HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Supplemental Material Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Pitt, J. J. Articles by Gorman, J. J. Search for Related Content PubMed PubMed Citation Articles by Pitt, J. J. Articles by Gorman, J. J. Social Bookmarking                      What's this?

J Biol Chem, Vol. 275, Issue 9, 6469-6478, March 3, 2000

Determination of the Disulfide Bond Arrangement of Newcastle Disease Virus Hemagglutinin Neuraminidase
CORRELATION WITH A -SHEET PROPELLER STRUCTURAL FOLD PREDICTED FOR Paramyxoviridae ATTACHMENT PROTEINS^*,

James J. Pitt, Elizabeth Da Silva, and Jeffrey J. Gorman

From the Biomolecular Research Institute, Parkville, Victoria 3052, Australia

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Disulfide bonds stabilize the structure and functions of the hemagglutinin neuraminidase attachment glycoprotein (HN) of Newcastle disease virus. Until this study, the disulfide linkages of this HN and structurally similar attachment proteins of other members of the paramyxoviridae family were undefined. To define these linkages, disulfide-linked peptides were produced by peptic digestion of purified HN ectodomains of the Queensland strain of Newcastle disease virus, isolated by reverse phase high performance liquid chromatography, and analyzed by mass spectrometry. Analysis of peptides containing a single disulfide bond revealed Cys⁵³¹-Cys⁵⁴² and Cys¹⁷²-Cys¹⁹⁶ linkages and that HN ectodomains dimerize via Cys¹²³. Another peptide, with a chain containing Cys¹⁸⁶ linked to a chain containing Cys²³⁸, Cys²⁴⁷, and Cys²⁵¹, was cleaved at Met²⁴⁹ with cyanogen bromide. Subsequent tandem mass spectrometry established Cys¹⁸⁶-Cys²⁴⁷ and Cys²³⁸-Cys²⁵¹ linkages. A glycopeptide with a chain containing Cys³⁴⁴ linked to a chain containing Cys⁴⁵⁵, Cys⁴⁶¹, and Cys⁴⁶⁵ was treated sequentially with peptide-N-glycosidase F and trypsin. Further treatment of this peptide by one round of manual Edman degradation or tandem mass spectrometry established Cys³⁴⁴-Cys⁴⁶¹ and Cys⁴⁵⁵-Cys⁴⁶⁵ linkages. These data, establishing the disulfide linkages of all thirteen cysteines of this protein, are consistent with published predictions that the paramyxoviridae HN forms a -propeller structural fold.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Newcastle disease virus (NDV)¹ and mumps virus belong to the rublavirus genus of the subfamily paramyxovirinae within the paramyxoviridae family of negative strand nonsegmented RNA viruses (1, 2). Rublaviruses are enveloped viruses with a lipid membrane bilayer containing two integral membrane glycoproteins that mediate the process of invasion of host cells by the virus through the two steps of attachment and fusion. Attachment is mediated by a type II integral membrane protein with hemagglutinin and sialidase activities, termed the hemagglutinin neuraminidase (HN). Fusion is caused principally by a type I integral membrane protein, termed the fusion protein (3, 4). There is evidence that the HN assists in the fusion process via interaction(s) with the fusion protein (3, 5-8).

Other members of the paramyxoviridae have similar membrane glycoproteins to orchestrate transmission of their nucleocapsids, containing their negative sense RNA genomes, into host cells. All members of the paramyxoviridae have fusion proteins with very similar structural motifs; however, the attachment proteins of the different subfamilies and genera exhibit different functional profiles and substantial structural variability. Members of the paramyxovirus genus (e.g. parainfluenza viruses) also have HN proteins, and members of the morbillivirus genus (e.g. measles virus) have a similar protein with hemagglutinin (H) activity and little (9) or no (10) sialidase activity. Whereas pneumovirus of mice exhibits hemagglutinin activity (11, 12), other members of the pneumovirinae subfamily lack both hemagglutinin and sialidase activities (13, 14), and their attachment proteins are structurally distinct from the other paramyxoviridae attachment proteins (15).

Influenza A and B viruses also exhibit the same functional characteristics as members of the rublaviruses such as NDV and paramyxoviruses such as the parainfluenza viruses (16). However, the attachment and fusion activities of these influenza viruses are associated with an H (17), and the sialidase activity is located on a separate protein, the neuraminidase (18). Three-dimensional structures of the influenza neuraminidases have been determined by x-ray crystallography (19-24). This structure has been described as a -sheet propeller structure with the blades representing -sheets connected by loops. The four -strands within each of these sheets are also connected by loops and arranged in an antiparallel fashion. No structure has been reported to date for any paramyxoviridae attachment protein. Despite this, structural predictions indicate that the HN has a three-dimensional fold similar to influenza neuraminidases (9, 25-27).

The HN ectodomains of the different strains of NDV are comprised of approximately 525 amino acids within which are usually twelve or thirteen cysteine residues and five or six consensus sites for N-linked glycosylation (28). A cysteine residue at position 123, which is variably present in different strains (28) of the virus, is believed to form an intermolecular disulfide with another HN ectodomain to form a covalent homodimer (29-32). There is evidence that the HN exists as a tetramer on virions (29, 32-34) but the association of covalent homodimers into homotetramers does not appear to rely on formation of additional covalent bonds (34). Predicted folding of the HN has suggested possible pairings between the remaining cysteines (9, 26, 27). Cysteine mutagenesis studies (31) have provided some experimental support for these predictions; however, the complete disulfide bond pattern of the protein has not been determined experimentally.

This report describes isolation of NDV HN ectodomains, generation and isolation of disulfide-linked peptides from the HN, and a comprehensive mass spectrometric analysis of the disulfide linkages between and within these peptides. The experimentally determined complete disulfide bond arrangement of the NDV HN is reported herein and compared with the disulfide bond patterns predicted by previous molecular modeling and cysteine mutagenesis studies.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- Dispase grade I (a neutral protease from Bacillus polymyxa), n-octylglucoside, pepsin, peptide-N-glycosidase F (PNGase F), trypsin and chymotrypsin were from Roche Molecular Biochemicals. Tributylphosphine was from Tokyo Kasei and tris-(2-carboxyethyl)phosphine (TCEP) and cyanogen bromide were from Pierce. 2'-(4-Methylumbelliferyl)--D-N-acetyl-neuraminic acid (MUNANA) was from Sigma. A vaccine preparation of the V4 isolate of the Queensland strain of NDV was obtained from Arthur Webster Pty. Ltd., Northmead, New South Wales, Australia. 2,6-Dihydroxyacetophenone from Fluka was recrystallized from 20% (v/v) ethanol while -cyano-4-hydroxy cinnamic acid from Aldrich was recrystallized from methanol. All other reagents were of either reverse-phase high performance liquid chromatography (RP-HPLC) grade or the highest quality analytical reagent grade.

Isolation of NDV HN Ectodomain Dimers-- NDV was propagated in 10-day-old embryonated chicken eggs and harvested as described previously (35, 36). The final pellet of virions was suspended to a concentration of approximately 10 mg/ml in 10 mM HEPES, 100 mM NaCl, 1 mM EDTA (HNE) buffer (pH 7.2). An equal volume of 4% (w/v) n-octylglucoside in H₂O was added to the virus preparation, and the mixture was allowed to stand at 22 °C for 30 min. Released nucleocapsids and any insoluble membrane aggregates were removed by centrifugation for 1.5 h at 110,000 × g and 15 °C. Dispase (10 mg/ml in HNE buffer (pH 7.2)) and CaCl₂ were added to the separated supernatant of the detergent-solubilized virions to final concentrations of 50 µg/ml and 10 mM, respectively. Dispase digestion was then conducted for 2 h at 37 °C, and the digest was cooled to 4 °C prior to adding 10 volumes of n-butanol at 20 °C. The n-butanol and aqueous phases were mixed gently until any signs of an emulsion disappeared, and precipitation of proteins was allowed to proceed for 16 h at 20 °C. Precipitated protein was recovered by centrifugation at 10 °C, washed twice with n-butanol at 20 °C and once with diethylether at 20 °C, and subsequently dried under a stream of high purity nitrogen gas. The dried protein preparation was resuspended in HNE buffer (pH 7.2), clarified by centrifugation, and filtered through a 0.22 µM membrane prior to application to a column of Superose 12 (Amersham Pharmacia Biotech) (2.6 cm × 95 cm). Separation of the proteins was achieved by elution at 2 ml/min and 22 °C with HNE buffer (pH 7.2) containing 0.02% (w/v) sodium azide. The effluent was continuously monitored at 280 nm for the presence of protein, and 5-ml fractions were collected. Detection of sialidase activity in column fractions was achieved by using MUNANA acid as a substrate. This involved incubation of 5 µl of each fraction with 50 µl of a solution containing 85 µM MUNANA, 8.5 mM CaCl₂, and 85 mM sodium cacodylate (pH 6.5). After incubation for 5 min at 22 °C in the dark the reaction was terminated by the addition of 200 µl of 133 mM glycine, 33 mM Na₂CO₃ (pH 10.7). Liberation of methylumbelliferone as a consequence of sialidase activity was determined by reading the resultant fluorescence in a Perkin-Elmer LS50B fluorimeter fitted with a plate reader and set for excitation at 365 nm and emission at 450 nm. The fractions of apparent homogeneity by polyacrylamide gel elctrophoresis in the presence of sodium dodecyl sulfate (SDS-PAGE) (37) and highest specific sialidase activity were pooled and concentrated at 4 °C to 10-12 mg/ml using an Amicon-stirred cell fitted with a YM 10 membrane. Concentrated protein preparations were stored at 4 °C.

Protein and Peptide Cleavages-- Pepsin digestion of HN ectodomains was conducted on ~400 µg of HN after precipitation of the protein from HNE buffer by the addition of 9 volumes of methanol at 20 °C and leaving the protein at 20 °C for 16 h. The resultant pellet was collected by centrifugation at 4 °C, washed with 20 °C methanol, and dried in vacuo. The dried pellet was dissolved in 100 µl of 100 mM acetic acid, 100 mM formic acid and incubated at 37 °C for 3 h after adding 20 µg of pepsin in 2 µl of 100 mM acetic acid, 100 mM formic acid.

PNGase F/trypsin digestion was performed after reconstitution of the contents of dried RP-HPLC fractions in 20 µl of 100 mM ammonium bicarbonate containing 1 mM N-ethylmaleimide, by the addition of 5 units of PNGase F and incubation at 37 °C for 2 h. Subsequent tryptic cleavage was achieved by the addition of 1 µg of trypsin and incubation at 37 °C for 24 h. Chymotryptic digestion was performed after RP-HPLC fractions were dried in vacuo and reconstituted in 10 µl of 100 mM ammonium bicarbonate by the addition of 0.5 µg of chymotrypsin solution and incubation at 37 °C for 3 h.

Cyanogen bromide treatment was performed on RP-HPLC fractions that had been dried in vacuo and reconstituted in 9 µl of 70% (v/v) trifluoroacetic acid. Cleavage was achieved by the addition of 1 µl of 1 M cyanogen bromide in acetonitrile and incubation at 22 °C for 13 h. The sample was evaporated in vacuo, and residual cyanogen bromide was removed by three cycles of reconstitution in 20 µl of 33% acetonitrile containing 0.1% (v/v) trifluoroacetic acid and evaporation. Prior to further analysis the sample was desalted on a microcolumn of C₁₈ silica.

Manual Edman degradation was performed in 1-ml tapered Reactivials. Dried samples were reconstituted in 20 µl of 50% (v/v) pyridine:water containing 1 mM N-ethylmaleimide; 5 µl of 50% (v/v) pyridine:water saturated with phenyl isothiocyanate was added, and the mixture was incubated at 50 °C for 20 min under nitrogen. The aqueous phase was extracted twice with 400 µl of heptane:ethyl acetate 2:1 (v/v) and then dried in vacuo. Cleavage was performed by adding 20 µl of trifluoroacetic acid and heating at 50 °C for 7 min under nitrogen. Trifluoroacetic acid was removed in vacuo, and the tube was washed with diethyl ether. After drying the tube in vacuo the truncated peptide was dissolved in 33% acetonitrile:water (v/v) containing 0.1% trifluoroacetic acid for subsequent analysis.

RP-HPLC-- A Hewlett Packard 1090M system was used for preparative RP-HPLC. Peptide separation was achieved at 0.15 ml/min using a 2.1 mm × 25 cm C₁₈ column (Vydac, catalog number 218TP52) with a linear gradient from 0.1% (v/v) aqueous trifluoroacetic acid to 32% (v/v) aqueous acetonitrile containing 0.1% (v/v) trifluoroacetic acid over 60 min and a subsequent linear increase in the acetonitrile concentration to 80% (v/v) over 20 min. A HP1090 photodiode array detector was used to continuously monitor the eluant at 214 and 280 nm. Fractions were collected into polypropylene tubes for subsequent analysis. Where required, peptic digests were reduced prior to RP-HPLC by the addition of 6 µl of 1 M ammonium bicarbonate to 30 µl of the digest, followed by 5 µl of 1% (v/v) tributylphosphine in isopropanol. The mixture was incubated at 37 °C for 45 min and then evaporated to dryness in vacuo. Traces of tributylphosphine were removed by two cycles of adding 20 µl of isopropanol and re-evaporation. The reduced digest was reconstituted in 50 µl of 0.2% (v/v) trifluoroacetic acid.

Matrix-assisted Laser Desorption/Ionization-Time-of-Flight-Mass Spectrometry (MALDI-TOF-MS)-- Protein molecular weight determination was performed in linear mode using a Bruker Reflex time-of-flight instrument fitted with a 337-nm nitrogen laser. Samples were ionized from a matrix comprised of 2,6-dihydroxyacetophenone containing diammonium hydrogen citrate (38), and bovine serum albumin was used for external calibration. Protein ions were accelerated at 20 kV and detected with a Bruker HIMAS detector. Peptides were analyzed in reflector mode with delayed extraction using an acceleration voltage of 20.0 kV and a reflectron voltage of 22.8 kV. Peptide data were generally collected after ionization from the above matrix at 500 MHz in the m/z = 500 to 8000 range with the instrument calibrated externally using a mixture of human angiotensin II, ACTH clip (residues 18-39), and bovine insulin. Peptide sequencing by post-source decay (PSD) employed an acceleration voltage of 26.3 kV and a series of 14 reflectron voltage steps from 30 to 0.95 kV. Individual spectra were obtained at each step by collecting ~50-100 laser shots. A pulsed gate mounted shortly after the ion source was used to select precursor ions. Bruker XMASS software was used to assemble resulting spectra onto a continuous true m/z scale. PSD spectra were calibrated using a calibration file created with fragments of ACTH residues 18-39 (39).

Reduction of disulfide-linked peptides prior to MALDI-TOF-MS was performed by mixing 4 µl of digest or RP-HPLC fraction with 0.5 µl of 1 M diammonium hydrogen citrate and 0.5 µl of 50 mM TCEP and incubation of the mixture at 37 °C for 60 min (38). For molecular weight determination of the reduced peptides, 1 µl of the reduction mixture was mixed with 1 µl of 2,6-dihydroxyacetophenone solution (10 mg/ml in 1:1 ethanol:acetonitrile (v/v)), and 1 µl of this mixture was dried briefly in vacuo on a stainless steel sample target. For PSD analysis of reduced peptides, 0.5 µl of the reduction mixture was mixed with 0.5 µl of 5% (v/v) trifluoroacetic acid, placed on a preprepared thin layer of -cyano-4-hydroxy cinnamic acid (40), and dried briefly in vacuo. The sample spot was then washed with two 1-µl aliquots of 1% (v/v) trifluoroacetic acid saturated with -cyano-4-hydroxy cinnamic acid, excess solvent being removed with a stream of nitrogen.

Electrospray Ionization-Mass Spectrometry (ESI-MS)-- A Hewlett Packard 1100 system was used for liquid chromatography-MS with separation performed on a 1 mm × 25 cm C₁₈ column (Vydac, catalog number 218TP51) at a flow rate of 40 µl/min using the same gradient as was used for preparative RP-HPLC except the initial and final trifluoroacetic acid concentrations were 0.05 and 0.045% (v/v), respectively. A HP 1100 photodiode array detector was used to continuously monitor the absorbance of the eluate at 214 and 280 nm, and the outlet of the flow cell was connected directly to the ESI source of a PE-Sciex 150EX liquid chromatography-MS system via a short length of fused silica tubing. The mass spectrometer was scanned from m/z = 100 to 3000 with an ion source orifice potential of 50 V. For experiments involving the production of in-source collisional fragmentation an orifice potential of 80 V was used. The mass spectrometer was scanned with a 0.25 atomic mass unit step and a dwell time of 0.40 or 0.50 ms/step.

Tandem mass spectrometry (MS/MS) experiments were performed using a PE-Sciex API-300 system. Samples, in a solvent of 50% (v/v) acetonitrile containing 0.1% (v/v) formic acid were continuously infused into the ion source via a Micro-ion Spray interface at a flow rate of 12 µl/hour maintained with a Harvard Model 55-1111 syringe pump. Precursor ions were selected at a resolution of 0.6 Da and fragmented by collision with nitrogen at an energy of 60 or 40 eV. Data were accumulated and averaged over a period of ~15 min.

N-terminal Sequence Analysis-- Automated Edman degradation was performed using a Hewlett Packard 241 (N + C) protein sequenator.

Nomenclature-- Disulfide linkages are indicated by slashes as follows: Cys^m/Cysⁿ indicates linkage of a peptide containing cysteine residue m in the protein sequence to a peptide containing cysteine residue n; Cys^w/(Cys^x, Cys^y, Cys^z/) indicates that a peptide containing cysteine residue w is linked to a peptide containing cysteines x, y, and z, which contains an additional intrachain disulfide; Cys^w/(Cys^x,Cys^y)/Cys^z indicates that separate peptides containing cysteines w and z, respectively, are linked to a peptide containing cysteines x and y but the linkage pattern is not defined; Cys^w/Cys^x, Cys^y/Cys^z indicates that a peptide containing cysteine residue w is linked via cysteine x to a peptide containing cysteines x and y which, is in turn linked to another peptide by linkage between cysteines y and z. Similarly, Xaa^j-Xaa^k/Xaa^p-Xaa^q describes a disulfide linkage between a peptide segment encompassing amino acid residues j to k and a peptide segment encompassing amino acid residues p to q and Xaa^a-Xaa^b/Xaa^c-Xaa^d/Xaa^e-Xaa^f indicates that peptide segments encompassing residues a to b and e to f, respectively, are linked by disulfide bonds to a peptide segment encompassing residues c to d.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Characterization of the Membrane Anchor-free NDV HN-- The supernatants from high speed centrifugation of detergent-solubilized NDV preparations showed a major Coomassie Blue staining component of 71 kDa upon SDS-PAGE under reducing conditions (Fig. 1A, lane 1). This component appeared to be converted to a band of 63 kDa upon exposure to Dispase (Fig. 1A, lane 2). Analysis of fractions produced by chromatographic separation of the Dispase-treated detergent extract on Superose 12 revealed that the majority of the sialidase activity was evident in a symmetrical peak with maximal absorbance in fraction 51 (Fig. 1B). A trace of sialidase activity co-eluted with a peak of absorbance in the void volume of the column. The major sialidase activity peak also contained the 63-kDa product of Dispase cleavage (Fig. 1A, lane 4). Selection of fractions 48-54 to pool for further processing was based on homogeneity estimations by SDS-PAGE gels and sialidase activity of individual fractions.

View larger version (51K): [in this window] [in a new window]   Fig. 1.   Isolation and characterization of soluble HN ectodomains. A, SDS-PAGE analysis (12% gels) of: lanes 1 and 2, high speed supernatant of detergent-solubilized NDV (10 µg) before and after Dispase treatment, respectively; and, lanes 3 and 4, 2.5 µg of nonreduced and reduced protein, respectively, from pooled fractions of Superose 12 eluate (B) that exhibited maximal neuraminidase activity. Lanes marked M indicate calibration mixtures containing proteins with the molecular mass to the left of each panel. M1 denotes Bio-Rad molecular weight markers dissolved in reducing sample buffer and M2 denotes Novex prestained markers dissolved in nonreducing sample buffer. B, Superose 12 chromatogram of Dispase-treated high speed supernatant of detergent-solubilized NDV. The protein was precipitated from 100 ml of treated supernatant with n-butanol, processed further, and applied to the column in 10 ml of buffer as described under "Experimental Procedures." The shaded area indicates fractions 48-54 that exhibited maximal neuraminidase activity and were pooled for further characterization.

N-terminal sequence analysis of the protein present in the pooled peak of sialidase activity from the Superose 12 column indicated that Dispase cleavage produced soluble HN ectodomains with ragged N termini corresponding to Ala¹¹⁷, Ala¹¹⁸, or Ser¹¹⁹ of the untruncated protein sequence (Fig. 2A). This finding indicates that an asparagine usually found at position 119 (28) was a serine in the isolate of NDV used in this study. SDS-PAGE analysis of the nonreduced protein revealed a molecular weight consistent with disulfide linkage of two monomers (Fig. 1A, lane 3). MALDI-TOF-MS of the product showed a broad base peak of m/z = 113,000 and a smaller peak at m/z = 56,615 consistent with the doubly charged version of this ion (supplemental Fig. S1). These data are also consistent with disulfide linkage of two truncated NDV HN ectodomains. Assuming that Dispase did not cleave the C terminus of the ectodomains, these data also indicate that glycosylation contributes approximately 12,700 Da to the mass of the dimer.

View larger version (47K): [in this window] [in a new window]   Fig. 2.   Sequence and topological model of the ectodomain of NDV HN showing experimentally determined disulfide linkages. A, sequence of the ectodomain of the Queensland strain of NDV HN. Two amino acid variations were detected in the ectodomain of the isolate of the virus used in this study compared with those published for the Queensland strain ((44), Swiss Prot accession number P13850, GenBankTM accession number J03911) namely N119S and I175M. Glycosylated N-linked consensus sites have solid underlines, whereas an unoccupied site at Asn538 has a dashed underline. Sites of Dispase cleavage to release soluble ectodomain heads are indicated by upward arrows. The major sites of cleavage by pepsin to produce disulfide-linked peptides used to define disulfide linkages are indicated by downward arrows. Dashed arrows indicate minor peptic cleavage sites. Disulfide bond linkages are indicated by bold lines. B, experimentally determined disulfide bonds are shown as bold lines linking structural units within a proposed model of HN. Demarcation of the sequence segments forming -stands of the six -sheets, labeled 1-6, short loops joining the strands in an antiparallel manner and longer loops connecting each sheet is based on the model of Langedijk et al. (9). The experimentally determined disulfide linkages differ from those proposed in the theoretical model in that the model had Cys344/Cys465 and Cys455/Cys461 linkages.

Isolation of Disulfide-linked Peptides-- RP-HPLC of nonreduced peptic digests of HN produced a chromatogram consisting of relatively broad and incompletely resolved peaks (Fig. 3A). A chromatogram was also obtained by reduction of the peptic digest before performing RP-HPLC (Fig. 3B). Careful comparison of the RP-HPLC chromatograms of the nonreduced and reduced samples revealed five major peaks in the nonreduced digest (P1-P5 in Fig. 3A) that were affected by reduction indicating that they contained disulfide-linked peptides. Several peaks appeared in the chromatogram of the reduced peptic digest that apparently corresponded to cysteine-containing peptides released from disulfide linkage by reduction. Peaks P1-P5 from the nonreduced digest were isolated by RP-HPLC of the nonreduced peptic digest for further analysis to determine the disulfide linkages in HN.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Chromatographic identification of disulfide-linked peptic peptides of NDV HN. A, chromatographic separation of ~21 µg of peptic digest of NDV HN using a Vydac C18 column (2.1-mm inner diameter × 250 mm). B, chromatogram obtained with the same amount of NDV HN peptic digest after reduction with tributylphosphine. The separation conditions were the same as in A. Major peaks affected by reduction are indicated by P1-P5 in A. Upward arrows in B indicate peaks that appear as a result of reduction.

Overview of Disulfide Linkage Determination-- The strategy for determination of the disulfide linkages of HN was to use MALDI-TOF-MS as the primary tool to analyze disulfide-linked peptic peptides in peaks P1-P5. In cases where this did not produce conclusive results, peptides were subjected to further treatment(s) involving deglycosylation and/or additional peptide bond cleavages. Peptides subjected to these supplementary treatments were also analyzed by MALDI-TOF-MS and, where necessary, ESI-MS/MS. The simplest cases, peptic peptides containing single inter- or intrachain disulfides, generally only required MALDI-TOF-MS to identify the cysteine residues involved in linkage of the chains. This usually involved analysis of the constituents of the relevant peak before and after reduction with TCEP and rationalization of mass changes that accompanied reduction with reference to cysteine-containing segments of the HN ectodomain sequence (Fig. 2A). In some cases this alignment process was aided by the presence of sequence tags originating from ragged peptic cleavage(s) of the linked chain(s). PSD analysis was also used to identify reduced chains. In the more complex cases, with peptides containing both intra- and interchain disulfides, MALDI-TOF-MS usually enabled identification of the linked chains; however, secondary peptide bond cleavage(s) was required to convert the intrachain bond into an interchain bond. When this conversion process produced peptides with single disulfide linkages it was possible to make disulfide bond assignments based on MALDI-TOF-MS analyses. When the conversion process produced peptides with three linked chains ESI-MS/MS provided the means of making the linkage assignments. Using this strategy it was possible to determine that Cys¹²³ is the residue involved in homodimerization of NDV ectodomains and to establish Cys¹⁷²/Cys¹⁹⁶, Cys¹⁸⁶/Cys²⁴⁷, Cys²³⁸/Cys²⁵¹, Cys³⁴⁴/Cys⁴⁶¹, Cys⁴⁵⁵/Cys⁴⁶⁵, and Cys⁵³¹/Cys⁵⁴² linkages (Fig. 2A). Data enabling these assignments are presented in Figs. 4 and 5, supplemental Figs. S2-S6, and Tables I-III. More detailed description of the data from the analysis of peaks P1-P5 is presented below.

View larger version (24K): [in this window] [in a new window]   Fig. 4.   ESI-MS/MS spectra of the peptide Tyr340-Asp349/Cys461-Val469/Phe447-Arg460. This peptide was derived by treatment of RP-HPLC peak P1 (Fig. 3) with PNGase F and trypsin and isolated by RP-HPLC as described in Fig. 3. A, ESI-MS/MS of the triply charged ion (m/z = 1185.0) using a collision energy of 60 eV and nitrogen as the collision gas. A series of peptide fragments corresponding to internal cleavages in the Tyr340-Asp349 chain is indicated. B, an expansion to show y series fragments between Cys461 and Cys465. C, the assignment of fragments and theoretical m/z values. Fragments are defined using standard nomenclature (53, 54). Average masses are used throughout and take into account the conversion of Asn341 to Asp as a result of deglycosylation.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   MALDI-TOF-MS of RP-HPLC peak P2 (Fig. 3). A, nonreduced peak P2 (100 laser shots). Dotted brackets indicate different forms of peptides containing the same disulfide linkage(s). B, after reduction with TCEP (125 laser shots). TCEP reduction and MALDI-TOF-MS were performed as described under "Experimental Procedures." Sequence tags resulting from ragged cleavage with pepsin are indicated in one-lettered code. Refer to Table I for identity of peptides.

Analysis of Peak P1-- A series of ions was observed in MALDI-TOF-MS spectra of nonreduced peak P1 (Fig. 3A) at around m/z = 5790 that were separated by approximately 203 and 146 Da (Fig. S2A) suggesting the attachment of a glycan. These ions disappeared upon reduction and there was a concomitant appearance of an intense ion at m/z = 2337.1 that corresponded to Phe⁴⁴⁷-Val⁴⁶⁹ containing cysteines 455, 461, and 465 in the reduced state (Fig. S2, A and B and Table I). This was confirmed by analysis of fragments formed by PSD of the ion at m/z = 2337.1, which produced data consistent with the sequence Phe⁴⁴⁷-Val⁴⁶⁹ (Table I). An ion at m/z = 2335.5 also disappeared upon reduction but ions at m/z = 2078.6 and m/z = 2108.4 were not affected by reduction (cf. Fig. S2, A and B). Minor ions at m/z = 2190.0 and m/z = 2238.0 also appeared upon reduction (Fig. S2B) that corresponded to Thr⁴⁴⁸-Val⁴⁶⁹ and Phe⁴⁴⁷-Gly⁴⁶⁸, respectively, produced by ragged cleavage (Fig. S2B and Table I). These data indicate that Phe⁴⁴⁷-Val⁴⁶⁹ is linked to a cysteine-containing glycopeptide. The ion at m/z = 2335.5 in the nonreduced peak may have been due to MALDI-induced cleavage of the interchain disulfide bond of this glycopeptide without reduction of the intrachain disulfide. The presence of glycosylation was confirmed by treatment of peak P1 with PNGase F (Table II), which resulted in appearance of ions corresponding to Phe⁴⁴⁷-Val⁴⁶⁹ linked via a disulfide to various lengths of peptide sequence containing Cys³⁴⁴ (Fig. S2C and Table II). The deglycosylated peptide was treated further with trypsin and analyzed by liquid chromatography-MS, which showed the presence of a predominant ion corresponding to Tyr³⁴⁰-Asp³⁴⁹/Cys⁴⁶¹-Val⁴⁶⁹/Phe⁴⁴⁷-Arg⁴⁶⁰. MALDI-TOF-MS of the equivalent fraction isolated by RP-HPLC revealed a series of ions consistent with this Cys³⁴⁴/(Cys⁴⁶¹,Cys⁴⁶⁵)/Cys⁴⁵⁵ arrangement (Fig. S2D and Table II). One round of manual Edman degradation on this isolated peptide followed by analysis of the product by MALDI-TOF-MS showed the presence of an ion of m/z = 2103.2 (Fig. S2E and Table II) that corresponded to Thr⁴⁴⁸-Arg⁴⁶⁰/Pro⁴⁶²-Val⁴⁶⁹, thus, establishing a Cys⁴⁵⁵/Cys⁴⁶⁵ linkage. This was complemented by an ion at m/z = 2250.2 that corresponded to Phe⁴⁴⁷-Arg⁴⁶⁰/Pro⁴⁶²-Val⁴⁶⁹, which may have been present due to incomplete coupling of phenylisothiocyanate to Phe⁴⁴⁷ during the single round of Edman degradation. This linkage arrangement was also indicated from liquid chromatography-MS experiments performed on the tryptic peptide with an elevated ion source orifice potential of 80 V to induce in-source fragmentation. Under these conditions, an ion was observed at m/z = 1126.2 that was rationalized in terms of a y-type cleavage between Cys⁴⁶¹ and Pro⁴⁶² to produce a fragment containing the Cys⁴⁵⁵/Cys⁴⁶⁵ linkage (Fig. 4C). This observation was corroborated by fragment ions formed during ESI-MS/MS of the deglycosylated tryptic peptide isolated by RP-HPLC, Tyr³⁴⁰-Asp³⁴⁹/Cys⁴⁶¹-Val⁴⁶⁹/Phe⁴⁴⁷-Arg⁴⁶⁰. A series of doubly charged y-type fragment ions was observed (Fig. 4, A and B), including the ion at m/z = 1126.2 (Fig. 4B), corresponding to cleavages between Cys⁴⁶¹ and Cys⁴⁶⁵ in the Cys⁴⁶¹-Val⁴⁶⁹ chain (Fig. 4C), which confirmed the Cys⁴⁵⁵/Cys⁴⁶⁵ and Cys³⁴⁴/Cys⁴⁶¹ linkages. Theoretical fragments that would result from the alternate linkages were not observed in this spectrum.

Analysis of Peaks P2-P4-- MALDI-TOF-MS of peak P2 confirmed the expectation that individual RP-HPLC peaks would contain multiple peptide species (Fig. 5A). Pseudomolecular ions were observed in this peak that could be rationalized as different forms of three separate species of disulfide-linked peptides. The ion at m/z = 3033.0 corresponded to the peptide Ile¹⁶³-Phe¹⁷⁸/Ser¹⁹⁴-Tyr²⁰⁵ representing a Cys¹⁷²/Cys¹⁹⁶ linkage. This assignment was based on the observation of an ion at m/z = 1622.8 in the reduced digest (Fig. 5B) that corresponded to Ile¹⁶³-Phe¹⁷⁸ containing Cys¹⁷² (Table I), which could be complemented with the theoretical mass of Ser¹⁹⁴-Tyr²⁰⁵ (1412.6 Da; not observed) to produce the ion at m/z = 3033.3. This Cys¹⁷²/Cys¹⁹⁶ linkage could also be deduced based on data from peak P4 (Fig. S3 and Table I). Ions were observed at m/z = 3245.9 and 3194.9 in the nonreduced digest (Fig. S3A and Table I) and at 1622.7 and 1998.7 in the reduced digest (Fig. S3B and Table I) that corresponded to peptides linked in this manner (Fig. S3B and Table I).

Ions at m/z = 3746.1, 3675.0, and 3604.1 in nonreduced peak P2 (Fig. 5A) appeared to be derived from the N terminus based on the correspondence of the mass differences between these ions with a sequence tag of Ala¹¹⁷-Ala¹¹⁸-Ser¹¹⁹ that represents the ragged N terminus of the protein observed by Edman degradation sequencing of the protein before pepsin digestion. Ions (m/z = 2202.0, 2130.9, and 2059.9) representing the same sequence tag were observed in the reduced digest (Fig. 5B). An ion corresponding to the difference between the reduced and nonreduced sequence tag ions (m/z = 1547.6) and encompassing the sequence Ala¹¹⁷-Tyr¹³² was also observed in the reduced digest (Fig. 5B). The data were assigned to represent homodimerization of the protein via a Cys¹²³/Cys¹²³ disulfide bond in the form of peptides Ala¹¹⁷-Glu¹³⁹/Ala¹¹⁷-Tyr¹³², Ala¹¹⁸-Glu¹³⁹/Ala¹¹⁷-Tyr¹³², and Ser¹¹⁹-Glu¹³⁹/Ala¹¹⁷-Tyr¹³² (Table I).

Ions were also observed in peaks P2 and P4 (Figs. 5 and S3) corresponding to a Cys¹⁸⁶/(Cys²³⁸,Cys²⁴⁷,Cys²⁵¹/) arrangement (Table I). Sequence tag information was evident in spectra of these peaks under both nonreduced and reduced conditions indicating the presence of peptides containing these cysteines (Table I). This Cys¹⁸⁶/(Cys²³⁸,Cys²⁴⁷,Cys²⁵¹/) arrangement could form in three alternate ways. Digestion of peak P4, which contained predominantly Cys¹⁸⁶/(Cys²³⁸,Cys²⁴⁷,Cys²⁵¹/) peptides (Table I), with chymotrypsin resulted in ions at m/z = 1172.4 and m/z = 2384.9 corresponding to Cys¹⁸⁶-Asn¹⁹⁰/Gly²⁴⁶-Leu²⁵⁰ and Asp²³⁰-Leu²⁴⁵/Cys²⁵¹-Glu²⁵⁶, respectively (Table III). Reduction of the chymotryptic digest yielded ions consistent with these assignments (Table III). Together these data are evidence for Cys¹⁸⁶/Cys²⁴⁷ and Cys²³⁸/Cys²⁵¹ linkages. Cyanogen bromide cleavage of Cys¹⁸⁶/(Cys²³⁸,Cys²⁴⁷,Cys²⁴⁷/) peptides in peak P4 resulted in the appearance of ions in the unfractionated cleavage mixture consistent with Cys¹⁸⁶/Cys²³⁸ or Cys¹⁸⁶/Cys²⁴⁷ linkages (Table III). The peptide Met¹⁸⁰-Asn¹⁹⁰/Asp²³⁰-X²⁴⁹/Leu²⁵⁰-Glu²⁵⁶, where X represents a homoserine lactone residue formed by treatment with cyanogen bromide, was analyzed by ESI-MS/MS (Fig. S4, A and B). Collisional-induced fragmentation was evident in all three of the disulfide-linked chains. In particular, a series of doubly and triply charged b- and y-type fragment ions were observed that corresponded to cleavage of peptides bonds between Cys²³⁸ and Cys²⁴⁷. Significantly, prominent y-type fragmentation on the N-terminal side of Pro²⁴⁴ was observed giving rise to doubly (m/z = 957.0) and triply (m/z = 638.2) charged ions. These ions were complemented by the doubly (m/z = 1136.2) and triply (m/z = 757.7) charged b-type ions produced by fragmentation of the same peptide bond. These data were also consistent with the linkages Cys¹⁸⁶/Cys²⁴⁷ and Cys²³⁸/Cys²⁵¹. Theoretical fragments that would be produced from alternative disulfide-linked versions of this peptide were not observed.

The MALDI-TOF-MS spectra obtained for peak P3 before and after TCEP reduction contained ions consistent with the presence of peptides representing Cys¹⁸⁶/(Cys²³⁸,Cys²⁴⁷,Cys²⁵¹/) and Cys¹⁷²/Cys¹⁸⁶ linkages (Table I). These ions were similar to those detected in peak P4 but with different relative intensities (Fig. S5), presumably the result of incomplete chromatographic separation of peptides contained in peaks P3 and P4.

Analysis of Peak P5-- Peak P5 produced a predominant ion at m/z = 2206.1 and minor associated sequence tag ions that corresponded to a Cys⁵³¹/Cys⁵⁴² linkage within Thr⁵²⁷-Ala⁵⁴⁶ and minor forms of the sequence lacking N-terminal threonine residues (Fig. S6A). This assignment was corroborated by an increase in mass of these ions upon reduction (Fig. S6B) and partial sequencing of the peptide by PSD analysis (Table I). Characterization of this peptide served to establish that the consensus glycosylation site at Asn⁵³⁸ was not occupied by a glycan.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Although HN primary sequences have been known for various members of the paramyxoviridae family of viruses for some time (41-43), there has been relatively little direct chemical analysis of the post-translational modifications that accompany maturation of these proteins into functionally active forms. Exceptions to this are documentation of the proteolytic activation and N-terminal acetylation of the HN proteins of some avirulent strains of NDV (44). Documentation of the disulfide arrangement of the NDV HN in the current study has provided further insights into the maturation of this protein.

It was possible to unambiguously assign Cys¹²³/Cys¹²³, Cys¹⁷²/Cys¹⁹⁶, and Cys⁵³¹/Cys⁵⁴² disulfide linkages by isolation of peptic peptides of NDV HN containing these linkages and using the established approach of determination of mass changes caused by reduction (45, 46). Pepsin was chosen as the primary protease for these studies as it has a low pH optimum, which makes it ideal for disulfide identification because of the lower possibility of disulfide exchange at acid pH (47). Peptic cleavage of HN resulted in complex mixtures of peptides because of relatively random cleavage at adjacent sites. This was a potential complication in aligning mass data with the known sequence of the protein; however, ragged peptic cleavage produced sequence tags that aided unambiguous alignment of mass data to the correct portion of the HN sequence.

Assignment of the disulfide linkages within the Cys³⁴⁴/(Cys⁴⁵⁵, Cys⁴⁶¹, Cys⁴⁶⁵/) and Cys¹⁸⁶/(Cys²³⁸,Cys²⁴⁷,Cys²⁵¹/) clusters was not straightforward as they were contained in two peptides each of which contained four cysteine residues. The topology of these two peptides was the same with two chains linked by a single interchain disulfide linkage and one chain having an additional intrachain disulfide linkage. With this topology three possible disulfide linkages are possible. In both cases, determination of the linkages was achieved by additional peptide bond cleavages between the intrachain disulfides and subsequent MS/MS of the resultant three chain peptides.

It was of interest to note the different apparent efficiencies of the proline-directed MS/MS cleavages in the two examples where this was used to establish or confirm disulfide linkages. In the Cys³⁴⁴/Cys⁴⁶¹,Cys⁴⁶⁵/Cys⁴⁵⁵ case, the intensity of the y ion produced as a result of N-terminal proline-directed cleavage between Cys⁴⁶¹ and Pro⁴⁶² was considerably weaker than for ions produced by proline-directed cleavage in the other two chains that constituted this peptide, most notably between Cys³⁴⁴ and Pro³⁴⁵ (Fig. 4). The proline-directed cleavage between Cys⁴⁶¹ and Pro⁴⁶² appears to be even weaker than fragmentation at nonproline peptide bonds of the other two chains. In contrast, proline-directed cleavage between Cys²³⁸ and Cys²⁴⁷, in the Cys¹⁸⁶/Cys²⁴⁷,Cys²³⁸/Cys²⁵¹ peptide, was much more pronounced than at other peptide bonds in the three chains of this peptide (Fig. S4). The relative difference in susceptibilities of the cleavages at proline residues between intrachain cysteine residues in these peptides may be because of the greater distance between Cys²³⁸ and Cys²⁴⁷ compared with Cys⁴⁶¹ and Cys⁴⁶⁵. Alternatively, the fact that Cys⁴⁶¹ is in linkage with Pro⁴⁶² may have constrained proline-directed cleavage; however, this did not appear to limit cleavage between Cys³⁴⁴ and Pro³⁴⁵ (Fig. 4). It has been noted previously that peptide bonds within intrachain cystine loops are resistant to MS/MS cleavage conditions (48-51). One interpretation is that this is due to steric constraints induced by the intrachain loop. The present data suggest that close juxtaposition of two cysteine residues in one peptide involved in interchain disulfide bonds with two other peptide chains may also serve to limit cleavage between the intrachain cysteines.

The experimentally determined disulfide linkages are summarized in Fig. 2. During the course of analyzing the peptic profile of HN many minor disulfide linked peptides were observed that served to corroborate the disulfide linkages ascertained for the major peaks P1-P5. No peptic peptides were observed reflecting alternative disulfide linkages. Thus, although it is not entirely possible to rule out the presence of alternate disulfide linkages, such linkages would be quantitatively minor based on assessment of the peptide peak heights from the RP-HPLC profile.

With the exception of Cys¹²³, the remaining cysteine residues in the ectodomain of the NDV HN are conserved in various strains of the virus (28, Swiss-Prot release 38). Variation of Cys¹²³ to Trp has been shown to be functionally acceptable but it abolishes covalent dimerization in the ectodomain (30-32). The data presented herein formally confirm that Cys¹²³ participates in covalent dimerization of NDV HN. Alignment of several paramyxoviridae HN sequences has revealed that most cysteine residues are highly conserved within this family (3, 4, 26). For example, the analogues of Cys¹⁷², Cys¹⁹⁶, Cys²³⁸, Cys²⁵¹, Cys³⁴⁴, Cys⁴⁵⁵, Cys⁴⁶¹, Cys⁴⁶⁵, Cys⁵³¹, and Cys⁵⁴² are all invariant in simian virus 5, Sendai virus, mumps virus, and human parainfluenza viruses 1, 2, and 3. This high degree of evolutionary conservation also suggests that these disulfides are likely to be important determinants of the fold and function of the HN protein. The analogues of Cys¹⁸⁶ and Cys²⁴⁷ are only found in simian virus 5 and human parainfluenza virus 2 but are found in concert suggesting that they are likely to be in linkage (26). Site-directed mutagenesis of cysteine residues and probing with conformationally sensitive antibodies have suggested possible Cys¹⁷²/Cys¹⁹⁶ and Cys⁵³¹/Cys⁵⁴² linkages and linkages within the Cys¹⁸⁶/(Cys²³⁸,Cys²⁴⁷,Cys²⁵¹/) and Cys³⁴⁴/(Cys⁴⁵⁵,Cys⁴⁶¹,Cys⁴⁶⁵/) clusters (31). An extension of this work (52) provided evidence that Cys⁵³¹/Cys⁵⁴² linkage may prevent glycosylation occurring at the intervening consensus site at Asn⁵³⁸. The experimentally determined disulfide bonding pattern in the present study confirms the suggestions arising from these alignment and mutagenesis studies, elucidates the exact linkages in the Cys¹⁸⁶/(Cys²³⁸,Cys²⁴⁷,Cys²⁵¹/) and Cys³⁴⁴/(Cys⁴⁵⁵,Cys⁴⁶¹,Cys⁴⁶⁵/) clusters, and confirms that a consensus site for N-linked glycosylation at Asn⁵³⁸ was not occupied (52).

The present results also impact on disulfide assignments made based on the proposed three-dimensional models of the paramyxoviridae HN. In the model of Epa (27), it was considered that Cys¹⁷²/Cys¹⁹⁶ and Cys⁵³¹/Cys⁵⁴² disulfide linkages were structurally feasible as previously predicted by Colman et al. (26) and as confirmed by our results. Colman et al. (26) had also predicted the Cys¹⁸⁶/Cys²⁴⁷ linkage based on their co-existence in NDV, simian virus 5, and parainfluenza virus 2 HN sequences. However, in the model of Epa (27) these two residues were considered to be too distant for linkage to be feasible although some uncertainty in the alignment of the model was noted in this region. The results of the present study clearly confirm the original prediction regarding this linkage (26) and indicate the need for further refinement of the subsequent model (27). The model of Langedijk et al. (9) predicted linkage Cys¹⁷²/Cys¹⁹⁶, Cys¹⁸⁶/Cys²⁴⁷, Cys²³⁸/Cys²⁵¹, and Cys⁵³¹/Cys⁵⁴² linkages, in accord with our experimentally determined linkages, but their prediction of Cys³⁴⁴/Cys⁴⁶⁵ and Cys⁴⁵⁵/Cys⁴⁶¹ linkages was incorrect.

The experimentally determined disulfide linkages of the NDV HN protein are shown in Fig. 2B mapped onto a depiction of the model of Langedijk et al. (9). Our data are generally consistent with the model of Langedijk et al. (9) in structural terms. For example, the Cys⁵³¹/Cys⁵⁴² linkage is consistent with the prediction that these residues are in close proximity on adjacent strands in the 6 sheet. The flexible loop regions of the model contain three of the experimentally determined disulfide linkages (Cys¹⁷²/Cys¹⁹⁶, Cys³⁴⁴/Cys⁴⁶¹, and Cys⁴⁵⁵/Cys⁴⁶⁵) indicating broad compatibility of the model with our data even though their predictions regarding linkages within the Cys³⁴⁴/(Cys⁴⁵⁵, Cys⁴⁶¹, Cys⁴⁶⁵/) cluster were incorrect. However, as noted by those authors, Cys⁴⁵⁵, Cys⁴⁶¹, and Cys⁴⁶⁵ are predicted to occur in a long, distended loop region between -sheets 4 and 5 making it difficult to predict the three-dimensional arrangement of these residues. The model is also plausible for the Cys¹⁸⁶/Cys²⁴⁷ linkage, which is predicted to occur between the second strand of the 1 sheet and the flexible loop connecting strands 1 and 2 of the 2 sheet. Less plausible is the mapping of the Cys²³⁸/Cys²⁵¹ linkage, which would appear to place these two residues at some distance. The model of Epa (27) is similar to that of Langedijk et al. (9) in how the experimentally determined disulfide linkages map onto the fold of the protein, the major differences being that the Cys²³⁸ and Cys²⁵¹ residues are now in a more plausible juxtaposition for linkage but, as noted above, Cys¹⁸⁶ and Cys²⁴⁷ are in a less feasible juxtaposition. Thus, whereas both models are broadly compatible with our experimentally determined disulfide linkages, some refinement of both models is required. The present study highlights the importance of direct determination of disulfide bond linkage as a means of evaluating theoretically predicted models of protein folding and indicate that determination of the three-dimensional structure of an HN will be required to provide the refined structural detail. Given the high degree of conservation of cysteine residues in the rublavirus and paramyxovirus HN sequences, there is every reason to expect that the HN of other viruses in this family should have the same arrangement of disulfide bonds as determined in the present study.

Based on the predicted HN structure, Langedijk et al. (9) extended their modeling studies to the H of the morbillivirus genus and concluded that the H also folds into a -sheet propeller structure (9). They further predicted the existence of a sialidase active site in H and produced limited data for a highly specific sialidase activity associated with the morbilliviruses, rinderpest virus, and peste des petits ruminants virus. Given the errors and some uncertainty regarding rationalization of the HN disulfide arrangement in terms of their predicted structure, it would be informative to determine whether their predicted disulfides in H are correct.

	ACKNOWLEDGEMENTS

We thank Greg Neumann and Phillip Strike for assistance with ESI-MS/MS and automated Edman degradation sequencing, respectively. The avirulent strain of NDV used in this study was prepared by Dennis Grix of the Victorian Institute of Animal Science, Attwood, Victoria, Australia. Critical review of this manuscript by Michael Lawrence and Neil McKern is appreciated. The participation of James Pitt in this study was made possible by generous provision of study leave from the Royal Children's Hospital, Parkville, Victoria, Australia.

	FOOTNOTES

^* This work was presented in part at the XIth International Congress of Virology, Sydney, Australia, August 9-13, 1999.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The on-line version of this article (available at http://www.jbc.org) contains supplemental material including mass spectra presented in Figs. S1-S6.

To whom correspondence should be addressed: Biomolecular Research Institute, 343 Royal Parade, Parkville, VIC 3052, Australia. Tel.: 61-3-9662-7100; Fax: 61-3-9662-7301; E-mail: jeff.gorman@bioresi. com.au.

	ABBREVIATIONS

The abbreviations used are: NDV, Newcastle disease virus; HN, hemagglutinin neuraminidase glycoprotein; H, hemagglutinin; PNGase F, peptide-N-glycosidase F; TCEP, tris(2-carboxyethyl)phosphine; MUNANA, 2'-(4-methylumbelliferyl)--D-N-acetyl-neuraminic acid; RP-HPLC, reverse-phase high performance liquid chromatography; HNE, HEPES-NaCl-EDTA; PAGE, polyacrylamide gel electrophoresis; MALDI-TOF-MS, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry; PSD, post-source decay; ESI, electrospray ionization; ESI-MS, electrospray ionization-mass spectrometry; MS/MS, tandem mass spectrometry.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				J. R. White, V. Boyd, G. S. Crameri, C. J. Duch, R. K. van Laar, L.-F. Wang, and B. T. Eaton Location of, immunogenicity of and relationships between neutralization epitopes on the attachment protein (G) of Hendra virus J. Gen. Virol., October 1, 2005; 86(10): 2839 - 2848. [Abstract] [Full Text] [PDF]

				A. Panda, S. Elankumaran, S. Krishnamurthy, Z. Huang, and S. K. Samal Loss of N-Linked Glycosylation from the Hemagglutinin- Neuraminidase Protein Alters Virulence of Newcastle Disease Virus J. Virol., May 15, 2004; 78(10): 4965 - 4975. [Abstract] [Full Text] [PDF]

				T. P. Wallis, C.-Y. Huang, S. B. Nimkar, P. R. Young, and J. J. Gorman Determination of the Disulfide Bond Arrangement of Dengue Virus NS1 Protein J. Biol. Chem., May 14, 2004; 279(20): 20729 - 20741. [Abstract] [Full Text] [PDF]

				Y.-M. She, S. Haber, D. L. Seifers, A. Loboda, I. Chernushevich, H. Perreault, W. Ens, and K. G. Standing Determination of the Complete Amino Acid Sequence for the Coat Protein of Brome Mosaic Virus by Time-of-Flight Mass Spectrometry. EVIDENCE FOR MUTATIONS ASSOCIATED WITH CHANGE OF PROPAGATION HOST J. Biol. Chem., June 1, 2001; 276(23): 20039 - 20047. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Supplemental Material Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Pitt, J. J. Articles by Gorman, J. J. Search for Related Content PubMed PubMed Citation Articles by Pitt, J. J. Articles by Gorman, J. J. Social Bookmarking                      What's this?