunches and they diffracted to 1.9 Å at CHESS. The crystal dimensions are mm. used to determine the structure by molecular replacement methods using the specific complex as the search model. The BamHI nonspecific structure will be an important step toward understanding how the enzyme achieves its elevated specificity. The cocrystals with oligonucleotides BA90 BA91 and BA92 BA93 indicate that it will be possible to solve the three-dimensional structure of BamHI bound to different nonspecific oligonucleotides. Several lessons emerge from our work that may aid future nonspecific DNA protein crystallization projects. For instance, a reduction in salt concentration was the most dramatic factor in our ability to crystallize the nonspecific complex. In general, nonspecific DNA binding is more salt-dependent than specific binding, with lower salt concentrations enhancing nonspecific binding. A common procedure in the crystallization of specific DNA protein complexes is to vary the length and sequence of the DNA (Aggarwal, 1990). A similar strategy proved essential in the optimization of the nonspecific crystals, with the best cocrystals obtained with a blunt-ended 11-mer. For BamHI, DNA binding is stimulated 240-fold at low Ca 2 concentrations, but, at high concentrations, it has the opposite effect (Engler et al., 1997). This may explain the sensitivity of our cocrystallization experiments on Ca 2 concentration, with the best cocrystals obtained in the presence of 5 10 mm CaCl 2 but deteriorating at higher Ca 2 concentrations. Taken together, the use of a low-salt concentration, a library of different length DNA oligomers, and the addition of divalent cations are features that could be incorporated into the crystallization of other nonspecific complexes. This work was supported by Grant GM44006 from the National Institutes of Health (A.K.A.). H.V. was supported by a Fullbright/ CONACYT Scholarship. REFERENCES Aggarwal, A. K. (1990) Crystallization of DNA binding proteins with oligodeoxynucleotides, Methods 1, Engler, L. E. (1998) Specificity Determinants in the BamHI Endonuclease DNA Interaction, Ph.D. thesis, p University of Pittsburgh, Pittsburgh. Engler, L. E., Welch, K. K., and Jen-Jacobson, L. (1997) Specific binding by EcoRV endonuclease to its DNA recognition site GATATC, J. Mol. Biol. 269, Jack, W. E., Greenough, L., Dorner, L. F., Xu, S. Y., Strzelecka, T., Aggarwal, A. K., and Schildkraut, I. (1991) Overexpression, purification and crystallization of BamHI endonuclease, Nucleic Acids Res. 19, Lesser, D. R., Kurpiewski, M. R., and Jen-Jacobson, L. (1990) The

5 CRYSTALLIZATION NOTE 85 energetic basis of specificity in the EcoRI endonuclease DNA interaction, Science 250, Newman, M., Strzelecka, T., Dorner, L. F., Schildkraut, I., and Aggarwal, A. K. (1995) Structure of BamHI endonuclease bound to DNA: Partial folding and unfolding on DNA binding, Science 269, Otwinoski, Z., and Minor, W. (1997) Processing of X-ray diffraction data collected in oscillation mode, Methods Enzymol. 276, Phillips, S. E. V., and Moras, D. (1999) Protein nucleic acids interactions, Curr. Opin. Struct. Biol. 9, Roberts, R. J., and Halford, S. E. (1993) Type II restriction endonucleases, in Linn, S. M., Lloyd, R. S., and Roberts, R. J. (Eds.), Nucleases, pp , Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Viadiu, H., and Aggarwal, A. K. (1998) The role of metals in catalysis by the restriction endonuclease BamHI, Nat. Struct. Biol. 5,