Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M312907200 on February 23, 2004

J. Biol. Chem., Vol. 279, Issue 20, 20729-20741, May 14, 2004
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
279/20/20729    most recent
M312907200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wallis, T. P.
Right arrow Articles by Gorman, J. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wallis, T. P.
Right arrow Articles by Gorman, J. J.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Determination of the Disulfide Bond Arrangement of Dengue Virus NS1 Protein*

Tristan P. Wallis{ddagger}, Chang-Yi Huang§, Subodh B. Nimkar¶, Paul R. Young§, and Jeffrey J. Gorman{ddagger}||

From the {ddagger}Institute for Molecular Bioscience and §Department of Microbiology and Parasitology, The University of Queensland, St. Lucia Queensland 4072, Australia and Applied Biosystems, Foster City, California 94404

Received for publication, November 26, 2003 , and in revised form, January 29, 2004.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
The 12 half-cystines of NS1 proteins are absolutely conserved among flaviviruses, suggesting their importance to the structure and function of these proteins. In the present study, peptides from recombinant Dengue-2 virus NS1 were produced by tryptic digestion in 100% H216O, peptic digestion in 50% H218O, thermolytic digestion in 50% H218O, or combinations of these digestion conditions. Peptides were separated by size exclusion and/or reverse phase high performance liquid chromatography and examined by matrix-assisted laser desorption ionization-time of flight mass spectrometry, matrix-assisted laser desorption ionization post-source decay, and matrix-assisted laser desorption ionization tandem mass spectrometry. Where digests were performed in 50% H218O, isotope profiles of peptide ions aided in the identification and characterization of disulfide-linked peptides. It was possible to produce two-chain peptides containing C1/C2, C3/C4, C5/C6, and C7/C12 linkages as revealed by comparison of the peptide masses before and after reduction and by post-source decay analysis. However, the remaining four half-cystines (C8, C9, C10, and C11) were located in a three-chain peptide of which one chain contained adjacent half-cystines (C9 and C10). The linkages of C8/C10 and C9/C11 were determined by tandem mass spectrometry of an in-source decay fragment ion containing C9, C10, and C11. This disulfide bond arrangement provides the basis for further refinement of flavivirus NS1 protein structural models.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Globally, Dengue virus is an important human pathogen, with perhaps half of the world's population geographically at risk of infection, and with an estimated 50–100 million new infections, 500,000 hospitalizations, and 25,000 deaths annually (1–5). Dengue virus belongs to the flavivirus genus, one of three genera within the Flaviviridae family of viruses, which include pestiviruses and hepaciviruses (6). Flavivirus virions consist of a single positive stranded RNA genome of ~11 kb (7). Flavivirus RNA encodes a single large polypeptide NH2-C-prM-E-NS1-NS2A-NS2B-NS3-NS4A-NS5-COOH (7), which is cotranslationally and post-translationally processed into mature viral proteins (7). The RNA is encapsidated within an icosahedral core of capsid (C)1 proteins (8, 9), and further enveloped in a host cell-derived lipid bilayer studded with membrane and envelope glycoproteins (10). The remaining proteins (NS1, NS2A, NS2B, NS3, NS4A, NS4B, and NS5) are non-structural and are involved in various aspects of viral replication (7).

Flavivirus NS1 proteins exhibit a high degree of sequence homology (11), and contain no amino- or carboxyl-terminal membrane anchoring sequences (12, 13). However, the precise roles of NS1 in the flavivirus life cycle remain unclear. Immature NS1 exists as a hydrophilic monomer in the endoplasmic reticulum lumen, and is rapidly processed into a stable hydrophobic non-covalent homodimer, with the subunits interacting via their carboxyl termini (14, 15). NS1 exists predominantly in the dimeric form, which is associated with intracellular and cell surface membranes (16). A component of this cell surface association has been shown to be covalently linked via a glycosylphosphatidylinositol anchor (17). NS1 is also secreted from infected cells as a soluble, detergent-labile hexamer (18, 19). Maturation of NS1 involves N-linked glycosylation with Dengue virus NS1 from all four serotypes containing two conserved N-linked glycosylation sites. When expressed in mammalian cells the cell-associated form has both of these sites occupied by high mannose moieties, whereas the secreted form comprises complex glycans on one site (implying post-translational glycan processing in the Golgi) with the other remaining in a high mannose form. Other flavivirus NS1 proteins may contain an additional occupied site (14, 16). Glycosylation of NS1 contributes to dimer stability and resultant membrane association because of increased hydrophobicity of the dimer (14, 20). Mutation of glycosylation sites has been shown to reduce viral RNA production and attenuate neurovirulence, suggesting a role for glycosylated NS1 dimers in viral RNA replication (20–25). NS1 dimers have been shown to interact with a number of other non-structural viral proteins and, via this association, with the viral RNA, and may be involved in assembly of the viral replicase complex and its localization to cytoplasmic membranes (23, 26–28). NS1 also induces a protective host immune response, however, anti-NS1 antibodies do not function in a neutralizing capacity as NS1 is not present in flavivirus virions. It is likely that antibodies targeted to cell surface-associated NS1 potentiate complement mediated lysis of infected cells (29).

The importance of NS1 to the flavivirus life cycle and immune response makes it an attractive target for development of immune-based and structure/function-based antiviral therapeutics. Mature Dengue virus NS1 contains 352 amino acid residues in a base polypeptide of ~40 kDa, with glycosylation increasing the apparent mass of the protein on SDS-PAGE (30). Dengue virus NS1 contains 12 cysteine residues that are absolutely conserved among all flavivirus NS1 proteins (7, 31, 32), indicating their importance to the structure and function of the protein. The ability to form intramolecular disulfide linkages, particularly in the carboxyl terminus of the protein, appears to be crucial for NS1 dimer formation and subsequent trafficking within and secretion from the cell (21). Disulfide-bonded NS1 monomers form non-covalent dimers that are stable to reduction but can be dissociated by heating, indicating that disulfide bonds are required to stabilize the structure of NS1 monomers for them to initially associate as dimers, but are not required for subsequent dimer stability (21). In the absence of crystallographic data, the disulfide bond configuration and glycosylation pattern may provide valuable information on protein folding and hence facilitate refinement of molecular models (33, 34).

The disulfide linkage arrangement of the NS1 protein of the related Murray Valley encephalitis (MVE) flavivirus has been partially described (35) by a combination of tryptic cleavage, reverse phase high performance liquid chromatography (rpHPLC) to isolate tryptic disulfide-linked peptides, and peptide analysis by Edman protein sequencing and/or electrospray ionization mass spectrometry (ESI-MS). Using this technique, the disulfide linkage pattern of the 6 half-cystines from the amino terminus of MVE NS1 was determined to be C1/C2, C3/C4, and C5/C6. The disulfide linkage pattern of the 6 halfcystines from the carboxyl terminus was not determined, because of the lack of amenable tryptic cleavage sites in this region of the protein, and a consequent inability to obtain disulfide-linked tryptic peptides.

Similar approaches were used in this study to confirm that the 3 amino-terminal disulfide linkages of Dengue virus NS1 were the same as for MVE NS1. In addition, pepsin and thermolysin were used to overcome the resistance of the carboxylterminal half of NS1 to tryptic digestion. Recombinant Dengue-2 virus NS1 was digested with these enzymes in buffers containing 50% H 182O to identify disulfide-linked peptide ions during mass spectrometry (MS), by virtue of their characteristic isotope profiles (36, 37). Disulfide-linked peptides were isolated by size exclusion HPLC (secHPLC), further purified where necessary by capillary HPLC (capHPLC), and analyzed by matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS), matrix-assisted laser desorption ionization post-source decay (MALDI-PSD), and matrix-assisted laser desorption ionization tandem mass spectrometry (MALDI-MS/MS). Together these data confirmed the previously determined disulfide linkages of the NS1 amino terminus and allowed determination of the remaining carboxyl-terminal disulfide linkages. These techniques enabled the definition of the complete disulfide linkage arrangement of Dengue virus NS1 and will contribute to the refinement of molecular models of this important protein.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Reagents
All reagents used in this study were of analytical or HPLC grade.

Plasmid Constructs
Generation of Baculovirus Shuttle Vector Encoding NS1—C6/36 insect cells were infected with Dengue-2 virus strain PR159 at a multiplicity of infection of 1.0 and incubated at 32 °C for 3 days. Infected cell RNA was then extracted with RNazol B (Tel-Test Inc.) according to the manufacturer's instructions. cDNA was generated from this infected cell RNA extract using the NS1RBglII reverse primer 5'-ATTAGATCTCAGGCTGTGACCAAGGAGTTGAC-3' (restriction sites are underlined), which was then used as template in a PCR employing this primer and NS1FBglII, 5'-ATTAGATCTCGGATAGTGGTTGCGTTGTGAGC-3'. The resulting PCR product comprised the full-length NS1 sequence flanked by BglII sites for cloning into the baculovirus shuttle vector pVTBac.His (generated and kindly provided by Dr. A. Khromykh). Cloning into the BglII site of this vector results in the expression by the final recombinant baculovirus of a fusion protein comprising a melittin signal sequence followed immediately by a 6-histidine tag and then the cloned insert. Such fusion proteins are targeted to the endoplasmic reticulum of infected insect cells for secretion as NH2-terminal His-tagged species. The PCR product was digested with BglII and ligated into BglII cut pVTBac.His. Selected recombinants were analyzed by automated sequence analysis to confirm correct orientation and sequence.

Generation of Recombinant Baculovirus, and Expression and Purification of NS1—The pVTBac.His.NS1 shuttle vector was co-transfected with linearized (digested with Bsu36I) BacPak6 DNA (Clontech) into Spodoptera frugiperda (Sf9) cells according to the manufacturer's instructions. Recombinant baculoviruses liberated into the culture media were isolated by plaque analysis and blue-white selection using 5-bromo-4-chloro-3-indolyl-{beta}-D-galactopyranoside (X-gal). Several plaque picks were amplified in culture and expression of NS1 assessed by immunoblot analysis using NS1-specific monoclonal antibodies with the final selected recombinant designated AcHis.D2.NS1. For high level recombinant protein expression, multiple 800-ml suspension cultures of HIGH FIVE (Invitrogen, Australia) cells in 2-liter flasks and at an initial concentration of 2 x 106 cells/ml were infected with AcHis.D2.NS1 virus at a multiplicity of infection of 1.0. Cells were grown in Hyclone HyQ SFX (Hyclone) serum-free insect cell media at 28 °C and shaken at 120 rpm on an orbital shaker. The cells were harvested and collected by centrifugation at 4,000 x g at 4 °C. Supernatant containing the secreted recombinant NS1 protein was filtered through a 0.2-µm Millipore filter unit (Millipore, Australia) at room temperature. The flow-through was then subjected to ultrafiltration and buffer exchange through two Vivaflow200 (Sartorius, Australia) units at room temperature. Final concentration and buffer exchange reduced the original harvest volumes that were in excess of 2 liters to 50 ml in a buffer comprising 1 mM imidazole, 0.15 M NaCl, and 20 mM Tris, pH 8.0. This concentrate was applied to a 2-ml nickel-nitrilotriacetic acid (Qiagen) column at a flow rate of 0.5 ml/min. The column was washed with 20 mM imidazole, 0.15 M NaCl, 20 mM Tris, pH 8.0, and then eluted with 100 mM imidazole, 0.15 M NaCl, and 20 mM Tris, pH 8.0, at room temperature. Elution fractions were analyzed for purity by 15% SDS-PAGE, pooled, and concentrated through a Centricon spin column. The yield of purified AcHis.D2.NS1 was routinely ~1 mg/liter.

Fast Protein Liquid Chromatography (FPLC)
Pooled fractions of nickel-nitrilotriacetic acid-purified NS1 were concentrated by Vivaspin with 50-kDa molecular mass cut-off (Sartorius, Germany) at room temperature and passed through a 0.2-µm filter before loading onto a Superdex 200 HR 10/30 column (Amersham Biosciences) linked to an ÄKTATM FPLCTM system (Amersham Biosciences). FPLC was performed in 20 mM Tris, 0.15 M NaCl, pH 8, with a flow rate of 0.6 ml/min at 4 °C. The column was calibrated in a separate run using High and Low FPLCTM standard molecular masses of 440, 232, 67, 43, 25, and 13.7 kDa (ferritin, catalase, bovine serum albumin, chymotrypsinogen A, and ribonuclease A, respectively, from Amersham Biosciences). The elution profile data were collected and analyzed by UNICORNTM version 2.31 (Amersham Biosciences). A standard plot was obtained by plotting the diffusion constant Kav, which was calculated from the equation Kav = (elution volume–void volume)/(column bed volume–void volume), versus log Mr (relative weight). The standard curve was computed using EXCELTM (Microsoft) and a standard formula was generated as log Mr = –3.2991 x (Kav) + 3.2041. The elution volume for NS1 in this column was 11.53 ml, which was used to obtain a Kav = 0.2404 and an Mr of 257,600.

Antibody Binding Assays
The secreted recombinant NS1 was assayed against an extended panel of both linear sequence and conformational specific monoclonal antibodies (38). Each binding experiment was done in triplicate and the volume used per well was 50 µl. Immulon 4 microtiter plates were coated with 0.01 mg/ml Protein A (Pharmacia) in coating buffer overnight at 4 °C. After washing the plates with 0.05% Tween 20 in phosphate-buffered saline (PBST), 150 µl of blocking solution of 1% gelatin in phosphate-buffered saline (PBS) was added to each well and left at room temperature for 1 h. The plates were washed with PBST and then 0.018 mg/ml rabbit anti-mouse IgG (H&L) in PBST, 0.25% gelatin was allowed to bind to the Protein A for 1 h at 37 °C. Following four washes with PBST, relevant monoclonal antibodies (diluted 1/100 in PBST, 0.25% gelatin) were added and incubated at 37 °C for 1 h. After removing monoclonal antibodies and washing, immunoaffinity purified 35S-labeled NS1 (2,000 cpm/well) in PBST, 0.25% gelatin was added to the plate and incubated at 37 °C for 1 h. The plate was then washed four times with PBST, and disruption solution (2% SDS, 1% 2-mercaptoethanol in SDS-PAGE sample buffer) was added to each well. After 10 min incubation at room temperature the disruption solution was removed individually from each well, added to 1.0 ml of scintillation fluid (Optiphase HiSafe II, Wallac), and counted in a scintillation counter (Amersham Biosciences).

Proteolytic Digestions
Trypsin Digestion—100 µg of NS1 was co-precipitated with 2 µg of trypsin (Roche modified sequencing grade) by addition of 10 volumes of methanol at –20 °C, and incubation at –20 °C overnight. The precipitate was pelleted by centrifugation at 12,000 x g for 10 min at 4 °C in a microcentrifuge, air dried, resuspended in 20 µl of freshly prepared 100 mM ammonium bicarbonate, pH 8.0, in H 162O, and incubated for 2 h at 37 °C. An additional 3 µg of trypsin was then added, to give a final trypsin:NS1 ratio of 1:20 (w/w), and the reaction incubated for a further 3 h at 37 °C. Digestion was terminated by storage at –20 °C. Tryptic subdigestions of isolated peptides were performed by resuspending dried peptides in 20 µl of 100 mM ammonium bicarbonate in H 162O containing trypsin at 5 ng/µl, and incubating at 37 °C for 3 h.

Pepsin Digestion—100 µg of NS1 was precipitated by addition of 10 volumes of methanol at –20 °C, and incubation was at –20 °C overnight. The precipitate was pelleted and air dried as described above, and resuspended in 20 µl of buffer containing 100 mM formic acid, 100 mM acetic acid, 1 µg of pepsin (Sigma number P6887) and H 182O (97%, Enritech, diluted to give a final concentration of 50% (v/v)). Digestion was conducted at 37 °C for 3 h and terminated by storage at –20 °C or immediate secHPLC as described below.

Thermolysin Digestion—100 µg of NS1 was precipitated, pelleted, and air dried as described above, and resuspended in 20 µl of buffer containing 100 mM ammonium bicarbonate, pH 8, 10 mM CaCl2, 5 µg of thermolysin (Roche number 161586) and H 182O at a final concentration of 50% (v/v). Digestion was conducted at 65 °C for 2 h, and terminated by storage at –20 °C.

HPLC
All peptide chromatography was performed on Agilent 1100 liquid chromatography systems running Chemstation software, with peptide absorbance monitored at 214 nm.

SecHPLC was performed using a 3.2 mm x 30-cm Superdex Peptide column (Amersham Biosciences). Peptides were separated over 90 min in 0.05% (v/v) aqueous trifluoroacetic acid in 6% (v/v) aqueous acetonitrile at 30 µl/min.

Peptides were separated by rpHPLC using a 90-min gradient from 0.05% (v/v) aqueous trifluoroacetic acid to 0.045% (v/v) trifluoroacetic acid in 80% (v/v) aqueous acetonitrile. Flow rates of 40 and 5 µl/min were used where rpHPLC was performed using a 1 mm x 25-cm C18 column (Vydac) and 0.5 mm x 15-cm C18 capHPLC column (Zorbax), respectively. A constant temperature of 25 °C was used for capHPLC. To prevent unwanted exchange of 18O from labeled peptides with aqueous buffers, collected peptides were immediately freeze dried following chromatography using a centrifugal evaporator (Heto) and stored dried at –20 °C.

Mass Spectrometry
MALDI-TOF-MS—Reduced and non-reduced samples were analyzed using a Bruker Reflex MALDI-TOF mass spectrometer in positive ion reflector mode. Data were acquired and analyzed using the Bruker XMass suite of software as described previously (39). 2,6-Dihydroxyacetophenone (DHAP, Fluka)/diammonium hydrogen citrate (Fluka) matrix was prepared as described previously (39). Samples were diluted 1:5 in 33% (v/v) acetonitrile, 0.1% (v/v) trifluoroacetic acid prior to analysis. 1–2 µl of diluted sample was mixed with an equivalent volume of matrix, 1 µl deposited on a Bruker Scout 26 MALDI target, and allowed to air dry for 10 min before analysis. Peptides were reduced with Tris(2-carboxyethyl)phosphine (TCEP, Pierce) and analyzed using a previously described modification of the DHAP/diammonium hydrogen citrate protocol (34, 39).

MALDI-PSD—Analysis was performed using the Bruker Reflex mass spectrometer and samples were prepared in {alpha}-cyano-4-hydroxycinnamic acid (CHCA) matrix essentially as described previously (40, 41). Variation to the previously described procedures involved use of a 100-ns delay before extraction of ions from the source, and data acquisition using a 2 GHz digitizer.

MALDI-MS/MS—Analysis was performed using a Q-STAR XL hybrid Quadrupole quadrupole (Qq)-TOF-MS equipped with an O-MALDI 2 source (Applied Biosystems MDS Sciex). All samples were dissolved in 50% (v/v) acetonitrile, 0.1% (v/v) trifluoroacetic acid and crystallized with CHCA matrix (Agilent) by spotting 0.5 µl of matrix + 0.5 µl of sample and allowing air drying at ambient temperature. TOF-MS data were acquired in the mass range of 1000–4000 atomic mass units. The mass spectrometer was calibrated using Glu-fibrinopeptide B in MS/MS mode. Subsequent MS and MS/MS data on the samples were acquired using external calibration. In this external calibration mode ~30 ppm or better mass accuracy was obtained with mass spectrometer resolution of ~10,000. The O-MALDI 2 source is equipped with a N2 laser that was operated at 7–8 µJ. The same laser power (energy) was used for MS and MS/MS experiments. MS/MS data were acquired with slightly open resolution (about 4-atomic mass unit wide open window for precursor selection in Quadrupole 1) to allow transmission of all the isotopes of the high mass peptide (2858 atomic mass units) for fragmentation. The collision energy (Q0–RO2) used varied depending on peptide mass and was 90 eV for the peptide at 2858 atomic mass units, 75 eV at 1700 atomic mass units, and around 70 eV at 1500 atomic mass units. Acquisition time (specified as accumulation time in the software) for one spectrum was 1 s for MS and MS/MS, and all the data were acquired in MCA mode (multiple channel addition or summed spectra) over a period of 30 s to 2 min. Analyst QS software was used for all the data acquisition software. Under the normal operating conditions of the O-MALDI 2 source, very little in-source fragmentation of tryptic peptides is observed. In this case of disulfide-linked peptides some insource fragmentation did occur and can be seen in MS spectra. Most of the disulfide-linked peptide remained intact for MS/MS analysis. For MS/MS/MS experiments slightly higher laser power (8–9 µJ) was used. Because of the orthogonal nature of this TOF instrument use of higher laser power has no effect on instrument resolution or mass accuracy.

Sequence Analysis and Disulfide Linkage Nomenclature
Peptide and fragment ion masses were localized to the NS1 sequence using PAWS freeware version 8.4 (Proteometrics), and Bioanalyst (Applied Biosystems) software. The 12 half-cystines of NS1 are indicated as C1 to C12 (starting from the amino proximal half-cystine), with interchain disulfide linkages indicated with a forward slash, C3/C4, C5/C6, and so on (34). Brackets are used to indicate that alternative disulfide linkages were possible. C8/[C9,C10]/C11 thus indicates a three-chain peptide where separate peptides containing C8 and C11 are linked to a peptide containing C9 and C10, although the exact disulfide linkage pattern is not defined.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Preliminary Analysis of NS1—When purified recombinant Dengue-2 virus NS1 used in this study was analyzed by gel filtration FPLC (Fig. S1A, Supplementary Materials), it eluted with an apparent Mr of 257,600, consistent with monomers of Mr 43,000 (Mr 40,000 base polypeptide plus glycans) associating as a hexamer. Examination of this material by electron microscopy showed 12-nm particles identical to those of native NS1 hexamers (data not shown). Subsequent analysis of hexameric recombinant NS1 by reducing SDS-PAGE, showed it to migrate with an apparent Mr of ~85,000 (Fig. S1B, Supplementary Materials). This is consistent with dissociation of the hexamers to a stable dimeric form of NS1, and indicated that the component monomers of the dimer interacted via a mechanism that was unaffected by reducing conditions at room temperature. A number of fainter lower molecular weight species were also observed by SDS-PAGE. Analysis of these bands by in-gel tryptic digestion and MALDI-TOF-MS (data not shown) yielded peptides corresponding to the NS1 sequence from the Mr 85,000 species and a lower molecular weight species of Mr 66,000. The species at approximately Mr 45,000 and Mr 40,000 failed to yield NS1 peptides, and apparently represented low-level contaminants. Boiling of the NS1 prior to analysis by SDS-PAGE caused the apparent dissociation of the dimer (data not shown), in agreement with other observations of the heat lability of NS1 dimers (15). In addition to being hexameric, secreted recombinant NS1 reacted with an extended panel of both linear sequence and conformational specific monoclonal antibodies (Fig. S1C, Supplementary Materials), in an equivalent manner to its native counterpart (38). The NS1 with the characteristics described above was used in the subsequent analyses without any further purification.

C3 Is Linked to C4 —NS1 was digested with trypsin in 100% H 162O, and the resultant peptide mixture was separated by rpHPLC. Individual rpHPLC fractions were collected, and examined by MALDI-TOF-MS in both non-reduced form and after reduction with TCEP. Only one fraction yielded data that could be reconciled with a defined disulfide linkage.

A peptide that eluted at ~58 min was observed by MALDI-TOF-MS as a series of ions of average m/z = 5199.0–5402.2–5605.3–5770.2 (Fig. 1A) in an non-reduced form. The mass spacing of m/z = 203–203 was indicative of a glycopeptide containing varying numbers of N-acetylhexosamine (Mr = 203) sugars. The m/z = 165 spacing is not indicative of any obvious glycosylation event, with hexose sugars such as mannose (Man) having the closest comparable Mr of 162. Following reduction, this ion series was replaced with a series of average m/z = 4117.3–4320.6–4523.7–4688.8, with the same peptide spacing of 203–203-165, and a single ion of m/z = 1084.4 (Fig. 1B). These data are consistent with tryptic peptide Ala59-Arg68 (C3), with a calculated monoisotopic MH+ of 1084.52, disulfide-linked to Met133-Arg158 (C4), and are in agreement with the C3/C4 pairing reported for MVE NS1. Met133-Arg158 spans the first NS1 glycosylation site, and is the only cysteinyl glycopeptide expected from tryptic cleavage of Dengue-2 virus NS1. Production of cysteinyl glycopeptides spanning the second glycosylation site at Asn217 would have required several missed tryptic cleavages resulting in peptides with masses greater than 4000. Unglycosylated Met133-Arg158 has a calculated MH+ of 2931.35. Occupancy of the N-linked glycosylation site of this peptide with a core glycan structure (HexNAc)2Man3 would yield a peptide with a calculated MH+ of 3823.40. A peptide at this m/z was not observed, however, addition of 2 fucose (Fuc) residues would produce a peptide with a calculated monoisotopic MH+ of 4115.41 (average mass 4117.9), in reasonable agreement with the observed peptide of average m/z = 4117.0. Additional GlcNAc and Man residues would contribute to the observed peptide spacings. This proposed glycan structure is consistent with known core and antennary fucosylation of glycoproteins produced by baculovirus-infected insect cells (42).



View larger version (22K):
[in this window]
[in a new window]
  		FIG. 1.
MALDI-TOF-MS analysis of disulfide-linked tryptic C3/C4 glycopeptide. The fraction that eluted at 58 min during rpHPLC of 100% H 16O tryptic digest of NS1 was analyzed by 2MALDI-TOF-MS using DHAP matrix before (A) and after (B) reduction with TCEP. The non-reduced ion series of interest and the sequence and putative glycan structure corresponding to the C3/C4 glycopeptide are indicated. {blacksquare}, N-acetylglucosame; , mannose; {blacktriangleup}, fucose.

 
C5 Is Linked to C6 —NS1 was digested with pepsin in 50% H 182O and the peptides separated by secHPLC (Fig. S3, Supplementary Materials). Eight individual secHPLC fractions were collected and analyzed by MALDI-TOF-MS. Potential disulfide-linked peptide ions were identified in non-reduced fractions by virtue of their characteristic mass spectrometric isotope profiles (36, 37), and their disulfide-linked status was confirmed by their disappearance following TCEP reduction. A number of putative disulfide-linked peptides were observed in secHPLC fractions 2, 3, and 5.

An ion was observed at m/z = 2859.2 in secHPLC fraction 2 with a MALDI-TOF-MS isotope profile consistent with it being a two-chain disulfide-linked peptide. This peptide was further purified by capHPLC (Fig. 2A), and was susceptible to reduction, yielding peptide ions at m/z = 1422.4 and 1439.5 (Fig. 2B). This behavior was consistent with the assignment Leu181-Lys199 (C5) with a calculated MH+ of 1439.69 linked to Val230-Leu241 (C6), with a calculated MH+ of 1422.72. This linkage was confirmed by MALDI-PSD analysis (Fig. 3). Limited peptide bond fragmentation was evident by PSD, however, strong peaks were observed that represented two overlapping ion triplets with intratriplet spacings of m/z = 32. These data were consistent with symmetrical and asymmetrical fragmentation of the disulfide bond between the two peptide chains observed by the reduction of the parent peptide (Fig. 2). The peptide bond fragment ions of approximate m/z = 657, 2205, 2515, and 2747 supported this assignment.



View larger version (19K):
[in this window]
[in a new window]
  		FIG. 2.
MALDI-TOF-MS analysis of putative disulfide-linked peptic C5/C6 peptide. A peptide isolated from a peptic digest of NS1 in 50% H 182O by capHPLC of secHPLC fraction 2 was analyzed by MALDI-TOF-MS using DHAP matrix before (A) and after (B) reduction with TCEP. The non-reduced parent ion, reduced daughter ions, and the sequence corresponding to the non-reduced C5/C6 peptide are indicated. Asterisks indicate polymer contaminants in the sample, which were derived from the microcentrifuge tubes used.

 



View larger version (29K):
[in this window]
[in a new window]
  		FIG. 3.
MALDI-PSD analysis of the putative disulfide-linked peptic C5/C6 peptide. The non-reduced ion at m/z = 2859.2 as described in the legend to Fig. 2 was isolated by timed ion selection after ionization using CHCA matrix and the resultant metastable ions characterized by PSD analysis. The inset shows a zoomed view of the overlapping triplets produced by symmetrical/asymmetrical fragmentation of the disulfide bond. MH+ values corresponding to the observed fragment ions are indicated.

 
An ion observed at m/z = 2374.0 from secHPLC fraction 3 also displayed an isotope profile indicative of a disulfide-linked peptide (Fig. S3A, Supplementary Materials). This peptide was purified by capHPLC and upon reduction yielded two product ions at m/z 1664.7 and 712.2 (Fig. S3B, Supplementary Materials). Two possible linkage assignments were calculated that would produce similar observed parent/daughter ions: Ile9-Leu23 (C1), MH+ = 1664.82, linked to Lys24-Phe30 (C2), MH+ = 711.34; or Phe188-Leu193 (C5), MH+ = 712.33, linked to Ile228-Leu241 (C6), MH+ = 1664.85. PSD of this ion yielded a low intensity 32 m/z spaced triplet series centered around an ion of approximate m/z = 1666 (Fig. S4, Supplementary Materials). This appeared to confirm the disulfide-linked peptide status, and other fragmentation events were observed that suggested fragmentation from the Ile228-Leu241 peptide chain (Fig. S4, Supplementary Materials). Because both potential peptide assignments contained sites for tryptic cleavage, the m/z = 2374.0 peptide was further digested with trypsin, analyzed by MALDI-TOF-MS (Fig. S5, Supplementary Materials) and the resulting peptides compared with expected theoretical tryptic peptides. Tryptic subdigestion of this disulfide-linked peptide would theoretically yield a peptide consisting of Phe188-Leu193 (C5) linked to Ser232-Lys237 (C6), with a calculated MH+ of 1466.66. The strong presence of a peptide at m/z = 1466.6 supported the assignment that the m/z = 2374.0 peptide ion contained the C5/C6 linkage. The isotope profiles of the peptides were consistent with the two-chain parent peptide with 18O incorporation at both carboxyl termini losing one of the 18O-labeled termini following trypsin digestion.

C8 Is Linked to C10, C9 Is Linked to C11—Surprisingly, secHPLC fraction 5 of the above peptic digest of NS1 in 50% H 182O contained a series of relatively large peptides at m/z = 2510.7–2639.7–2740.8–2839.8–2938.9 (Fig. 4A), all with similar isotope patterns characteristic of disulfide-linked peptides. The presence of these higher mass peptides in a later eluting secHPLC fraction suggested that the peptides were in a highly compact conformation because of extensive disulfide linkages, which decreased their hydrodynamic volume and thus increased their retention time. Interaction between the peptides and the secHPLC matrix may also have contributed to the later elution time, although secHPLC was performed in 6% acetonitrile to reduce hydrophobic interactions. The m/z differences between these peptides of 129.0–101.1–99.0–99.1 correspond to a sequence tag of ETVV, resulting from ragged peptic cleavage. This sequence occurs in the NS1 sequence as 296VVTE299, which indicated that the observed peptide peaks represented cleavage variants of the same disulfide-linked peptide containing this amino-terminal sequence on one peptide chain.



View larger version (17K):
[in this window]
[in a new window]
  		FIG. 4.
MALDI-TOF-MS analysis of the disulfide-linked peptic peptide series putatively representing the C8/[C9,C10]/C11 linkage arrangement. SecHPLC fraction 5 of the peptic digest of NS1 in 50% H 182O was analyzed by MALDI-TOF-MS using DHAP matrix before (A) and after (B) reduction with TCEP. Insets show zoomed views of isotope profiles of non-reduced parent and reduced daughter peptide ions. The two possible disulfide linkage configurations for these peptides are shown, with arrowheads indicating the sites of ragged peptic cleavage that gave rise to the observed non-reduced peptide ion series. The 2 possible disulfide linkage configurations (C8/C9, C10/C11 or C8/C10, C9/C11) are represented by a "forked" disulfide bond.

 
Attempts to separate the components of this peptide series by capHPLC were not successful. Confirmation of the disulfide-linked peptide status was therefore achieved by reduction of the secHPLC peptide mixture to yield a single intense MALDI-TOF-MS peak at m/z = 1597.4, which displayed an isotope profile of a single chain peptide with 18O incorporation at one terminus only (Fig. 4B). The most likely sequence assignment for this peptide is Trp321-Tyr333 (C9, C10, and C11), with calculated MH+ = 1597.74. No ions representing the complementary daughter peptide series resulting from the ragged peptic cleavage of the other peptide chain were observed (Fig. 4B). The most likely sequence assignments for the expected complementary daughter peptide ion series were: Asp300-Leu308 (C8), MH+ = 918.40; Glu299-Leu308 (C8), MH+ = 1047.49; Thr298-Leu308 (C8), MH+ = 1148.49; Val297-Leu308 (C8), MH+ = 1247.56; and Val296-Leu308 (C8), MH+ = 1346.63.

MALDI-PSD (Fig. 5) of the m/z = 2510.7 and 2639.7 peptides in both cases resulted in a triplet of peaks of approximate m/z = 1563–1595-1628, corresponding to the oxidized form of Trp321-Tyr333 (C9, C10, C11). In addition to these common fragments, the 2510.74 and 2639.73 parents yielded 32 m/z spaced triplets centered around m/z = 918 and 1048, respectively. These data were consistent with symmetrical and asymmetrical disulfide bond cleavage, in confirmation of the disulfide-linked peptide status and in agreement with the theoretical and observed reduction data. Assuming that 2 of the 3 cysteines in the Trp321-Tyr333 (C9, C10, C11) peptide formed an intrachain disulfide link as well as the interchain link to C8, the calculated MH+ of the resulting disulfide-linked peptides is 2511.14 and 2640.23. Had the cysteine residues been unpaired, the masses of the resultant peptides would have been observed at 2 atomic mass units higher.



View larger version (25K):
[in this window]
[in a new window]
  		FIG. 5.
MALDI-PSD analysis of two putative disulfide-linked C8/[C9,C10]/C11 peptic peptides. The peptide ions at m/z = 2510.7 and 2639.7 as described in the legend to Fig. 4 were isolated by timed ion selection after ionization using the CHCA matrix and the resultant metastable ions were characterized by PSD analysis. A, MALDI-PSD of m/z = 2639.73 peptide; B, MALDI-PSD of m/z = 2510.74 peptide. Symmetrical/asymmetrical disulfide bond fragmentation events giving rise to the observed fragment ions are shown. The two possible disulfide linkage configurations (C8/C9, C10/C11 or C8/C10, C9/C11) are represented by a forked disulfide bond.

 
The closeness of the observed and theoretical peptide masses indicated that the 3 cysteine residues in the sequence of 321WC9C10RSC11TLPPLRY333 were all involved in disulfide bonds, therefore additional analysis was required to determine which were involved in intrachain and which were involved in interchain bonds. Conformational constraints would prevent a linkage between the adjacent C9 and C10 residues. The two possible linkages representing the observed peptides were therefore either C8/C9, C10/C11 or C8/C10, C9/C11. Differentiating between the two conformations required cleavage between C9 and C10. To facilitate this process the peptides were subjected to tryptic cleavage to convert them from two-chain peptides, in which one chain contained an internal disulfide loop, to three-chain peptides with C9 and C10 on one peptide chain. These three-chain peptides were all expected to contain a variable length A-chain containing C8, linked to a central WC9C10R B-chain, which is also linked to the SC11TLPPLRY C-chain (Fig. S6, Supplementary Materials).

Tryptic digestion of secHPLC fraction 5 in 100% H 162O produced peptides with masses and isotope profiles consistent with conversion of the m/z = 2510.7–2938.9 series of peptides to three-chain peptides (Fig. S6, Supplementary Materials). It was evident that peptides were consistently cleaved between C10 and C11, to yield three-chain peptides with 18O incorporation at the termini of the A- and C-chains because of the original peptic digests in 50% H 182O. Where cleavage also occurred at the arginine residue on the A-chain an additional population of three-chain peptides with 18O incorporation only at the carboxyl terminus of the C-chain was produced. The observed m/z of the cleavage products of the m/z = 2510.7 peptide (two-chain, two 18O) were 2529.1 (three-chain, two 18O) and 2175.0 (three-chain, one 18O), which corresponded well to the calculated MH+ of 2511.14 (two-chain, two 18O) cleaved to 2529.14 (three-chain, two 18O) and 2174.94 (three-chain, one 18O). These cleavage product masses would be generated regardless of whether C9 or C10 was involved in the initial interchain linkage. Expected three-chain cleavage products were observed for all of the peptides in the m/z = 2510.7–2838.9 series, providing additional evidence supporting the initial assignments for these peptides.

Initial attempts to fragment these tryptic peptides by ESI-MS/MS were not successful. In common with their peptic parent peptides, they could not be individually purified by capHPLC. Furthermore, their nanoelectrospray ionization behavior was not conductive to acquisition of ESI-MS and ESI-MS/MS spectra (data not shown). Because the peptides had been shown to ionize well when analyzed by MALDI-TOF-MS, a Qq-TOF mass spectrometer with a MALDI source was used to obtain fragmentation data. Analysis of the unfractionated peptide series derived by combined peptic and tryptic cleavage using the MALDI-Qq-TOF-MS with no collisional fragmentation revealed the presence of the same three-chain peptides as observed by MALDI-TOF-MS (Fig. S7, Supplementary Materials). These three-chain peptide ions also exhibited a tendency to undergo in-source decay (ISD) fragmentation around both of the disulfide bonds as evidenced by the presence of 32 m/z triplet peaks centered around m/z = 1049.53 and m/z = 1613.70 (Fig. S7, Supplementary Materials). These triplets, respectively, represented the common C and B + C chain products of symmetrical and asymmetrical fragmentation around either disulfide bond. A number of peptides representing the corresponding unique A and A + B chain fragments were also observed. By themselves these data did not provide any further information on the precise disulfide configuration, as the same fragment ion masses were expected for either of the two possible disulfide bond configurations, however, they did provide still further confirmation of the gross peptide chain assignment.

Collisional activated decomposition of the three-chain ion at m/z = 2858.25 was performed using MALDI-Qq-TOF-MS/MS, in an effort to fragment between C9 and C10 on the B-chain. The resulting mass spectrum was dominated by the four triplets with spacings of 32 m/z centered on m/z = 1049.52, 1247.53, 1613.69, and 1811.72, resulting from symmetrical and asymmetrical fragmentation of the two disulfide bonds, with no informative peptide bond fragments observed (Fig. 6).



View larger version (23K):
[in this window]
[in a new window]
  		FIG. 6.
MALDI-Qq-TOF-MS/MS analysis of the three-chain disulfide-linked C8/[C9,C10]/C11 peptide produced by successive cleavage with pepsin and trypsin. The three-chain peptide ion at m/z = 2858.2 as detailed in Fig. S6 (Supplementary Materials) was analyzed by MALDI-Qq-TOF-MS/MS using the CHCA matrix. Insets show zoomed views of the isotope patterns and the chain compositions and calculated MH+ values of the parent peptide and symmetrical disulfide cleavage products. The two possible disulfide linkage configurations (C8/C9, C10/C11 or C8/C10, C9/C11) are represented by a forked disulfide bond.

 
Together the MALDI-TOF-MS, MALDI-Qq-TOF-MS, and MALDI-Qq-TOF-MS/MS data showed the propensity of the three-chain peptides to undergo ISD and collisional activated decomposition fragmentation of the disulfide bonds rather than at peptide bonds. Thus, the two-chain ISD fragment ions were selected for further MS/MS fragmentation (ISD-MS/MS). The ISD fragment ion of m/z = 1581.73, representing the smaller asymmetrical B + C-chain fragment, was susceptible to fragmentation around the disulfide bond, as evidenced by an ISD-MS/MS spectrum showing the appearance of two triplet series with 32 m/z spacings, centered around m/z = 533.30 and 1049.53 (Fig. 7). These series represented B- and C chain fragments, respectively, produced by symmetrical/asymmetrical fragmentation of the B–C chain disulfide bond, with an additional series of ions centered around m/z = 886.45 representing C-chain fragments with peptide fragmentation of the carboxyl-terminal tyrosine residue.



View larger version (27K):
[in this window]
[in a new window]
  		FIG. 7.
MALDI-Qq-TOF-MS/MS of the two-chain [C9,C10]/C11 ISD fragment ion derived from the C8/[C9,C10]/C11 peptide produced by combined peptic and tryptic cleavage. The asymmetrical two-chain MALDI-Qq-TOF-MS ISD fragment ion of m/z = 1581.73, common to all C8/[C9,C10]/C11 peptic plus tryptic fragments (Fig. S7, Supplementary Materials) was selected as a parent ion for MS/MS. Symmetrical/asymmetrical disulfide cleavage events and the corresponding fragment ions are indicated. Additional fragmentation events corresponding to observed ions are also indicated. Ions corresponding to fragmentation of the peptide backbone are also indicated. Ions marked with an (i) represent internal cleavage events, see text for details.

 
In addition to disulfide fragmentation, a number of ions representing peptide bond fragmentation were evident. Peaks at 628.34 and 645.35 represented z and y ions, respectively, derived from proline directed fragmentation of PPLRY from the carboxyl terminus of the C-chain. The ion at m/z = 422.27 likely represented a C-chain fragment ion SCTL. The peak at m/z = 1425.65 represented c type fragmentation of the carboxyl-terminal Arg residue from the B-chain, similarly the ion at m/z = 1396.65 represented y type fragmentation of the aminoterminal Trp from the B-chain. Most importantly, the peak at m/z = 1356.65 represented a c type ion resulting from fragmentation between the C9 and C10 residue of the B-chain. These data indicated that C9 was linked to C11 in the C-chain, which allowed the disulfide linkage of this complex peptide to be determined as C8/C10, C9/C11.

C7 Is Linked to C12—NS1 was digested with thermolysin in 50% H 182O at 65 °C and the resultant peptide mixture was separated by secHPLC. Two peptides were observed by MALDI-TOF-MS with isotope profiles characteristic of two-chain disulfide-linked peptides in two of the fractions collected. One of these was apparent at m/z = 2836.1 in fraction 4 (data not shown), and the other at m/z = 2121.2 in fraction 11 (Fig. 8A). The very late elution time of this second peptide suggests some interaction with the secHPLC column matrix. The disulfide-linked peptide status of these peptides was confirmed by reduction and MALDI-TOF-MS analysis. The m/z = 2836.1 peptide was reduced to two peptides of m/z = 1332.2 and 1508.0 (data not shown). Assuming that all cysteines were involved in interchain and intrachain disulfide bonds, Val295-Ser307 (C8, calculated MH+ 1332.61) disulfide-linked to Ile318-Pro330 (C9, C10, C11, calculated MH+ 1508.66) would yield a disulfide-linked peptide of MH+ 2836.28, in agreement with the observed peptides and in further confirmation of the peptic C8, C9, C10, C11 data above.



View larger version (24K):
[in this window]
[in a new window]
  		FIG. 8.
MALDI-TOF-MS analysis of a disulfide-linked thermolytic peptide corresponding to a C7/C12 linkage. The secHPLC fraction 11 of the thermolytic digest of NS1 in 50% H 182O was analyzed by MALDI-TOF-MS using DHAP matrix before (A) and after (B) reduction with TCEP. Insets show zoomed views of the isotope patterns of the non-reduced parent ion and reduced daughter peptide ions. Regularly spaced peaks marked with an asterisk represent polymer contaminants.

 
The peptide at m/z = 2121.2 disappeared following reduction, with the concomitant appearance of two peptides of m/z = 1204.9 and 918.9 (Fig. 8B). The isotope profiles of the non-reduced parent and reduced daughter peptides suggested a two-chain peptide with 18O incorporation at each carboxyl terminus giving rise to two single-chain peptides with 18O incorporation at their carboxyl termini. The most likely assignment for these data was Phe287-Thr294 (C7, calculated MH+ 919.34) linked to Tyr333-Gly342 (C12, calculated MH+ 1205.46) to give a disulfide-linked peptide of calculated MH+ 2121.80. Production of these fragments was consistent with the expected thermolytic cleavage sites producing half-cystinyl peptides from NS1.

To strengthen the assignment of the m/z = 2121.2 peptide as representing Phe287-Val294 linked to Tyr333-Gly342, the peptide was subjected to MALDI-PSD (Fig. S8, Supplementary Materials). Interestingly, the 32 m/z triplets indicating ISD or PSD of the disulfide bond, which were commonly observed with NS1 peptic peptides, were not observed with this thermolytic peptide. However, asymmetric fragmentation around the disulfide bond was apparent, with ions observed at m/z ~1174 and ~1239. A number of PSD fragment ions from both of the peptide chains were also observed that were consistent with the assignment. In particular the ion at m/z ~622 represented the b ion YRGED derived from the amino terminus of Tyr333-Gly342. Similarly, the ions at m/z ~2023 and ~1860 represented c ions resulting from fragmentation of Thr and GTT residues, respectively, from the carboxyl terminus of Phe287-Thr294.

C1 Is Linked to C2 (Indirect Observations)—Peptides representing the linkage of C1 and C2 were not directly observed in tryptic or thermolytic digests. Initial peptic digests of NS1 in 50% H 182O resulted in a peptide ion of m/z = 2255.5, which was observed in a number of secHPLC fractions. After reduction, this peptide ion was observed with an m/z = 2257.5. This shift of m/z = 2 is characteristic of reduction of an intrachain disulfide linkage, and is consistent with reduction of Asp11-Phe30 (calculated reduced MH+ 2257.04), containing C1 and C2. However, with subsequent refinement of the digestion and secHPLC conditions, this peptide was not observed consistently in peptic digests that precluded subsequent confirmation of its identity. The 6-histidine tag at the amino terminus of the recombinant NS1 used in this study may have caused the failure to consistently observe C1/C2 peptides. C1 peptides generated by cleavage events that retain this tag may adhere to surfaces and be less amenable to ionization by MALDI. This proposal was supported by the fact that peptides corresponding to the 6-histidine amino terminus were not observed by MALDI-TOF-MS analysis of tryptic digests of reduced and alkylated NS1 (data not shown).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
The present report represents the first description of the complete disulfide bond arrangement for a flavivirus NS1 protein. Several of the disulfide bonds were identified by the fairly standard procedure of proteolytic cleavage into two-chain peptides followed by mass analysis before and after reduction (34, 36, 43–45) with reconciliation of mass data in terms of possible fragments from the known sequence of the protein. Support for the assignments was obtained by use of MALDI-PSD to generate fragmentation of the disulfide bond (46–48). Some of these disulfide bonds were previously characterized for the MVE NS1 protein (35). However, attempts to define the complete disulfide bond arrangement of MVE NS1 were limited by the approach used, especially the use of trypsin for proteolysis. Unfavorable sites for tryptic cleavage in the carboxyl-terminal half of the protein prevented production of suitable disulfide-linked peptides from this region. When a similar approach was attempted in this study with Dengue virus NS1, only one unequivocal disulfide bond assignment, C3/C4, was made, although the NS1 sequence reasonably suggests that assignments for C1/C2 and C5/C6 were possible by this method. Examination of the Dengue NS1 sequence indicated that the observed disulfide linkages of the 6 carboxyl-terminal cysteines would preclude generation of discrete two-chain disulfide-linked tryptic peptides (Fig. S8, Supplemental Materials). Tryptic digestion would instead be expected to produce a large multichain disulfide-linked peptide containing all 6 carboxylterminal cysteines, which would not yield disulfide linkage information (49) by simple reduction analysis.

Pepsin has a lower sequence specificity than trypsin, with cleavage likely on the amino side of Leu, Asp, Glu, Phe, Trp, and Tyr residues, with other cleavages also possible. The sequence of Dengue NS1 contains a large number of potential cleavage sites for pepsin, and pepsin digestion indeed yielded disulfide-linked peptides from the protein carboxyl terminus. he lower specificity of pepsin resulted in a complex mixture of NS1 peptides, which were partially resolved by secHPLC into size related fractions. Peptides related by "ragged" peptic cleavage of the same sequence co-eluted, and the mass differences evident by MALDI-TOF-MS constituted sequence tags that aided identification. This was exemplified in the case of the C8/[C9,C10]/C11 peptide series observed in secHPLC fraction 5, where the mass differences observed between these peptides represented an ETVV sequence tag that localized one of the peptide chains to 295VVTEDC9GNRGPSL308. Additionally, the elution of these peptides in a later secHPLC fraction than expected for their mass suggested that the peptides were in a highly compact form because of disulfide linkages. These peptides could not be further separated by capHPLC because of their very similar hydrophobicities, however, individual peptide ions were readily selected from mixtures by ion gating for MALDI-PSD and MALDI-Qq-TOF-MS/MS.

By performing the peptic digestions of NS1 in 50% H 182O, two-chain peptides incorporated 18O into both termini in a way that allowed them to be identified/characterized as disulfide-linked peptides. For example, C5/C6 and C8/[C9/C10]/C11 peptides produced by peptic digestion of NS1 in 50% H 182O exhibited isotope profiles that differentiated them from single chain peptides of similar mass (36, 37). The disulfide-linked peptide status of these peptides was confirmed by reduction and ISD/ collisional activated decomposition fragmentation of the disulfide bonds. However, determination of the precise disulfide bond configuration of the C8/[C9/C10]/C11 disulfide-linked peptide required fragmentation between the adjacent C9 and C10 residues. Analysis of disulfide-linked peptides containing adjacent cysteine residues represents a particular challenge, because the peptide bond between the residues is not readily amenable to enzymatic cleavage. Methods have been described that use partial reduction and cyanylation to facilitate basic cleavage between adjacent cysteines (50), and Edman sequencing also offers the potential to examine the linkages of adjacent cysteines (51). However, both techniques rely on obtaining sufficient peptide to provide meaningful data after the multiple reaction steps involved. We instead chose to exploit the sensitivity and peptide fragmentation ability of MS to attempt fragmentation of the adjacent cysteine residues of the C8/[C9,C10]/C11 peptide.

Failure to observe significant peptide bond fragmentation of the initially isolated C8/[C9,C10]/C11 peptide by MALDI-PSD was because of the extreme lability of the interchain disulfide bond (Fig. 5). This may have been because of the restriction of cleavage of the peptide backbone between C9 and C11 because of its location in a disulfide loop in the two-chain peptide (51), which directed cleavage to the interchain disulfide bond by PSD. However, the selective lability of the disulfide linkages to fragmentation was still observed even when the two-chain peptic peptide had been converted to a three-chain peptide by tryptic cleavage. The interchain disulfide bonds of the resultant three-chain peptide were extremely labile to ISD. Selective lability of disulfide bonds to ISD and PSD has been observed previously (46–48, 50) although not universally. We were able to exploit this phenomenon with MALDI-Qq-TOF-MS/MS by using ISD to generate a two-chain fragment containing [C9,C10]/C11, and then affecting further MS/MS fragmentation between the adjacent C9/C10 residues of this ISD fragment. ISD-MS/MS fragment ions were observed that were consistent with linkage of C9 to C11, demonstrating the linkage of C9/C11 and C8/C10.

In addition to detection of disulfide-linked peptides by altered isotope profiles, 18O labeling may also facilitate detailed analysis of complex intra- and interchain disulfide-linked peptides by providing information on the number of 18O-incorporated termini of peptide fragment ions produced by MS/MS. This is supported by previous MS/MS studies in our laboratory where 18O isotope labeling aided the detailed analysis of a cystine noose peptide from the central subdomain of respiratory syncitial virus G protein (37). Although isotope profiles of unfragmented peptides analyzed by MALDI-TOF-MS and MALDI-Qq-TOF-MS were reliable indicators of 18O incorporation and disulfide linkage status in the present study, 18O labeling did not facilitate interpretation of complex fragmentation protocols. Peptide ions resulting from fragmentation of the disulfide bond as observed in MALDI-Qq-TOF-MS and MALDI-Qq-TOF-ISD-MS/MS exhibited isotope profiles that were distorted to the extent that they could not always be used to reliably confirm their 18O content (data not shown). In these cases it was apparent that processes additional to 18O incorporation were contributing to the change in isotope profile. The incompletely understood mechanisms involved in the symmetrical and asymmetrical dissociation of disulfide bonds by ISD, PSD, and collisional activated decomposition can potentially result in populations of disulfide fragment ions with differing numbers of protons and hence an altered isotope profile. Disulfide-linked peptides may dissociate via protonation and reduction of the disulfide bond (47, 48, 53), or via electron capture dissociation of already protonated peptides (52, 53), or in the case of the MALDI-Qq-TOF instrument in this study, a combination of both mechanisms. This resulted in the isotope patterns of disulfide fragment ions being altered by both 18O incorporation and differential protonation. The multiple peptide ion gating mechanisms required to produce MALDI-Qq-TOF-ISD-MS/MS data from parent ions with broad isotope profiles produced by 18O labeling may have also contributed to distortion of the expected isotope profile. While this phenomenon represents a potential limitation of isotope profile analysis of peptide fragments produced by fragmentation of the disulfide bond by ISD and ISD-MS/MS, it proved to be no impediment to the ultimate analysis of the C8/[C9,C10]/C11 peptide.

The high degree of sequence homology and conservation of the 12 cysteine residues of flavivirus NS1 proteins strongly suggests that they have a common disulfide bond arrangement. The previous assignment of the MVE NS1 amino-terminal disulfide linkages of C1/C2, C3/C4, and C5/C6 was confirmed by this study. The direct observation of C3/C4 and C5/C6 linkages in Dengue NS1 implied the C1/C2 linkage, even though this was only tentatively observed. Similarly, the direct observation of C8/[C9,C10]/C11 linkages in pepsin, 50% H 182O digests strongly implied the linkage of C7/C12 even though disulfide-linked peptides representing this linkage were not observed in peptic digests. To confirm this implied linkage NS1 was digested with thermolysin in 50% H 182O. Analysis of the NS1 sequence suggested favorable sites for thermolytic cleavage flanking C7 and C12. Although thermolysin has an alkaline pH optimum, the potential for alkaline pH-mediated disulfide exchange (36, 37, 54) was minimized by the fact that the 12 NS1 cysteines were all involved in disulfide linkages, with no free cysteine residues to initiate exchange. Furthermore, thermolysin cleaves at the amino side of its target residues (Leu, Ile, Phe, Trp, Tyr, Val, Pro, and others). Enzymes such as trypsin that cleave on the carboxyl side of target residues can undergo multiple cycles of acyl-enzyme formation with the carboxyl terminus of peptides, and subsequent hydrolytic deacylation (36, 37, 55). In the presence of 50% H 182O this can result in additional 18O incorporation events and generate false positive single-chain peptides with 18O profiles indicative of more than one 18O incorporation. Last, thermolytic digests were carried out at 65 °C, at which the NS1 structure was potentially unfolded, with a concomitant increase in the accessibility of the enzyme to previously highly folded regions of the protein. The C7/C12 peptide was identified in thermolytic digests of NS1 by the 18O profile, and confirmed by reduction and MALDI-PSD.

Determination of the complete disulfide linkage arrangement of Dengue virus NS1 has a number of implications. Because disulfide linkages within NS1 monomers have been shown to be important to non-covalent dimerization of NS1 (21), the fact that the recombinant NS1 used in this study formed stable non-covalent dimers, which themselves associated as NS1 hexamers (Fig. S1, A and B, Supplementary Materials), indicated that it was folded correctly and stabilized by disulfide bonds analogous to those of the native Dengue virus NS1. This assumption is re-enforced by the observation that all members of an extensive panel of monoclonal antibodies specific for native NS1 (38) were reactive for this recombinant species (Fig. S1C, Supplementary Materials). Furthermore, the observation of complex glycosylation (Fig. 1) at the first NS1 glycosylation site indicated that the recombinant NS1 was secreted via the Golgi and subject to post-translational glycan processing analogous to the secreted native form (14, 20). Experimental determination of disulfide bonds can be used as indicators of potential protein folds (33). Thus the disulfide bonds of NS1 may be used to confirm and refine molecular models of this protein. The disulfide bond assignments of Dengue virus NS1 may be complemented with secondary structure calculations and epitope mapping studies (56) to develop such a model.

The complete determination of the disulfide bond arrangement of Dengue virus NS1 required separate proteolysis with trypsin, pepsin, and thermolysin, and analysis of the resultant disulfide-linked peptides with various combinations of TCEP reduction, MALDI-TOF-MS, MALDI-PSD, and MALDI-Qq-TOF-MS/MS. The process was greatly facilitated by the use of 18O labeling to identify and characterize disulfide-linked peptides from complex peptide mixtures. Given the identical role played by NS1 proteins in the flavivirus life cycle, and the fact that the disulfide bond arrangements of the amino-terminal half of MVE and Dengue virus NS1 were shown to be identical, the disulfide linkage C1/C2, C3/C4, C5/C6, C7/C12, C8/C10, and C9/C11 (Fig. 9) determined for Dengue virus NS1 is likely to apply to all flavivirus NS1 proteins. These proteins are also likely to share a similar tertiary structure. The validity of this hypothesis awaits experimental determination of the complete disulfide bond configuration of other flavivirus NS1 proteins, and the three-dimensional structure of a representative NS1 protein.



View larger version (31K):
[in this window]
[in a new window]
  		FIG. 9.
Amino acid sequence and the disulfide-linkage arrangement of the Dengue virus 2 NS1 protein. Disulfide linkages of the 12 cysteines of NS1 are indicated with dotted lines. N-Linked glycosylation sites are underlined. The amino-terminal 6-histidine tag is presented in italics.

 

   FOOTNOTES
 
* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

The on-line version of this article (available at http://www.jbc.org) contains Figs. S1–S8. Back

|| To whom correspondence should be addressed: Molecular and Cellular Proteomics Laboratory, Institute for Molecular Bioscience, The University of Queensland, St. Lucia, Queensland 4072, Australia. Tel.: 61-7-3346-2995; Fax: 61-7-3346-2101; E-mail: j.gorman{at}imb.uq.edu.au.

1 The abbreviations used are: C, capsid protein; NS, non-structural; capHPLC, capillary high performance liquid chromatography; CHCA, {alpha}-cyano-4-hydroxycinnamic acid; DHAP, 2,6-dihydroxyacetophenone; FPLC, fast protein liquid chromatography; Fuc, fucose; ISD, in-source decay; MALDI-TOF, matrix-assisted laser desorption ionization time-of-flight; MVE, Murray Valley encephalitis; PBS, phosphate-buffered saline; PSD, post-source decay; q, collision quadrupole; Q, quadrupole; rpHPLC, reverse phase high performance liquid chromatography; secHPLC, size exclusion high performance liquid chromatography; TCEP, tris(2-carboxyethyl)phosphine. Back


   ACKNOWLEDGMENTS
 
We thank Gary Shooter and John Holland for critical review of this manuscript.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Halstead, S. B. (1988) Science 239, 476–481[Abstract/Free Full Text]
  2. Jacobs, M. G., and Young, P. R. (1998) Curr. Opin. Infect. Dis. 11, 319–324
  3. Gubler, D. J. (1999) in Encyclopedia of Virology (Granoff, A., and Webster, R. G., eds) 2nd Ed., pp. 375–384, Academic Press, San Diego, CA
  4. Gubler, D. J. (2002) Trends Microbiol. 10, 100–103[CrossRef][Medline] [Order article via Infotrieve]
  5. Clarke, T. (2002) Nature 416, 672–674[CrossRef][Medline] [Order article via Infotrieve]
  6. Ruggli, N., and Rice, C. M. (1999) Adv. Virus Res. 53, 183–207[Medline] [Order article via Infotrieve]
  7. Rice, C. M. (1996) in Fields Virology (Fields, B. N., Knipe, D. M., and Howley, P. M., eds) 3rd Ed., pp. 931–959, Lippencott-Raven, Philadelphia, PA
  8. Schlesinger, S., and Schlesinger, M. J. (1990) in Replication of Togaviridae and Flaviviridae (Fields, B. N., and Knipe, D. M., eds) Vol. 1, pp. 697–711, Lippincott-Raven, New York
  9. Kuhn, R. J., Zhang, W., Rossmann, M. G., Pletnev, S. V., Corver, J., Lenches, E., Jones, C, T., Mukhopadhyay, S., Chipman, P. R., Strauss, E. G., Baker, T. S., and Strauss, J. H. (2002) Cell 108, 717–725[CrossRef][Medline] [Order article via Infotrieve]
  10. Russell, P. K., Brandt, W. E., and Dalrymple, J. M. (1980) in The Togaviruses: Biology, Structure and Replication (Schlesinger, R. W., ed) pp. 503–529, Academic Press, New York
  11. Mackow, E., Makino, Y., Zhao, B., Zhang, Y-M., Markoff, I., Buckler-White, A., Guiler, M., Chanock, R., and Lai, C-J. (1987) Virology 159, 217–228[CrossRef][Medline] [Order article via Infotrieve]
  12. Speight, G., Coia, A., Parker, M. D., and Westaway, E. G. (1988) J. Gen. Virol. 69, 23–24[Abstract/Free Full Text]
  13. Wright, P. J., Cauchi, M. R., and Ng, M. L. (1989) Virology 171, 61–67[CrossRef][Medline] [Order article via Infotrieve]
  14. Winkler, G., Randolph, V. B., Cleaves, G. R., Ryan, T. E., and Stollar, V. (1988) Virology 162, 187–196[CrossRef][Medline] [Order article via Infotrieve]
  15. Parrish, C. R., Woo, W. S., and Wright, P. J. (1991) Arch. Virol. 117, 279–286[CrossRef][Medline] [Order article via Infotrieve]
  16. Winkler, G., Maxwell, S. E., Ruemmler, C., and Stoller, V. (1989) Virology 171, 302–305[CrossRef][Medline] [Order article via Infotrieve]
  17. Jacobs, M. G., Robinson, P. J., Bletchly, C., Mackenzie, J. M., and Young, P. R. (2000) FASEB J. 14, 1603–1610[Abstract/Free Full Text]
  18. Flamand, M., Megret, F., Mathieu, M., Lepault, J., Rey, F. A., and Deubel, V. (1999) J. Virol. 73, 6104–6110[Abstract/Free Full Text]
  19. Crooks, A. J., Lee, J. M., Easterbrook, L. M., Tinofeev, A. V., and Stephenson, J. R. (1994) J. Gen. Virol. 75, 3453–3460[Abstract/Free Full Text]
  20. Pryor, M. J., and Wright, P. J. (1994) J. Gen. Virol. 75, 1183–1187[Abstract/Free Full Text]
  21. Pryor, M. J., and Wright, P. J. (1993) Virology 194, 769–780[CrossRef][Medline] [Order article via Infotrieve]
  22. Pryor, M. J., Gualano, R. C., Lin, B., Davidson, A., and Wright, P. J. (1998) J. Gen. Virol. 79, 2631–2639[Abstract]
  23. Muylaert, I. A., Chambers, T. J., Galler, R., and Rice, C. M. (1996) Virology 222, 159–168[CrossRef][Medline] [Order article via Infotrieve]
  24. Pletnev, A. G., Bray, M., and Lai, C. J. (1993) J. Virol. 67, 4956–4963[Abstract/Free Full Text]
  25. Mackenzie, J., Jones, M., and Young, P. R. (1996) J. Virol. Methods 220, 232–240
  26. Muylaert, I. R., Galler, R., and Rice, C. M. (1997) J. Virol. 71, 291–298[Abstract]
  27. Lindenbach, B. D., and Rice, C. M. (1997) J. Virol. 71, 9608–9617[Abstract]
  28. Lindenbach, B. D., and Rice, C. M. (1999) J. Virol. 73, 4611–4621[Abstract/Free Full Text]
  29. Schlesinger, J. J., Brandriss, M. W., Putnak, J. R., and Walsh, E. E. (1990) J. Gen. Virol. 71, 593–599[Abstract/Free Full Text]
  30. Deubel, V., Kinney, R. M., and Trent, D. W. (1988) Virology 165, 234–244[CrossRef][Medline] [Order article via Infotrieve]
  31. Gibson, C. A., Schlesinger, J. J., and Barrett, A. D. T. (1988) Vaccine 6, 7–9[CrossRef][Medline] [Order article via Infotrieve]
  32. Wang, P., Geng, L., Qin, E., Tu, M., and Zhao, W. (2001) J. Biochem. Mol. Biol. 17, 148–154
  33. Wedemeyer, W. J., Welker, E., Narayan, M., and Scheraga, H. A. (2000) Biochemistry 39, 4207–4216[CrossRef][Medline] [Order article via Infotrieve]
  34. Pitt, J. J., Da Silva, E., and Gorman, J. J. (2000) J. Biol. Chem. 275, 6469–6478[Abstract/Free Full Text]
  35. Blitvich, B. J., Scanlon, D., Shiell, B. J., Mackenzie, J. S., Pham, K., and Hall, R. A. (2001) J. Gen. Virol. 82, 2251–2256[Abstract/Free Full Text]
  36. Wallis, T. P., Pitt, J. J., and Gorman, J. J. (2001) Protein Sci. 10, 2251–2271[CrossRef][Medline] [Order article via Infotrieve]
  37. Gorman, J. J., Wallis, T. P., and Pitt, J. J. (2002) Mass Spectrom. Rev. 21, 183–216[CrossRef][Medline] [Order article via Infotrieve]
  38. Falconar, A. K., and Young, P. R. (1991) J. Gen. Virol. 72, 961–965[Abstract/Free Full Text]
  39. Gorman, J. J., Ferguson, B. L., and Nguyen, T. B. (1996) Rapid Commun. Mass. Spectrom. 10, 529–536[CrossRef][Medline] [Order article via Infotrieve]
  40. Gorman, J. J., Ferguson, B. L., Speelman, D., and Mills, J. (1997) Protein Sci. 6, 1308–1315[Medline] [Order article via Infotrieve]
  41. Lopaticki, S., Morrow, C. J., and Gorman, J. J. (1998) J. Mass. Spectrom. 33, 950–960[CrossRef][Medline] [Order article via Infotrieve]
  42. Rudd, P. M., Downing, A. K., Cadene, M., Harvey, D. J., Wormald, M. R., Weir, I., Dwek, R. A., Rifkin, D. B., and Gleizes, P. E. (2000) Biochemistry 39, 1596–1603[CrossRef][Medline] [Order article via Infotrieve]
  43. Morris, H. R., and Pucci, P. (1985) Biochem. Biophys. Res. Commun. 126, 1122–1128[CrossRef][Medline] [Order article via Infotrieve]
  44. Yazdanparast, R., Andrews, P., Smith, D. L., and Dixon, J. E. (1986) Anal. Biochem. 153, 348–353[CrossRef][Medline] [Order article via Infotrieve]
  45. Yazdanparast, R., Andrews, P., Smith, D. L., and Dixon, J. E. (1987) J. Biol. Chem. 262, 2507–2513[Abstract/Free Full Text]
  46. Zhou, J., Poppe-Schriemer, N., Standing, K. G., and Westmore, J. B. (1993) Int. J. Mass. Spectrom. Ion Proc. 126, 115–122[CrossRef]
  47. Patterson, S. D., and Katta, V. (1994) Anal. Chem. 66, 3727–3732[Medline] [Order article via Infotrieve]
  48. Crimmins, D. L., Saylor, M., Rush, J., and Thoma, R. S. (1995) Anal. Biochem. 226, 355–361[CrossRef][Medline] [Order article via Infotrieve]
  49. Cauchi, M. R., Henchal, E. A., and Wright, P. J. (1991) Virology 180, 659–667[CrossRef][Medline] [Order article via Infotrieve]
  50. Yang, Y., Wu, J., and Throck Watson, J. (1998) J. Am. Chem. Soc. 120, 5834–5835[CrossRef]
  51. Haniu, M., Acklin, C., Kenney, W. C., and Rohde, M. F. (1994) Int. J. Pept. Protein Res. 43, 81–86[Medline] [Order article via Infotrieve]
  52. Bean, M. F., and Carr, S. A. (1992) Anal. Biochem. 201, 216–226[CrossRef][Medline] [Order article via Infotrieve]
  53. Schnaible, V., Wefing, S., Resemann, A., Suckau, D., Bücker, A., Wolf-Kümmeth, S., and Hoffman, D. (2002) Anal. Chem. 74, 4980–4988[Medline] [Order article via Infotrieve]
  54. Ryle, A. P., and Sanger, F. (1955) Biochem. J. 60, 535–540[Medline] [Order article via Infotrieve]
  55. Schnolzer, M., Jedrzejewski, P., and Lehmann, W. D. (1996) Electrophoresis 17, 945–953[CrossRef][Medline] [Order article via Infotrieve]
  56. Hall, R. A., Broom, A. K., Hartnett, A. C., Howard, M. J., and Mackenzie, J. S. (1995) J. Virol. Methods 51, 201–210[CrossRef][Medline] [Order article via Infotrieve]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 Mol. Cell. ProteomicsHome page
K. A. Dave, F. Whelan, C. Bindloss, S. G. B. Furness, A. Chapman-Smith, M. L. Whitelaw, and J. J. Gorman
Sulfonation and Phosphorylation of Regions of the Dioxin Receptor Susceptible to Methionine Modifications
Mol. Cell. Proteomics, April 1, 2009; 8(4): 706 - 719.
[Abstract] [Full Text] [PDF]

Home page
 Exp. Biol. Med.Home page
H.-J. Cheng, C.-F. Lin, H.-Y. Lei, H.-S. Liu, T.-M. Yeh, Y.-H. Luo, and Y.-S. Lin
Proteomic Analysis of Endothelial Cell Autoantigens Recognized by Anti-Dengue Virus Nonstructural Protein 1 Antibodies
Experimental Biology and Medicine, January 1, 2009; 234(1): 63 - 73.
[Abstract] [Full Text] [PDF]

Home page
 J. Virol.Home page
K. M. Chung, G. E. Nybakken, B. S. Thompson, M. J. Engle, A. Marri, D. H. Fremont, and M. S. Diamond
Antibodies against West Nile Virus Nonstructural Protein NS1 Prevent Lethal Infection through Fc {gamma} Receptor-Dependent and -Independent Mechanisms
J. Virol., February 1, 2006; 80(3): 1340 - 1351.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
279/20/20729    most recent
M312907200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wallis, T. P.
Right arrow Articles by Gorman, J. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wallis, T. P.
Right arrow Articles by Gorman, J. J.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2004 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement