- Email: [email protected]
Surface Features of Potato Virus X from Fiber Diffraction
Virology 300, 291–295 (2002) doi:10.1006/viro.2002.1483
Surface Features of Potato Virus X from Fiber Diffraction Lauren Parker,* Amy Kendall,† and Gerald Stubbs† ,1 *Department of Chemistry and †Department of Biological Sciences, Vanderbilt University, Nashville, Tennessee 37235 Received February 26, 2002; accepted March 19, 2002 Fiber diffraction patterns have been obtained from oriented sols of potato virus X. Orientation in the sols was greatly improved by a combination of centrifugation and exposure to very high magnetic fields. Diffraction patterns were also improved by using a very finely collimated synchrotron X-ray beam. The diffraction patterns show that there are 8.9 subunits in each turn of the viral helix and that intersecting sets of deep grooves mark the viral surface, with one set running almost longitudinally and the other following the simple viral helix. © 2002 Elsevier Science (USA) Key Words: potato virus X; potexviruses; virus structure; fiber diffraction; magnetic orientation.
The quality of a fiber diffraction pattern depends primarily on the degree of orientation of the particles in the diffracting specimen. Namba’s group (Yamashita et al., 1998a,b) has had considerable success in improving the quality of fiber diffraction patterns from bacterial flagella by centrifuging sols and exposing them to very high magnetic fields. We have used modifications of their procedures and considerably improved the orientation in PVX sols. The resulting diffraction patterns allow us to describe the surface of the PVX virion and to define the helical symmetry with greatly improved accuracy.
INTRODUCTION Structural studies of flexible filamentous viruses are difficult. The structures of many spherical viruses have been determined crystallographically, and several of the rigid tobamoviruses have yielded to high-resolution fiber diffraction studies (Wang et al., 1997), but no flexible filamentous virus structure has yet been determined in molecular detail. Since flexible filaments do not usually crystallize, and electron microscopy of macromolecules rarely reaches molecular resolution, fiber diffraction, with all its inherent difficulties (Chandrasekaran and Stubbs, 2001), must be the method of choice for such studies. Fiber diffraction studies have been made of a number of potexviruses, as well as a few isolated members of other groups (reviewed by Tollin and Wilson, 1988; Stubbs, 2001). These studies have, however, generally been limited to the determination of the helical parameters of the viruses, or even simply to the helical pitch. Potato virus X (PVX) is a particularly interesting filamentous virus. As the type member of the potexvirus genus, it is well studied. It is of considerable significance to both agriculture (White et al., 1994) and biotechnology (Santa Cruz et al., 1996). Fiber diffraction studies (Tollin et al., 1980; Stubbs et al., 2000) have shown that it has a helical structure, with just under nine subunits per helical turn. Until now, however, diffraction patterns have not been sufficiently well ordered to obtain more information than this. Realization of the full potential of PVX as a model system and in biotechnology will require considerably more details of the subunit structure.
RESULTS AND DISCUSSION Orientation Sols of PVX were centrifuged and exposed to high magnetic fields. Orientation in most of the sols appeared to improve in the magnetic field, as judged by the degree of extinction between crossed polarizers. Most of the orientation occurred during exposures to the highest fields used, 14 and 18.8 T, even though exposures to lower fields were much longer (several weeks in some cases). Examination of the best diffracting sol under a polarizing microscope showed that only a small region at the top of the collapsed region (see Oriented sols below) had oriented. The degree of perfection in the final oriented sol is increased by high concentration (Gregory and Holmes, 1965; Yamashita et al., 1998a), but high concentration also increases viscosity and thus makes orientation more difficult. Much longer exposure times, particularly to the highest available magnetic fields, should extend orientation into the more concentrated collapsed regions of sols; both the longer exposure and the higher concentration should increase the likelihood of improved orientation.
1 To whom correspondence and reprint requests should be addressed at Department of Biological Sciences, Vanderbilt University, Box 1634, Station B, Nashville, TN 37235. Fax: 615 343 6707. E-mail: [email protected].
291
0042-6822/02 $35.00 © 2002 Elsevier Science (USA) All rights reserved.
292
PARKER, KENDALL, AND STUBBS
Diffraction patterns A diffraction pattern from the best oriented sol is shown in Fig. 1. Measurements of several well-resolved diffraction maxima give mean disorientations between 4 and 5°. Disorientation is defined as the standard deviation of a Gaussian approximating the profile of the diffracted intensity along an arc centered at the origin of the diffraction pattern (Holmes and Barrington Leigh, 1974; Yamashita et al., 1998a). The sols are not as well oriented as sols of tobacco mosaic virus (Gregory and Holmes, 1965) or flagellin (Yamashita et al., 1998a), both of which achieve disorientations of 1° or less. They are, however, much better oriented than any previously oriented flexible virus. The pattern is dominated by a series of near-meridional layer lines spaced at intervals of 1/34.5 Å ⫺1, representing a viral helix pitch of 34.5 ⫾ 0.5 Å, in good agreement with the value of 34.3 Å determined by Tollin et al. (1980). Unlike previously recorded patterns, however, the pattern also shows diffraction maxima away from these layer lines (Fig. 1). The most intense of these maxima, and one of most noticeable features in the pattern, is the very strong off-equatorial intensity near the center of the pattern. In previously reported PVX diffraction patterns, this intensity could not be distinguished from the equator. In this pattern, however, it clearly belongs to the first layer line. This observation has only been made possible by the combination of improved orientation and the very small X-ray beam obtainable at the BioCAT synchrotron facility. Layer lines were fitted to the pattern. An excellent fit was obtained visually using a layer line spacing of 1/345 Å ⫺1. Again, this fit is only possible because of the clearly defined first layer line. The first near-meridional layer line, at 1/34.5 Å ⫺1, is the tenth layer line, rather than the eighth as suggested by electron microscopic observations (Horne et al., 1975; Tollin et al., 1980). That is, the viral helix has close to an exact repeat after 10 turns. The difference between the fiber diffraction results and the electron microscopic observations can easily be accounted for by a very slight twisting of the viral helix under the conditions of negative staining for electron microscopy. The number of subunits per helical turn for PVX, similar to all potexviruses, is believed to be a little less than nine (Richardson et al., 1981; Tollin and Wilson, 1988). Our results therefore show that there are 8.90 ⫾ 0.01 subunits per turn of the viral helix, close to the previously determined figure (Tollin et al., 1980) of 8.875. Analysis The remarkable intensity of the first layer line suggests very strong contrast in some feature of the corresponding projection of the viral structure. The first layer line represents the projection along the nine-start family of helices (Klug et al., 1958). These helices may be visualized by considering a point in any subunit of the virus as
an origin, taking a line through the corresponding point nine subunits along the basic viral helix, and continuing the line along the length of the virus. Since there are 8.9 subunits per turn, the ninth subunit is almost immediately above the origin subunit (with the viral axis vertical). Thus the nine-start helices run almost longitudinally along the virus (Fig. 2). The only feature consistent with macromolecular structure that could give rise to an intensity maximum at such low resolution would be a marked groove running along the nine-start helices. We can make a rough estimate of the depth of this groove. The intensity at a point R on a layer line depends on J n (2 Rr), where J n is the Bessel function of order n, R is the radial coordinate in reciprocal space (the distance along the layer line from the meridian), and r is the radius of the structural feature giving rise to the diffracted intensity (Cochran et al., 1952; Chandrasekaran and Stubbs, 2001). The value of n depends on the symmetry of the diffracting helix. For the first layer line in the PVX diffraction pattern, n ⫽ 9. The first maximum in J 9 ( x) is at x ⫽ 10.7, and the observed maximum intensity on the first layer line is at R ⫽ 0.031 Å ⫺1, so if x ⫽ 2 Rr, the average radial coordinate of the groove, r, is estimated to be 55 Å. The actual depth of the groove will depend upon its detailed shape, but given a maximum viral radius of 65 Å (Tollin and Wilson, 1988), and assuming a simple shape for the groove, the groove is probably about 20 Å deep. A similar argument can be made using data from the first near-meridional layer line (the tenth layer line, as discussed above). For this layer line, n ⫽ 1, and the first observed intensity maximum is at R ⫽ 0.005 Å ⫺1. Again, the strength of this maximum suggests a distinct groove, this time following the one-start helix, the basic viral helix. The average radial coordinate of this groove is estimated to be 56 Å, that is, this groove is probably also about 20 Å deep. The calculations in the previous paragraphs make many assumptions. Some indication of their validity may be obtained by making the same calculations for tobacco mosaic virus (TMV), whose three-dimensional structure is known (Namba et al., 1989). The first observed peak on the first layer line in the TMV fiber diffraction pattern is at R ⫽ 0.035. This peak is not nearly as intense as the corresponding peak in PVX, but following the calculations above, it suggests the presence of an almost longitudinal groove (16-start, because of the symmetry of TMV) about 15 Å deep. Similarly, the first observed peak on the third layer line of TMV, at R ⫽ 0.0036, predicts a very pronounced one-start groove following the basic viral helix, of about the same depth. Examination of the structure of TMV (as shown, for example, in Figure 1 of Namba et al., 1989) shows that these grooves are indeed present, both having close to the predicted depths, and, consistent with the relative intensities of the diffraction maxima, the one-start groove is much more pronounced than the 16-start.
FIG. 1. (a) Diffraction pattern from an oriented sol of PVX. The white arrow indicates layer line 10, the first in the series of near-meridional layer lines. The black arrows indicate the off-meridional layer lines 1 and 21. (b) Diffraction pattern from (a), projected in reciprocal space so that layer lines are straight and distances in reciprocal space are not distorted (Fraser et al., 1976). Layer lines 0 (the equator), 10, 20, 30, and 40 are marked (black lines), as are the off-meridional layer lines 1 and 21 (white lines). 293
294
PARKER, KENDALL, AND STUBBS
of TMV has a radius of 75 Å. The deep grooves, and particularly the one-start groove, probably contribute significantly to the flexibility of PVX. MATERIALS AND METHODS Virus purification
FIG. 2. Schematic drawing of a segment of the PVX virion, containing about 89 subunits in 10 turns of the viral helix. Each surface protrusion represents a subunit. Two sets of grooves are shown: the 9-start grooves are almost vertical, while the 1-start groove is almost horizontal, following the viral helix.
PVX thus has a grooved surface (Fig. 2) similar to that of TMV. Such grooving was predicted by Tollin et al. (1967), who observed by electron microscopy a reduction in interparticle distances on drying, suggesting interpenetration of particles as seen in TMV (Franklin and Klug, 1956). There are, however, two significant differences between the grooves in PVX and TMV. In PVX, the nearlongitudinal grooves are much more pronounced than in TMV, as shown by the intensity of the first diffraction maximum on the first layer line. Furthermore, relative to the radius of the virus, the grooves in PVX are much deeper than those of TMV. At 20 Å deep, the PVX grooves occupy 30% of the viral radius, reducing the radius of the core of the virus to 45 Å. In contrast, the 15 Å deep TMV grooves occupy only 18% of the viral radius, and the core
PVX was purified from infected tobacco plants (Nicotiana clevelandii) using a procedure modified from that of Shepard (1972). In the purification described here, a protease inhibitor cocktail (PIC) was used. Two stock solutions were used: the first was 0.1 M benzamidine and 0.1 mg/ml leupeptin in water. The second was 1.0 mg/ml pepstatin A in methanol. “1% by volume PIC” means that 1 ml of each stock solution was included in 100 ml of the final solution. All procedures were carried out at 4°C unless otherwise noted. Harvested plants were homogenized in a blender with 4 ml buffer per gram of tissue. The buffer was 0.5 M sodium borate, pH 8.2, made 10 mM in -mercaptoethanol and 1% by volume PIC. The extract was strained through a triple layer of cheesecloth, mixed with an equal volume of chloroform, and shaken for 1 min. The resulting emulsion was centrifuged at 1700 g for 20 min, and then the aqueous phase was centrifuged at 87,000 g for 90 min. The pellet was resuspended in 4% sodium chloride, 1% PIC, using 0.2 ml of solution per gram of plant tissue. The resulting solution was centrifuged at 9000 g for 10 min to remove insoluble debris. A 20% solution of polyethylene glycol (PEG), molecular weight 5000–7000, 1% in PIC, was added to the supernatant, using 0.105 ml PEG solution per gram of plant tissue, and stirred for 1 h. The suspension was then centrifuged at 10,000 g for 10 min. The pellet was again resuspended in sodium chloride solution; the PEG precipitation was repeated, and the pellet was resuspended in sodium chloride a third time. The solution was centrifuged through a cushion of 30% sucrose in 50 mM borate buffer, pH 8.2, 1% PIC, at 63,000 g for 3 h. Each aliquot of virus solution (about 20 ml) was centrifuged through a 5-ml sucrose cushion. The final pellets were resuspended in 10 mM Tris–HCl pH 8, 5 mM in ethylenediamine tetraacetic acid (EDTA), and 1% in PIC. Oriented sols Oriented sols for fiber diffraction were prepared by the method of Yamashita et al. (1998a). Soft virus pellets were prepared by centrifuging 1 ml viral solutions in Eppendorf tubes at 12,000 g (17,000 rpm in a Beckman TLA100.3 rotor) for 21 h. The most successful experiments used solutions 6 mg/ml in virus, 50 mM in NaCl, 10 mM Tris–HCl pH 8, 5 mM EDTA, and 1% PIC. A column of virus about 1 cm long was drawn into a clean glass capillary of diameter 0.5 mm (Charles Supper Co., Natick, MA) and moved back and forth by aspiration (Gregory
SURFACE FEATURES OF POTATO VIRUS X
and Holmes, 1965). The capillaries were sealed by flaming and with epoxy resin and then centrifuged in a Sorvall HB-6 swinging-bucket rotor at 5000 g (7000 rpm) for about 60 h. This treatment typically separates the sol into a dilute region, a concentrated region, and a “collapsed” region (Yamashita et al., 1998a), with the collapsed region at the bottom of the capillary. The dilute region and a significant portion of the concentrated region (typically about half the column) were removed, and the capillary was resealed. The capillaries were exposed to high magnetic fields for several weeks. The magnetic fields varied, since it was necessary to use fields from NMR magnets (Bruker, Germany) opportunistically as demand for the NMR spectrometers allowed. The specimen whose diffraction pattern is shown in Fig. 1 was exposed to a field of 18.8 T (inside an 800 MHz NMR instrument) for 44 h, about 9 T (outside the 800 MHz instrument) for 66 h, and 14 T (inside a 600 MHz instrument) for 18 h. Preliminary assessments of orientation were made under a polarizing microscope (Gregory and Holmes, 1965). Fiber diffraction Fiber diffraction data were collected at the BioCAT beam line of the Advanced Photon Source synchrotron, Argonne National Laboratory. The specimens were mounted with the fiber axis vertical. The beam had dimensions of 35 m vertically and 60 m horizontally at the detector, and because of the long focal length of the mirrors, was very close to this size at the specimen. Data were recorded using a CCD detector approximately 100 mm from the specimen. The detector measured approximately 85 mm horizontally and 50 mm vertically, with an effective pixel size of 24 m square (the actual pixel size was 12 m square, but pixels were binned in 2 ⫻ 2 arrays). The X-ray wavelength was 1.03 Å. Data were analyzed using the public domain NIH Image program (developed at the U.S. National Institutes of Health, http://rsb.info.nih.gov/nih-image/) for direct measurements and the program WCEN (H. Wang and G. Stubbs, unpublished observations) for measurements in reciprocal space, including the fitting of layer lines. ACKNOWLEDGMENTS We thank David Baulcombe for providing the initial inoculum of PVX, Markus Voehler and the Vanderbilt University Center for Structural Biology for access to high-field magnets, Kelly Lynch, Nicholas Taraska, Tom Irving, and the staff of BioCAT for help with data collection, and Keiichi Namba, Mitzi Dunagan, and Greg Ferrell for discussions about specimen preparation. This work was supported by Grants MCB-9809879 and DBI-9604789 from the National Science Foundation. Use of the Advanced Photon Source was supported by the U.S. Department of Energy, Basic Energy Sciences, Office of Science, under contract No. W-31-109-ENG-38. BioCAT is a National Institutes of Health supported Research Center RR-08630.
295
REFERENCES Chandrasekaran, R., and Stubbs, G. (2001). Fibre diffraction. In “International Tables for Crystallography, Vol. F: Crystallography of Biological Macromolecules” (M. G. Rossman and E. Arnold, Eds.), pp. 444–450. Kluwer Academic Publishers, The Netherlands. Cochran, W., Crick, F. H. C., and Vand, V. (1952). The structure of synthetic polypeptides. I. The transform of atoms on a helix. Acta Cryst. 5, 581–586. Franklin, R. E., and Klug, A. (1956). The nature of the helical groove on the tobacco mosaic virus particle. Biochim. Biophys. Acta 19, 403– 416. Fraser, R. D. B., Macrae, T. P., Miller, A., and Rowlands, R. J. (1976). Digital processing of fibre diffraction patterns. J. Appl. Cryst. 9, 81–94. Gregory, J., and Holmes, K. C. (1965). Methods of preparing oriented tobacco mosaic virus sols for X-ray diffraction. J. Mol. Biol. 13, 796–801. Holmes, K. C., and Barrington Leigh, J. (1974). The effect of disorientation on the intensity distribution of non-crystalline fibres. I. Theory. Acta Cryst. A30, 635–638. Horne, R. W., Hobart, J. M., and Ronchetti, I. P. (1975). Application of the negative staining-carbon technique to the study of virus particles and their components by electron microscopy. Micron 5, 233–261. Klug, A., Crick, F. H. C., and Wyckoff, H. W. (1958). Diffraction by helical structures. Acta Cryst. 11, 199–213. Namba, K., Pattanayek, R., and Stubbs, G. (1989). Visualization of protein-nucleic acid interactions in a virus; refinement of intact tobacco mosaic virus structure at 2.9 Å resolution by fiber diffraction. J. Mol. Biol. 208, 307–325. Richardson, J. F., Tollin, P., and Bancroft, J. B. (1981). The architecture of the potexviruses. Virology 112, 34–39. Santa Cruz, S., Chapman, S., Roberts, A. G., Roberts, I. M., Prior, D. A. M., and Oparka, K. J. (1996). Assembly and movement of a plant virus carrying a green fluorescent protein overcoat. Proc. Natl. Acad. Sci. USA 93, 6286–6290. Shepard, J. F. (1972). Gel-diffusion methods for the serological detection of potato viruses X, S and M. Bulletin 662 of the Montana Agricultural Experimental Station, Bozeman, MT. Stubbs, G. (2001). Fibre diffraction studies of filamentous viruses. Rep. Progr. Phys. 64, 1389–1425. Stubbs, G., Ferrell, G., Reams, M., and Fletcher, N. (2000). Fibre diffraction and diversity in filamentous plant virus structure. Fib. Diffract. Rev. 9, 24–28. Tollin, P., and Wilson, H. R. (1988). Particle structure. In “The Plant Viruses, Vol. 4: The Filamentous Plant Viruses” (R. G. Milne, Ed.), pp. 51–83. Plenum, New York. Tollin, P., Wilson, H. R., Young, D. W., Cathro, J., and Mowat, W. P. (1967). X-ray diffraction and electron microscope studies of narcissus mosaic virus, and comparison with potato virus X. J. Mol. Biol. 26, 353–355. Tollin, P., Wilson, H. R., and Bancroft, J. B. (1980). Further observations on the structure of particles of potato virus X. J. Gen. Virol. 49, 407–410. Wang, H., Culver, J. N., and Stubbs, G. (1997). Structure of ribgrass mosaic virus at 2.9 Å resolution: Evolution and taxonomy of tobamoviruses. J. Mol. Biol. 269, 769–779. White, K. A., Rouleau, M., Bancroft, J. B., and Mackie, G. A. (1994). Potexviruses. In “Encyclopedia of Virology” (R. G. Webster and A. Ganoff, Eds.), Vol. 2, pp. 1142–1147. Academic Press, London. Yamashita, I., Suzuki, H., and Namba, K. (1998a). Multiple-step method for making exceptionally well-oriented liquid-crystalline sols of macromolecular assemblies. J. Mol. Biol. 278, 609–615. Yamashita, I., Hasegawa, K., Suzuki, H., Vonderviszt, F., Mimori-Kiyosue, Y., and Namba, K. (1998b). Structure and switching of bacterial flagellar filaments studied by X-ray fiber diffraction. Nat. Struct. Biol. 5, 125–132.
Surface Features of Potato Virus X from Fiber Diffraction Lauren Parker,* Amy Kendall,† and Gerald Stubbs† ,1 *Department of Chemistry and †Department of Biological Sciences, Vanderbilt University, Nashville, Tennessee 37235 Received February 26, 2002; accepted March 19, 2002 Fiber diffraction patterns have been obtained from oriented sols of potato virus X. Orientation in the sols was greatly improved by a combination of centrifugation and exposure to very high magnetic fields. Diffraction patterns were also improved by using a very finely collimated synchrotron X-ray beam. The diffraction patterns show that there are 8.9 subunits in each turn of the viral helix and that intersecting sets of deep grooves mark the viral surface, with one set running almost longitudinally and the other following the simple viral helix. © 2002 Elsevier Science (USA) Key Words: potato virus X; potexviruses; virus structure; fiber diffraction; magnetic orientation.
The quality of a fiber diffraction pattern depends primarily on the degree of orientation of the particles in the diffracting specimen. Namba’s group (Yamashita et al., 1998a,b) has had considerable success in improving the quality of fiber diffraction patterns from bacterial flagella by centrifuging sols and exposing them to very high magnetic fields. We have used modifications of their procedures and considerably improved the orientation in PVX sols. The resulting diffraction patterns allow us to describe the surface of the PVX virion and to define the helical symmetry with greatly improved accuracy.
INTRODUCTION Structural studies of flexible filamentous viruses are difficult. The structures of many spherical viruses have been determined crystallographically, and several of the rigid tobamoviruses have yielded to high-resolution fiber diffraction studies (Wang et al., 1997), but no flexible filamentous virus structure has yet been determined in molecular detail. Since flexible filaments do not usually crystallize, and electron microscopy of macromolecules rarely reaches molecular resolution, fiber diffraction, with all its inherent difficulties (Chandrasekaran and Stubbs, 2001), must be the method of choice for such studies. Fiber diffraction studies have been made of a number of potexviruses, as well as a few isolated members of other groups (reviewed by Tollin and Wilson, 1988; Stubbs, 2001). These studies have, however, generally been limited to the determination of the helical parameters of the viruses, or even simply to the helical pitch. Potato virus X (PVX) is a particularly interesting filamentous virus. As the type member of the potexvirus genus, it is well studied. It is of considerable significance to both agriculture (White et al., 1994) and biotechnology (Santa Cruz et al., 1996). Fiber diffraction studies (Tollin et al., 1980; Stubbs et al., 2000) have shown that it has a helical structure, with just under nine subunits per helical turn. Until now, however, diffraction patterns have not been sufficiently well ordered to obtain more information than this. Realization of the full potential of PVX as a model system and in biotechnology will require considerably more details of the subunit structure.
RESULTS AND DISCUSSION Orientation Sols of PVX were centrifuged and exposed to high magnetic fields. Orientation in most of the sols appeared to improve in the magnetic field, as judged by the degree of extinction between crossed polarizers. Most of the orientation occurred during exposures to the highest fields used, 14 and 18.8 T, even though exposures to lower fields were much longer (several weeks in some cases). Examination of the best diffracting sol under a polarizing microscope showed that only a small region at the top of the collapsed region (see Oriented sols below) had oriented. The degree of perfection in the final oriented sol is increased by high concentration (Gregory and Holmes, 1965; Yamashita et al., 1998a), but high concentration also increases viscosity and thus makes orientation more difficult. Much longer exposure times, particularly to the highest available magnetic fields, should extend orientation into the more concentrated collapsed regions of sols; both the longer exposure and the higher concentration should increase the likelihood of improved orientation.
1 To whom correspondence and reprint requests should be addressed at Department of Biological Sciences, Vanderbilt University, Box 1634, Station B, Nashville, TN 37235. Fax: 615 343 6707. E-mail: [email protected].
291
0042-6822/02 $35.00 © 2002 Elsevier Science (USA) All rights reserved.
292
PARKER, KENDALL, AND STUBBS
Diffraction patterns A diffraction pattern from the best oriented sol is shown in Fig. 1. Measurements of several well-resolved diffraction maxima give mean disorientations between 4 and 5°. Disorientation is defined as the standard deviation of a Gaussian approximating the profile of the diffracted intensity along an arc centered at the origin of the diffraction pattern (Holmes and Barrington Leigh, 1974; Yamashita et al., 1998a). The sols are not as well oriented as sols of tobacco mosaic virus (Gregory and Holmes, 1965) or flagellin (Yamashita et al., 1998a), both of which achieve disorientations of 1° or less. They are, however, much better oriented than any previously oriented flexible virus. The pattern is dominated by a series of near-meridional layer lines spaced at intervals of 1/34.5 Å ⫺1, representing a viral helix pitch of 34.5 ⫾ 0.5 Å, in good agreement with the value of 34.3 Å determined by Tollin et al. (1980). Unlike previously recorded patterns, however, the pattern also shows diffraction maxima away from these layer lines (Fig. 1). The most intense of these maxima, and one of most noticeable features in the pattern, is the very strong off-equatorial intensity near the center of the pattern. In previously reported PVX diffraction patterns, this intensity could not be distinguished from the equator. In this pattern, however, it clearly belongs to the first layer line. This observation has only been made possible by the combination of improved orientation and the very small X-ray beam obtainable at the BioCAT synchrotron facility. Layer lines were fitted to the pattern. An excellent fit was obtained visually using a layer line spacing of 1/345 Å ⫺1. Again, this fit is only possible because of the clearly defined first layer line. The first near-meridional layer line, at 1/34.5 Å ⫺1, is the tenth layer line, rather than the eighth as suggested by electron microscopic observations (Horne et al., 1975; Tollin et al., 1980). That is, the viral helix has close to an exact repeat after 10 turns. The difference between the fiber diffraction results and the electron microscopic observations can easily be accounted for by a very slight twisting of the viral helix under the conditions of negative staining for electron microscopy. The number of subunits per helical turn for PVX, similar to all potexviruses, is believed to be a little less than nine (Richardson et al., 1981; Tollin and Wilson, 1988). Our results therefore show that there are 8.90 ⫾ 0.01 subunits per turn of the viral helix, close to the previously determined figure (Tollin et al., 1980) of 8.875. Analysis The remarkable intensity of the first layer line suggests very strong contrast in some feature of the corresponding projection of the viral structure. The first layer line represents the projection along the nine-start family of helices (Klug et al., 1958). These helices may be visualized by considering a point in any subunit of the virus as
an origin, taking a line through the corresponding point nine subunits along the basic viral helix, and continuing the line along the length of the virus. Since there are 8.9 subunits per turn, the ninth subunit is almost immediately above the origin subunit (with the viral axis vertical). Thus the nine-start helices run almost longitudinally along the virus (Fig. 2). The only feature consistent with macromolecular structure that could give rise to an intensity maximum at such low resolution would be a marked groove running along the nine-start helices. We can make a rough estimate of the depth of this groove. The intensity at a point R on a layer line depends on J n (2 Rr), where J n is the Bessel function of order n, R is the radial coordinate in reciprocal space (the distance along the layer line from the meridian), and r is the radius of the structural feature giving rise to the diffracted intensity (Cochran et al., 1952; Chandrasekaran and Stubbs, 2001). The value of n depends on the symmetry of the diffracting helix. For the first layer line in the PVX diffraction pattern, n ⫽ 9. The first maximum in J 9 ( x) is at x ⫽ 10.7, and the observed maximum intensity on the first layer line is at R ⫽ 0.031 Å ⫺1, so if x ⫽ 2 Rr, the average radial coordinate of the groove, r, is estimated to be 55 Å. The actual depth of the groove will depend upon its detailed shape, but given a maximum viral radius of 65 Å (Tollin and Wilson, 1988), and assuming a simple shape for the groove, the groove is probably about 20 Å deep. A similar argument can be made using data from the first near-meridional layer line (the tenth layer line, as discussed above). For this layer line, n ⫽ 1, and the first observed intensity maximum is at R ⫽ 0.005 Å ⫺1. Again, the strength of this maximum suggests a distinct groove, this time following the one-start helix, the basic viral helix. The average radial coordinate of this groove is estimated to be 56 Å, that is, this groove is probably also about 20 Å deep. The calculations in the previous paragraphs make many assumptions. Some indication of their validity may be obtained by making the same calculations for tobacco mosaic virus (TMV), whose three-dimensional structure is known (Namba et al., 1989). The first observed peak on the first layer line in the TMV fiber diffraction pattern is at R ⫽ 0.035. This peak is not nearly as intense as the corresponding peak in PVX, but following the calculations above, it suggests the presence of an almost longitudinal groove (16-start, because of the symmetry of TMV) about 15 Å deep. Similarly, the first observed peak on the third layer line of TMV, at R ⫽ 0.0036, predicts a very pronounced one-start groove following the basic viral helix, of about the same depth. Examination of the structure of TMV (as shown, for example, in Figure 1 of Namba et al., 1989) shows that these grooves are indeed present, both having close to the predicted depths, and, consistent with the relative intensities of the diffraction maxima, the one-start groove is much more pronounced than the 16-start.
FIG. 1. (a) Diffraction pattern from an oriented sol of PVX. The white arrow indicates layer line 10, the first in the series of near-meridional layer lines. The black arrows indicate the off-meridional layer lines 1 and 21. (b) Diffraction pattern from (a), projected in reciprocal space so that layer lines are straight and distances in reciprocal space are not distorted (Fraser et al., 1976). Layer lines 0 (the equator), 10, 20, 30, and 40 are marked (black lines), as are the off-meridional layer lines 1 and 21 (white lines). 293
294
PARKER, KENDALL, AND STUBBS
of TMV has a radius of 75 Å. The deep grooves, and particularly the one-start groove, probably contribute significantly to the flexibility of PVX. MATERIALS AND METHODS Virus purification
FIG. 2. Schematic drawing of a segment of the PVX virion, containing about 89 subunits in 10 turns of the viral helix. Each surface protrusion represents a subunit. Two sets of grooves are shown: the 9-start grooves are almost vertical, while the 1-start groove is almost horizontal, following the viral helix.
PVX thus has a grooved surface (Fig. 2) similar to that of TMV. Such grooving was predicted by Tollin et al. (1967), who observed by electron microscopy a reduction in interparticle distances on drying, suggesting interpenetration of particles as seen in TMV (Franklin and Klug, 1956). There are, however, two significant differences between the grooves in PVX and TMV. In PVX, the nearlongitudinal grooves are much more pronounced than in TMV, as shown by the intensity of the first diffraction maximum on the first layer line. Furthermore, relative to the radius of the virus, the grooves in PVX are much deeper than those of TMV. At 20 Å deep, the PVX grooves occupy 30% of the viral radius, reducing the radius of the core of the virus to 45 Å. In contrast, the 15 Å deep TMV grooves occupy only 18% of the viral radius, and the core
PVX was purified from infected tobacco plants (Nicotiana clevelandii) using a procedure modified from that of Shepard (1972). In the purification described here, a protease inhibitor cocktail (PIC) was used. Two stock solutions were used: the first was 0.1 M benzamidine and 0.1 mg/ml leupeptin in water. The second was 1.0 mg/ml pepstatin A in methanol. “1% by volume PIC” means that 1 ml of each stock solution was included in 100 ml of the final solution. All procedures were carried out at 4°C unless otherwise noted. Harvested plants were homogenized in a blender with 4 ml buffer per gram of tissue. The buffer was 0.5 M sodium borate, pH 8.2, made 10 mM in -mercaptoethanol and 1% by volume PIC. The extract was strained through a triple layer of cheesecloth, mixed with an equal volume of chloroform, and shaken for 1 min. The resulting emulsion was centrifuged at 1700 g for 20 min, and then the aqueous phase was centrifuged at 87,000 g for 90 min. The pellet was resuspended in 4% sodium chloride, 1% PIC, using 0.2 ml of solution per gram of plant tissue. The resulting solution was centrifuged at 9000 g for 10 min to remove insoluble debris. A 20% solution of polyethylene glycol (PEG), molecular weight 5000–7000, 1% in PIC, was added to the supernatant, using 0.105 ml PEG solution per gram of plant tissue, and stirred for 1 h. The suspension was then centrifuged at 10,000 g for 10 min. The pellet was again resuspended in sodium chloride solution; the PEG precipitation was repeated, and the pellet was resuspended in sodium chloride a third time. The solution was centrifuged through a cushion of 30% sucrose in 50 mM borate buffer, pH 8.2, 1% PIC, at 63,000 g for 3 h. Each aliquot of virus solution (about 20 ml) was centrifuged through a 5-ml sucrose cushion. The final pellets were resuspended in 10 mM Tris–HCl pH 8, 5 mM in ethylenediamine tetraacetic acid (EDTA), and 1% in PIC. Oriented sols Oriented sols for fiber diffraction were prepared by the method of Yamashita et al. (1998a). Soft virus pellets were prepared by centrifuging 1 ml viral solutions in Eppendorf tubes at 12,000 g (17,000 rpm in a Beckman TLA100.3 rotor) for 21 h. The most successful experiments used solutions 6 mg/ml in virus, 50 mM in NaCl, 10 mM Tris–HCl pH 8, 5 mM EDTA, and 1% PIC. A column of virus about 1 cm long was drawn into a clean glass capillary of diameter 0.5 mm (Charles Supper Co., Natick, MA) and moved back and forth by aspiration (Gregory
SURFACE FEATURES OF POTATO VIRUS X
and Holmes, 1965). The capillaries were sealed by flaming and with epoxy resin and then centrifuged in a Sorvall HB-6 swinging-bucket rotor at 5000 g (7000 rpm) for about 60 h. This treatment typically separates the sol into a dilute region, a concentrated region, and a “collapsed” region (Yamashita et al., 1998a), with the collapsed region at the bottom of the capillary. The dilute region and a significant portion of the concentrated region (typically about half the column) were removed, and the capillary was resealed. The capillaries were exposed to high magnetic fields for several weeks. The magnetic fields varied, since it was necessary to use fields from NMR magnets (Bruker, Germany) opportunistically as demand for the NMR spectrometers allowed. The specimen whose diffraction pattern is shown in Fig. 1 was exposed to a field of 18.8 T (inside an 800 MHz NMR instrument) for 44 h, about 9 T (outside the 800 MHz instrument) for 66 h, and 14 T (inside a 600 MHz instrument) for 18 h. Preliminary assessments of orientation were made under a polarizing microscope (Gregory and Holmes, 1965). Fiber diffraction Fiber diffraction data were collected at the BioCAT beam line of the Advanced Photon Source synchrotron, Argonne National Laboratory. The specimens were mounted with the fiber axis vertical. The beam had dimensions of 35 m vertically and 60 m horizontally at the detector, and because of the long focal length of the mirrors, was very close to this size at the specimen. Data were recorded using a CCD detector approximately 100 mm from the specimen. The detector measured approximately 85 mm horizontally and 50 mm vertically, with an effective pixel size of 24 m square (the actual pixel size was 12 m square, but pixels were binned in 2 ⫻ 2 arrays). The X-ray wavelength was 1.03 Å. Data were analyzed using the public domain NIH Image program (developed at the U.S. National Institutes of Health, http://rsb.info.nih.gov/nih-image/) for direct measurements and the program WCEN (H. Wang and G. Stubbs, unpublished observations) for measurements in reciprocal space, including the fitting of layer lines. ACKNOWLEDGMENTS We thank David Baulcombe for providing the initial inoculum of PVX, Markus Voehler and the Vanderbilt University Center for Structural Biology for access to high-field magnets, Kelly Lynch, Nicholas Taraska, Tom Irving, and the staff of BioCAT for help with data collection, and Keiichi Namba, Mitzi Dunagan, and Greg Ferrell for discussions about specimen preparation. This work was supported by Grants MCB-9809879 and DBI-9604789 from the National Science Foundation. Use of the Advanced Photon Source was supported by the U.S. Department of Energy, Basic Energy Sciences, Office of Science, under contract No. W-31-109-ENG-38. BioCAT is a National Institutes of Health supported Research Center RR-08630.
295
REFERENCES Chandrasekaran, R., and Stubbs, G. (2001). Fibre diffraction. In “International Tables for Crystallography, Vol. F: Crystallography of Biological Macromolecules” (M. G. Rossman and E. Arnold, Eds.), pp. 444–450. Kluwer Academic Publishers, The Netherlands. Cochran, W., Crick, F. H. C., and Vand, V. (1952). The structure of synthetic polypeptides. I. The transform of atoms on a helix. Acta Cryst. 5, 581–586. Franklin, R. E., and Klug, A. (1956). The nature of the helical groove on the tobacco mosaic virus particle. Biochim. Biophys. Acta 19, 403– 416. Fraser, R. D. B., Macrae, T. P., Miller, A., and Rowlands, R. J. (1976). Digital processing of fibre diffraction patterns. J. Appl. Cryst. 9, 81–94. Gregory, J., and Holmes, K. C. (1965). Methods of preparing oriented tobacco mosaic virus sols for X-ray diffraction. J. Mol. Biol. 13, 796–801. Holmes, K. C., and Barrington Leigh, J. (1974). The effect of disorientation on the intensity distribution of non-crystalline fibres. I. Theory. Acta Cryst. A30, 635–638. Horne, R. W., Hobart, J. M., and Ronchetti, I. P. (1975). Application of the negative staining-carbon technique to the study of virus particles and their components by electron microscopy. Micron 5, 233–261. Klug, A., Crick, F. H. C., and Wyckoff, H. W. (1958). Diffraction by helical structures. Acta Cryst. 11, 199–213. Namba, K., Pattanayek, R., and Stubbs, G. (1989). Visualization of protein-nucleic acid interactions in a virus; refinement of intact tobacco mosaic virus structure at 2.9 Å resolution by fiber diffraction. J. Mol. Biol. 208, 307–325. Richardson, J. F., Tollin, P., and Bancroft, J. B. (1981). The architecture of the potexviruses. Virology 112, 34–39. Santa Cruz, S., Chapman, S., Roberts, A. G., Roberts, I. M., Prior, D. A. M., and Oparka, K. J. (1996). Assembly and movement of a plant virus carrying a green fluorescent protein overcoat. Proc. Natl. Acad. Sci. USA 93, 6286–6290. Shepard, J. F. (1972). Gel-diffusion methods for the serological detection of potato viruses X, S and M. Bulletin 662 of the Montana Agricultural Experimental Station, Bozeman, MT. Stubbs, G. (2001). Fibre diffraction studies of filamentous viruses. Rep. Progr. Phys. 64, 1389–1425. Stubbs, G., Ferrell, G., Reams, M., and Fletcher, N. (2000). Fibre diffraction and diversity in filamentous plant virus structure. Fib. Diffract. Rev. 9, 24–28. Tollin, P., and Wilson, H. R. (1988). Particle structure. In “The Plant Viruses, Vol. 4: The Filamentous Plant Viruses” (R. G. Milne, Ed.), pp. 51–83. Plenum, New York. Tollin, P., Wilson, H. R., Young, D. W., Cathro, J., and Mowat, W. P. (1967). X-ray diffraction and electron microscope studies of narcissus mosaic virus, and comparison with potato virus X. J. Mol. Biol. 26, 353–355. Tollin, P., Wilson, H. R., and Bancroft, J. B. (1980). Further observations on the structure of particles of potato virus X. J. Gen. Virol. 49, 407–410. Wang, H., Culver, J. N., and Stubbs, G. (1997). Structure of ribgrass mosaic virus at 2.9 Å resolution: Evolution and taxonomy of tobamoviruses. J. Mol. Biol. 269, 769–779. White, K. A., Rouleau, M., Bancroft, J. B., and Mackie, G. A. (1994). Potexviruses. In “Encyclopedia of Virology” (R. G. Webster and A. Ganoff, Eds.), Vol. 2, pp. 1142–1147. Academic Press, London. Yamashita, I., Suzuki, H., and Namba, K. (1998a). Multiple-step method for making exceptionally well-oriented liquid-crystalline sols of macromolecular assemblies. J. Mol. Biol. 278, 609–615. Yamashita, I., Hasegawa, K., Suzuki, H., Vonderviszt, F., Mimori-Kiyosue, Y., and Namba, K. (1998b). Structure and switching of bacterial flagellar filaments studied by X-ray fiber diffraction. Nat. Struct. Biol. 5, 125–132.