COMMUNICATION

AFM Study of Potato Virus X Disassembly Induced byMovement Protein

Olga I. Kiselyova1,2, Igor V. Yaminsky1,2, Olga V. Karpova3

Nina P. Rodionova3, Stanislav V. Kozlovsky3, Marina V. Arkhipenko3

and Joseph G. Atabekov3*

1Faculty of Physics, Faculty ofChemistry of Moscow StateUniversity, Moscow 119992Russia

2Advanced Technologies CenterMoscow 119331, Russia

3Department of Virology ofMoscow State UniversityMoscow 119992, Russia

Recently we have reported that a selective binding of potato virus X(PVX)-coded movement protein (termed TGBp1 MP) to one end of apolar coat protein (CP) helix converted viral RNA into a translatableform and induced a linear destabilization of the whole helical particle.Here, the native PVX virions, RNase-treated (PVXRNA-DEG) helical particleslacking intact RNA and their complexes with TGBp1 (TGBp1–PVX andTGBp1–PVXRNA-DEG), were examined by atomic force microscopy (AFM).When complexes of the TGBp1 MP with PVX were examined by meansof AFM in liquid, no structural reorganization of PVX particles wasobserved. By contrast, the products of TGBp1-dependent PVX degra-dation termed “beads-on-string” were formed under conditions of AFMin air. The AFM images of PVXRNA-DEG were indistinguishable from imagesof native PVX particles; however, the TGBp1-dependent disassembly ofthe CP-helix was triggered when the TGBp1–PVXRNA-DEG complexeswere examined by AFM, regardless of the conditions used (in air or inliquid). Our data supported the idea that binding of TGBp1 to one end ofthe PVX CP-helix induced linear destabilization of the whole helicalparticle, which may lead to its disassembly under conditions of AFM.

q 2003 Elsevier Ltd. All rights reserved.

Keywords: atomic force microscopy; potato virus X; linear destabilization;movement protein*Corresponding author

Potato virus X (PVX) is a filamentous positive-strand RNA plant virus.1 About 1300 identicalprotein subunits in a filamentous PVX particle(modal length of about 515 nm, 13.5 nm indiameter) form a helical array (3.6 nm pitch) withthe 2.1 £ 106 Da RNA packed between the turns ofthe helix. There are 8.9 subunits per turn ofprimary helix,2 and five nucleotides are associatedwith each subunit. The molecular mass of PVXcoat protein (CP) calculated from the nucleotidesequence of the PVX CP gene3 is 25 kDa, although,using SDS-PAGE, a mass of about 29 kDa was cal-culated due to PVX CP anomalous electrophoreticbehaviour.4 Three partially overlapping genes inPVX RNA termed triple gene block (TGB) code forthree movement proteins (referred to as TGBp1,

TGBp2 and TGBp3 MPs) that are involved in cell-to-cell movement of PVX.5 We have recentlyreported that encapsidated PVX RNA was comple-tely non-translatable in a cell-free translation sys-tem. However, PVX particles could be convertedinto a fully translatable form after interaction withthe 25 kDa TGBp1 MP. TGBp1-dependent transla-tional PVX activation was rapid and most efficientat the molar TGBp1:PVX ratios from 25 to 100.However even at the calculated ratio as low as 1:1,the conversion of the PVX into a translatable formcould be detected after two hours (but not onehour) of translation.6 Experiments on binding ofthe 14C-labeled TGBp1 to PVX indicated that notmore than 60% of the TGBp1 preparation rep-resented the protein molecules capable of formingTGBp1–PVX complexes.6 The inability of 40% ofTGBp1 molecules to interact with PVX could bedue to improper folding of a certain portion ofbacterially expressed TGBp1 molecules in thecourse of purification. This indicates that the real

0022-2836/$ - see front matter q 2003 Elsevier Ltd. All rights reserved.

E-mail address of the corresponding author:atabekov@genebee.msu.su

Abbreviations used: PVX, potato virus X; TGB, triplegene block; AFM, atomic force microscopy.

doi:10.1016/S0022-2836(03)00835-0 J. Mol. Biol. (2003) 332, 321–325

molar TGBp1:PVX ratios were somewhat lowerthan the ratios calculated from the total concen-trations of TGBp1 preparations.

It was particularly noteworthy that the TGBp1molecules visualized by means of immunoelectronmicroscopy were bound to only one end of PVXvirions.6 This indicated that a certain domain ofterminal subunits exposed on only one of the twosurfaces of the polar helical PVX particles wasrecognized.

There is accumulating evidence that a selectivebinding of the TGBp1 to one extremity of the PVXCP-helix resulted in a linear destabilization of thewhole helical particle and its conversion in a statetermed “metastable” (our unpublished results).The TGBp1-mediated destabilization of the CP-helix was fully reversible and PVX particles couldbe reverted into a native (non-translatable) formby TGBp1 removal from the TGBp1–PVX complex.Binding of TGBp1 to the end of the PVX virion wasessential, but not sufficient for the CP-helix disas-sembly. The influence of an additional factor wasrequired to induce disassembly. Thus, no disas-sembly was observed, unless the TGBp1–PVXcomplex was translated, whereas rapid disassem-bly was triggered in a cell-free translation systemat an early stage of translation (our unpublishedresults). Similarly, no disassembly could bedetected, unless the complexes of TGBp1 withRNase-treated particles (PVXRNA-DEG) lacking intactRNA (TGBp1–PVXRNA-DEG) were centrifuged,whereas rapid disassembly occurred at accelera-tions of 14,000g and higher (our unpublishedresults). In the present work, the TGBp1-dependentdisassembly of the PVX CP-helix in the TGBp1–PVX and TGBp1–PVXRNA-DEG complexes was visu-alized by atomic force microscopy (AFM).

AFM is a high-resolution technique7 that hasbeen used in studies of biological macromolecules,in particular, for examination of mechanicalproperties and dimensions of virus particles8,9 andviral nucleoprotein complexes.10,11 In preliminaryexperiments, PVX virions were analyzed in air orin a liquid environment. The diameter of particles(about 13.5 nm) and their length distribution(major component of about 500 nm) determinedby AFM were in good agreement with the resultsof electron microscopy.1 Furthermore, the com-plexes of 25 kDa TGBp1 MP with native PVX wereexamined. It is well known that in AFM images,horizontal dimensions of objects are always overes-timated due to the effect of the geometry of theAFM tip, which has finite dimensions. As a conse-quence, the objects appear broader than their realdimensions.9 Therefore, the fine structure of thePVX CP-helix composed of individual 25 kDa sub-units was not resolved with AFM. Likewise, thepresence of the 25 kDa TGBp1 molecule(s) boundto one end of the PVX particle could not berevealed on AFM images (Figures 1 and 2). NoTGBp1-induced changes were detected in the geo-metry of PVX particles when complexes of TGBp1with native PVX formed at molar TGBp1: PVX

ratios varying from 10:1 to 100:1 were examined ina liquid environment (Figure 1). Similarly, no mor-phological distinctions were revealed by AFM inair when native PVX and TGBp1–PVX complexesformed at low (1:1 and 10:1) TGBp1:PVX ratioswere visualized in air (Figure 2a and b). By con-trast, dramatic changes in PVX structure weredetected by AFM in air when the complexes wereformed at the molar TGBp1:PVX ratios of 25:1 and100:1 (Figure 2c and d). The structures shown inFigure 2c and d, referred to as “beads-on-string”,represent the products of PVX particles degra-dation mediated by TGBp1 binding. The degra-dation was strictly TGBp1-dependent, since nosigns of the PVX disassembly could be detectedby AFM in the absence of TGBp1. Theseobservations correlated with our results onTGBp1-induced translational activation of PVXRNA showing that encapsidated PVX RNA couldbe converted into a translatable form after 60 min-utes incubation with TGBp1 at the molar TGBp1:PVX ratios of 25:1 and higher, rather than at 1:1and 10:1.6 Importantly, the length distribution of

Figure 1. AFM images of PVX TGBp1–PVX complexesin liquid; the complexes were formed at the molarTGBp1:PVX ratio of 100:1 (a); b, cross section madealong the line marked in a. For immobilization, freshlycleaved mica was treated with aminopropylsilatran(APS-mica)13 AFM measurements were made with aNanoscopee IIIa multimode scanning probe microscope(Digital Instruments, USA) in tapping mode. Forimaging in water, the tapping mode liquid cell (DigitalInstruments, USA) was mounted onto the sample priorto drying. Standard silicon nitride cantilevers(Nanoprobee) with a spring constant of 0.6 N/m wereused. The cantilever oscillation frequency was8–10 kHz. The scale bar represents 1000 nm.

322 MP-dependent PVX Disassembly

the beads-on-string particles (the modal length ofabout 500 nm) did not differ from that of PVXvirions. Figure 3 shows that positions of maximallengths are similar for PVX and TGBp1–PVXcomplexes, i.e. no elongation occurred. Thisobservation allows to suggest that detachment ofthe CP subunits from PVX particles occurs aftertheir adsorption and fixation on the substrate. Theprecise structure and the mechanism of theTGBp1-induced PVX conversion into a beads-on-string form remains obscure. Presumably, theglobules (“beads”) represent clusters of proteinsubunits bound to RNA and separated by protein-free RNA segments of varying length; however,RNA could not be visualized. The mean height of

beads-on-string particles was about half that ofnative PVX. Apparently, this conversion is trig-gered after the TGBp1–PVX complex has adheredto the substrate surface (APS-mica). One mayspeculate that the beads comprised the CP sub-units located at the side of the PVX CP helix thatadhered tightly to the substrate. It has beenevidenced9 that “AFM may be useful in assessingthe damage caused by various procedures” in avirus structure. In liquid environment, the TGBp1-dependent destabilization did not result in struc-tural reorganization of PVX particles (Figure 1),whereas the degradation of PVX particles involvedin TGBp1–PVX complexes was triggered underconditions of AFM in air (Figure 2c and d). The

Figure 2. AFM images of the TGBp1–PVX complexes formed at different TGBp1:PVX ratios. Bacterially expressed(His)6–TGBp1 MP and conditions for in vitro TGBp1–PVX complexes formation were the same as described earlier.6,14

Samples dried in air were imaged with 125 mm NanoSensor tapping mode etched silicon probes. The cantileveroscillation frequency was 300–350 kHz, the setpoint ratio (loaded versus free oscillations amplitude) was 0.9. TheTGBp1–PVX complexes were formed at the molar TGBp1:PVX ratios of: a, 1:1; b, 10:1; c, 25:1; and d, 100:1. The scalebars represent 500 nm.

MP-dependent PVX Disassembly 323

stability of TGBp1–PVX complexes observed inliquid may be due to the fact that the conditionsmost favorable for the biological specimen are pro-vided when AFM is performed in a liquidenvironment.9 Taking into account that no degra-dation could be observed in liquid (Figure 1), wesuggest that destabilization of PVX particlescaused by TGBp1 binding resulted subsequentlyin their conversion into beads-on-string particles,which occurred during dehydration of the speci-men on the surface of APS-mica. It is difficult toconceive that PVX destabilized by TGBp1 bindingmay be damaged directly by the mechanical forceapplied by the cantilever to the surface of the sample.

We have recently shown that, contrary to TMV,encapsidated PVX RNA was not protected fromRNase attack and that the bulk of RNA in RNase-

treated (PVXRNA-DEG) particles was represented byproducts shorter than five to six nucleotides (ourunpublished results). Electron microscopy andAFM (Figure 4a) indicated that PVXRNA-DEG par-ticles were morphologically similar to native PVX.However, it was not unexpected that, despite mor-phological integrity, degradation of the virion PVXRNA resulted in a considerable destabilization ofPVXRNA-DEG particles. It is significant that promi-nent TGBp1-dependent disassembly of TGBp1–PVXRNA-DEG complexes was observed under AFMconditions: instead of PVX-like particles presentedin Figure 4a, the structures shown in Figure 4bwere invariably revealed, indicating that entire dis-assembly of the CP helix occurred. It is noteworthythat disassembly of these preparations wasobtained both in air and in liquid environments.

Figure 3. Distribution of particle length for PVX (a) and TGBp1–PVX (b) complexes formed at the TGBp1:PVX ratioof 100:1. The number of particles measured for each histogram is 290.

Figure 4. AFM images of the PVXRNA-DEG particles and their degradation induced by binding of TGBp1 MP. a, RNase-treated PVXRNA-DEG particles lacking intact RNA. To obtain the PVXRNA-DEG particles, the purified virus was incubatedfor 25 minutes at room temperature with a mixture of RNase A (0.25 mg) and RNase T1 (two units) per 1 mg of virus;b, the major component (low molecular mass form of the PVX CP) revealed after PVXRNA-DEG incubation with TGBp1(at the molar TGBp1: PVXRNA-DEG ratio of 100:1). The scale bars represent 500 nm.

324 MP-dependent PVX Disassembly

Only a negligible proportion (if any) of the beads-on-string type particles could be found in TGBp1–PVXRNA-DEG preparations. Three components havebeen identified previously by sedimentationanalysis of CP preparations isolated from PVX,including the smallest component with sedimen-tation coefficient of about 1.8–2.8 S correspondingto a monomer or dimer of CP subunits and twostable intermediate oligomeric aggregates withs20,v of 3–5 S and 10–15 S.12 The globular particlesproduced upon degradation of TGBp1–PVXRNA-DEG

complexes (Figure 4b) may represent the PVX CPmonomers and their aggregates. The measuredheights, and, therefore, the diameters of globularparticles, were not uniform and varied from 3.5 nmto 20.0 nm. Possibly, the smaller particles (3.5–4.5 nm in height) corresponded to monomers andthe bigger ones represented 3–5 S, 10–15 S and lar-ger aggregates. Taken together, our results suggestedthat binding of TGBp1 molecule(s) to one end of thePVX or PVXRNA-DEG particles6 induced a linear desta-bilization of the whole CP-helix, although, PVX par-ticles did not undergo disassembly as a result ofinteraction with TGBp1 per se. No disassembly wasdetected when TGBp1–PVX complexes were visual-ized by AFM in a liquid environment, whereas entiredissociation of the TGBp1–PVXRNA-DEG particles wasrevealed by AFM both in liquid and in air. On theother hand, only partial disassembly of PVX virion(beads-on-string type particle formation) wasdetected by AFM in air after formation of theTGBp1–PVX complexes. This difference could bedue to the presence of intact RNA molecules in PVXparticles bound to TGBp1. Collectively, the data pre-sented here and not shown (our unpublished results)support the hypothesis that binding of 25 kDaTGBp1 MP to one end of the PVX CP-helix6 resultsin linear destabilization of the whole filamentousparticle, presuming that intersubunit conformationalchanges may be transferred from one end of aparticle along the CP-helix as a “wire”.

Acknowledgements

This work was supported, in part, by the Rus-sian Foundation for Basic Research (grants 00-04-55020 and 03-04-48833), INTAS 01-0045 and by theFogarty International Center grant R03 TW01239-01.

References

1. Tollin, P. & Wilson, H. R. (1988). Particle structure. In

The Plant Viruses (Milne, R. C., ed.), vol. 4, pp. 51–83,Plenum Press, New York.

2. Parker, L., Kendall, A. & Stubbs, G. (2002). Surfacefeatures of potato virus X from fiber difractiom.Virology, 300, 291–295.

3. Skryabin, K. G., Morozov, S. Yu., Kraev, A. S.,Rozanov, M. N., Chernov, B. K., Lukasheva, L. I. &Atabekov, J. G. (1988). Conserved and variableelements in RNA genomes of potexviruses. FEBSLetters, 240, 33–40.

4. Koenig, R., Stegemann, M. E., Francksen, H. & Paul,H. L. (1970). Protein subunits of potato virus Xgroup. Determination or the molecular weights bypolyacrylamide electrophoresis. Biochim. Biophys.Acta, 207, 184–189.

5. Chapman, S. N., Hills, G., Watts, J. & Baulcombe,D. C. (1992). Mutational analysis of coat proteingene of potato virus X. Effects on virion morphologyand viral pathogenicity. Virology, 191, 223–230.

6. Atabekov, J. G., Rodionova, N. P., Karpova, O. V.,Kozlovsky, S. V. & Poljakov, V. Yu. (2000). The move-ment protein-triggered in situ conversion of potatovirus X virion RNA from a nontranslatable into atranslatable form. Virology, 271, 259–263.

7. Binnig, G., Quate, C. F. & Gerber, Ch. (1986). Atomicforce microscope. Phys. Rev. Letters, 56, 930–933.

8. Drygin, Yu. F., Bordunova, O. A., Gallyamov, M. O.& Yaminsky, I. V. (1998). Atomic force microscopyexamination of TMV and virion RNA. FEBS Letters,425, 217–221.

9. Kuznetsov, Yu. G., Malkin, A. J., Lucas, R. W., Plomp,M. & McPherson, A. (2001). Imaging of viruses byatomic force microscopy. J. Gen. Virol. 82, 2025–2034.

10. Kiselyova, O. I., Yaminsky, I. V., Karger, E. M.,Frolova, O. Yu., Dorokhov, Yu. L. & Atabekov, J. G.(2001). Visualization by atomic force microscopyof tobacco mosaic virus movement protein–RNAcomplexes formed in vitro. J. Gen. Virol. 82,1503–1508.

11. Nurkiyanova, K. M., Ryabov, E. V., Kalinina, N. O.,Fan, Y., Andreev, I., Fitzgerald, A. G. et al. (2001).Umbravirus-encoded movement protein inducestubule formation on the surface of protoplasts andbinds RNA incompletely and non-cooperatively.J. Gen. Virol. 82, 2579–2588.

12. Kaftanova, A. S., Kiselev, A. N., Novikov, V. K. &Atabekov, J. G. (1975). Structure of products of pro-tein reassembly and reconstitution of potato virus X.Virology, 65, 283–287.

13. Shlyakhtenko, L. S., Potaman, V. N., Sinden, R. R. &Lyubchenko, Yu. L. (1998). Structure and dynamicsof supercoil-stabilized DNA crusiforms. J. Mol. Biol.280, 61–72.

14. Karpova, O. V., Ivanov, K. I., Rodionova, N. P.,Dorokhov, Yu. L. & Atabekov, J. G. (1997). Non-translatlbility and dissimilar behavior in plants andprotoplasts of viral RNA and movement proteincomplexes formed in vitro. Virology, 230, 11–21.

Edited by M. Moody

(Received 30 January 2003; received in revised form 2 June 2003; accepted 23 June 2003)

MP-dependent PVX Disassembly 325