Deamidation in ricin studied by capillary zone electrophoresis- and liquid chromatography–mass spectrometry

Accepted Manuscript Title: Deamidation in ricin studied by capillary zone electrophoresis- and liquid chromatography- mass spectrometry ˚ Fredriksson Calle Author: Tomas Bergstr¨om Sten-Ake ˚ Nilsson Crister Astot PII: DOI: Reference:

S1570-0232(14)00649-7 http://dx.doi.org/doi:10.1016/j.jchromb.2014.10.015 CHROMB 19164

To appear in:

Journal of Chromatography B

Received date: Revised date: Accepted date:

9-7-2014 24-9-2014 7-10-2014

˚ Fredriksson, C. Nilsson, C. Please cite this article as: T. Bergstr¨om, S.-A. ˚ Astot, Deamidation in ricin studied by capillary zone electrophoresis- and liquid chromatography- mass spectrometry, Journal of Chromatography B (2014), http://dx.doi.org/10.1016/j.jchromb.2014.10.015 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Deamidation in ricin studied by capillary zone electrophoresis- and liquid chromatography- mass

cr

ip t

spectrometry

us

Authors:

an

Tomas Bergström, Sten-Åke Fredriksson, Calle Nilsson and Crister Åstot*

M

Affiliation:

pt

*Corresponding author:

ed

Swedish Defence Research Agency, CBRN Defence and Security, SE-901 82 Umeå, Sweden

Ac ce

Crister Åstot, tel: +46 90 106808, fax: +46 90 106803. e-mail: [email protected]

Keywords:

Ricin, deamidation, capillary zone electrophoresis, liquid chromatography, mass spectrometry, isoelectric diversity.

1 Page 1 of 37

ABSTRACT Deamidation in ricin, a toxin present in castor beans from the plant Ricinus communis, was investigated using capillary zone electrophoresis (CZE) and liquid chromatography coupled to

ip t

high resolution mass spectrometry. Potential sites for deamidation, converting asparagine

(Asn) into aspartic or isoaspartic acid (Asp or isoAsp), were identified in silico based on the

cr

protein sequence motifs and tertiary structure. In parallel, CZE- and LC-MS-based screening

us

were performed on the digested toxin to detect deamidated peptides. The use of CZE-MS was critical for the separation of small native/deamidated peptide pairs. Selected peptides were

an

subjected to a detailed analysis by tandem mass spectrometry to verify the presence of deamidation and determine its exact position. In the ricin preparation studied, deamidation

M

was confirmed and located to three asparagine residues: Asn54 in the A-chain, and Asn42 and

ed

Asn60 in the B-chain. Possible in vitro deamidation occurring during sample preparation was monitored using a synthetic peptide with a known and rapid rate of deamidation. Finally, we

Ac ce

deamidation.

pt

showed that the isoelectric diversity previously reported in ricin is related to the level of

2 Page 2 of 37

1. Introduction Ricin is a highly toxic protein that can be isolated from castor beans, the seeds of the ricin plant Ricinus communis. Ricin plants are naturally occurring in tropical climates and castor beans are produced throughout the world as a source of castor oil, which is mainly used in

ip t

various industrial applications. Annual world production of castor beans exceeds 1 million

cr

tons [1]. In addition, ricin plants are widely cultivated as ornamentals due to their attractive colour and shape. The extraction, storage and use of ricin is regulated under the Chemical

us

Weapons Convention (CWC) and must be reported to national authorities [2].The toxin content in the seed is up to 1% and production of crude toxin from the castor bean is

an

unfortunately fairly simple. Also, the easy access to castor beans has contributed to frequently

M

occurring “white powder” incidents [3,4,5]. Therefore, an increased need for forensic attribution information relating to the origin of the sample has evolved. Reported forensic

ed

markers are sequence information for different varieties of ricin, differences or similarities in the content of co-extracted proteins, variation in the extent of glycosylation and also content

pt

of peptide biomarkers [6,7,8,9,10,11,12]. Recently, isotope ratio comparison has also been shown to be useful for this purpose [13].

Ac ce

The ricin toxin is a 66 kDa plant lectin of RIP 2-type (ribosome-inactivating protein) that induces toxicity by inhibiting protein synthesis inside a target cell, eventually leading to cell death [14]. Ricin contains two polypeptide chains. The larger B-chain exhibits the lectin properties, thereby mediating binding to a target cell surface and enabling the toxin to be incorporated into the cell. Inside the cell, the B-chain is released and the free A-chain acts as an enzymatic ribosomal inhibitor [14,15]. Deamidation in proteins is a posttranslational process that converts asparagine to aspartic and isoaspartic acid, and glutamine to glutamic acid. These transformations alter the protein charge and affect the pI of the proteins. Resulting conformational changes can lead to protein 3 Page 3 of 37

dysfunction, including enzymatic deactivation, loss or altered affinity for various ligands and reduced stability [16]. However, deamidation can function as a timed signal for protein events such as activity switching and has an important role as a molecular clock in biological systems [17,18,19]. Deamidation is a commonly occurring process that has been studied in

ip t

eye lens crystalline proteins [20], proteins extracted from potato tubers [21], antibodies [22], toxicity factors of anthrax [23] and synthetic peptide products [24]. The standard model for

cr

nonenzymatic deamidation of asparagine is outlined in Fig. 1. This reaction has been found to

us

be about 100 times faster than glutamine deamidation, mainly because of the larger, less favourable intermediate formed in the latter case [25,26,27,28]. The smaller and more flexible

an

the adjacent residues are, especially on the carboxyl side, the more prone asparagines are to undergo spontaneous deamidation. Therefore, an arginine-glycine (NG) motif located in a

M

flexible region of a protein has a high probability of becoming deamidated [29,30,31]. In such a reaction, both aspartic acid and isoaspartic acid are formed. In neutral or alkaline solutions,

ed

formation of isoaspartic acid is favoured with a typical isoAsp:Asp ratio of 3:1, whereas in acidic solutions, only Asp is formed [29]. As displayed in Fig. 1, when isoaspartic acid is

pt

formed, the peptide backbone is extended by one carbon. The elongation of the peptide

Ac ce

backbone can severely disrupt the protein structure and function, and thus isoAsp formation has been widely studied in the field of protein stability [32,33]. To study deamidation at the amino acid level, one of the most feasible methods is to digest the protein of interest into peptides and then separate and analyse the peptides by liquid chromatography coupled to mass spectrometry (LC-MS) [34]. However, deamidation induces only a small shift in peptide hydrophobicity and size, hampering separation of the peptide pairs (native and deamidated) by reversed phase chromatography systems. In addition, deamidation is detected by mass spectrometry as a ~1 Da shift in the peptide mass. More accurately, the mass shift is +0.9840 Da, making the mass difference between the (M+H)+

4 Page 4 of 37

monoisotopic ion of the deamidated peptide and the [(M+neutron)+H]+ isotopologue ion of the native peptide to be only 0.0193 Da (based on 13C contribution). Thus, the isotope clusters of the native and deamidated peptide pairs overlap and co-elution makes detection and identification of the deamidated peptide difficult.

ip t

As a complementary technique for the analysis of proteins, peptides and small molecules, on-line capillary zone electrophoresis (CZE)-MS has gained increasing attention

cr

[35,36,37,38,39,40]. The separation principle of CZE is based on the charge-to-size ratio of

us

the analytes, making it a complement to the hydrophobicity-based separations in reversed phase-LC. However, the coupling of CZE with MS limits the choice of available buffer

an

systems due to the requirement for volatile components. Furthermore, stability problems are usually worse for CZE-MS than for LC-MS. On the other hand, if the drawbacks can be

M

controlled, CZE-MS is a powerful analytical tool with high separation efficiency that has proven useful in various applications. One example is the separation of small peptides, i.e.

ed

shorter than five amino acids, as these are difficult to retain in reversed phase-LC [41]. In addition, the technique is useful for resolving the native and deamidated peptide pairs of short

pt

peptides, as both the difference in charge and mobility significantly contribute to their

Ac ce

separation. The migration order in both CZE-MS and LC-MS is usually Asn→Asp→isoAsp. The deamidated, aspartic/isoaspartic acid-containing peptides have lower mobility and slightly higher hydrophobicity than the corresponding native peptide and the effect is strongest for the isoaspartic acid-peptide due to the induced extension of the peptide backbone by one carbon.

Protein deamidation has been studied at the peptide level for various proteins. Zhang et al. [42] determined the deamidation positions in recombinant human interleukin-11 after digestion, separation and Edman degradation. Terreaux et al. used N-terminal sequencing to identify sites of deamidation in synthetic gliadin peptide [43]. In addition, LC-MS/MS has

5 Page 5 of 37

been used to locate deamidation in IgG1 antibodies [44,45,46] and Bacillus anthracis protective antigen [23]. The aim of this study was to investigate deamidation in ricin for potential use in forensic applications. The complete workflow from prediction of sites in intact ricin to verification and

ip t

exact determination of the position of these sites using CZE-MS/MS and LC-MS/MS in both tryptic and chymotryptic ricin peptides is reported. Finally, the effect of deamidation on the pI

cr

pattern of the toxin was demonstrated.

us

2. Experimentals 2.1. Safety considerations

an

Ricin is listed in Schedule 1 of the Chemical Weapons Convention (CWC), which states that countries that have ratified the CWC are required to declare the production (extraction),

M

possession and usage of ricin to the Organisation for the Prohibition of Chemical Weapons, OPCW [2,47]. Ricin is a toxin that inhibits protein synthesis in cells. Due to its high toxicity,

ed

it is recommended that all work performed with intact toxin is carried out in specially designated laboratories. All contact with the active toxin should be avoided. In this study, all

pt

the laboratory materials used, including single-use materials, were decontaminated by

Ac ce

submersion in 2 M NaOH. 2.2. Chemicals

Methanol and acetonitrile (LC-MS grade) were obtained from Merck (Germany). Formic acid (MS grade) was obtained from Fluka (Switzerland), NaOH (CE grade) and H20 (ultrapure) from Agilent (USA). CE-capillaries of bare fused silica type were obtained from Genetec (Sweden). Porcine trypsin (MS grade) was obtained from Promega (USA), and bovine pancreas chymotrypsin (TLCK treated) and ammonium bicarbonate were obtained from Sigma-Aldrich (USA). Peptides and proteins used for developing the CE and LC methods were also obtained from Sigma-Aldrich (USA). The CIEF ampholyte solution (pH

6 Page 6 of 37

range 3-10), CIEF gel, anolyte phosphoric acid and catholyte NaOH were all obtained from Beckman Coulter (USA). 2.3. Preparation of ricin The toxin of isoform D was isolated from Ricinus communis of zanzibariensis variety

ip t

(Rara Växter, Stockholm, Sweden) and purified in-house according to previously described methods [7]. Sequence information on ricin (isoform D) was obtained from Uniprot [48]. The

cr

three-dimensional structure of ricin was obtained from the Protein Data Bank [49]. The

us

structure was supplied by Rutenber et al. [50]. 2.4. Prediction method

an

Predicted rates of each asparagine deamidation in ricin were calculated using the algorithm of Robinson and Robinson [51]. Details of the calculated deamidation coefficient (CD) and

M

how it is determined for each asparagine of a submitted protein has been described elsewhere [52]. The predicted half-life in days of a specific asparagine in a defined environment (37 °C,

2.5. Synthetic peptides

ed

pH 7.4, 0.15 M Tris buffer) can be calculated as 100*CD.

pt

The following synthetic peptides were purchased from the CASLO Laboratory ApS

Ac ce

(Denmark): SLNGEWR (E. coli derived peptide sequence); ENIELGDGPLEEAISALYYYSTGGTQLPTLAR (tryptic peptide #12 in the A-chain of ricin, deamidated at the GDG motif); SNGK (tryptic peptide #9 in the B-chain of ricin); SDGK (deamidated variant of tryptic peptide #9 in the B-chain of ricin). 2.6. Deamidation control

The E. coli derived synthetic peptide SLNGEWR was used as a deamidation control. This peptide has a documented half-life with respect to deamidation of its Asn of approximately 8 h in trypsin (and chymotrypsin) digestion milieu [53]. By monitoring this peptide after different digestion times (optimised conditions), we found that the first indications of

7 Page 7 of 37

deamidation (< 5 percentage) were observed after 1 h. Thus, subsequent ricin digestions were carried out for 30 minutes and all reactions were spiked with the control peptide to monitor any undesirable deamidation. 2.7. Digestion protocol

ip t

Digestions were made on organic solvent-denatured and precipitated toxin using the following procedure: Purified ricin was denatured and precipitated with 5 volumes of ice-cold

cr

methanol in a microcentrifuge tube and kept at -20 °C for approx. 10 minutes. The sample

us

was then subjected to centrifugation at 14000 g for 2 minutes, after which the excess liquid was removed and the pellet dried at room temperature. Digestion of the pellet was carried out

an

in 50 mM ammonium bicarbonate, trypsin or chymotrypsin were used in a 1:20 weight ratio relative to ricin and the control peptide was added to a 1:5 molar ratio relative to ricin. The

M

reaction vial was placed in a heating block set to 43 °C. The reaction was stopped after 30 minutes by addition of formic acid (pH <4). The digested samples were kept at 4 °C and

ed

analysed on the same day as the reaction was performed. The resulting tryptic (T) and chymotryptic (C) peptides were annotated by their theoretical enumeration from the N-

pt

terminal to the C-terminal for the A and B chain, respectively (Table 1 and 2).

Ac ce

2.8. Capillary zone electrophoresis mass spectrometry method development For capillary zone electrophoresis (CZE) separations, a G1600A CE instrument from Agilent (USA) was used. Agilent ChemStation software was used for instrument control and data collection. In order to minimize peptide interactions with the capillary inner wall, standard bare fused silica capillaries and low pH buffers were employed. For the CZE-MS application, the method development was focused on volatile buffer systems. The CZE-MS conditions, including the make-up flow properties and spray voltage settings, were also selected to obtain a stable electrospray and avoid high currents that might result in capillary overheating. The conditions were not further optimized in order to improve the

8 Page 8 of 37

electrophoretic efficiency of the CZE separation since the performance was sufficient for the separation of small native and deamidated peptide pairs in this study and CZE peak widths < 4 s were not desired given the minimal MS/MS-scan time of the QTOF instrument used. 2.9. CZE-MS setup

ip t

The CZE-capillary was kept as short as possible (60 cm) by guiding it directly to the mass spectrometer, without passing the UV-window. The air-cooling of the system was extended

cr

all the way to the spray device using an Agilent CZE-MS cassette. The CZE-MS interface

us

was a sheath flow device from Waters (USA), modified so that the thin 200 µm O.D. capillary stretched from the vial to the spray head and there acted as the spray tip itself. The spray

an

device comprised the CZE capillary in the centre with the polyimide coating removed at the tip. Slightly withdrawn and surrounding the fused silica tip was a steel capillary delivering the

M

make-up flow and supporting the electrospray. The auxiliary pump of a Waters nanoAcquityUPLC system was used to deliver the make-up flow. Desolvation gas was not used.

ed

The settings found to be the most suitable for this setup of CZE-MS was; bare fused silica capillary (200 µm O.D. /75 µm I.D), separation using 1 M formic acid in 20%

pt

acetonitrile/H2O and a make-up flow of 50% methanol/0.1% acetic acid mixture delivered at

Ac ce

1 µl/min. The electrospray voltage on the mass spectrometer was set at 3.3 kV and the CE voltage at 24.3 kV (ramped during the initial 0.4 min of the separation), giving an effective separation voltage of 21 kV. Using a capillary length of 60 cm resulted in electric field strength of 350 V/cm. The mass spectrometer spray voltage was applied after the CE voltage had been ramped more than halfway. The CE capillary was preconditioned before each run by flushing with 0.5 M NaOH, 1M formic acid in 20% acetonitrile/H2O and H2O for 5 min each. The spray device was withdrawn from the interface during the precondition step. Samples were injected using 50 mbar pressure for 5-10 seconds to load typically 200-400 fmol of synthetic peptides and 2-4 pmol of protein digests.

9 Page 9 of 37

2.10.

LC and MS settings

A Waters nanoAcquity-UPLC system was used for all LC-separations. Samples of volume 1-5 µl were injected using the partial loop setting. The samples were trapped on a 180 µm×20 mm C18 column (5 μm Symmetry C18, Waters) and then separated on a 75 µm×150 mm

ip t

analytical column (1.7 µm BEH 130 C18, Waters). The flow was set to 0.4 µl/min and the gradient settings were as follows: starting at 97% A (H2O/0.2% formic acid) and 3% B

cr

(acetonitrile/0.2% formic acid), then ramping from 3% to 60% B over 25 minutes, then up to

us

90% B during 1 minute and maintaining this setting for 4 minutes before switching back to the initial settings for a further 10 minutes. The total run time was 40 minutes. A Waters

an

quadrupole time-of-flight (Q-TOF) Ultima mass spectrometer equipped with a nanospray ion source was used for the MS analysis. Picotip fused silica spray tips (New Objective, USA)

M

were used. The mass range in MS mode was m/z 300−1800, whereas in MS/MS mode, the range was adjusted depending on the size and charge of the parent ion. In both modes, data

2.11.

ed

were acquired at a frequency of one spectrum per second. CIEF, ricin deamidation experiment

pt

Two ricin D samples (zanzibariensis) were prepared in 50 mM ammonium bicarbonate

Ac ce

buffer. One was kept at 37 °C and the other at 4 °C, both for 20 days. After incubation, the ammonium bicarbonate buffer was removed by repeatedly washing the samples placed on a 10 kDa cut off microcentrifuge filter (Millipore, USA) with ultra-pure water. The samples were analysed using both capillary isoelectric focusing (CIEF) on intact toxin and CZE-MS on trypsin digested toxin. Settings for the digestion and CZE-MS were as described above. For the CIEF analysis, the same instrument and software as for CZE was used. The samples were mixed with 50% of ampholyte/CIEF gel mixture (1:150) to give a final sample concentration of approx. 2 mg/ml. A 38 cm neutral capillary (Beckman Coulter) was used for the analysis. The anolyte and catholyte solutions were 10 mM phosphoric acid and 10 mM

10 Page 10 of 37

NaOH, respectively. The samples were pressure injected to fill the entire capillary before the focusing voltage (400 V/cm) was applied. To enhance protein focusing, injection of a blank sample comprising ultra-pure water and the ampholyte/gel mixture was made just prior to each ricin sample injection. After 4 minutes of sample focusing, a 50 mbar pressure was

ip t

applied to slowly move the focused sample through the detector slit in the capillary. For detection, an Agilent CE instruments standard diode array detector was employed, modified

cr

so that a 280 nm UV-lens (Beckman Coulter) was placed between the UV-lamp and the

us

capillary to reduce spectral noise. The capillary temperature was set at 12 °C throughout the CIEF analysis.

an

3.1. Prediction of deamidation

Deamidation rates have been studied previously in short synthetic peptides to evaluate the

M

effect of adjacent amino acids on the reaction rate [54,55,56,57,58]. Based on these findings, an algorithm has been constructed that also takes the three-dimensional protein structure into

ed

account [51]. This algorithm was applied to the R. communis sequence data to predict the deamidation rate of the present asparagines. The results are summarized in Table 1 and 2,

pt

which rank the Asn containing motifs in the ricin A- and B-chain, respectively, according to

Ac ce

the predicted half-life in days of the amide. However, differences between the in silico predictions and experimental observations have been reported earlier [23]. Therefore, a detailed investigation of all potential ricin deamidation sites was conducted in order to compare theoretical and empirical data. 3.2. Enzymatic digestion

Deamidation is a process that may occur in both the native protein and during sample preparation and protein digestion. In studies of deamidation at the protein level, any artificial deamidation of peptides is undesirable. In many previous studies, the contribution of sample treatment to the overall deamidation was not considered. A lack of stabilising structures in

11 Page 11 of 37

proteins under preparation (i.e. in denatured, reduced or digested proteins) has been shown to increase the rate of deamidation [59,60,61,62]. For ricin, we have observed that denaturation, reduction and digestion protocols significantly affect the extent of deamidation (data not shown). To reduce the risk of artificial deamidation, we modified our previously published

ip t

rapid digestion protocol [63].We found that most of the ricin peptides were present after only a few minutes of trypsin or chymotrypsin reaction (data not shown). In addition, a small

cr

increase in the digestion temperature favoured the rapid formation of the desired peptides. We

us

therefore used slightly elevated digestion temperatures and digestion times of 30 minutes. To monitor artificial deamidation, we added the synthetic E. coli peptide SLNGEWR, which has

an

a reported Asn half-life of 8 hours [52]. Under the optimized conditions used in this study, no deamidation of this control peptide was observed.

M

3.3. Separation of deamidated/non-deamidated peptide pairs

In the tryptic ricin digest, we focused on two short peptides containing potential

ed

deamidation sites, TB9 and TA22 (SNGK and NGSK). To resolve the [(M+neutron)+H]+ isotopologue peaks (hereafter termed the second isotope) of these peptides from the (M+H)+

pt

monoisotopic peak (the first isotope) of their deamidated analogues, a mass resolution >21000

Ac ce

is required. Thus, co-eluting peptide pairs make the analysis of deamidation difficult due to the overlapping isotopic clusters and separation of the deamidated from the native peptide poses a particular challenge. The combination of LC-MS and CZE-MS has the potential to solve this separation problem. Large peptide pairs can normally be separated by LC, whereas smaller peptides pairs tend to be unresolved. CZE, in contrast, is well suited to resolve short and hydrophilic peptide pairs. The short TB9 and TA22 ricin peptides discussed above contain the deamidation motif (NG) and neither of these peptides is retained in reversed phase LC. Deamidation of Asn60 in the TB9 peptide would result in the SNGK/SDGK peptide pair and

12 Page 12 of 37

in the CZE-MS analysis of the synthetic peptide pair, baseline separation was observed (Fig. 2). 3.4. Screening and identification of deamidated peptides To obtain mass spectrometry data on peptides covering all asparagines in ricin, two

ip t

analytical techniques were used in combination with digestion by two different proteolytic enzymes. Ricin D contains 38 asparagines (Fig. 3) distributed over 25 tryptic or 30

cr

chymotryptic theoretical peptides. In the combined CZE-MS and LC-MS analyses, 22 tryptic

us

and 20 chymotryptic peptides were detected, covering all asparagines in the protein. These peptides were screened for the presence of their deamidated analogues, and the tryptic

an

peptides TA7 TB6 and TB9 were selected for further analysis (Table 1 and 2). 3.5. Locating the sites of deamidation using MS/MS

M

For deamidated peptides, clean product ion spectra requires baseline separation of the peptide pairs. This was achieved by LC-MS for TA7 and TB6 and by CZE-MS for TB9. The

ed

chromatograms of the TA7 peptide (VGLPINQR) and its deamidated counterpart are shown in Fig. 4. The daughter ion spectrum of the deamidated peptide is shown in Fig. 5. The mass of

pt

the y2 and y3 fragments locates the deamidation to Asn54. For the short peptide TB9 (SNGK),

Ac ce

the y2 and y3 fragments locate the deamidation to Asn60 (Supplementary Fig S1). The high frequency of asparagines in ricin gives rise to several tryptic and chymotryptic peptides with more than one asparagine. The sequence of the TB6 peptide (SNTDANQLWTLK) contains two possible deamidation sites, Asn42 and Asn46, where the first is predicted to be more susceptible by a factor of seven according to the prediction model (Table 2). The y10, y11 and b2 fragments obtained in the LC-MS/MS analysis of the deamidated peptide showed that the location of deamidation was exclusively at Asn42 (Fig. 6 and Supplementary Fig. S2). 3.6. Discrepancies between detected and predicted deamidation

13 Page 13 of 37

In this study, the experimental deamidation data on ricin differed from the predicted values. The A-chain Asn54 was found to be deamidated, although it was predicted to be stable with an expected half-life of 100 years (ranked 31st of all Asn sites in the toxin). For the Bchain, i.e. residues Asn60 and Asn42, there was better agreement between the experimental

ip t

and predicted deamidation data (Table 2). Perhaps most noteworthy from Table 2 was that Asn141 in the A-chain, the top deamidation candidate with a predicted half-life of 23 days,

cr

did not appear to be deamidated according to the experiments. To investigate if this

us

deamidated peptide was genuinely absent or undetectable by the techniques used, we verified that the synthetic deamidated TA12-peptide could be detected by both LC-MS and CE-MS

an

(data not shown).

Similar discrepancies between observed and predicted deamidation have been shown by

M

Powell et al. [23], who studied the in vivo deamidation of Bacillus anthracis recombinant PA (rPA) protein. From their data, it was concluded that artificially induced deamidation was

ed

more in accordance with the predictive model than in vivo deamidation, suggesting that there are two classes of deamidation: in vitro deamidations that occur at approximately the

pt

predicted rates and in vivo deamidations that differ significantly from the predictions. In line

Ac ce

with the conclusions of Powell et al., we suggest that the discrepancies we observed may be due to several factors, e.g. in vivo enzymatic or non-enzymatic reactions, which increase the degree of deamidation at specific sites and thus cannot be successfully predicted by a general model. Correspondingly, a lower degree of deamidation than predicted is suggested to be caused by stabilising effects not taken into account by the prediction model. This could be the reason for the low deamidation of Asn141 in contrast to the prediction. The three-dimensional structure of ricin [49] suggests that Asn141 may be stabilised by a hydrogen bond to Thr13. In addition, Asn141 is close to a rigid alpha-helix structure and consequently the likelihood of deamidation is reduced.

14 Page 14 of 37

3.7. Deamidation effect on CIEF pattern of ricin Ricin extracted from different varieties of R. communis show large isoform variations. These variations are thought to mainly reflect the differences in sequence and abundance of the ricin D and E isoforms [7]. Smaller isoform differences were observed when analysing

ip t

intact D or E ricin by CIEF (Fig. 7, top). This agrees with an earlier report by Despeyroux et al., who discussed the possibility of amino acid sequence contribution to the heterogeneities

cr

observed using different CE techniques [6]. Based on studies showing protein pI

us

heterogeneity and shifts due to deamidation of asparagines [64,65,66,67], we hypothesised that the pI differences observed for ricin isoforms could be linked to the extent of

an

deamidation. In our study (Fig 7a/b), the differences were 0.3-0.4 pI units based on measurements with calibration proteins (data not shown), which were in accordance with

M

these previous studies. Each converted asparagine in a protein contributes to lowering its pI, i.e. the protein becomes neutral at a slightly lower pH than before. In CIEF analysis, a

ed

deamidation would then result in a peak shift towards lower pI and the intensity of that second peak would reflect the number of protein molecules modified by one deamidation. A third

pt

peak may also be present showing the amount of protein molecules with two deamidations, at

Ac ce

any position in the protein, lowering the proteins pI by two steps, etc. To investigate this assumption, a ricin D-sample was kept at elevated temperature to induce deamidation. Compared to the control sample kept in fridge, the CIEF results of the induced sample showed a shift of peak intensities towards lower pI (Fig. 7a/b). Deamidation was evaluated on the digested samples using CZE-MS and LC-MS. The TB9 peptide showed a distinct increase of the two deamidated variants (containing Asp and isoAsp) in the induced sample and a corresponding decrease of the Asn-containing peptide compared to the control sample (Fig. 8). The results for the incubated samples verified that the observed CIEF pattern

15 Page 15 of 37

in a sequence-homogenous ricin sample is due to different charge states, originating from deamidations. 4. Conclusions By using a combined CZE- and LC-MS approach, three deamidation sites were identified

ip t

in a ricin preparation. The observed discrepancies between the predicted and observed deamidation sites indicate that experimental data are required for a conclusive deamidation

cr

analysis. We also concluded that the variation in the CIEF analysis pattern of ricin can be

us

explained in terms of deamidation. In forensic applications, the deamidation heterogeneity could potentially be used for matching ricin samples to specific sources.

an

Acknowledgments

We acknowledge Fredrik Ekström for valuable discussions and Elisabeth Artursson for

M

assistance with the ricin preparation. This work was supported by the Swedish Civil

Supplementary data

ed

Contingencies Agency.

Supplementary data associated with this article can be found, in the online version, at

Ac ce

pt

http://dx.doi.org/XXXXXX

16 Page 16 of 37

Figure legends Fig. 1. Standard model for nonenzymatic deamidation of asparagine (modified from [68]).

ip t

The peptide backbone (thick line) is extended by one carbon when isoaspartic acid is formed.

Fig. 2. Overlaid CZE-MS electropherograms from the separation of the synthetic peptides a)

an

* the peak for the second isotope of a).

us

tryptic peptide #9 from the B-chain of ricin toxin (TB9).

cr

SNGK (m/z 405.21) and b) SDGK (m/z 406.19), illustrating the deamidation of Asn60 in the

Fig. 3. Amino acid sequence of Ricin D sequence with the 38 asparagines highlighted in bold

ed

M

and confirmed deamidated asparagines underlined.

Fig. 4. Overlaid LC-MS chromatograms displaying the three variants of the ricin peptide TA7

pt

(VGLPI-(N/D/isoD)-QR). [M+2H]2+ of a) the intact peptide, m/z 448.77 and the deamidated peptide, m/z 449.26, containing b) aspartic acid, and c) isoaspartic acid.

Ac ce

* the peak for the second isotope of a)

Fig. 5. Product ion spectrum of the deamidated ricin peptide TA7 (VGLPIDQR, m/z 449.262+). The y2 and y3 fragments confirmed the aspartic acid position.

Fig. 6. Product ion spectrum of the deamidated ricin peptide TB6 (SDTDANQLWTLK, m/z 696.342+). The y10 and y11 fragments confirmed the aspartic acid position, further supported by the b2 fragment.

17 Page 17 of 37

Fig. 7. CIEF analysis of incubated ricin. Lower trace: sample kept in fridge; upper trace: sample incubated at 37 °C for 20 days.

Fig. 8. Analysis of induced deamidation in ricin. Extracted m/z 406.2 ion electropherograms

ip t

from CZE-MS analysis of peptides from incubated (upper) and control (lower) samples. The peaks correspond to a) the second isotope of the intact TB9 peptide and the deamidated

us

cr

peptide containing b) aspartic acid, and c) isoaspartic acid. Arrows indicate intensity changes.

Fig. S1. Product ion spectrum of the deamidated ricin peptide TB9 (SDGK, m/z 406.19+). The

an

y2 and y3 fragments confirm the aspartic acid position.

M

Fig. S2. Overlaid LC-MS chromatograms displaying the three variants of the ricin peptide TB6 (S-(N/D/isoD)-TDANQLWTLK). [M+2H]2+of a) the intact peptide, m/z 695.85, and the

ed

deamidated peptides, m/z 696.34, containing b) aspartic acid, and c) isoaspartic acid.

Ac ce

pt

* the peak for the second isotope of a).

18 Page 18 of 37

References [1] Food and Agriculture Organization of the United Nations, FAOSTAT, http://faostat3.fao.org/home/index.html#DOWNLOAD (accessed October 19 2012).

ip t

[2] Convention on the prohibition of the development, production, stockpiling and use of chemical weapons and

cr

on their destruction (www.opcw.org/chemical-weapons-convention).

us

[3] S. Worbs, K.Köhler, D. Pauly, M.A. Avondet, M. Schaer, M.B. Dorner, B.G. Dorner, Ricinus communis intoxications in human and veterinary medicine-a summary of real cases, Toxins 10 (2011) 1332-72.

an

[4] J. M. Bale, A. Bhattacharjee, E. Croddy, R. Pilch, Ricin Found in London: An al-Qa`ida Connection?, Center

M

for Nonproliferation Report (2003).

[5] D.M. Schieltz, S.C. McGrath, L.G. McWilliams, J. Rees, M.D. Bowen, J.J. Kools, L.A. Dauphin, E. Gomez-

ed

Saladin, B.N. Newton, H.L. Stang, M.J. Vick, J. Thomas, J.L. Pirkle, J.R. Barr, Analysis of active ricin and castor bean proteins in a ricin preparation, castor bean extract, and surface swabs from a public health

pt

investigation, Forensic Sci. Int. 209 (2011) 1-3.

Ac ce

[6] D. Despeyroux, N. Walker, M. Pearce, M. Fisher, M. McDonnell, S.C. Bailey, G.D. Griffiths, P. Watts, Characterization of ricin heterogeneity by electrospray mass spectrometry, capillary electrophoresis, and resonant mirror, Anal. Biochem. 279 (2000) 23-36.

[7] S-A. Fredriksson, A.G. Hulst, E. Artursson, A.L. de Jong, C. Nilsson, B.L. van Baar, Forensic identification of neat ricin and of ricin from crude castor bean extracts by mass spectrometry, Anal. Chem. 77 (2005) 1545-55.

[8] B.H. Woo, J.T. Lee, D.H. Na, K.C. Lee, Sepharose-unbinding ricin E as a source for ricin A chain immunotoxin, J. Immunol. Methods 249 (2001) 91-8.

19 Page 19 of 37

[9] P. Thullier, G. Griffiths, Broad recognition of ricin toxins prepared from a range of Ricinus cultivars using immunochromatographic tests, Clin. Toxicol. 47 (2009) 643-50.

[10] D.S. Wunschel, A.M. Melville, C.J. Ehrhardt, H.A. Colburn, K.D. Victry, K.C. Antolick, J.H. Wahl, K.L. Wahl, Integration of gas chromatography mass spectrometry methods for differentiating ricin preparation

ip t

methods, Analyst 137 (2012) 2077-85.

cr

[11] S.P. Ovenden, S-A. Fredriksson, C.K. Bagas, T. Bergström, S.A. Thomson, C. Nilsson, D.J. Bourne, De

us

novo sequencing of RCB-1 to -3: peptide biomarkers from the castor bean plant Ricinus communis, Anal. Chem.

an

81 (2009) 3986-96.

[12] F.C. Nogueira, G. Palmisano, V. Schwämmle, E.L. Soares, A.A. Soares, P. Roepstorff, G.B. Domont, F.A. Campos, Isotope labeling-based quantitative proteomics of developing seeds of castor oil seed (Ricinus

M

communis L.), J. Proteome Res. 12 (2013) 5012-24.

ed

[13] H.W. Kreuzer, J.B. West, J.R. Ehleringer, Forensic applications of light-element stable isotope ratios of

pt

Ricinus communis seeds and ricin preparations, J. Forensic Sci. 58 (2013) S43–S51.

65-70.

Ac ce

[14] S.M. Bradberry, K.J. Dickers, P. Rice, G.D. Griffiths, J.A. Vale, Ricin poisoning, Toxicol. Rev. 22 (2003)

[15] M.J. Lord, N.A. Jolliffe, C.J. Marsden, C.S. Pateman, D.C. Smith, R.A. Spooner, P.D. Watson, L.M. Roberts, Ricin. Mechanisms of cytotoxicity, Toxicol. Rev. 22 (2003) 53-64.

[16] A. Sun, K.U. Yuksel, R.W. Gracy, Terminal marking of triosephosphate isomerase: consequences of deamidation, Arch. Biochem. Biophys. 322 (1995) 361–8.

[17] T. Solstad, T. Flatmark, Microheterogeneity of recombinant human phenylalanine hydroxylase as a result of nonenzymatic deamidations of labile amide containing amino acids. Effects on catalytic and stability properties, Eur. J. Biochem. 267 (2000) 6302-10.

20 Page 20 of 37

[18] T. Flatmark, K. Sletten, Multiple forms of cytochrome c in the rat. Precursor-product relationship between the main component Cy I and the minor components Cy II and Cy 3 in vivo, J. Biol. Chem. 243 (1968) 1623-9.

[19] C.F. Midelfort, A.H. Mehler, Deamidation in vivo of an asparagine residue of rabbit muscle aldolase, Proc.

ip t

Natl. Acad. Sci. USA 69 (1972) 1816-9.

cr

[20] P.G. Hains, R.J. Truscott. Age-dependent deamidation of lifelong proteins in the human lens, Invest.

us

Ophthalmol. Vis. Sci. 51 (2010) 3107-14.

potato seed-tubers, Plant Physiol. 119 (1999) 89-100.

an

[21] G.N. Kumar, R.L. Houtz, N.R. Knowles, Age-induced protein modifications and increased proteolysis in

M

[22] A.I. Lukash, A.P. Shepelev, N.V. Pushkina, I.E. Tsybul'skiĭ, D.V. Al'perovich, Effect of non-enzymatic

ed

desamidation of gamma-globulin on its properties and biological activity, Vopr. Med. Khim. 31 (1985) 104-8.

[23] B.S. Powell, J.T. Enama, W.J. Ribot, W. Webster, S. Little, T. Hoover, J.J. Adamovicz, G.P. Andrews,

pt

Multiple asparagine deamidation of Bacillus anthracis protective antigen causes charge isoforms whose

Ac ce

complexity correlates with reduced biological activity, Proteins 68 (2007) 458-79.

[24] L.H. Silvertand, J.S. Toraño, G.J. de Jong, W.P. van Bennekom, Development and characterization of cIEFMALDI-TOF MS for protein analysis, Electrophoresis 30 (2009) 1828-35.

[25] A.B. Robinson, C.J. Rudd, Deamidation of glutaminyl and asparaginyl residues in peptides and proteins, Curr. Top. Cell Regul. 8 (1974) 247-95.

[26] N.E. Robinson, Z.W. Robinson, B.R. Robinson, A.L. Robinson, J.A. Robinson, M.L. Robinson, A.B. Robinson, Structure-dependent nonenzymatic deamidation of glutaminyl and asparaginyl pentapeptides, J. Pept. Res. 63 (2004) 426-36.

21 Page 21 of 37

[27] H.T. Wright, Nonenzymatic deamidation of asparaginyl and glutaminyl residues in proteins, Crit. Rev. Biochem. Mol. Biol. 26 (1991) 1-52.

[28] H. Lindner, W. Helliger, Age-dependent deamidation of asparagine residues in proteins, Exp. Gerontol. 36

ip t

(2001) 1551-63.

[29] R. Tyler-Cross, V. Schirch, Effects of amino acid sequence, buffers, and ionic strength on the rate and

us

cr

mechanism of deamidation of asparagine residues in small peptides, J. Biol. Chem. 266 (1991) 549-56.

[30] H.T. Wright, Sequence and structure determinants of the nonenzymatic deamidation of asparagine and

an

glutamine residues in proteins, Protein Eng. 4 (1991) 283-94.

M

[31] N.E. Robinson, A.B. Robinson, Molecular clocks, Proc. Natl. Acad. Sci. USA 98 (2001) 944-9.

[32] L. Böhme, J.W. Bär, T. Hoffmann, S. Manhart, H.H. Ludwig, F. Rosche, H.U. Demuth, Isoaspartate

pt

53.

ed

residues dramatically influence substrate recognition and turnover by proteases, Biol. Chem. 389 (2008) 1043-

[33] K.J. Reissner, D.W. Aswad, Deamidation and isoaspartate formation in proteins: unwanted alterations or

Ac ce

surreptitious signals?, Cell. Mol. Life Sci. 60 (2003) 1281-95.

[34] L. Huang, J. Lu, V.J. Wroblewski, J.M. Beals, R.M. Riggin, In vivo deamidation characterization of monoclonal antibody by LC/MS/MS, Anal. Chem. 77 (2005) 1432-9.

[35] R. Haselberg, G.J. de Jong, G.W. Somsen, Capillary electrophoresis-mass spectrometry for the analysis of intact proteins, J. Chromatogr. A 1159 (2007) 81-109.

[36] D.C. Simpson, R.D. Smith, Combining capillary electrophoresis with mass spectrometry for applications in proteomics, Electrophoresis 26 (2005) 1291-305.

22 Page 22 of 37

[37] J. Hernández-Borges, C. Neusüss, A. Cifuentes, M. Pelzing, On-line capillary electrophoresis-mass spectrometry for the analysis of biomolecules, Electrophoresis 25 (2004) 2257-81.

[38] M. Herrero, E. Ibañez, A. Cifuentes, Capillary electrophoresis-electrospray-mass spectrometry in peptide

ip t

analysis and peptidomics, Electrophoresis 29 (2008) 2148-60.

[39] W.F. Smyth, Recent applications of capillary electrophoresis-electrospray ionisation-mass spectrometry in

us

cr

drug analysis, Electrophoresis 27 (2006) 2051-62.

[40] D. Fliser, S. Wittke, H. Mischak, Capillary electrophoresis coupled to mass spectrometry for clinical

an

diagnostic purposes, Electrophoresis 26 (2005) 2708-16.

M

[41 ] Y. Li, M.M. Champion, L Sun, P.A. Champion, R Wojcik, N.J. Dovichi, Capillary zone electrophoresiselectrospray ionization-tandem mass spectrometry as an alternative proteomics platform to ultraperformance liquid chromatography-electrospray ionization-tandem mass spectrometry for samples of intermediate complexity, Anal Chem. 84 (2012) 1617-22.

ed

[42] W. Zhang, J.M. Czupryn, P.T. Boyle Jr, J. Amari, Characterization of asparagine deamidation and aspartate

pt

isomerization in recombinant human interleukin-11, Pharm. Res. 19 (2002) 1223-31.

[43] C. Terreaux, T. Walk, Y. van de Wal, F. Koning, G. Jung, B. Fleckenstein, Increased HLA-DQ2-affinity of

Ac ce

a synthetic gliadin peptide by acid-induced deamidation of glutamine residues, Bioorg. Med. Chem. Lett. 8 (1998) 2039-44.

[44] L. Huang, J. Lu, V.J. Wroblewski, J.M. Beals, R.M. Riggin, In vivo deamidation characterization of monoclonal antibody by LC/MS/MS, Anal. Chem. 77 (2005) 1432-9.

[45] K. Diepold, K. Bomans, M.Wiedmann, B. Zimmermann, A. Petzold, T. Schlothauer, R. Mueller, B. Moritz, J.O. Stracke, M. Mølhøj, D. Reusch, P. Bulau, Simultaneous assessment of Asp isomerization and Asn deamidation in recombinant antibodies by LC-MS following incubation at elevated temperatures, PLoS One 7 (2012).

23 Page 23 of 37

[46] Y.D. Liu, J.Z. van Enk, G.C. Flynn, Human antibody Fc deamidation in vivo, Biologicals 37 (2009) 313-22.

[47] OPCW; the Technical Secretariat of the Organisation for the Prohibition of Chemical Weapons, Scientific Advisory Board, 1999; SAB-II/-I.

ip t

[48] www.uniprot.org, using entry P02879.

us

cr

[49] www.rcsb.org, using the PDB identifier 2AAI.

[50] E. Rutenber, B.J. Katzin, S. Ernst, E.J. Collins, D. Mlsna, M.P. Ready, J.D. Robertus, Crystallographic

an

refinement of ricin to 2.5 A, Proteins 10 (1991) 240-50.

M

[51] N.E. Robinson, A.B. Robinson, www.deamidation.org.

ed

[52] N.E. Robinson, Protein deamidation, Proc. Natl. Acad. Sci. USA 99 (2002) 283-8.

[53] O.V. Krokhin, M. Antonovici, W. Ens, J.A. Wilkins, K.G. Standing, Deamidation of -Asn-Gly- sequences

Ac ce

78 (2006) 6645-50.

pt

during sample preparation for proteomics: Consequences for MALDI and HPLC-MALDI analysis, Anal. Chem.

[54] A.B. Robinson, J.H. McKerrow, P. Cary, Controlled deamidation of peptides and proteins: an experimental hazard and a possible biological timer, Proc. Natl. Acad. Sci. USA 66 (1970) 753-7.

[55] A.B. Robinson, J.H. McKerrow, M. Legaz, Sequence dependent deamidation rates for model peptides of cytochrome C, Int. J. Pept. Protein Res. 6 (1974) 31-5.

[56] A.B. Robinson, J.W. Scotchler, Sequence dependent deamidation rates for model peptides of histone IV, Int. J. Pept. Protein Res. 6 (1974) 279-82.

[57] N.E. Robinson, A.B. Robinson, Molecular clocks, Proc. Natl. Acad. Sci. USA 98 (2001) 944-9.

24 Page 24 of 37

[58] D. Chelius, D.S. Rehder, P.V. Bondarenko, Identification and characterization of deamidation sites in the conserved regions of human immunoglobulin gamma antibodies, Anal. Chem. 77 (2005) 6004–11.

[59] S. Capasso, S. Salvadori, Effect of the three-dimensional structure on the deamidation reaction of

ip t

ribonuclease A, J. Pept. Res. 54 (1999) 377-82.

cr

[60] W.J Chazin, A.A. Kossiakoff, The role of secondary and tertiary structure in intramolecular deamidation of

us

proteins, in: Deamidation and Isoaspartate Formation in Peptides and Proteins, CRC Press (1995) 193–206.

an

[61] S. Sinha, L. Zhang, S. Duan, T.D. Williams, J. Vlasak, R. Ionescu, E.M. Topp, Effect of protein structure on deamidation rate in the Fc fragment of an IgG1 monoclonal antibody, Protein Sci. 18 (2009) 1573-84.

M

[62] D. Ren, G.D. Pipes, D. Liu, L.Y. Shih, A.C. Nichols, M.J. Treuheit, D.N. Brems, P.V. Bondarenko, An

(2009) 12-21.

ed

improved trypsin digestion method minimizes digestion-induced modifications on proteins, Anal. Biochem. 392

pt

[63] A. Ostin, T. Bergström, S-A. Fredriksson, C. Nilsson, Solvent-assisted trypsin digestion of ricin for forensic

Ac ce

identification by LC-ESI MS/MS, Anal. Chem. 79 (2007) 6271-8.

[64] M. Perkins, R. Theiler, S. Lunte, M. Jeschke, Determination of the origin of charge heterogeneity in a murine monoclonal antibody, Pharm. Res. 17 (2000) 1110-7.

[65] H. Sarioglu, F. Lottspeich, T. Walk, G. Jung, C. Eckerskorn, Deamidation as a widespread phenomenon in two-dimensional polyacrylamide gel electrophoresis of human blood plasma proteins, Electrophoresis 21 (2000) 2209-18.

[66] N. Li, K. Kessler, L. Bass, D. Zeng, Evaluation of the iCE280 Analyzer as a potential high-throughput tool for formulation development, J. Pharm. Biomed. Anal. 43 (2007) 963-72.

25 Page 25 of 37

[67 ] K. Shimura, M. Hoshino, K. Kamiya, M. Enomoto, S. Hisada, H. Matsumoto, M. Novotny, K. Kasai, Estimation of the deamidation rates of major deamidation sites in a Fab fragment of mouse IgG1-κ by capillary isoelectric focusing of mutated Fab fragments, Anal. Chem. 85 (2013) 1705-10.

[68 ] J.J. Cournoyer, C. Lin, M.J. Bowman, P.B. O'Connor, Quantitating the relative abundance of isoaspartyl

Ac ce

pt

ed

M

an

us

cr

ip t

residues in deamidated proteins by electron capture dissociation, J. Am. Soc. Mass Spectrom. 18 (2007) 48-56.

26 Page 26 of 37

*Highlights (for review)

Research highlights Deamidation in ricin studied by capillary zone electrophoresis- and liquid chromatography- mass spectrometry, Bergström et al.

We studied deamidation in ricin toxin from the plant Ricinus communis.

Three aspargines in ricin were confirmed as being susceptible for deamidation.

ip t

Native / deamidated peptide pairs were analyzed by CZE- and LC-MS.

Ac

ce pt

ed

M

an

us

cr

The deamidation level in ricin was linked to previously reported isoelectric diversity.

Page 27 of 37

Ac ce p

te

d

M

an

us

cr

ip t

Figure 1

Page 28 of 37

Ac ce p

te

d

M

an

us

cr

ip t

Figure 2

Page 29 of 37

Figure 3

Ricin A chain:

ip t

IFPKQYPIINFTTAGATVQSYTNFIRAVRGRLTTGAD VRHEIPVLPNRVGLPINQRFILVELSNHAELSVTLAL DVTNAYVVGYRAGNSAYFFHPDNQEDAEAITHLFTDV QNRYTFAFGGNYDRLEQLAGNLRENIELGNGPLEEAI SALYYYSTGGTQLPTLARSFIICIQMISEAARFQYIE GEMRTRIRYNRRSAPDPSVITLENSWGRLSTAIQESN QGAFASPIQLQRRNGSKFSVYDVSILIPIIALMVYRC APPPSSQF

Ricin B chain:

an

us

cr

ADVCMDPEPIVRIVGRNGLCVDVRDGRFHNGNAIQLW PCKSNTDANQLWTLKRDNTIRSNGKCLTTYGYSPGVY VMIYDCNTAATDATRWQIWDNGTIINPRSSLVLAATS GNSGTTLTVQTNIYAVSQGWLPTNNTQPFVTTIVGLY GLCLQANSGQVWIEDCSSEKAEQQWALYADGSIRPQQ NRDNCLTSDSNIRETVVKILSCGPASSGQRWMFKNDG TILNLYSGLVLDVRASDPSLKQIILYPLHGDPNQIWL PLF

Ac

ce pt

ed

M

Fig. 3.

Page 30 of 37

Ac ce p

te

d

M

an

us

cr

ip t

Figure 4

Page 31 of 37

Ac

ce

pt

ed

M

an

us

cr

i

Figure 5

Page 32 of 37

Ac

ce

pt

ed

M

an

us

cr

i

Figure 6

Page 33 of 37

Ac ce p

te

d

M

an

us

cr

ip t

Figure 7

Page 34 of 37

Ac ce p

te

d

M

an

us

cr

ip t

Figure 8

Page 35 of 37

Table 1

Table 1.

Adjacent amino acids and predicted half-life for each Asn site in the A-chain of ricin.

GLY-ASN141-GLY ARG-ASN236-GLY SER-ASN222-GLN SER-ASN64-HIS ILE-ASN10-PHE TYR-ASN195-ARG ASP-ASN97-GLN THR-ASN78-ALA GLY-ASN132-LEU GLU-ASN209-SER GLN-ASN113-ARG PRO-ASN47-ARG GLY-ASN88-SER THR-ASN23-PHE GLU-ASN136-ILE ILE-ASN54-GLN GLY-ASN122-TYR

23 267 341 361 846 1174 2401 3408 4398 4407 6740 8249 9792 19072 29331 35566 44901

TA12 TA22 TA20 TA8 TA2 CA36 TA9 TA8 TA11 TA19 CA18 TA6 TA9 TA2 TA12 TA7 TA10

us an

M

ed

Deamidated Asn

ip t

Peptide analyseda

glycosylatedb + -

cr

Predicted half-life (days)

ce pt

a

Chain A primary sequence

Peptides selected to map deamidation. The tryptic (T) and chymotryptic (C) peptides were

annotated according to their theoretical enumeration from the N-terminal to the C-terminal of

b

Ac

the A and B chain, respectively.

Glycosylated asparagines cannot be deamidated

Page 36 of 37

Table 2

Table 2.

Adjacent amino acids and predicted half-life for each Asn site in the B-chain of ricin.

ALA-ASN155-SER SER-ASN60-GLY ARG-ASN17-GLY HIS-ASN30-GLY GLY-ASN113-SER SER-ASN42-THR ASP-ASN189-CYS GLY-ASN32-ALA CYS-ASN81-THR SER-ASN196-ILE LYS-ASN220-ASP PRO-ASN255-GLN ILE-ASN100-PRO THR-ASN123-ILE LEU-ASN226-LEU ASN-ASN136-THR ALA-ASN46-GLN GLN-ASN186-ARG THR-ASN135-ASN ASP-ASN95-GLY ASP-ASN55-THR

270 343 399 404 1024 1133 1784 2099 2321 2893 9526 10559 17268 22053 29361 60250 80656 93619 95589 199548 289664

CB24 TB9 TB3-ss-TB5 TB3-ss-TB5 CB16 TB6 TB14-ss-TB16 TB3-ss-TB5 TB10 TB14-ss-TB16 TB18 TB20 TB11 TB12 TB18 CB19 TB6 TB13 TB12 TB11 TB8

us

an

M

ed

ce pt

Deamidated Asn

ip t

Peptide analyseda

+ + glycosylatedb glycosylatedb -

cr

Predicted half-life (days)

Peptides selected to map deamidation. The tryptic (T) and chymotryptic (C) peptides were

Ac

a

Chain B primary sequence

annotated according to their theoretical enumeration from the N-terminal to the C-terminal for the A and B chain, respectively.

b

Glycosylated asparagines cannot be deamidated

Page 37 of 37