N-Terminal segment of potato virus X coat protein subunits is glycosylated and mediates formation of a bound water shell on the virion surface Lyudmila A. Baratova 1 , Nataliya V. Fedorova 1 , Eugenie N. Dobrov 1 , Elena V. Lukashina 1 , Andrey N. Kharlanov 2 , Vitaly V. Nasonov 3 , Marina V. Serebryakova 4 , Stanislav V. Kozlovsky 1 , Olga V. Zayakina 1 and Nina P. Rodionova 1 1 A. N. Belozersky Institute of Physico-Chemical Biology, Moscow State University, Russia; 2 Department of Chemistry, Moscow State University, Russia; 3 M. M. Shemyakin and Yu. A. Ovchinnikov Institute of Bioorganic Chemistry, Russian Academy of Sciences, Moscow, Russia; 4 V. N. Orekhovich Institute of Biomedical Chemistry, Russian Academy of Medical Sciences, Moscow, Russia The primary structures of N-terminal 19-mer peptides, released by limited trypsin treatment of coat protein (CP) subunits in intact virions of three potato virus X (PVX) isolates, were analyzed. Two wild-type PVX strains, Russian (Ru) and British (UK3), were used and also the ST mutant of UK3 in which all 12 serine and threonine residues in the CP N-terminal segment were replaced by glycine or alanine. With the help of direct carbohydrate analysis a nd MS, it w as found that the acetylated N-terminal peptides of both wild- type strains are glycosylated by a single monosaccharide residue (galactose or fuco se) at NAcSer in the first position of the CP sequence, whereas the acetylated N-termin al seg- ment of the ST mutant CP is unglycosylated. Fourier transform infrared spectra in the 1000–4000 cm )1 region were m easured for fi lms o f the intact and in situ trypsin- degraded PVX preparations at low a nd high humidity. These spectra revealed the presence of a broad-ban d in the region of valent vibrations of OH bonds (3100–3700 cm )1 ), which can be represented b y superposition of three bands corresponding to tightly bound, weakly bound, and free OH groups. On calculating difference (ÔwetÕ minus ÔdryÕ) spectra, it was f ound that the intact w ild-type PVX virions are characterized by high water-absorbing capacity and the ability to o rder a large number of water molecules on the virus particle. This effect w as much weaker for the ST mutant and completely absent in the trypsin-treated PVX. It is proposed that the s urface-located and glycosylated N-terminal CP segments of intact PVX virions induce the formation of a columnar-type shell from bound water molecules around the virions, w hich probably play a major role in maintaining the virion surface structure. Keywords: bound water; coat protein; Fourier transform infrared spectroscopy; glycosylation; potato virus X. Potato virus X (PVX) is the type member of the potexvirus group of filamentous plant viruses [1]. Its coat protein (CP) was extensively studied and it h as been shown that the CP participation in the PVX infection cycle is not limited to its role in virion formation. The PVX CP has b een shown to be involved in processes of ge nomic RNA accumulation and infection transport in plants [2–4]. It is also r esponsible for induction of the R x resistance system in potato plants [5,6], and has been recently shown to play a m ajor part in regulation of virion translational activity at different stages of the infection process [7–10]. These (and many others) studies demonstrate the importance of potexvirus CP at all stages of virus–plant interactions. Thus, the question arises which features of t he CP structure d etermine the different kinds of its activity. It is well known t hat, on SDS/PAGE, i ntact P VX CP displays anomalously slow electrophoretic mobility. This mobility corresponds for different PVX strains to molecular masses of 27–29 kDa, instead of 25 kDa as determined from the primary structure [11,12]. In 1994 Tozzini et al. [13] found that PVX CP contains O-linked carbohydrates. It was the first report on the presence of glycosyl residues in coat proteins of plant viruses. However, the exact nature of the g lycosyl residues, their location in the PVX CP sequence, and their functional importance remain unknown. Thus, a more detailed analysis is needed, all the more so, as it is now generally accepted that glycosylation can induce s ignificant alterations in biopolymer structure and flexibility [14]. High-resolution X-ray (fiber) diffraction data for potex- viruses are not available and only one, more or less detailed, model of P VX CP structure in the virion (based on tritium planigraphy and s econdary-structure prediction data) has been suggested [15]. The N-terminal region of potexvirus CP is surface- located [15,16], highly sensitive to the action o f plant sap proteases [12], and can be easily removed by mild trypsin treatment without disruption of the virion structure [17]. To Correspondence to L. A. Baratova, Department of Chromatography, A.N. Belozersky Institute of Physico-Chemical Biology, Moscow State University, Moscow 119992, Russia. Fax: + 7095 9393181, Tel.: + 7095 9395408, E-mail: baratova@belozersky.msu.ru Abbreviations: CP, coat protein; CPY, carboxypeptidase Y; FTIR, Fourier transform infrared; PVX, potato virus X. (Received 24 January 2004, revised 15 May 2004, accepted 3 June 2004) Eur. J. Biochem. 271, 3136–3145 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04243.x elucidate the role of the PVX CP N-terminal region in determining the physicochemical properties of the virion surface, preservation of the virion integrity, and the anomalous electrophoretic mobility, we used a special PVX CP mutant. In this mutant (designated ST) all serine and threonine residues (potential glycosylation s ites) in the N-terminal segment are substituted by glycine or alanine residues [7]. The aim of this work was t o determine the presence and nature of carbohydrate residues in the peptides removed by a limited trypsin hydrolysis from the intact PVX virions of two wild-type s trains (Russian and British) a nd the S T mutant. The N-terminal amino-acid sequences of these three PVX isolate CPs are shown in Fig. 1 [7,18,19]. It is widely known that many polyoxy molecules (including proteins) exist in water solution in a hydrated state, and the presence of carbohydrate residues may greatly increase the bound water content [20]. To estimate bound water content in the i ntact a nd trypsin-treated P VX virions, w e u sed Fourier transform infrared (FTIR) spectroscopy. A possible role for the surface-located and glycosylated PVX CP N-terminal segment in preserving the structural and func- tional integrity of the PVX virion s is discussed. Materials and methods Reagents and chemicals Trypsin treated with 1-chloro-4-phenyl-3- L -toluene-p-sulfo- namidobutan-2-one (TPCK-trypsin) was obtained from Sigma. Trifluoroacetic acid was from Perkin-Elmer. Aceto- nitrile was from Criochrom (St Petersburg, Russ ia). Rea- gents for preparing gels were from Bio-Rad Laboratories. Water was obtained using a Milli-Q System (Millipore). All other chemicals were analytical grade. For carbohydrate analysis, only f reshly prepared reagents thrice-distilled in quartz glassware water were used. Meth- anol was additionally purified by distillation over magnes- ium methylate, and ethanol by distillation over calcium oxide. Pyridine was d istilled twice over sodium hydroxide and once over barium oxide. HCl and trifluoroacetic acid were additionally purified by distillation in borosilicate glassware. Virus preparations The R ussian (Ru) and British (UK3) strains of PVX were purified from systemically infected leaves of Nicotiana bentham iana and Datura stramonium as described previ- ously [9]. The ST mutant of the UK3 strain with the wild- type N-terminal sequence, SAPASTTQPIGSTTSTTTKT, changed to AAPAGGAQPIGAGGAAGAKA was obtained as described by Kozlovsky et al.[7]. Limited tryptic digestion of the virus preparations Limited tryptic digestion of PVX virions (500 lgvirus preparation in 0.5–1 mL 0.025 M Tris/HCl buffer, pH 8.0) was carried out at an enzyme/substrate ratio of 1 : 500 (w/w) for 2 h at 37 °Cin0.2 M Tris/HCl buffer, pH 8.0. In the case of the ST mutant, the tryptic digestion was carried out at an enzyme/substrate ratio of 1 : 2000 (w/w). Then the virus particles were pelleted by high-speed centrifugation (50.2Ti rotor, Beckman L5-50; 105 000 g,4°C, 1.5 h), and pellets and p eptide-containing supernatants were used for further analysis. SDS/PAGE SDS/PAGE (8–20% gels) was carried out essentially a s described by Laemmli [21]. The protein bands were visual- ized by staining with Coomassie Brilliant Blue (G-250). HPLC equipment and conditions HPLC analyses were performed on a narrow-bore column (Milichrom A-02; EnviroChrom LC, Chromatography Institute ECONOVA, Novosibirsk, Russia; 75 · 2 mm) packed with 5-lm p articles of Nucleosil C 18 , pore size 120 A ˚ (Macherey-Nagel, Duren, Germany). Separations were performed at 25 °C; a dual wavelength (214 nm and 280 nm) detector was used. The elution gradient profile was as follows. The elution solvents were A (0.1% trifluoroacetic acid in water, pH 2.2) and B (acetonitrile with 0.1% trifluoroacetic acid). The linear gradient was 0–60% B in 60 min and then 60–80% B in 10 min; the flow rate was 8 0 lLÆmin )1 . Fractions were collected for s ubse- quent analysis using a Gilson 201 fraction collector. Peptide yields were 30–50%. The conditions used allowed separ- ation of all full-size PVX CP tryptic peptides. Identification of carbohydrates in tryptic peptides The m ethod for d etermining the monosaccharide c ompo- sition of glycoconjugates (glycopeptides and glycoproteins) involved derivatization of the monosaccharides, released on acid hydrolysis, into N-(4-methylcoumarin-7-yl)glycamines (AMC-sugars) and their subsequent analysis by reverse- phase (RP) HPLC with fluorimetric detection [22]. The authentic AMC-sugars were prepared by reductive N-alkylation of 7-amino-4-methylcoumarin with the f ol- lowing monosaccharides in the presence o f N aCNBH 3 : D -Glc, D -Gal, D -Man, L -Fuc, D -GlcNAc, D -GalNAc, D -ManNAc. HPLC analyses were performed on a Du Pont 8800 chromatograph equipped with fluorescence detector . Columns with Ultrasphere ODS (Beckman; 250 · 4.6 mm internal diameter) w ere used. AMC-sugars were separated at 25 °C using 17.5% ethanol in water with 0.1% trifluoro- acetic acid, pH 2.5–2.6. The flow rate was 0.75 mLÆmin )1 . Carboxypeptidase Y (CPY) hydrolysis of N-terminal PVX CP tryptic peptide For CPY hydrolysis, peptide-containing solutions were dried to eliminate trifluoroacetic acid and dissolved in water Fig. 1. N-Ter minal amino-acid sequences of the coat proteins of Russian (Ru) and British (UK3) PVX strains and the ST m utant of UK3 PVX. Ó FEBS 2004 N-Terminal glycosylation of potato virus X protein (Eur. J. Biochem. 271) 3137 to a concentration of  0.1 mgÆmL )1 .CPY(Sigma)was added to this solution to a n enzyme/peptide ratio of 1 : 5 (w/w). Hydrolysis was c arried out at 22 °C. To obtain MALDImassspectra,0.5-lL samples were removed at different times from the hydrolysis start. Mass spectrometry Mass spectra were obtained using a MALDI-TOF mass spectrometer (Reflex III model; Bruker Analytic GmbH, Bremen, Germany) with 337 nm UV laser. A study sample solution (0.5 lL) was mixed with an equal volume of 2,5-dihydroxybenzoic acid (Aldrich; 10 mgÆmL )1 in 20% acetonitrile in water with 0.1% trifluoroacetic acid), and the mixture was dried in the air. Mass spectra of material from HPLC fractions and CPY hydrolysates of NAcSAPAS-peptide were obtained in reflectron mode with positive ion detection (mass peak accuracy 0.015%). Mass spectra of CPY h ydrolysates of the N-terminal PVX C P t ryptic peptide and its C-terminal fragment were obtained in reflectron mode with negative ion detection (peak accuracy 0.02%). FTIR spectroscopy IR spectra were acquired with a FTIR spectromete r Equinox 55/S (Bruker Analytic GmbH, Bremen, Germany) in the wavenumber range 1000–4000 cm )1 . To obtain thin films, 10 lL of t he virus s uspensions ( 6mgÆmL )1 )in 20 m M Tris/HCl buffer, pH 7.8, were applied to B aF 2 plates and dried in a vacuum desiccator (at 0.13 Pa) over P 2 O 5 at 20 °C, with visual control o f the film homogeneity. T o create 100% humidity, distilled water was placed at the bottom of the desiccator. Analytical methods Tryptic peptides were hydrolyzed as described by Tsugita & Scheffler [23], and amino-acid analysis was carried out on a Hitachi-835 analyzer (Tokyo, Japan) in the standard mode for protein hydrolysate analysis with cation-exchange separation and ninhydrin postcolumn derivatization. The short N-terminal amino-acid sequences of the tryptic peptides were determined by Edman degradation on the automated Procise cLC Protein Sequencing System (model 491; PE Applied Biosystems). Phenylthiohydantoin deriva- tives of amino acids were identified with the PTH Analyzer (model 120A; PE Applied Biosystems). Results Electrophoretic analysis of wild-type and mutant PVX CP On SDS/PAGE of PVX preparations, it was found that the intact wild-type and ST mutant CPs differ i n their electrophoretic mobility, and this difference disappears after mild trypsin treatment resulting in r emoval of the N-terminal CP peptide (Fig. 2). T his indicates t hat the anomalous PVX CP electrophoretic mobility is determined by the N-terminal segment. The relatively large difference in electrophoretic mobili- ties between t he intact wild-type a nd ST mutant CPs can hardly be explained by minor differences in m olecular mass and may be due to the absence of certain post-translational modifications in the mutant protein (see below). HPLC analysis of the PVX CP peptides released on limited trypsin hydrolysis After the trypsin treatment, PVX virions were pelleted by high-speed centrifugation, and the peptide-containing sup- ernatants were subjected to RP-HPLC. The surface location of the N-terminal CP peptide in PVX virions [15,16] suggests that the peptides released from the PVX virions on mild trypsin hydrolysis correspond to the N-terminal regions of the PVX CP subunits. As can be seen in Fig. 3, the supernatant f rom t he ST mutant contains one major peptide fraction and the supernatants from the UK3 and Ru strains contain two. [The differences in peak mobilities for the UK3 and Ru peptides are probably due to differences in their amino-acid seque nces: the former has p roline and isoleucine at positions 9 and 10, respectively, and the latter has alanine and threon ine (Fig. 1).] The elution of the ST mutant peptide a t h igher a cetonitrile concentrations than the UK3 peptide may be due to the substitution of 11 serine and threonine residues with g lycine and alanine, which would increase the peptide hydrophobicity. The minor amounts of other peptides observed in the chromatographic profiles in Fig. 3 were not analysed further. Determination of primary structure of trypsin-cleaved peptides The HPLC-purified peptide fractions were used for a mino- acid analysis, microsequencing and MALDI MS. The only peptide released on mild trypsinolysis of the ST mutant (Fig. 3 C) was the acetylated N-terminal peptide w ith a molecular mass of 1535 Da (Table 1), corresponding exactly t o the calculat ed mass of the first 19 amino acids of the ST mutant CP with a n acetyl group. The trypsin- cleaved peptides o f the wild-type (Ru and UK3) PVX strains each gave two peaks (1 and 2) on RP-HPLC (Fig. 3 A,B). For both viruses, the amino-acid compositions of the material in the two peaks were identical and coincided with that predicted for the corresponding N-terminal (19 residues) sequence (Fig. 1). Microsequencing results were negative for all four peaks, confirming that material in all Fig. 2. Anal ysis of wild-type (UK3) PVX and ST mutant preparations by SDS/PAGE (8–20% gels). Lanes1and2,UK3PVX;lanes3and4, ST mutant; lanes 1 and 3, CPs from intact virions; lanes 2 and 4, CPs from trypsin-treated virions. 3138 L. A. Baratova et al.(Eur. J. Biochem. 271) Ó FEBS 2004 the peaks was N-blocked, in a ccordance with our p revious results [15]. Possible reasons for the differences in c hroma- tographic mobility of peak 1 and peak 2 material were revealed by MS (Table 1). The molecular m asses for both Ru and UK3 peptides turned ou t to be significantly higher than expected. In peak 1 (of both strains) this difference was 162 Da, corresponding exactly to the addition of a hexose residue. In peak 2 (again for both viruses), a more complex picture was observed. Some material had the same mass as in peak 1, but there was also material with a molecular mass that differed from that expected by 146 Da (Table 1). This may correspond to the addition of a deoxyhexose residue. No peaks corresponding to more than a singly glycosylated peptide (i.e. for instance, 2041 + 162 or 2041 + 146 in the case of the UK3 strain peak 1) were detected in the mass spectrograms. These r esults forced us to turn to direct carbohydrate analysis. Chromatographic analysis of the PVX CP N-terminal peptide carbohydrates Direct determination of carbohydrates in the P VX CP N-terminal peptides involved acid hydrolysis, derivatization of released monosaccharides to AMC-sugars, and identifi- cation of derivatives by RP-HPLC with fluorescence detection [22]. This analysis re vealed (for both Ru a nd UK3 strains) the presence of a galactose residue (the additional 162 Da) in t he material of peak 1 (Fig. 3A,B) and both galactose and fucose (the a dditional 1 46 Da) residues in peak 2 (th e data for the Ru strain are shown in Fig. 4). These results correlated closely with the r esults of MS (Table 1). No carbohydrates were found in the CP N-terminal peptide of t he ST mutant, also in accordance with the MS. Thus, from the results of both MS and direct carbo- hydrate determination it follows that the surface-located N-terminal CP segments of the two wild-type PVX strains studied are glycosylated and contain a sugar residue (galactose or fucose) O-linked to serine or threonine residues, which are absent from the N-terminal CP peptide of the ST mutant. The presence of galactose in both peak 1 and peak 2 may be explained by the glycoconjugate stereochemistry. It is known that carbohydrate diastereomers differ in their physicochemical properties, and therefore they may have Fig. 3. RP-HPLC separation of PVX CP preparations after mild trypsin hydrolysis of intact virions. (A) Russian strain (Ru); (B) British strain (UK3); (C) ST m utant. Table 1. R esults of MS analysis of trypsin-cleaved PVX CP N-terminal fragments. PVX strain Peptide fraction number (Fig. 3) Peptide molecular mass (Da) Observed Calculated Difference Russian (Ru) 1 2003 1841 162 2 1987 146 2003 162 British (UK3) 1 2041 1879 162 2 2025 146 2041 162 ST mutant 1 1535 1535 – Ó FEBS 2004 N-Terminal glycosylation of potato virus X protein (Eur. J. Biochem. 271) 3139 different chromatographic mobility on highly selective sorbents. The question a rises: are a ll CP subunits in the P VX virions N-terminally glycosylated? On one hand, we found n o o ther major peaks on HPLC other than peaks 1 and 2 (Fig. 3A,B), which indicates that the vast majority of the 1 300 CP molecu les in t he PVX v irion are glycosylated. On the other, on m ass spectrograms of tryptic h ydrolysates of t he isolate d full-size UK 3 a nd Ru CPs some material with molecular masses corresponding to unglycosylated peptides (1879 and 1841 Da, respect- ively) could be s een (data not shown). T his may mean that PVX v irions contain a proportion (not more than 10%) of CP molecules that a re not N-terminally glycosy- lated, assum ing that it is not the r esult of partial peptide deglycosylation in t he course of MS . We d id not observe any peaks corresponding to deacetylated PVX CP N-terminal peptides. Identification of glycosylation site(s) in the PVX CP N-terminal segment To locate glycosylation site(s) in the N-terminal segment of PVX CP, we obtained spectra of MS fragmentation of material from chromatographic peaks 1 and 2, prepared by mild trypsin treatment of UK3 PVX virions (Fig. 3). Standard MS fragmentation methods should lead to formation o f i ons of the p eptide C-terminal fragments, because the dominant protonation site (Lys19) is located at the C-terminus. In MALDI postsource decay spectra [24], we ob served only masses corresponding to ions of C-terminal fragments (y-ions) of our peptide without carbohydrate residues. On the basis of these results, i t may be suggested that the glycosylation site is located on the N-terminal serine of the peptide. However, it is well known that, on MALDI fragmentati on, intensive deglycosyla- tion takes place [25], making unequivocal conclusions Fig. 4. RP-H PLC analysis of carbohydrate content of the Ru strain CP N-terminal peptide. Monosaccharides were analyzed as AMC derivatives. (A) Analysis of a blank sample ( eluate fraction between peaks in chromatograp hic profile shown in Fig. 3). (B) Analysis of a stan dard mixture containing Glc, Gal, Man, Fuc, GlcNAc, GalNAc, ManNAc. (C) Analysis of peak 1 (Fig. 3A). (D) A nalysis of peak 2 (Fig. 3A). 3140 L. A. Baratova et al.(Eur. J. Biochem. 271) Ó FEBS 2004 impossible. MS s equencing using electr ospray ionization also resulted in deglycosylation. Therefore we decided to obtain a series of MALDI spectra of PVX CP N -terminal tryptic peptides shortened from the C-terminus by CPY hydrolysis [26]. For these experiments, we modified the procedure for preparing the N-terminal PVX CP segment: time, temperature and trypsin/protein r atio were kept the same, but the PVX virion (and enzyme) concentration w as increased about fivefold. On HPLC of the p eptide-containing sample, three major p eaks were eluted before peaks 1 and 2 (Fig. 5A). Previously, we sometimes obtained similar chromato- graphic profiles but did not analyze them further supposing that they resulted from overhydrolysis. However, this time we performed MALDI MS analysis of all five peaks eluted at the beginning of the chromatogram (peaks 1*, 2*, 3*, 1 and 2 in Fig. 5A). For peaks 1 and 2, molecular masses of 2025 and 2041 Da were obtained as before (Table 1). The molecular mass of peak 3* was 1424 Da, corresponding to the unglycosylated C-terminal part of the PVX CP N-terminal tryptic peptide (Thr6–Lys19; Fig. 1). Figure 5B shows a combined MALDI mass spectrum of peptides from p eaks 3*, 1 and 2 after partial CPY hydrolysis. The observed mass values of fragments p roduced by seq uential removal of the C-terminal residues (up to Ile10) from the initial tryptic peptide confirm the localization of a carbohydrate residue in the nonfragmented N-terminal part of the analyzed CP peptide. Moreover, the mass spectrum of peak 3* material treated with CPY confirmed our suggestion about its Fig. 5. Stage s of glycosylation site identification. (A) Initial part of RP-HPLC profile of PVX preparation after modification of the trypsin treatment procedure (see text). (B) Combined MALDI mass spectrum of material from peaks 3*, 1 and 2 after partial CPY hydrolysis. (C) M ALDI mass spectra of p eak 2* mater ial before (upper part) and after (lower part) CPY treatment (642 Da, sodium salt of deoxyhexose-containing peptide; 658 Da, potassium salt of the same peptide and/or sodium salt of hexose-containing peptide; 674 Da, potassium salt of hexose-containing peptide; additional peaks in lower spectrum correspond to salts from CPY solution). Ó FEBS 2004 N-Terminal glycosylation of potato virus X protein (Eur. J. Biochem. 271) 3141 primary structure. Thus, carbohydrate residues could be linked only to S er1 o r Ser5 of o ur peptide N-terminal pentameric part NAcSAPAS Peaks 1* and 2* were shown to contain only these modified N -terminal p entapeptides: in peak 1 *, hexose was linked to the peptide, and in peak 2* (just as in peak 2) either hexose or deoxyhexose. With the help of CPY hydrolysis, we managed to unambiguously localize t he glycosylation site in the PVX CP N-terminal segment. In Fig. 5C, MALDI mass spectra of peak 2* material before and after CPY treatment are s hown (upper and lower parts, respect- ively). Mass values in the upper p art correspond to ions of sodium and po tassium salts of the glycosylated peptide NAcSAPAS. After CPY treatment, the unglycosylated C-terminal serine was r eleased (lower p art of F ig. 5C), confirming our suggestion that a hexose or deoxyhexose residue is linked to the acetylated N-terminal serine residue (NAcSer1) of the PVX C P. As far as w e know, t his type o f glycosylation has not previously been observed in plant virus proteins. FTIR spectroscopy of wild-type and mutant virus preparations Figure 6 shows FTIR spectra in the 1000–4000 cm )1 region for films of t he intact and trypsin-treated Ru PVX and the intact ST mutant preparations. To compare the water- absorbing capacity of the different PVX variants, we measured their FTIR spectra in dry films and after saturation of the film with water at 100% relative humidity. Infrared band assignments were taken from Parker’s book [27]. In t he FTIR spectra, peaks cor responding t o valent CH-bond vibrations ( 2900 cm )1 ), valent C¼O bond vibrations (amide I,  1650 cm )1 ), defo rmational amide group vibrations (amide II,  1550 cm )1 , and amide III,  1300 cm )1 ) can be seen (Fig. 6). A broad-band in the 3100–3700 cm )1 region corresponds to superposition of several absorption p eaks. Here a complex band o f valent water OH-bond vibrations is overlapped with p eaks of valent OH-bond and NH-bond vibr ations of PVX CP. Thus, to estimate the state of water molecules in the virus preparations, we calculated FTIR difference spectra (the samples at 100% humidity minus dry samples). The broad 3100–3700 cm )1 band in the difference spectra (Fig. 7 ) can be represented by s uperposition o f three bands corresponding to absorption of OH bonds in tightly bound OH groups (the band a t  3240 cm )1 ), in weakly bound OH groups ( 3440 cm )1 ) and in free OH groups of absorbedwatermolecules( 3600 cm )1 )[27].Fromthe data in Fig. 7, it can be seen that the intact wild-type P VX differs from the other virus preparations by a greatly increased water-absorbing capacity. W hat is e ven more important, on s aturation of the wild-type PVX film with water, an intense 3240 cm )1 band was observed in the difference spectrum, supporting the o rdering o f a large number of water mole cules on the virus particle. This effe ct was much weaker for the ST mutant and completely absent in the case of the trypsin-treated P VX. Fig. 6. FTI R absorbance spectra of the intact and trypsin-treated Ru PVX and of the intact ST mutant preparations in the 1000–4000 cm -1 region, for dry films and after saturation of the film with water at 100% relative humidity. 3142 L. A. Baratova et al.(Eur. J. Biochem. 271) Ó FEBS 2004 Discussion The role of CP glycosylation in plant virus life cycles remains unclear, although data o n the effects of glycosyla- tion on the conformation and dynamics of O-linked glycoproteins are accumulating [28–30]. First, we were interested in the presence of glycosyl modification(s) in the t rypsin-cleaved PVX CP N-terminal segment. Structural analysis of the peptides cleaved from the wild-type protein revealed the presence of two monosac- charides, galactose and fucose, alternatively linked to NAcSer in the first position of the CP sequence. It was also shown that almost all PVX CP subunits in the wild-type PVX virion contain sugar residues in their N-terminal peptides. Small amounts of unglycosylated CP in the virus samples may originate from the virio n ends, where t he subunit conformation may be unsuitable f or glycosylation. What is the importance of this CP N-terminal g lycosy- lation for PVX virion structure and function? According to our model [15], the PVX CP N-terminal segment is located on the v irion surface and forms a super-secondary structure, a three-stranded b-sheet. We presume that removal of this structure leads to drastic changes in the physicochemical properties of the PVX virion surface. It is widely known that water in close proximity to the protein surface is fundamental to protein f olding, stability, recognition and activity, and thus understanding protein hydration is crucial in unraveling protein functions [31]. Virus particles with their unique highly regular morphology may be especially interesting in this respect. Different types of macromolecule s tructures a re known to have different types of hydration she ll: columnar or sheet-like. Cylindrical structures (which helical viruses have) usually induce the columnar type of hydration [20]. The results of our comparative FTIR spectroscopy study of the intact wild-type and ST mutant PVX virions and the trypsin-treated wild-type virus particles suggest the pres- ence of a large number of ordered water molecules ( water shell) around intact wild-type v irions. Thus, a PVX virion may be considered as an electric c able with several layers of insulation. The virus RNA is packed into t he ordered helical protein shell. The surface-located CP subunit N-terminal segments with their fixed super-secondary structure of three b-strands and glycosyl residues linked to the N-terminal serines form the next ordered layer of the cable structure. The su rface layer of the cable is formed by a shell consisting of ordered water molecules. NMR and ESR data show [32] that two water molecules are usually bound per sugar ring, i.e. about 2600 molecules per PVX virion. Besides the presence of sugar residues, the wild-type PVX CP N-terminal segment is also characterized by exception- ally h igh content of hydroxyl-containing amino acids (11 serine and threonine residues in the 19-residue sequence). T his, as well as the predicted existence of a three-stranded b structure in t his segment [15], may facilitate tightly bound water shell formation. Moreover the presence of deep grooves on the surface of helical PVX particles was recently demonstrated by fiber X-ray diffraction analysis [33], an d, a s shown by Falconi et al. [31], water hydration sites are mainly located around protein cavities and clefts. Raman optical activity spectra also indicate a high degree of hydration in PVX virions [34]. Fig. 7. FTIR difference (‘wet’ minus ‘dry’) spectra of the intact and trypsin-treated Ru PVX and the intact ST mutant preparations in the 2500– 4000 cm )1 region. Ó FEBS 2004 N-Terminal glycosylation of potato virus X protein (Eur. J. Biochem. 271) 3143 The PVX virions without N-terminal CP peptide do not simply lose the surface layer of water molecules; this loss would lead to drastic changes in physicochemical properties of the virion s urface. We propose that the absence of t he outer ordered water layer explains the greater sensitivity to trypsin of the ST mutant virions compared with the wild- type PVX particles [7]. Other data [2,17] support our suggestion of critical changes in PVX virions on cleavage (or changing) of the CP N-terminal peptide. In their 1972 electron microscopy study, Tremaine & Agrawal [17] observed unusual twisting of trypsin-treated PVX particles. In 1992 Chapman et al.[2] reported that a PVX CP deletion mutant, devoid of 19 N-terminal amino-acid residues, produces virions with abnormal morphology. We suggest that the presence of an ordered water column around PVX virions, by itself and/or through formation of a water-mediated net of hydrogen bonds between (or inside) CP subunits, strongly affects the structure and properties o f the externally lo cated region s of the virus protein coat. However, it cannot be excluded that the structu re of internally located parts of the virion protein subunits and the structure of t he vir ion its elf are affected by changes i n the state of the virion water shell. and this state may be altered by changes in glycosylation, phosphorylation and other t ypes of modification of t he externally located virus CP regions. In this way the structure of a whole virion may be changed by a signal molecule binding to the virion surface. Recent observations [7–10] that the ST mutant intravirus RNA, in contrast with that o f the wild-type PVX, is accessible to ribosomes in the intact virions and can be effectively t ranslated in cell-free systems may be an example of such a structural transition. We plan to continue to study structural alterations in the PVX CP subunits in virions induced by N-terminal segment modification or c leavage. Acknowledgements We thank Professor J. G. Atab ekov for valuable discussions. The work was partially supported by Russian Foundation for Basic Rese arch (grants 02-04-48651 and 03-04-48833). References 1. Koenig, R. & Lesemann, D E. (1989) Potato X virus. Descript. Plant Vir. 354. 2. Chapman, S., Hills, G ., Watts, J. & Baulcombe, D. (1992) Mutational analysis of the c oat protein gene of potato virus X: effects on virion morphology a nd viral pathogenicity. Vi rology 191, 223–230. 3. Goulden, M.G., Kohm, B.A., Santa Cruz, S., Kavanagh, T.A. & Baulcombe, D.C. (1993) A feature of the coat protein of po tato virus X affects both induced virus resistance in potato and viral fitness. Virology 197, 293–302. 4. Fedorkin, O.N., Solovyev, A.G., Yelina, N.E., Zamyatnin, A.A. Jr, Zinovkin, R.A., Makinen, K., Sch ieman n, J. & Morozov, S.Yu (2001) Cell-to-cell movement of potato virus X involves distinct functions of the coat protein. J. General Virol. 82, 449–458. 5. Kohm, B.A., Goulden, M.G., Gilbert, J.E., Kavanagh, T.A. & Baulcombe, D.C. (1993) A potato virus X resistance gene mediates an induced, nonspecific r esistance in protoplasts. Plant Cell 5, 913–920. 6. Goulden, M.G. & Baulcombe, D.C. (1993) Functionally homo- logous ho st components recogniz e potato virus X in Gomphrena globosa and potato. Plant Cell 5, 921–930. 7. Kozlovsky, S.V., Karpova, O.V., Arkhipenko, M.V., Z ayakina, O.V., Rodionova, N.P. & Atabekov, J.G. (2003) The influence of potato virus X coat protein N-terminal region on virion structure. Dokl. Acad. Nauk (Russia) 391, 189–191. 8. Atabekov, J .G ., Rodionova, N.P., Karpo va, O.V., Ko zlo vsky, S.V.,Novikov,V.K.&Arkhipenko,M.V.(2001)Translational activation of encapsidated potato virus X RNA by coat protein phosphorylation. Virology 286, 466–474. 9. Atabekov, J .G ., Rodionova, N.P., Karpo va, O.V., Ko zlo vsky, S.V. & Poljakov, V.Y. (2000) T he movement protein-triggered in situ conversion of po tato virus X virion RN A from a non- translatable into a translatable form. Virology 271, 259–263. 10. Rodionova, N.P., Karpova, O. V., K ozlovsky, S.V., Zayakina, O.V., Ark hipenko, M.V. & Atabeko v, J.G. (2003) Linear remodeling of helical virus by movement protein binding. J. Mol. Biol. 333, 565–572. 11. Koenig, R. (1972) Anomalous behavior of the coat proteins of potato virus X and cactus virus X during electrophoresis in dodecyl sulfate-containing polyacrylamide gels. Virology 50, 263– 266. 12. Koenig, R ., Tre maine, J.H. & Shepard, J.F. (197 8) In situ degradation of the protein chain of potato virus X at the N- and C-termini. J. General Virol. 38, 329–337. 13. Tozzini, A.C., Ek, B., Palva, E.T. & H opp, H.E. (1994) Potato virus X coat protein: a glycoprotein. Virology 202, 651–658. 14. Bourne, Y. & Cambillau, Ch (1993) The role of structural water molecules in protein-saccharide complexes. In Water and Biological Macromolecules (To pics in Molecular and S tructural Biology) (Westhof, E., ed.), pp. 321–337. CRC Press, Boca Raton. 15. Baratova, L.A., Grebenshchikov , N.I., Dobrov, E.N., Gedrovich, A.V., Kashirin, I.A., Shishk ov, A.V., Efimov, A .V., Jarvekulg, L., Radavsky, Y.L. & Saarma, M. (1992) The organization of potato virus X coat proteins in virus particles studied by tritium plani- graphy and model building. Virology 188, 175–180. 16. Sober, J., Jarvekulg, L., Toots, I., Radavsky, J., Villems, R. & Saarma, M. (1988) Antigenic characterization of potato virus X with monoclonal antibodies. J. General Virol. 69, 1799–1807. 17. Tremaine, J.H. & Agrawal, H.O. (1972) Limited proteolysis of potato virus X by tryp sin and plant proteases. Virology 49, 735– 744. 18. Skryabin, K.G., Kraev, A.S., Morozov, S.Yu, Rozanov, M.N., Chernov, B.K., Lukasheva, L. I. & Atabekov, J.G. (198 8) The nucleotide sequen ce of potato virus X RNA. Nucleic Acids Res. 16, 10929–10930. 19. Kavanagh, T., Goulden, M., Santa Cruz, S., Chapman, S., Bar- ker, I. & Baulcombe, D. (1992) Molecular analysis o f a resistance- breaking strain of potato virus X. Virology 189, 609–617. 20. Peres, S. (1993) Polysaccharide interactions with w ater. In Water and Biological Macromolecules (Topics in Molecular and Structural Biology) (Westhof, E., ed.), pp. 295–320. CRC Press, Boca Raton. 21. Laemmli, U. K. (1970) Cleavage of struct ural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680– 685. 22. Khorlin, A.Ya., Shiyan, S.D., Markin, V.A., Nasonov, V.V. & Mirzayanova, M.N. (1986) Fluorescent derivatives of carbo- hydrates in structural studies of glycocon jugates. N-(4-methyl- coumarin-7-yl) glycamines: synthesis, characterization and use in carbohydrate analysis. Bioorg. Khimiya (Russia) 12, 1203– 1212. 23. Tsugita, A. & Scheffler, J.J. (1982) A rapid method for acid hydrolysis of protein with a mixture of trifluoroacetic acid and hydrochloric acid. Eur. J. Biochem. 124, 585–588. 3144 L. A. Baratova et al.(Eur. J. Biochem. 271) Ó FEBS 2004 24. Chaurand, P., Lu etzenkirchen ,F.&Spengler,B.(1999)Peptide and protein identificati on by m atrix-assisted laser d esorption ionization (MALDI) and MALDI-post-source decay time-of- flight mass spectrometry. J.Am.Soc.MassSpectrom.10, 91–103. 25. Harvey, D.J. ( 1996) M atrix-assisted la ser d esorption/ionisation mass spectrom etry of oligosac charides and glycoconjugates. J. Chromatogr. A 720, 429–446. 26.Thiede,B.,Wittmann-Liebold,B.,Bienert,M.&Krause,E. (1995) MALDI-MS for C-terminal sequence d etermination of peptides and proteins degraded by c arbo xypeptidase Y and P. FEBS Lett. 357, 6 5–69. 27. Parker, F.S. (1983) Applications of Infrared, Raman, and Resonance Raman Spectroscopy in Biochemistry. Plen um Publications, Corp. 28. Hart, G.W., Haltiwanger, R.S., Holt, G.D. & Kelly, W.G. (1989) Glycosylation in the nucleus and cytoplasm. Annu. Rev. Biochem. 58, 841–874. 29. Haltiwanger, R.S., Kelly, W.G., Roquemore, E.P., Blomberg, M.A.,Dong,L.Y.,Kreppel,L.,Chou,T.Y.&Hart,G.W.(1992) Glycosylation of nuclear and cytoplasmic proteins is ubiquitous and dynamic. Biochem. Soc. Trans. 20, 2 64–269. 30. Gerken, T .A., Butenhof, K.J. & Shogren, R. (1989) Effects of glycosylation on the conformation and dynamics of O-linked glycoproteins: carbon-13 NMR studies of ovine submaxillary mucin. Biochemistry 28, 5536–5543. 31. Falconi, M., Brunelli, M., Pesce, A., Ferrario, M., Bolognesi, M. & Desideri, A. (200 3) Static and dynamic water molecules in Cu,Zn superoxide dismutase. Proteins 51, 607–615. 32. Lusse, S. & Arnold, K. (1998) Water binding of polysaccharides – NMR and ESR studies. Macromolecules 31, 6891–6897. 33. Parker, L., Kendall, A. & Stubbs, G. (2002) Surface features of potato virus X from fiber diffraction. Virology 300, 291–295. 34. Blanch, E.W., Robinson, D.J., Hecht,L.,Syme,C.D.,Nielsen,K. & Barron, L.D. (2002) Solution structures of potato virus X and narcissus mosaic virus from Raman optical activity. J. Genera l Virol. 83, 241–246. Ó FEBS 2004 N-Terminal glycosylation of potato virus X protein (Eur. J. Biochem. 271) 3145 . N-Terminal segment of potato virus X coat protein subunits is glycosylated and mediates formation of a bound water shell on the virion surface Lyudmila. urface-located and glycosylated N-terminal CP segments of intact PVX virions induce the formation of a columnar-type shell from bound water molecules around the