Glycans that modify proteins are heterogeneous, and glycoform analysis at the protein level may be possible with purified materials, but it is generally still quite difficult. Simultaneous analysis of complex mixture of such glycoforms is thought to be impossible. Therefore, glycoproteomics is limited to digesting protein mixtures, enriching glycopeptides, performing LC/MS analysis, obtaining glycoforms at the peptide level, and inferring the overall picture in a bottom-up manner. Moreover, glycans are often ambiguous in their composition information (e.g., Hex5HexNAc4Fuc1NeuAc2, etc.), and many details of their structure are unknown. However, even this level of analysis is not easy. The reasons are multiple, including the diversity and heterogeneity of glycans and the intrinsic structural characteristics that induce misidentification. For example, the building blocks that make up a glycopeptide include elements that differ in mass by one oxygen, such as Ala-Ser, Phe-Tyr, Met-its oxide, Hex-dHex (Fuc), and NeuAc-NeuGc. Even if these blocks are replaced, the mass is exactly the same (e.g., M(Met+Hex) = M(Met(ox) + Fuc)), making it difficult to assign a specific structure based on mass. If there is no clear evidence of the presence of a specific motif, it can lead to incorrect assignments. Also, glycopeptides have relatively large masses, and when they exceed 5000, the ionization efficiency decreases, the intensity of the monoisotopic signal also decreases, and it is easy to misassign. If this is mistaken and the mass is detected as 1 larger, in the case of glycopeptides, M(NeuAc) + 1 = M(Fuc) × 2, so the MS2 fragments easily match by changing the glycan composition. There are many other factors that can lead to misidentification. Figure 5.1 is a higher-energy collision-induced dissociation (HCD) MS2 spectrum estimated for a hypothetical glycopeptide like the one in the inset. This HCD breaks the glycosidic bonds of the glycans, producing glycan fragments (B ions, also called diagnostic ions) and peptide side ions (Y ions) and also, albeit weakly, peptide part fragments (b and y ions). Software that searches for glycopeptides from MS2 spectra exploits these commonly observed features (signals) in an attempt to somehow identify the original glycopeptide, but it has its pros and cons. The Human Proteome Organization (HUPO) is promoting better software development by comparing the performance of current multiple identification software as part of the activities of the international collaborative research Human Glycoproteomics Initiative (HGI) [1]. The low ionization efficiency of glycopeptides, which was a problem in the previous version of this paper, has been compensated for by improvements in equipment sensitivity, but considering that proteome analysis has become possible at the single cell level and that more than 10,000 proteins can be identified in one analysis, glycoproteomics can still be said to be difficult. The analysis of proteins having a large number of mucin-type glycans is even more difficult.

Glycoproteomics

章节
发表日期：

2026年

查看全文

Hiroyuki Kaji

摘要

Glycans that modify proteins are heterogeneous, and glycoform analysis at the protein level may be possible with purified materials, but it is generally still quite difficult. Simultaneous analysis of complex mixture of such glycoforms is thought to be impossible. Therefore, glycoproteomics is limited to digesting protein mixtures, enriching glycopeptides, performing LC/MS analysis, obtaining glycoforms at the peptide level, and inferring the overall picture in a bottom-up manner. Moreover, glycans are often ambiguous in their composition information (e.g., Hex5HexNAc4Fuc1NeuAc2, etc.), and many details of their structure are unknown. However, even this level of analysis is not easy. The reasons are multiple, including the diversity and heterogeneity of glycans and the intrinsic structural characteristics that induce misidentification. For example, the building blocks that make up a glycopeptide include elements that differ in mass by one oxygen, such as Ala-Ser, Phe-Tyr, Met-its oxide, Hex-dHex (Fuc), and NeuAc-NeuGc. Even if these blocks are replaced, the mass is exactly the same (e.g., M(Met+Hex) = M(Met(ox) + Fuc)), making it difficult to assign a specific structure based on mass. If there is no clear evidence of the presence of a specific motif, it can lead to incorrect assignments. Also, glycopeptides have relatively large masses, and when they exceed 5000, the ionization efficiency decreases, the intensity of the monoisotopic signal also decreases, and it is easy to misassign. If this is mistaken and the mass is detected as 1 larger, in the case of glycopeptides, M(NeuAc) + 1 = M(Fuc) × 2, so the MS2 fragments easily match by changing the glycan composition. There are many other factors that can lead to misidentification. Figure 5.1 is a higher-energy collision-induced dissociation (HCD) MS2 spectrum estimated for a hypothetical glycopeptide like the one in the inset. This HCD breaks the glycosidic bonds of the glycans, producing glycan fragments (B ions, also called diagnostic ions) and peptide side ions (Y ions) and also, albeit weakly, peptide part fragments (b and y ions). Software that searches for glycopeptides from MS2 spectra exploits these commonly observed features (signals) in an attempt to somehow identify the original glycopeptide, but it has its pros and cons. The Human Proteome Organization (HUPO) is promoting better software development by comparing the performance of current multiple identification software as part of the activities of the international collaborative research Human Glycoproteomics Initiative (HGI) [1]. The low ionization efficiency of glycopeptides, which was a problem in the previous version of this paper, has been compensated for by improvements in equipment sensitivity, but considering that proteome analysis has become possible at the single cell level and that more than 10,000 proteins can be identified in one analysis, glycoproteomics can still be said to be difficult. The analysis of proteins having a large number of mucin-type glycans is even more difficult.