METHOD FOR ANALYZING SIALYL SUGAR CHAIN

Abstract

Provided is a method for efficiently analyzing the structure of a sialylated glycan, including the linkage type of a sialic acid linked to a terminal of the glycan. After the sialic acid residues in a sample (glycan) to be analyzed are modified in a linkage-type-specific way (S2), a mass spectrum for the sample is obtained by mass spectrometry (S3). Based on the masses of the modified glycans estimated from the m/z values of the peaks observed on the mass spectrum, all possible monosaccharide compositions are exhaustively estimated under specific conditions including the kinds of monosaccharides and the range of the number of occurrences of each monosaccharide (S4). Monosaccharide composition candidates listed for each peak are narrowed down by using the mass difference between different modifications of a sialic acid residue generated by the linkage-type-specific modification (S5). Specifically, peaks whose mass differences correspond to the mass difference between different modifications of the sialic acid residue generated by the linkage-type-specific modification are located and assumed as a cluster of peaks originating from the same glycan including linkage isomers. For each peak belonging to this cluster, the corresponding monosaccharide composition candidates are narrowed down by excluding candidates that do not satisfy specific conditions, such as the presence or absence of a modified sialic acid residue and the number of occurrences of the modified sialic acid residue.

Claims

1. A sialylated glycan analysis method using mass spectrometry for analyzing a structure of a sialylated glycan to which a sialic acid is linked, the method comprising: a) a modification step in which a linkage-type-specific modification is performed on a sialic acid residue included in a sialylated glycan to be analyzed; b) an analysis execution step in which the sialylated glycan after the modification by the modification step is subjected to a mass spectrometric analysis to obtain mass spectrum information; c) a candidate estimation step in which a candidate of a monosaccharide composition is estimated for a peak observed on the mass spectrum obtained in the analysis execution step, based on mass information of the peak; and d) a candidate-narrowing step in which a plurality of peaks observed on the mass spectrum and having a mass difference corresponding to a technique of the linkage-type-specific modification are assumed to be a cluster of peaks corresponding to sialylated glycans including sialic acid residues of a same kind with different linkage types and being identical in the monosaccharide composition exclusive of the sialic acid residues, and monosaccharide composition candidates obtained for those peaks are narrowed down by judging each monosaccharide composition candidate from at least one of following aspects: presence or absence of a sialic acid residue; presence or absence of a specific type of modified sialic acid residue, or number of occurrences of the specific type of modified sialic acid residue; and identity of the monosaccharide composition exclusive of the sialic acid residue and the modified sialic acid residue.

2. The sialylated glycan analysis method according to claim 1, wherein: an analysis of a glycan structure is performed in which 2,3-sialic acid residue and 2,6-sialic acid residue are distinguished from each other as the sialic acid residues with different linkage types.

3. The sialylated glycan analysis method according to claim 2, wherein: an isopropylamide modification and a methylamide modification are performed as the linkage-type-specific modification in the modification step.

4. The sialylated glycan analysis method according to claim 1, wherein: in the candidate-narrowing step, for a sialylated glycan having a sialic acid linked to one or more terminals of the glycan, a plurality of peaks corresponding to a mass difference between different types of modified sialic acid residue generated by the linkage-type-specific modification are located, and the monosaccharide composition candidates are narrowed down under a condition that the monosaccharide composition corresponding to a peak on a lower-mass side among the plurality of peaks should include at least one modified sialic acid residue having a smaller mass while the monosaccharide composition corresponding to a peak on a higher-mass side should include at least one modified sialic acid residue having a larger mass.

5. The sialylated glycan analysis method according to claim 4, wherein: in the candidate-narrowing step, two peaks corresponding to N times the mass difference M between the different types of modified sialic acid residue generated by the linkage-type-specific modification (where N is an integer equal to or greater than one) are located, and the monosaccharide composition candidates are narrowed down under a condition that the monosaccharide composition corresponding to the peak on the lower-mass side should include the modified sialic acid residue having the smaller mass and occurring at N locations.

6. The sialylated glycan analysis method according to claim 5, wherein: the modified sialic acid residue having the smaller mass and occurring at N locations is a modified 2,3-sialic acid residue.

7. The sialylated glycan analysis method according to claim 4, wherein: in the candidate-narrowing step, two peaks corresponding to N times the mass difference M between the different types of modified sialic acid residue generated by the linkage-type-specific modification (where N is an integer equal to or greater than one) are located, and the monosaccharide composition candidates are narrowed down under a condition that the monosaccharide composition corresponding to the peak on the higher-mass side should include the modified sialic acid residue having the larger mass and occurring at N locations.

8. The sialylated glycan analysis method according to claim 7, wherein: the modified sialic acid residue having the larger mass and occurring at N locations is a modified 2,6-sialic acid residue.

9. The sialylated glycan analysis method according to claim 1, wherein: an estimation of a monosaccharide composition candidate in the candidate estimation step is performed by computing a combination of monosaccharides which are consistent with masses of the peaks under search conditions which specify kinds of potentially contained monosaccharides and a range of the number of occurrences of each monosaccharide.

10. The sialylated glycan analysis method according to claim 1, wherein: the mass spectrometric analysis in the analysis execution step performed in a negative-ion mode.

Description

BRIEF DESCRIPTION OF DRAWINGS

[0040] FIG. 1 is a flow chart schematically showing an analytical procedure in a sialylated glycan analysis method as one embodiment of the present invention.

[0041] FIG. 2 is a diagram explaining a method for narrowing down monosaccharide composition candidates in the sialyl glycan analysis method according to the present embodiment (in the case where the number of sialic acid residues is two).

[0042] FIG. 3 is a diagram explaining a method for narrowing down monosaccharide composition candidates in the sialyl glycan analysis method according to the present embodiment (in the case where the number of sialic acid residues is three).

[0043] FIG. 4 is an illustration showing a measured mass spectrum for an N-linked glycan derived from bovine fetuin modified in a way that is non-specific to the linkage type of the sialic acid, as well as the result of a search for monosaccharide compositions for four major peaks.

[0044] FIG. 5 is an illustration showing a measured mass spectrum for an N-linked glycan derived from bovine fetuin modified in a way that is specific to the linkage type of the sialic acid, as well as the result of a search for monosaccharide compositions for nine major peaks.

[0045] FIG. 6 is an enlarged view of a section around m/z=2500 in a measured mass spectrum for an N-linked glycan derived from bovine fetuin modified in a way that is specific to the linkage type of the sialic acid, as well as the result of a process in which the result of the search for the monosaccharide compositions for the peaks detected within the enlarged section has been narrowed down by the method according to the present embodiment.

[0046] FIG. 7 is an enlarged view of a section around m/z=2800 in a measured mass spectrum for an N-linked glycan derived from bovine fetuin modified in a way that is specific to the linkage type of the sialic acid, as well as the result of a process in which the result of the search for the monosaccharide compositions for the peaks detected within the enlarged section has been narrowed down by the method according to the present embodiment.

[0047] FIG. 8 is an enlarged view of a section around m/z=3150 in a measured mass spectrum for an N-linked glycan derived from bovine fetuin modified in a way that is specific to the linkage type of the sialic acid, as well as the result of a process in which the result of the search for the monosaccharide compositions for the peaks detected within the enlarged section has been narrowed down by the method according to the present embodiment.

[0048] FIG. 9 is an enlarged view of a section around m/z=3500 in a measured mass spectrum for an N-linked glycan derived from bovine fetuin modified in a way that is specific to the linkage type of the sialic acid, as well as the result of a process in which the result of the search for the monosaccharide compositions for the peaks detected within the enlarged section has been narrowed down by the method according to the present embodiment.

DESCRIPTION OF EMBODIMENTS

[0049] A sialylated glycan analysis method according to one embodiment of the present invention is hereinafter described in detail with reference to the attached drawings. FIG. 1 is a flow chart schematically showing an analytical procedure in the sialylated glycan analysis method according to the present embodiment. FIGS. 2 and 3 are conceptual diagrams for explaining a method for narrowing down the monosaccharide composition candidates in the sialyl glycan analysis method according to the present embodiment.

[0050] Referring to FIGS. 1-3, the following description initially provides an overview of the sialylated glycan analysis method according to the present embodiment as well as the characteristic method for narrowing down monosaccharide composition candidates.

[0051] Initially, a sample is prepared from a glycoprotein including glycans to be analyzed (sialylated glycans) by cutting out the glycans from the glycoprotein by a known method and then purifying the cut glycans (Step S1).

[0052] Subsequently, the sialic acids linked to the glycans contained in the prepared sample are modified by a linkage-type-specific derivatization (Step S2). Two particularly important kinds of linkage types, i.e. 2,3-linkage and 2,6-linkage, are considered in the present embodiment, although there are also other known forms for sialic acids to be linked to glycans, such as the 2,8-linkage. As noted earlier, there are several commonly known modification methods for distinguishing between the 2,3-sialic acid residue and 2,6-sialic acid residue (for example, see Non-Patent Literature 1-5). Any modification (derivatization) method may be used as long as it can modify sialic acid residues in a linkage-type-specific way. The combination of isopropyl amide modification and methyl amide modification disclosed in Non-Patent Literature 5 is used in the present embodiment. In this case, the 2,6-sialic acid residue turns into an isopropyl amide modification, while the 2,3-sialic acid residue turns into a methyl amide modification. In the following description, the sialic acid modified with isopropyl amide is called the iPA modification, while the sialic acid modified with methyl amide is called the MA modification.

[0053] As noted earlier, there are sialic acids other than Neu5Ac, such as Neu5Gc and KDN. In the present embodiment, Neu5Ac is considered, which is the most abundant. The masses to be mentioned in the following description will basically be expressed in nominal mass (an integer mass of an isotope having the highest natural abundance ratio) in order to avoid complication of the description. The mass of Neu5Ac is 291, while those of the iPA and MA modifications are 332 and 304, respectively. The two modifications have a mass difference of 28 Da.

[0054] After the sialic acid residues included in the glycans in the sample have been modified in the linkage-type-specific way as described earlier, the sample is subjected to a mass spectrometric analysis to obtain a mass spectrum covering a predetermined range of mass-to-charge ratios (Step S3). The mass accuracy should preferably be high so that the monosaccharide composition can be estimated based on mass information as will be described later. Accordingly, it is preferable to use a mass spectrometer capable of an analysis with high accuracy and high sensitivity, such as a time-of-flight mass spectrometer or Fourier transform mass spectrometer.

[0055] After the mass spectrum has been obtained, the peaks observed on the mass spectrum are extracted. For each extracted peak, the monosaccharide composition is estimated from the mass value of the peak. The estimation of the monosaccharide composition is specifically achieved by an exhaustive search for all possible combinations of monosaccharides which are consistent with the mass value of the measured peak (or practically, which fall within a predetermined numerical range) under the search conditions specified beforehand by an analysis operator, including the kinds of monosaccharaides and the range of the number of occurrences of each monosaccharide, with the permissible mass error and other items of information as analytical parameters (Step S4). Such a search can be performed on a computer, for example, using the Glyco-Peakfinder, which is a piece of glycan search software freely available on an Internet website (see Non-Patent Literature 6). Needless to say, other software products may also be used to perform similar processing.

[0056] In estimating the monosaccharide composition, it is necessary to specify the kinds of monosaccharides and the range of the number of occurrences of each monosaccharide in a slightly broader way in order to avoid an omission of the correct monosaccharide composition. Therefore, two or more monosaccharide composition candidates will inevitably be listed for each peak. In the case of the already described example of FIG. 5, ten monosaccharide composition candidates are listed for a peak of m/z=3529.4. Accordingly, the next task is to narrow down the monosaccharide composition candidates for each peak, using the information on the mass difference corresponding to the technique used for the linkage-type-specific modification process performed in Step S2 (Step S5). The method for this narrowing process is hereinafter described with reference to FIGS. 2 and 3.

[0057] Now, consider the case where the sialyl glycan has a two-antenna (two-branched) structure with one sialic acid linked to the terminal of each chain. In this case, as shown in boxes (a)-(c) in FIG. 2, there are three possible combinations of the 2,3-sialic acid and 2,6-sialic acid. The three combinations have the same mass and cannot be distinguished from each other by their masses. After being subjected to the linkage-type-specific modification as described earlier, the 2,3-sialic acid residue becomes the MA modification, while the 2,6-sialic acid residue becomes the iPA modification. Accordingly, there are three possible combinations of the modifications as shown in boxes (d)-(f) in FIG. 2. As noted earlier, the MA and iPA modifications have a mass difference of 28 Da. Therefore, the masses of the three combinations of the modifications differ from each other in units of 28 Da. That is to say, the sialyl glycans which have undergone the linkage-type-specific modification have distinguishable masses which depend on the linkage types of the sialic acid residues.

[0058] The three aforementioned combinations will appear as three peaks P1, P2 and P3 at intervals of 28 Da on a mass spectrum obtained by a mass spectrometric analysis of the sialylated glycans which have been modified in the linkage-type-specific way. Accordingly, if there are a plurality of peaks observed at intervals of 28 Da on a mass spectrum, those peaks are most likely to have originated from glycans which merely differ from each other in the kind of modification of the sialic acid residue, i.e. which are identical in the monosaccharide composition exclusive of the sialic acid residues as well as in the number of sialic acid residues. Such a group of peaks are assumed to be an isomer cluster peak, i.e. a group of peaks which are identical in the monosaccharide composition exclusive of the sialic acid residues as well as in the number of sialic acid residues while being merely different from each other in the linkage type of the sialic acid residues.

[0059] The correspondence relationship between the three combinations of the modifications and the three peaks in FIG. 2 demonstrates the following facts:

[0060] (A1) The monosaccharide compositions corresponding to the three peaks must include a modified sialic acid residue.

[0061] (A2) The monosaccharide composition corresponding to the smaller-mass peak of any two peaks located at an interval of 28 Da includes one or more MA modifications, i.e. the 2,3-sialic acid residues.

[0062] (A3) The monosaccharide composition corresponding to the larger-mass peak of any two peaks located at an interval of 28 Da includes one or more iPA modifications, i.e. the 2,6-sialic acid residues.

[0063] (A4) For a glycan having two sialic acid residues, if three peaks are observed at intervals of 28 Da, the monosaccharide composition corresponding to the peak having the largest mass includes only the iPA modification, while the monosaccharide composition corresponding to the peak having the smallest mass includes only the MA modification.

[0064] (A5) The glycans are identical in the monosaccharide composition exclusive of the sialic acid residues.

[0065] Monosaccharide composition candidates which do not satisfy those conditions (A1) to (A5) can be excluded from the candidates. Thus, the narrowing of the candidates can be achieved.

[0066] Next, consider the case of a sialylated glycan having a three-antenna (three-branched) structure with one sialic acid linked to the terminal of each chain. In this case, as shown in boxes (a)-(d) in FIG. 3, there are four possible combinations of the 2,3-sialic acid and 2,6-sialic acid. The four combinations have the same mass and cannot be distinguished from each other by their masses. After the linkage-type-specific modification is performed as described earlier, there are four possible combinations of the modifications as shown in boxes (e)-(h) in FIG. 3. As noted earlier, the MA and iPA modifications have a mass difference of 28 Da. Therefore, the masses of the four combinations of the modifications differ from each other in units of 28 Da. That is to say, the sialyl glycans which have undergone the linkage-type-specific modification have distinguishable masses which depend on the linkage types of the sialic acid residues.

[0067] The four aforementioned combinations will appear as four peaks P1, P2, P3 and P4 at intervals of 28 Da on a mass spectrum obtained by a mass spectrometric analysis of the sialyl glycans which have been modified in the linkage-type-specific way. Accordingly, similar to the previous case, if there are a plurality of peaks observed at intervals of 28 Da on a mass spectrum, those peaks can be assumed to be an isomer cluster peak. Similar to the previous case in which the number of sialic acid residues is two, the following conditions can be applied to the isomer cluster peak.

[0068] (B1) The monosaccharide compositions corresponding to the four peaks must include a modified sialic acid residue.

[0069] (B2) The monosaccharide composition corresponding to the smaller-mass peak of any two peaks located at an interval of 28 Da includes one or more MA modifications, i.e. the 2,3-sialic acid residues.

[0070] (B3) The monosaccharide composition corresponding to the larger-mass peak of any two peaks located at an interval of 28 Da includes one or more iPA modifications, i.e. the 2,6-sialic acid residues.

[0071] (B4) The monosaccharide composition corresponding to a peak along with N peaks located at intervals of 28 Da on the larger-mass side includes N or more MA modifications, i.e. the 2,3-sialic acid residues.

[0072] (B5) The monosaccharide composition corresponding to a peak along with N peaks located at intervals of 28 Da on the smaller-mass side includes N or more iPA modification, i.e. the 2,6-sialic acid residues.

[0073] Monosaccharide composition candidates which do not satisfy those conditions (B1) to (B5) can be excluded from the candidates. Thus, the narrowing of the candidates can be achieved.

[0074] As described to this point, the monosaccharide composition candidates which are related to a plurality of peaks belonging to one isomer cluster peak identified by the mass difference of the peaks on a mass spectrum can be appropriately narrowed down by excluding candidates which are inconsistent with specific judgment conditions, such as the presence or absence of a modified sialic acid residue, presence or absence of a specific type of modified sialic acid residue, number of occurrences of the modification of the specific type of modified sialic acid residue, or identity of the monosaccharide composition exclusive of the sialic acid residues. Such a narrowing process can also be performed on a computer. The monosaccharide composition candidates which remain after the narrowing process are presented to the analysis operator as the analysis result, for example, on the screen of a display unit (Step S6).

EXAMPLE

[0075] An experimental example in which the sialyl glycan analysis method according to the present embodiment was applied is hereinafter described in detail.

[0076] [Step S1] Release and Purification of Glycans

[0077] In the experiment, a sample was prepared in the following manner by releasing glycans from a glycoprotein which includes both 2,3-linked and 2,6-linked sialyl glycan.

[0078] Initially, bovine fetuin was dissolved in a mixed liquid containing ammonium bicarbonate in a concentration of 20 mM, dithiothreitol (DTT) in a concentration of 10 mM and sodium dodecyl sulfate (SDS) in a concentration of 0.02%. The solution was heated at 100 C. for three minutes to denature and reduce the fetuin. Subsequently, the solution was cooled to room temperature. After PNGase F (Peptide-N-Glycosidase F) enzyme was added, the solution was incubated at 37 C. overnight to release glycans from the peptides. Then, the solution was heated at 100 C. for three minutes to deactivate PNGase F and thereby discontinue the enzymatic reaction.

[0079] The glycans released by the enzymatic reaction were subsequently desalted and purified using a carbon column. This carbon column was a self-build microchip similar to a purification column. It was prepared by filling a 200-L column chip with a stack of active-carbon discs of approximately 1 mm in diameter punched from a carbon disk (Empore disk carbon). Initially, 100 L of acetonitrile (ACN) was put into this carbon column, and the solution was centrifugally discharged. Similar operations were also performed using each of the following solutions: 1M sodium hydroxide (NaOH), 1M hydrochloric acid (HCl), 60% ACN, 0.1% trifluoroacetic acid (TFA) solution, and water, in order to wash and equilibrate the column adsorbent. The solution obtained through the enzymatic reaction was introduced into this column, and the solution was centrifugally discharged. An operation of adding 200 L of water and centrifugally discharging the liquid was repeated three times for washing. Finally, an operation of adding 20 L of a mixed liquid of 60% ACN and 0.1% TFA solution and centrifugally collecting the solution was repeated two times to elute the glycans. The entire eluate obtained through this two-time operation was put together, and the solvent was removed by a centrifugal vacuum concentrator to obtain the sample in a dried form.

[0080] [Step S2] Linkage-Type-Specific Modification of Sialic Acid Residues

[0081] To the dried sample, 10 L of a 4M solution of isopropylamine hydrochloride dissolved in dimethyl sulfoxide (DMSO) was added. Then, 10 L of a solution prepared by dissolving diisopropyl carbodiimide (DIC) and 1-hydroxy benzotriazole (HOBt) in DMSO so that each solute was contained in a concentration of 500 mM was added as a dehydration condensation agent. The mixed solution was stirred at room temperature for two minutes and made to react with each other at 37 C. for one hour. After the reaction, the solution was diluted with 120 L of 93.3% ACN and 0.13% TFA solution. An excessive amount of reagent was removed by GL-Tip Amide, manufactured by GL Sciences Inc. This was specifically performed as follows: An operation of adding 100 L of water to GL-Tip Amide and centrifugally discharging the solution was repeated three times to wash the tip. Next, an operation of adding 100 L of 90% ACN and 0.1% TFA solution and centrifugally discharging the solution was repeated three times to equilibrate the tip. Then, the diluted solution after the reaction was entirely added and centrifugally processed to make the glycans be adsorbed onto the adsorbent. Subsequently, an operation of adding 200 L of 90% ACN and 0.1% TFA solution and centrifugally discharging the solution was repeated three times to wash the tip. Finally, an operation of adding 10 L of water and centrifugally discharging the water was repeated two times to elute the glycans. The entire eluate obtained through this two-time operation was put together, and the solvent was removed by a centrifugal vacuum concentrator to obtain the sample in a dried form.

[0082] Subsequently, 10 L of a 2M solution of methylamine hydrochloride dissolved in DMSO was added to the dried sample. Then, 10 L of a solution of tripyrrolizinophosphonium hexafluorophosphate (PyBOP) dissolved in 30% N-methylmorpholine (NMM) to a molar concentration of 500 mM was added as a dehydration condensation agent. The mixed solution was stirred at room temperature for one hour to promote the reaction. After the reaction, 120 L of 93.3% ACN and 0.13% TFA solution was added to the solution. The previously described operations for the purification and elution using GL-Tip Amide were similarly performed. The collected eluate was processed by the centrifugal vacuum concentrator to obtain the sample in a dried form.

[0083] Through the previously described processes, a sample in which the sialic acids linked to the glycans are modified in a linkage-specific way can be prepared.

[0084] [Step S3] Execution of Mass Spectrometric Analysis

[0085] The dried sample was re-dissolved in an appropriate amount of water. A fraction (1 L) of the obtained solution was dropped on a MALDI focus plate, to which 0.5 L of a 50% ACN solution containing 100 mM 3AQ/CA and 2 mM ammonium sulfate was added as the matrix. The plate was placed on a heat block at 75 C. and left intact for 1.5 hours. This is to promote the reaction of labelling the reducing end of the glycans with 3AQ. After the reaction was completed, the plate was cooled to room temperature, and a mass spectrometric analysis in the negative-ion mode was performed using a MALDI-QIT-TOFMS (AXIMA Resonance, manufactured by Shimadzu Corporation and Kratos Analytical Ltd.) as the mass spectrometer, to obtain a mass spectrum over a predetermined range of mass-to-charge ratios. For the measurement, a technique called On-Target 3AQ Labeling was used, which is a high-sensitivity glycan detection method proposed by the applicant in Non-Patent Literature 7, 8 and other documents. The measurement technique is not limited to this one.

[0086] [Step S4] Estimation of Monosaccharide Composition

[0087] For the estimation of the monosaccharide composition, Glyco-Peakfinder was used, which is a piece of software for searching for monosaccharide composition candidates which are consistent with the mass-to-charge-ratio values detected on a mass spectrum. This software basically does not use any database; it performs an exhaustive search for computing all possible combinations of the known masses of the monosaccharides and the number of occurrences of each monosaccharide. The aforementioned software in its original setting does not consider modifications of sialic acids. Therefore, the masses of the modified sialic acid residues were additionally set as optional monosaccharide residues for the search.

[0088] The kinds of monosaccharides and the range of the number of occurrences of each monosaccharide specified as the search conditions were as follows: [0089] Hexose (Hex): 3 to 15 [0090] N-acetylhexosamine (HexNAc): 2 to 24 [0091] Deoxyhexose (dHex): 0 to 4 [0092] N-acetylneuraminic acid (Neu5Ac): 0 to 5 (or 0 to 5 for each mass if Neu5Ac in the modified form has a different mass depending on the linkage type) [0093] Sulfate (S): 0 to 2

[0094] The search result was as shown in FIG. 5, which has already been mentioned.

[0095] [Step S5] Narrowing of Monosaccharide Composition Candidates

[0096] <In the Case of Biantennary Glycan+Two Sialic Acid Residues>

[0097] FIG. 6 shows an enlarged view of a section of the mass spectrum around m/z=2500 and a list showing the result of a monosaccharide composition search within the enlarged section. Candidates which have been excluded by the narrowing process are indicated in faint-colored letters in the list showing the result of the monosaccharide composition search in FIG. 6 (as well as in FIGS. 8 and 9).

[0098] The peak observed at m/z=2471.9 in the mass spectra shown in FIGS. 4 and 5 can be identified, from its mass-to-charge-ratio value, as a biantennary glycan with two sialic acid residues added, as with the glycan structure shown in FIG. 6. In the mass spectrum shown in FIG. 6, the peak of m/z=2499.9 is accompanied by two peaks on both sides, with each peak distanced from the central peak by 28 Da which is based on the difference in mass-to-charge ratio between the 2,3-sialic acid residue and the 2,6-sialic acid residue (i.e. between the MA and iPA modifications). These three peaks can be considered as forming an isomer cluster peak. As described earlier, a glycan which corresponds to three peaks belonging to an isomer cluster peak must include a sialic acid residue. Therefore, among the monosaccharide composition candidates related to those three peaks, those which include no sialic acid residue should be excluded from the candidates. In the example of FIG. 6, the first and second candidates among the four candidates related to peak P3 include no sialic acid residue and should be excluded.

[0099] The monosaccharide which corresponds to a peak accompanied by another peak with a mass difference of 28 Da on the larger-mass side on the mass spectrum among the three aforementioned peaks must have one or more MA modifications in its composition. Accordingly, among the monosaccharide composition candidates related to that peak, those which include no MA modification should be excluded from the candidates. In the example of FIG. 6, the second candidate among the two candidates related to peak P2 includes no 2,3-sialic acid residue and should be excluded.

[0100] Conversely, the monosaccharide which corresponds to a peak accompanied by another peak with a mass difference of 28 Da on the smaller-mass side on the mass spectrum must have one or more iPA modification in its composition. Accordingly, among the monosaccharide composition candidates related to that peak, those which include no iPA modification should be excluded from the candidates. In the example of FIG. 6, the second candidate among the two candidates related to peak P1 includes no 2,6-sialic acid residue and should be excluded.

[0101] Additionally, the monosaccharide corresponding to peak P3 must include two or more iPA modifications in its composition, since peak P2 is located at a position corresponding to a mass difference of 56 Da, which equals two times 28 Da, on the smaller-mass side from peak P3. Therefore, among the two remaining candidates related to peak P3, the fourth candidate which includes only one 2,6-sialic acid residue should be excluded.

[0102] The monosaccharide composition candidates can be narrowed down by such a process. In the example of FIG. 6, the monosaccharide composition candidates related to the three peaks can be narrowed down to a single candidate for each peak.

[0103] <In the Case of Triantennary Glycan+Two Sialic Acid Residues>

[0104] FIG. 7 shows an enlarged view of a section of the mass spectrum around m/z=2800 and a list showing the result of a monosaccharide composition search within the enlarged section. In the present case, there are three estimated monosaccharide compositions for the peak of m/z=2865.1. However, there is no isomer cluster peak observed with a significant intensity, i.e. with an intensity equal to or higher than a predetermined threshold. Therefore, in the present case, it is impossible to narrow down the candidates using an isomer cluster in the previously described manner.

[0105] <In the Case of Triantennary Glycan+Three Sialic Acid Residues>

[0106] FIG. 8 shows an enlarged view of a section of the mass spectrum around m/z=3150 and a list showing the result of a monosaccharide composition search within the enlarged section.

[0107] The peak located at m/z=3141.1 can be identified, from its mass-to-charge-ratio value, as a triantennary glycan with three sialic acid residues added, as with the glycan structure shown in FIG. 8. There are four monosaccharide composition candidates listed for this peak of m/z=3141.1. There are two peaks observed at positions of +28 Da and +56 Da (=282) from this peak, respectively. Accordingly, it is possible to consider that these three peaks form an isomer cluster peak.

[0108] Among the monosaccharide composition candidates corresponding to the peak of m/z=3141.1 belonging to this isomer cluster peak, the first candidate which includes no 2,3-sialic acid residue (Neu5Ac(2,3)) and the fourth candidate which includes only one 2,3-sialic acid residue can be excluded. The peak of m/z=3169.1 located at the center of the three peaks has seven monosaccharide composition candidates, among which the first through third candidates which include no sialic acid residue can be excluded. The sixth and seventh candidates can also be excluded, since any candidate for this peak should include at least one 2,3-sialic acid residue (Neu5Ac(2,3)) and at least one 2,6-sialic acid residue (Neu5Ac(2,6)). The peak located at the highest mass-to-charge ratio of m/z=3197.3 among the three peaks has six monosaccharide composition candidates, among which the first through third candidates as well as the sixth candidate can be excluded, since any candidate for this peak should include two or more 2,6-sialic acid residues (Neu5Ac(2,6)).

[0109] Thus, the number of monosaccharide composition candidates can be decreased from 17 to 6 by assuming that the three peaks having a mass difference of 28 Da form an isomer cluster peak.

[0110] <In the Case of Triantennary Glycan+Four Sialic Acid Residues>

[0111] FIG. 9 shows an enlarged view of a section of the mass spectrum around m/z=3500 and a list showing the result of a monosaccharide composition search within the enlarged section.

[0112] The peak located at m/z=3501.4 can be identified, from its mass-to-charge-ratio value, as a three-antenna glycan with four sialic acid residues added, as with the glycan structure shown in FIG. 9. Since there are four sialic acid residues, there are four possible combinations of the 2,3-sialic acid residue and the 2,6-sialic acid residue. However, only three peaks having a mass-to-charge-ratio difference of 28 Da will be observed on the mass spectrum. Even in this case, it is possible to exclude monosaccharide composition candidates based on similar rules to the previously described ones.

[0113] That is to say, for the peak of m/z=3501.4, a peak is observed at a position of +28 Da. Therefore, among the ten monosaccharide composition candidates corresponding to the peak of m/z=3501.4, the first, third, fifth and tenth candidates which include no 2,3-sialic acid residue should be excluded. Additionally, another peak is observed at a position of 28 Da, although its intensity is low. The presence of this peak allows for the exclusion of the second, fourth and sixth candidates which include no 2,6-sialic acid. Thus, the candidates can be narrowed down to three.

[0114] Similarly, among the ten monosaccharide composition candidates corresponding to the peak of m/z=3529.4, the seventh candidate which includes no sialic acid residue can be excluded. The presence of a peak at a position of 28 Da allows for the exclusion of the first, third and fifth candidates which include no 2,6-sialic acid residue. Furthermore, the presence of another peak at a position of 56 Da allows for the exclusion of the second, fourth and sixth candidates which include only one 2,6-sialic acid residue. Thus, the candidates can be narrowed down to three. Any of the monosaccharide composition candidates remaining after the narrowing process includes four sialic acid residues. In this manner, the candidates can be satisfactorily narrowed down even when all peaks that should belong to an isomer cluster peak are not completely observed.

[0115] In the narrowing method used in the previously described experiment, candidates are excluded based on the presence or absence of the sialic acid residue or its modification as well as the number of occurrences of the sialic acid residue or its modification. As described earlier, the narrowing of the candidates of a glycan corresponding to a peak belonging to an isomer cluster peak can also be achieved by excluding candidates which do no satisfy the condition that the monosaccharide composition exclusive of the modified sialic acid residues is the same, i.e. by excluding candidates having different monosaccharide compositions.

[0116] In the previously described embodiment, the isopropyl amide modification and the methyl amide modification are used as the linkage-type-specific modifications. A lactone modification may be used in place of the methyl amide modification. In that case, the two modifications have an nominal mass difference of 59 Da. Accordingly, an isomer cluster peak can be located by searching for peaks located at intervals of 59 Da or its integral multiple.

[0117] In the case of the linkage-type-specific modification method described in Non-Patent Literature 1 (lactone modification and methyl ester modification), the nominal mass difference between the two modifications is 32 Da. In the case of the linkage-type-specific modification method described in Non-Patent Literature 2 (lactone modification and ethyl ester modification), the nominal mass difference between the two modifications is 48 Da. In the case of the linkage-type-specific modification method described in Non-Patent Literature 3 (lactone modification and amide modification, followed by permethylation), the nominal mass difference between the two modifications is 13 Da. In the case of the linkage-type-specific modification method described in Non-Patent Literature 4 (lactone modification and dimethyl amide modification), the integer mass difference between the two modifications is 45 Da. In any of these cases, an isomer cluster peak can similarly be located by searching for peaks located at intervals of the mass difference or its integral multiple.

[0118] In the previously described embodiment, MALDI is used as the ionization method. It is naturally possible to apply the present invention in the case where a different ionization method is used, such as an electrospray ionization in which multiply-charged ions are generated. In the case where the sample is detected as a multiply-charged ion due to the use of such an ionization method, attention should be paid to the fact that the mass difference between the modifications to be considered must be divided by the number of the charge of the ion in order to locate a correct isomer cluster peak. Recent high-performance mass spectrometers can instantly determine the number of the charge of an ion based on the intervals of the isotopic masses calculated from an obtained mass spectrum. If such a mass spectrometer is used, the mass of a modified glycan can be automatically estimated from the determined charge. For the detection of an isomer cluster peak, the masses of the modified glycans estimated from mass-to-charge ratios may be used in place of the mass-to-charge ratios.

[0119] Although only the N-acetylneuraminic acid is considered as the sialic acid in the previously described embodiment, it is evident that the present invention is also applicable to other kinds of sialic acids, such as the N-glycolylneuraminic acid. It is also evident that the present invention is also applicable to any linkage type other than the 2,3- or 2,6-linkage as long as the sialic acid can be modified in a way that is specific to the linkage type concerned.

[0120] It should also be noted that the previously described embodiment is a mere example of the present invention, and any change, modification, addition or the like appropriately made within the spirit of the present invention other than the previously described variations will also naturally fall within the scope of claims of the present application.