Comparison of the Three-dimensional Structures of Human, Yeast ...

THEJOURNAL OF BIoLOClCAL CHEMISTRY 0 1987 by The American Society of Biological Chemists, Inc

Vol. 262, No. 13, Issue of May 5, pp. 63964399.1987 Printed in U.S.A.

Comparison of the Three-dimensional Structures of Human, Yeast, and Oat Ubiquitin" (Received for publication, January 6, 1987)

Senadhi Vijay-Kumar$Q,Charles E. Bugg$§lI, Keith D. Wilkinsonll, RichardD. Vierstra**, Peggy M. Hatfield**, and William J. Cook$§$$@ From the Departments of $$Pathology and qBiochemistry, $Center for Macromolecular Crystallography,and §Comprehensive Cancer Center, University of Alabama at Birmingham, Birmingham, Alabama 35294, the I(Department of Biochemistry, Emory Universitv Schoolof Medicine. Atlanta. Georgia 30322. and the **Department of Horticulture, University of Wisconsin-Madison, Madison,"Wisconsin42706

-

'

The crystal structure of human ubiquitin has been solved by x-ray diffraction methods and refined by standard procedures to a convgntional crystallographic R factor of 0.176 at 1.8-A resolution (VijayKumar, S., Bugg, C. E., and Cook, W.J. (1987) J. Mol. Biol. 194, 525-538). Crystals of yeast and oat ubiquitin have been grown using human ubiquitin crystals as seeds. Diffraction data for yeast and oat ubiquitin have been collected to a resolution of 1.9 and 1.8 A, respectively. DifferenceFourier electron-density maps reveal that the structures of yeast and oat ubiquitin are quite similar tohuman ubiquitin. All the amino acid changes are clustered in two small patches on one surface of the molecule. This surface is probably not involved in conjugation with proteins destined for ATP-dependent proteolysis.

Ubiquitin is a small protein that is probably present in all eukaryotic cells (1).It consists of a single 8565-dalton polypeptide chain of 76 amino acids and it appears to be one of the most conserved of all eukaryotic proteins. Ubiquitin has been isolated and sequenced from a variety of sources, and the primary structures from all animal sources examined thus far are identical (2). The only ubiquitin variants identified to date are from two plant sources, yeast (3) and oat (4). The amino acid sequences of yeast and oat ubiquitin differ by 3 residues from that of ubiquitin in animals and by only 2 residues from each other. One possible explanation for this difference is that the functions of ubiquitin may vary between plants and animals. Ubiquitin has been identified in the nucleus, in the cytoplasm, and on the cell-surface membrane, but it appears that its primary role is in intracellular ATP-dependent protein degradation (2). Protein breakdown in this pathway requires the formation of covalent conjugates in which the carboxyl terminus of ubiquitin becomes attached to thetarget protein via amide linkages to t-amino groups of their lysine residues (to yield branched ubiquitin-protein conjugates) and/or to the NH, terminus. Ubiquitin is also present in the nucleus, where * This work wassupported in part by National Institutes of Health Grants GM-30308 (to K. D. W.), GM-27144 (to W. J . C.), CA-13148 and DE-02670 (to C. E. B.), and United States Department of Agriculture Competitive Research Grants Office Grant 85-CRCR-11547 (to R. D. V.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Recipient of National Institutes of Health Career Development Award CA-00696.

it is conjugated to histone 2A (5), and it has been suggested that this conjugate may be involved in the transcription of active genes (6). Most recently ubiquitin has been identified on the cell surface as part of a lymphocyte homing receptor (7) and as part of the receptor for platelet-derived growth factor (8),but its function in these locations is still unclear.The structure of human erythrocytic ubiquitin at 2.8-A resolution hasobeen reported (9), and the model has been refined to 1.8 A (10). We now report the crystal data for yeast and oat ubiquitin and describe the changes in conformation from human ubiquitin. The relationship of the sites of amino acid substitution to theother portions of the molecule is also described. EXPERIMENTALPROCEDURES

CrystallizationProcedures-Yeast and oatubiquitins were purified as described (3, 11).Efforts to crystallize yeast and oat ubiquitin under the same de novo conditions as human (12) were unsuccessful. Therefore, crystals of human ubiquitin were used for seeding. For seeding experiments, the hanging drop vapor-diffusion method was used. Lyophilized yeast or oatubiquitin was dissolved at a concentration of 20 mg/ml in distilled water. The drop was formed by mixing 3 rl of the ubiquitin solution with 3 r l of a solution of 40% polyethylene glycol 4000 (PEG-4000)' (w/v) in 0.05 M cacodylate buffer, pH 5.6. The coverslip was then inverted andset over the reservoir containing 1.0 ml of 24% PEG-4000 (w/v) in 0.05 M cacodylate buffer, pH 5.6. The drops were placed in a 4 "C cold room, and seeding was carried out after 2-4 h. A single crystal of human ubiquitin (0.05 X 0.1 X 0.2 mm) was washed in a stabilizing solution of 20% (w/v) PEG-4000 in 0.05 M cacodylate buffer, pH 5.6, and then transferred with as little liquid as possible to the drop. Introduction of the seed crystal into the drop was accomplished by using a 0.5-mm capillary and a micromanipulator. The coverslip was then reinverted over the reservoir, and the traywas kept in the cold room. After 2-4 days at 4 "C, large crystals were obtained (Fig. 1). Two applications of the seeding technique were required to obtain good crystals. In the first attempt, seed crystals of human ubiquitin were used and these produced large imperfect crystals within which the seed was still clearly visible (Fig. la). However, small (0.1 X 0.1 X 0.2 mm) perfect crystals of the yeast or oat ubiquitin were also formed in the same drop. For the second seeding attempt, these new crystals were used as seeds, and large perfect rectangular prisms were obtained (Fig. lb). The crystals seem to be stable indefinitely at 4 "C, and they can be kept for weeks a t room temperature in a stabilizing solution of 32% (w/v) PEG-4000 in 0.05 M cacodylate buffer, pH 5.6. Data Collection-Intensity data were collected with a Picker FACS1 diffractometer at room temperature using an omega step-scan procedure (scan width, 0.6-0.8") and nickel-filtered CuKa radiation. The reflections were divided into 20 shells containing 100 reflections each and were collected from high to low resolution. Diffraction data for yeast and oat ubiquitin were collected to a resolution of 1.9 and 1.8 A, respectively. Two crystals of each protein were used for data collection. Data for one crystal were collected from 1.8 (or 1.9 A) to

6396

The abbreviation used is: PEG-4000, polyethylene glycol 4000.

Comparison of Human, Yeast, I

j I

, !

FIG. 1. Crystals of yeast ubiquitin obtained by seeding. 0, yeast ubiquitincrystalgrownfromhumanuhiquitin seed; b, yeast ubiquitin crystal grown from yeast ubiquitin seed. 2.8 A, and data for the second were collected from 2.5 A to infinity. To monitor and correct for decomposition effects, six standard reflections with a wide spread across reciprocal space were measured periodically. The overlapping data in the resolution range 2.5-2.8 A were used to scale the two data sets together. The merging R index calculated from the overlapping reflections in these two data sets was 0.07 for yeast ubiquitin and 0.06 for oat ubiquitin. The empirical absorption correction due to North et al. (13) was used to correct the anisotropy in x-ray transmission as a function of 4. Lorentz and polarization corrections were applied, For yeast ubiquitin, there was a total of 5600 reflections for 1.9 A data; of these, 4304 had intensities exceeding 2.50. For oat ubiquitin, 5493 of the 6200 reflections collected to 1.8-A resolution had intensities exceeding 2.50. The diffraction amplitudes were scaled together and compared by R = S(IF(human)- F(plant)ll/Z F(human), where F values are the moduli of structure factors. The value of 5 increases smoothly from 9% at a resolution of 6.0 A to 62% at 1.9 A for yeast ubiquitin and from 8% at6.0 A to 30% at 1.8 A for oat ubiquitin.

and Oat Ubiquitin

6397

least squares to an R index of 0.176 at a resolution of 1.8 A. The details of the structure solution and refinement have been described (9,lO). Changes in the yeast and ubiquitiv oat structures were revealed by difference Fourier maps a t 3.0-A resolution. The difference maps were calculated using modified phases of human ubiquitin and theobserved amplitudes of yeast or oat ubiquitin minus the modified calculated amplitudes of human ubiquitin. In each case, the phase angles were calculated after replacing the three substituted amino acids with glycine residues. Fig. 2 shows the quality of the difference electron density maps a t residues 18-21 for yeast and oat ubiquitin. Difference Fourier peaks at the sites of substitution were approximately 4 times the root mean square value of the difference electron density. Most of the remaining peaks of this size corresponded to changes in the positions of water molecules in the inner shell of hydration. Other than the slight amount of solvent rearrangement, there were several peaks that corresponded to small movements of some of the hydrophobic side chainsin the core of the molecule. However, except for the three amino acid substitutions in each case, the surface of the molecule was essentially unchanged. The most prominent secondary structural features of ubiquitin are shown in Fig. 3. A full description of the secondary structure of ubiquitin hasbeen given (lo), but several features should be emphasized here. There is a mixed @-sheet that contains five strands. Two of the inner strands,composed of residues 1-7 and 64-72, are parallel. The other three strands, which are composed of residues 10-17,40-45, and 48-50, run in an antiparallel direction. There is an a-helix that includes residues 23-34; this helix fits into the concavity formed by the sheet. There is also a short piece of 310-helixthat includes residues 56-59. All of the amino acid changes identified in yeast and oat ubiquitin are clusteredtwo in small patches on one surface of the molecule (Fig. 3). In yeast ubiquitintwo of the substitutions,Glu-to-Asp at position 24 and Ala-to-Ser a t position 28, are one turn apart in the a-helix. Both occur on the surface of the molecule, and one would not expect these changes to have large effects on the structure. Similarly, in oat ubiquitin,two of the substitutions,Glu-to-Asp. a t position . 24 and Ser-to-Ala at position 57, occur in helical segments and are located on the surface of the molecule. The third substitution in each plant ubiquitin occurs at position 19, where proline is replaced by serine. In human ubiquitin proline 19 is the 2nd residue ina reverse turn thatoccurs between an outer strand of the @-sheet and thebeginning of the ahelix. One might have suspected that this substitutionwould cause changes in the relative orientation of the helix to the @-sheet, butthereappearsto be no significant effect. In summary, the secondary structure of ubiquitin is conserved in each case in spite of the three aminoacid substitutions. DISCUSSION

It is noteworthy that themajor hydrogen bonding interactions within human ubiquitin are maintained in the yeast and oat structures. One portion of the molecule displays an unusual region that involves a symmetric arrangement of the RESULTS two helices and two Type I reverse turns. In humanubiquitin the main chainN atoms of the first 2 residues of the a-helix The crystalsof all three ubiquitins are orthorhombic, space group P212121,and the unitcell dimensions for yeast and oat (isoleucine 23 and glutamate 24) form hydrogen bonds to the ubiquitin are almost identical to those of human ubiquitin carbonyl oxygen atoms of the 2nd and 4th residues in the (10). Themajor difference is in the length of the c axis, which turn involving residues 51-54. Similarly, the main chain N is slightly shorter in the plant ubiquitins (0.3% for yeast and atoms of the first 2 residues of the 3lo-helix (leucine 56 and serine 57) form hydrogenbonds to the carbonyl oxygen atoms 0.7% for oat). The structure of human ubiquitin has been solved in this of the 2nd and 4th residues in the turninvolving residues 18laboratory with x-ray diffraction techniques and refined by 21. There is almost 2-fold symmetry in this portion of the

6398

Comparison of Human, Yeast,

Oatand

Ubiquitin

FIG. 2. Stereo drawing of residues18-21 with the superimposed electron density contour surfaces. Difference maps were calculated by using the modified refined phases of human ubiquitin and the amplitudes of yeast or oat ubiquitin minus the modified calculated amplitudes of human ubiquitin. a, yeast ubiquitin; b, oat ubiquitin.

Comparison of Human, and Yeast,

Oat Ubiquitin

6399

FIG. 3. Stereo drawing of the a-carbon backbone of human ubiquitin. The side chains of the residues involved in amino acid substitutions (at positions 19, 24, 28, and 57) are included.

structure. It is interesting that, even though two of the three amino acid changes in yeast ubiquitin and all three of the amino acid changes in oat ubiquitinoccur in this region, the hydrogen bonding pattern and the secondary structural features are preserved. These interactions may make important contributions to the unusual stability of the molecule. Most studies suggest that ubiquitin functions primarily in intracellularATP-dependentprotein degradation. Protein breakdown in this pathway requires the formation of covalent conjugates in which carboxyl termini of ubiquitin molecules become attached to the target protein via amide linkages. Comparison of human and yeast ubiquitin demonstrated no difference in the abilityof these two proteins to stimulate proteindegradation (3). Kineticanalysis of oatubiquitin demonstrated only 50% of the activity of human ubiquitin (11),but this difference may reflect COOH-terminal heterogeneity, secondary to partial proteolysis or modification by plant polyphenols during extraction.’ The extreme evolutionary stability of ubiquitin suggests that the structural constraints necessary for catalytic activity orrecognition of ubiquitin by other proteins are quite strict. Since only thesequence changes identified thus far are confined to two small regions on one surface of the molecule, it seems unlikely that these regions are directly involved in the function of ubiquitin in ATP-dependent proteolysis.

* R. D. Vierstra, unpublished data.

REFERENCES 1. Goldstein, G., Scheid, M., Hammerling, U., Boyse, E. A., Schlesinger, D. H. & Niall, H. D. (1975) Proc. Nutl. Acud. Sci. U. S. A. 72, 11-15 2. Finley, D. & Varshavsky, A. (1985) TrendsBiochem. Sci. 1 0 , 343-347 3. Wilkinson, K. D., Cox, M. J., O’Connor, L. B. & Shapira, R. (1986) Biochemistry 25,4999-5004 4. Vierstra, R. D., Langan, S . M. & Schaller, G. E. (1986) Biochemistry 25,3105-3108 5 . Busch, H. & Goldknopf, I. L. (1981) Mol. Cell. Bwchem. 40,173187 6 . Levinger, L. & Varshavsky, A. (1982) Cell 28,375-385 7. Siepelman. M., Bond. M.W.. Gallatin. W. M.. St.John. T.. Smith, H. T:, Fried, V.A. & Weissman, I. L.’(1986) Science 23 1,823-829 8. Yarden, Y., Escobedo, J. A., Kuang, W-J., Yang-Feng, T. L., Daniel, T. O., Tremble, P. M., Chen, E. Y., Ando,M.E., Harkins, R. N., Francke, U., Fried, V. A., Ullrich, A. & Williams, L. T. (1986) Nature 323,226-232 9. Vijay-Kumar, S., Bugg, C. E., Wilkinson, K. D. & Cook, W. J. (1985) Proc. Nutl. Acud. Sci. U. S. A. 82, 3582-3585 10. Vijay-Kumar, S., Bugg, C. E. & Cook, W. J. (1987) J. Mol. Biol. 194,525-538 11. Vierstra, R. D., Langan, S. M. & Haas, A. L. (1985)J.Biol. Chem. 260, 12015-12021 12. Cook, W. J., Suddath, F. L., Bugg, C. E. & Goldstein, G. (1979) J. Mol. Bi01. 130, 353-355 13. North, A.C. T., Phillips, D. C. & Mathews, F. S. (1968) Acta Crystullogr. A24, 351-359