Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2023 May 26;95(23):8789–8797. doi: 10.1021/acs.analchem.2c05599

Identification of the High Mannose N-Glycan Isomers Undescribed by Conventional Multicellular Eukaryotic Biosynthetic Pathways

Chia Yen Liew †,‡,§, Hong-Sheng Luo †,∥, Ting-Yi Yang †,∥, An-Ti Hung †,⊥, Bryan John Abel Magoling †,#,∇, Charles Pin-Kuang Lai †,∇,%, Chi-Kung Ni †,§,⊥,*
PMCID: PMC10267891  PMID: 37235553

Abstract

graphic file with name ac2c05599_0006.jpg

N-linked glycosylation is one of the most important post-translational modifications of proteins. Current knowledge of multicellular eukaryote N-glycan biosynthesis suggests high mannose N-glycans are generated in the endoplasmic reticulum and Golgi apparatus through conserved biosynthetic pathways. According to conventional biosynthetic pathways, four Man7GlcNAc2 isomers, three Man6GlcNAc2 isomers, and one Man5GlcNAc2 isomer are generated during this process. In this study, we applied our latest mass spectrometry method, logically derived sequence tandem mass spectrometry (LODES/MSn), to re-examine high mannose N-glycans extracted from various multicellular eukaryotes which are not glycosylation mutants. LODES/MSn identified many high mannose N-glycan isomers previously unreported in plantae, animalia, cancer cells, and fungi. A database consisting of retention time and CID MSn mass spectra was constructed for all possible MannGlcNAc2 (n = 5, 6, 7) isomers that include the isomers by removing arbitrary numbers and positions of mannose from canonical N-glycan, Man9GlcNAc2. Many N-glycans in this database are not found in current N-glycan mass spectrum libraries. The database is useful for rapid high mannose N-glycan isomeric identification.

Introduction

The asparagine linked (N-linked) glycosylation is one of the most important post-translational modifications of proteins in eukaryotes.1,2 The biosynthetic process of N-glycans in multicellular eukaryotes has been studied in detail for eukaryotes3−9 and can be divided into three distinct stages. The first stage is the preassembly of lipid dolichol-phosphate linked oligosaccharide, Glc3Man9GlcNAc2, followed by the transfer of the oligosaccharide moiety to proteins. The second stage involves trimming of Glc3Man9GlcNAc2 to Man5GlcNAc2 by various glycosidases in the endoplasmic reticulum (ER) and Golgi apparatus.3−9 In the final stage, Man5GlcNAc2 is converted to hybrid and complex N-glycans.

The conventional second stage biosynthetic pathways of multicellular eukaryotes are illustrated in Figure 1. It begins with removal of the terminal Glcα1–2 and Glcα1–3 residues by two glucosidases to generate Man9GlcNAc2, followed by the generation of isomers of Man8GlcNAc2, 8E1 and 8E2, by α-1,2-mannosidases in ER. Here 8E1 and 8E2 are the labels of N-glycans. The definition of N-glycan labels is shown in Figure 1. The other enzyme, endo-α-mannosidase, acts on Glc1Man9GlcNAc2 to form the third Man8GlcNAc2 isomer, 8G1. After the generation of Man8GlcNAc2, various Golgi-localized α-1,2-mannosidases remove all the mannoses connected by an α-1 → 2 glycosidic bond and convert Man8GlcNAc2 to a single isomer of Man5GlcNAc2 (5E1). According to these biosynthetic pathways, there are three possible isomers of Man8GlcNAc2 (8E1, 8E2, 8G1), four possible isomers of Man7GlcNAc2 (7D3, 7E1, 7E2, 7G1), three possible isomers of Man6GlcNAc2 (6D3, 6F1, 6F2), and only one Man5GlcNAc2 isomer (5E1).

Figure 1.

Figure 1

Open in a new tab

Conventional second stage biosynthetic pathways of N-glycans in multicellular eukaryotes. N-glycans not labeled by red stars are predicted by the current biosynthetic pathways. N-glycans labeled by red stars are not predicted by the current biosynthetic pathways but were found in this study. Black arrows represent the current biosynthetic pathways. Red (α-1,6-mannosidases), orange (α-1,3-mannosidases), and gray (α-1,2-mannosidases) arrows represent the proposed biosynthetic pathways based on the observed N-glycans. Additional enzymatic information is required to confirm the proposed pathways. Symbols of the SNFG system are used. The N-glycan labels are represented by two numbers and a letter (e.g., 8E1, 8E2, and 8G1 for Man8GlcNAc2). The first number represents the number of hexose in N-glycans; the letter in the middle represents the group of the isomers classified by LODES. The second number represents the sequence of the isomer in the group.

An important technique for structure determination of N-glycans is nuclear magnetic resonance (NMR) spectroscopy.10,11 However, NMR requires milligrams of samples and NMR spectroscopy may become very complicated when oligosaccharides exceed a certain size. Mass spectrometry has high sensitivity and is capable of detecting oligosaccharides of low abundance.12−15 However, conventional mass spectrometry cannot distinguish anomeric configuration and stereoisomers; additional techniques are needed for precise structural identification of N-glycans. These additional methods include modification of oligosaccharides (e.g., methylation), enzyme digestion,18,19 comparison to the existing mass spectrum database of oligosaccharides or N-glycan standards,20−26 and the deduction from the currently understood N-glycan biosynthetic pathways.16,17 These additional methods are confined by the current knowledge of N-glycan biosynthesis, the availability of enzymes, and an incomplete database of oligosaccharides and N-glycan standards. For example, the current multicellular eukaryote biosynthetic pathway suggests there is only one isomer of Man5GlcNAc2 (5E1). Therefore, if an ion of Man5GlcNAc2 is found in the mass spectra of multicellular eukaryote samples, it is typically assumed to be the isomer 5E1 without further structural identification.16,17

We recently developed a mass spectrometry method, namely, logically derived sequence (LODES) tandem mass spectrometry (MSn), for oligosaccharide structural determination.27−34 LODES/MSn involves sequential low energy collision induced dissociation (CID) of oligosaccharides in multistage tandem mass spectrometry. LODES/MSn does not require the mass spectrum database of oligosaccharides or N-glycan standards, and it can determine the linkage positions, anomericities, and stereoisomers of monosaccharides in oligosaccharides. Thus, LODES/MSn is particularly useful for the structural determination of the N-glycan not discovered before. In this study, we applied LODES/MSn to determine the structures of high mannose N-glycans of various species which are not glycosylation mutants. Many N-glycan isomers newly found in plantae, animalia, cancer cells, and fungi cannot be described by the conventional biosynthetic pathways.35 Using the N-glycans we found, we constructed a high mannose N-glycan database consisting of the retention time and CID mass spectra of all possible MannGlcNAc2 (n = 5, 6, 7) isomers. Many N-glycans in this database are not found in current N-glycan libraries. The database is very useful for rapid isomeric identification of high mannose N-glycans, in particularly useful for the structural identification of the high mannose N-glycans beyond the conventional biosynthetic pathways.

Experimental Section

Sources of materials and detailed experimental methods are described in the Supporting Information. In brief, beef, pork, and mushroom (Agaricus bisporus) were lyophilized and ground into powder separately; black bean (Phaseolus vulgaris) was ground into powder directly; red bean (Vigna angulariz) was ground into powder directly or roasted in an oven before grinding into powder; bovine lactoferrin powder was used directly; cell membrane proteins were extracted using the differential ultracentrifugation protocol.36 The N-glycans were released from proteins using the ammonia-catalyzed reaction37 or PNGase F digestion, followed by the N-glycan purification using ethanol precipitation, solid phase extraction, and size exclusion chromatography. The N-glycans were first separated into several fractions according to their sizes by high-performance liquid chromatography (HPLC) with a TSKgel Amide-80 column. The fractions collected from HPLC were then sent into another HPLC with a PGC Hypercarb column separately for isomer separation. For the N-glycans generated by enzyme digestion of large N-glycans, they ware purified using preconditioned ZipTip C4 and the corresponding elution was sent into the HPLC with a PGC column for isomer separation. The HPLC conditions for the TSKgel Amide-80 column (150 mm × 2.0 mm, particle size of 5 μm) were as follows: the flow rate was 0.2 mL/min, the gradient was changed linearly from A (H2O) = 35% and B (acetonitrile) = 65% at t = 0 to A = 45% and B = 55% at t = 50 min. The HPLC conditions for the PGC Hypercarb column (2.1 mm × 100 mm, particle size of 3 μm) were as follows: the flow rate was 0.2 mL/min; the gradient was changed linearly from A = 100% and B = 0% at t = 0 to A = 82% and B = 18% at t = 30 min.

In the construction of the N-glycan database, a PGC column was coupled to a linear ion trap mass spectrometer, and the eluents from the PGC column were sent into a mass spectrometer directly to record the retention time and MS2 and MS3 CID spectra. For the N-glycan structure determination, isomers separated by the PGC column were collected separately. They were sent into a nanoelectrospray ionization instrument coupled to a linear ion trap mass spectrometer for the multistage collision-induced dissociation (CID) tandem mass spectrometry (MSn) measurement and structure determination. Details of N-glycan structural identification were described in our previous study32 and the Supporting Information.

Results

N-Glycan Structural Determination

Here we used Man5GlcNAc2N-glycans extracted from black beans as examples to demonstrate the processes of structural determination using two orthogonal methods, LODES/MSn and enzyme digestion. The N-glycans released from glycoproteins were purified by solid phase extraction and size exclusion chromatography. Next, the N-glycans were separated according to their sizes by HPLC using a TSKgel Amide-80 column; the chromatogram is shown in Figure 2(a). The separation of N-glycans by different sizes avoids the interference in structural determination of small N-glycans by large N-glycans due to the ESI in-source decay.38 For example, Figure 2(a) shows that the retention time of 19–23 min mainly consists of Man6GlcNAc2 (m/z 1419) and a small amount of Man5GlcNAc2 (m/z 1257). The Man5GlcNAc2 in this retention time likely was produced by the ESI in-source decay of Man6GlcNAc2; thus, this part of Man5GlcNAc2 was not analyzed in the structural determination of Man5GlcNAc2. The Man5GlcNAc2 collected from retention time t = 15–18 min was sent into the second HPLC using a PGC column for isomer separation. The chromatogram is illustrated in Figure 2(b). Notably, the N-glycans are intact N-glycans (i.e., no reduction); thus, there are two anomers (α- and β-anomers of the GlcNAc at the reducing end) for each N-glycan isomer. The PGC Hypercarb column is known for the separation of isomers as well as anomers. Each isomer may appear as two peaks in the chromatogram if two anomers of the same isomer are separated. There are eight peaks in Figure 2(b), indicating at least four isomers in this chromatogram. To identify which two peaks belong to the same isomer, the eluents from the PGC column were collected every 30 s such that different peaks were collected in different tubes. These fractions were stored at room temperature for 6 h before they were concentrated and reinjected into the same PGC column individually. If two peaks in Figure 2(b) belong to the same isomer, the reinjection of the eluents into the same PGC column would show two peaks again in the chromatogram, although only one peak was collected initially for each fraction. The relative intensities and the retention times of these two peaks must remain the same as those in Figure 2(b). This is because the α and β anomers of the same isomer change to each other through mutarotation, which typically takes about 0.5–2 h in aqueous solution at room temperature to reach equilibrium. The chromatograms of the reinjection, illustrated in Figure 2(c), show four pairs of peaks, highlighted by blue, green, red, and orange bars, representing four isomers.

Figure 2.

Figure 2

Open in a new tab

(a) Chromatograms of the N-glycans extracted from black beans, separated by HPLC using an amide column. The m/z values represent the sodium ion adducts of MannGlcNAc2 (n = 5–8). (b) Chromatograms of the Man5GlcNAc2 eluents collected from t = 15 to 18 min in chromatogram (a), separated by HPLC using a PGC column. (c) Chromatograms of the eluents collected from chromatogram (b) every 30 s, separated by HPLC using a PGC column. (d) CID sequences for differentiation of Man5GlcNAc2 and GlcMan4GlcNAc2 isomers. Half circles and three-quarter circles represent cross-ring dissociation and dehydration, respectively. (e–g) CID spectra of 5E1. Chromatogram of 5E1 before (h) and 3 min after (i) adding enzyme Canavalia ensiformis. Chromatogram of Man4GlcNAc2 (j) 3 min after adding enzyme Canavalia ensiformis to 5E1, synthesized 4D2 (k), and 4E1 (l). All Y axes in the mass spectra represent selected ion monitoring (SIM), and the width of the mass selected window is 1 Th.

The N-glycan structural determination by LODES/MSn involves multistage collision induced dissociation in tandem mass spectrometry. In each stage of CID, a fragment ion was selected as the precursor ion for the next stage of CID. The selection of fragment ions was determined by carbohydrate dissociation mechanisms.39−42 These CID sequences are illustrated in Figure 2(d) for Hex5GlcNAc2 isomers. The isomers in Figure 2(d) include the removal of arbitrary numbers and positions of glucose and mannose from canonical Glc3Man9GlcNAc2, the N-glycan before any removal of glucose and mannose by enzyme in ER and Golgi. These isomers include the isomers predicted and not predicted by conventional biosynthesis and thus provide the opportunity to discover new N-glycans. In Figure 2(d), the CID sequence, 1257 → 851 →, classifies isomers into D, E, F, and G groups, according to fragments m/z 761, (791, 761, 731), 599, and 437, respectively; 1257 → 833 → 689 → 527 → differentiates isomers 5F1 and 5F2; 1257 → 833 → 527 → differentiates isomers in group E; 1257 → 833 → 365 → differentiates isomers 5D1 and 5D2.

Here we used the isomer in tube 50 to demonstrate how LODES/MSn determines the N-glycan structures. The CID spectra of the isomer in tube 50 are illustrated in Figure 2(e–g). The MS2 CID spectrum in Figure 2(e) represents the dissociation of precursor ion, Hex5GlcNAc2 sodium ion adduct (m/z 1257). Fragments m/z 1239 (dehydration product) and m/z 1156 (cross-ring dissociation product) suggest a HexNAc is located at the reducing end and it is connected to the other sugars by 1 → 4 or 1 → 6 linkage, according to the HexNAc dissociation mechanism.42 Fragment m/z 833 represents the elimination of two HexNAc, indicating five hexoses of Hex5GlcNAc2 are connected without HexNAc between them. The MS3 spectrum [Figure 2(f)] through the sequence 1257 → 851 → classifies isomers into four groups, D, E, F, and G [Figure 2(d)]. Fragment m/z 599 in Figure 2(f) suggests the isomer in tube 50 belongs to group E, according to Figure 2(d). The MS4 spectrum [Figure 2(g)], through the sequence 1257 → 833 → 527 →, differentiates the isomers in group E. Fragment m/z 275 in Figure 2(g) indicates the isomer is 5E1, according to Figure 2(d). Detailed structural determination procedures using LODES/MSn are described in the Supporting Information.

The structure of the isomer in tube 50 was crosschecked using enzyme digestion. The chromatogram of ion m/z 1257 (Hex5GlcNAc2 sodium ion adduct) before and 3 min after enzyme (α-mannosidase of Canavalia ensiformis) digestion are illustrated in Figure 2(h) and (i), respectively. The decrease in intensity indicates the degradation of Hex5GlcNAc2 by enzyme digestion. The chromatogram of ion m/z 1095 (Hex4GlcNAc2 sodium ion adduct) produced from the degradation of Hex5GlcNAc2 by enzyme is illustrated in Figure 2(j). The retention time of the four peaks in Figure 2(j) agrees with that of synthesized Man4GlcNAc2 isomers 4D2 and 4E1 (structures of synthesized 4D2 and 4E1 crosschecked using LODES/MSn are shown in the Supporting Information). The CID MS2, MS3, and MS4 spectra of the Hex4GlcNAc2 produced from the isomer in tube 50 by enzyme digestion are also the same as those of the synthesized Man4GlcNAc2 isomers 4D2 and 4E1 (Supporting Information). The structural determination of the other isomers is illustrated in the Supporting Information.

Databases of MannGlcNAc2 (n = 5, 6, 7) Isomers

To facilitate the rapid N-glycan structural determination of various biological samples, we first constructed the N-glycan database which includes high performance liquid chromatography (HPLC) retention time and multistage collision-induced dissociation (MSn CID) mass spectra. The sources of these N-glycans include (1) commercial products, (2) isolation and purification from biological samples, and (3) generation from large N-glycans by enzymic degradation. Various N-glycan databases have been reported previously.43−47 The differences between our new database and the existing databases include the following: (1) Our database consists of all possible high mannose N-glycan isomers by removing arbitrary numbers and positions of mannoses from Man9GlcNAc2 (9D1). (2) We used intact N-glycans for structural analysis. No derivatization (permethylation, reduction, or labeling at the reducing end) is required. Therefore, the potential interference by the products generated by the side reactions during derivatization is avoided.

In the construction of the N-glycan database, the structures of commercial N-glycans which have been determined by manufacturers were crosschecked using LODES/MSn. The N-glycans isolated from biological samples were separated by HPLC chromatography and structurally determined by LODES/MSn. The structures of parts of N-glycans isolated from biological samples were crosschecked using enzyme digestion. Details are presented in the Supporting Information.

In our database, some isomers were generated by the enzyme degradation of large synthesized N-glycans. There may be more than one isomer produced by enzyme degradation. Fortunately, we had the retention time and CID spectra of all the potential isomers produced by enzyme degradation except the one we wanted to generate by enzymatic degradation; therefore, we were able to assign these isomers unambiguously. For example, both isomers 5F1 and 5F2 could be generated by the enzyme degradation of Man6GlcNAc2 isomers 6G1. Because we have the retention time of isomer 5F1, we assigned the enzyme generated isomer for which the retention time is different from that of isomer 5F1 to isomer 5F2.

After structural determination of all Man5GlcNAc2 isomers, a database consisting of the chromatograms and CID spectra of these isomers was constructed (Figure 3). The chromatograms and CID spectra of Man7GlcNAc2 and Man6GlcNAc2 in the database are illustrated in the Supporting Information.

Figure 3.

Figure 3

Open in a new tab

Database of Man5GlcNAc2 isomers. (a) HPLC chromatograms of ion m/z 1257 (sodium ion adduct), (b) CID spectra through sequence 1257 →, (c) CID spectra through sequence 1257 → 851 → fragments, (d) CID spectra through sequence 1257 → 833 → , (e) CID spectra through sequence 1257 → 833 → 527 →. Isomers 5E1 and 5E2 have close retention time and similar MS2 and MS3 spectra. MS4 spectra are useful for differentiation of these two isomers when the HPLC separation is not very good. These N-glycans were in intact form (no reduction), and there are two peaks in the chromatogram for each isomer. All Y axes represent the total ion count (TIC) of the fragment ions produced from precursor ion in arbitrary units.

High Mannose N-Glycans of Various Biological Samples

The structural assignments of N-glycans extracted from biological samples were made by comparing to the database on the similarities of the following three properties: (1) HPLC retention time in the chromatogram of the selected m/z value, (2) one MS2 and two MS3 CID mass spectra at the corresponding retention time, (3) the relative intensity of two peaks (resulting from α and β anomers of the GlcNAc at the reducing end) in the chromatogram of the selected isomer.

Figure 4(a) and (b) shows the HPLC chromatograms and structural assignments of Man7GlcNAc2 isomers extracted from bovine lactoferrin. The isomers with large abundances belong to the N-glycan isomers, 7D3, 7E1, 7E2, and 7G1, predicted by conventional biosynthesis. Isomer 7D2, one of the N-glycans not predicted by conventional biosynthetic pathways (denoted as unusual N-glycans), was found, although the abundance is low. Here we used two methods, enzyme PNGase F [Figure 4(a)] and the ammonia-catalyzed reaction [Figure 4(b)], to release N-glycans from bovine lactoferrin. The unusual N-glycan isomer 7D2 was found using both methods, although the relative abundances of isomers were different between these two methods.

Figure 4.

Figure 4

Open in a new tab

Chromatograms and structural assignments of (a, b) Man7GlcNAc2 (TIC of m/z 1581, sodium ion adduct) and (c–l) Man6GlcNAc2 (TIC of m/z 1419, sodium ion adduct) N-glycan isomers extracted from various samples using enzyme PNGase F or the ammonia-catalyzed reaction to release N-glycans from proteins. N-glycans labeled by red stars represent the N-glycans not predicted by conventional biosynthesis but found in this study. Peaks labeled by black stars represent impurities. All Y axes represent the total ion count (TIC) of the fragment ions produced from precursor ion in arbitrary units.

Figure 4(c–l) shows the HPLC chromatograms of Man6GlcNAc2 isomers extracted from bovine lactoferrin, red bean, beef, pork, and mushroom. Both N-glycan release methods were used for bovine lactoferrin, red bean, and beef. The relative abundances of isomers are slightly different between these two methods. Only the ammonia-catalyzed reaction was used in pork and mushroom. The unusual Man6GlcNAc2 isomers were found in all the samples we studied. None of these unusual N-glycans have the largest abundance, but they are not negligible.

The HPLC chromatograms of Man5GlcNAc2 extracted from lactoferrin, beef, red bean, human cell line M10, breast cancer cell line MDA-MB-231, pork, mushroom, and the structural assignments are illustrated in Figure 5. The N-glycans not predicted by conventional biosynthetic pathways were found in all the samples we studied. Isomer 5E1 which is the only isomer predicted by conventional biosynthesis has the largest abundance in most samples, but the abundances of unusual N-glycans found in beef and pork are not small. Interestingly, the abundances of unusual N-glycans found in red bean are as large as those of isomer 5E1. The high mannose N-glycans observed in various samples are summarized in Supporting Information Table S1.

Figure 5.

Figure 5

Open in a new tab

Chromatograms (TIC of m/z 1257, sodium ion adduct) and the structural assignments of Man5GlcNAc2N-glycan isomers extracted from various biological samples.

The conventional biosynthesis suggests all the mannoses connecting by a Manα-1 → 2 glycosidic bond were removed from Man9GlcNAc2 by α-1,2-mannosidases before the cleavage of the Manα-1 → 3 or Manα-1 → 6 glycosidic bond by other enzymes. These unusual N-glycans indicate that there are enzymes efficiently removing Manα-1 → 3 and Manα-1 → 6 from MannGlcNAc2 (n = 8, 7, 6) before the complete removal of Manα-1 → 2. One possibility is that the enzymes are not located in the ER and Golgi, and these unusual N-glycans are produced during cell rupture in the N-glycan extraction process such that glycoproteins in the ER or Golgi encounter the enzymes not located in the ER and Golgi. To exclude this possibility, we compared the red bean powder samples prepared by two different methods. In one method, red beans were grounded into powder by a grinder directly. In the other method, red beans were roasted in an oven at 90 °C for 20 min and then immersed in liquid nitrogen for 20 min, followed by reiteration of roasting in an oven and immersing in liquid nitrogen twice before grounding into powder. The process inactivates the potential enzymes that could contact glycoproteins during cell rupture. Both enzyme PNGase F and the ammonia-catalyzed reaction were used to release N-glycans from red bean powder prepared in both methods. In the ammonia-catalyzed reaction, red bean powder was added into 25% ammonia solution. The high pH value (>11) of ammonia solution ensures that the potential enzymes which are active at low pH value, such as the enzymes in lysosome, are not involved in the generation of the unusual N-glycans. Figure 5(e–h) shows the chromatograms of the N-glycans released by PNGase F and the ammonia-catalyzed reaction from red beans and roast red beans, respectively. The relative abundances of isomers between two sample preparation methods and two N-glycans release methods are similar. The results suggest the unusual isomers found in this study were not produced by the artifact during the sample preparation.

Discussion

The high mannose N-glycans beyond the conventional multicellular eukaryote biosynthetic pathways have been reported in previous studies.48−54 Most of these unusual N-glycans were found in glycosylation mutants or in low abundance or the abundances of these unusual N-glycans were not reported. Consequently, Man5GlcNAc2 and Man6GlcNAc2 found in the mass spectra of multicellular eukaryote samples are usually assumed to be isomer 5E1 and isomers 6D3, 6F1, and 6F2, respectively, and no further structural identification was made in most studies.16,17

In the first stage of biosynthesis, Man5GlcNAc2 isomer 5E4, Man6GlcNAc2 isomer 6E1, and Man7GlcNAc2 isomer 7D2 are generated and connected to the lipid dolichol-phosphate located on the cytoplasmic face of the ER membrane. The release of these oligosaccharides from dolichol typically requires acidic solution.43 The observation of isomers 7D2, 6E1, and 5E4 is not likely due to the release from dolichol during the sample preparation. This is because (1) the pH values of ammonia solution and PNGase F buffer solution are 11 and 7.5, respectively, and (2) isomers 6E1 and 5E4 are not the only unusual N-glycan isomers of Man6GlcNAc2 and Man5GlcNAc2 we found in various samples.

The unusual high mannose N-glycans found in this study in various biological samples which are not glycosylation mutants suggest additional biosynthetic pathways are involved in the generation of high mannose N-glycans, and these additional pathways in some biological systems, such as beef, pork, and red beans, are quite efficient. One possible biosynthetic pathway is the direct transfer of Man7GlcNAc2 7D2 from dolichol-phosphate to proteins. It has been shown that unglucosylated oligosaccharides, Man7GlcNAc2 7D2, are transferred from lipid to protein in congenital disorders of glycosylation type Ig55 and F9 tetratocarcinoma cells.56 The sequential removal of terminal Man-α-(1 → 2) from 7D2 N-glycan generates parts of the unusual N-glycans (6E1, 6D2, 5E4, 5D2, and 5E3). Another possible biosynthetic pathway is enzymes that can act on the terminal Man-α-(1 → 3) and Man-α-(1 → 6) before the complete removal of terminal Man-α-(1 → 2) from N-glycans. Enzymes related to the degradation of high mannose N-glycans in the ER, cytoplasm, and lysosomes are potential candidates, although these endoplasmic reticulum-associated degradation pathways typically start from trimming one GlcNAc at the reducing end which results in the N-glycans being structurally different from what we reported here. However, it was suggested that part of the enzymes related to degrading high-mannose N-glycans, such as ER degradation-enhancing α-mannosidase-like proteins and lysosomal α-mannosidase secreted via a trans-Golgi network, were responsible for the generation of MannGlcNAc2 (n = 6–8) in the HEK293 cells which disrupted all α-1,2-mannosidase genes.57 Another possible enzyme is the α-mannosidase IIx which is known as a trimming enzyme that digests the Man-α-(1 → 3) or Man-α-(1 → 6) linkage and produces Man4GlcNA2, Man-α-(1 → 2)-Man-α-(1 → 3)-[Man-α-(1 → 6)]-Man-β-(1 → 4)-GlcNA2, from Man6GlcNAc2 isomer 6D3.58 However, whether α-mannosidase IIx acts on Man7GlcNAc2 or Man8GlcNAc2 is not clear.

Conclusions

A new mass spectrometry method, logically derived sequence tandem mass spectrometry, enables determination of the structures of N-glycans without using standard compounds. Many high mannose N-glycans not described by the conventional multicellular eukaryotic biosynthetic pathways were found in many biological samples that are not glycosylation mutants, including plantae, animalia, cancer cells, and fungi. Our current findings warrant future investigation to identify enzymes and pathways responsible for their synthesis in multicellular eukaryotic cells. Before these enzymes and pathways are identified, one should be very careful in the high mannose N-glycan structural identification by using the current biosynthetic pathways.

A N-glycan database which consists of HPLC retention time and CID mass spectra is constructed. It includes all possible isomers of MannGlcNAc2 (n = 5, 6, 7) by removing arbitrary positions and numbers of mannose from the canonical Man9GlcNAc2, an oligosaccharide transferred from dolichol-phosphate to proteins before removing any mannose by enzymes in the ER. This database is useful for rapid isomeric high mannose N-glycan identification.

Acknowledgments

This work was financially supported by iMATE project (AS-iMATE-111-32) and Grand Challenge Seed Program (AS-GC-109-06) of Academia Sinica, Taiwan.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.analchem.2c05599.

  • Details of sample extraction, structural determination using LODES/MSn, and electronic version of CID spectra (PDF)

The authors declare the following competing financial interest(s): C.Y.L. and C.-K.N. are co-inventors of a United States patent (US 10,796,788 B2) that part of the method described in the patent to determine the carbohydrate structure was used in this work. All other authors declare no competing interests.

Supplementary Material

ac2c05599_si_001.pdf (4.7MB, pdf)

References

  1. Shental-Bechor D.; Levy Y. Effect of glycosylation on protein folding: a close look at thermodynamic stabilization. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 8256–8261. 10.1073/pnas.0801340105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Live D. H.; Kumar R. A.; Beebe X.; Danishefsky S. J. Conformational influences of glycosylation of a peptide: a possible model for the effect of glycosylation on the rate of protein folding. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 12759–12761. 10.1073/pnas.93.23.12759. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Schachter H.Processing of protein-bound N-glycans. In Comprehensive glycoscience from chemistry to systems biology; Kamerling J. P., Ed.; Elsevier Science Ltd: Oxford, U.K., 2007; Vol. 3, pp 11–25. [Google Scholar]
  4. Stanley P.; Taniguchi N.; Aebi M.. N-glycans. In Essentials of Glycobiology, 3rd ed.; Varki A.; et al., Eds.; Cold Spring Harbor Laboratory Press: New York, 2017; pp 99–112. 10.1101/glycobiology.3e.009. [DOI] [Google Scholar]
  5. Parodi A.; Cummings R. D.; Aebi M.. Glycans in glycoprotein quality control. In Essentials of Glycobiology, 3rd ed.; Varki A.; et al., Eds.; Cold Spring Harbor Laboratory Press: New York, 2017; pp 503–512. 10.1101/glycobiology.3e.039. [DOI] [Google Scholar]
  6. Kornfeld R.; Kornfeld S. Assembly of asparagine-linked oligosaccharides. Annu. Rev. Biochem. 1985, 54, 631–664. 10.1146/annurev.bi.54.070185.003215. [DOI] [PubMed] [Google Scholar]
  7. Schenk B.; Fernandez F.; Waechter C. J. The ins(ide) and out(side) of dolichyl phosphate biosynthesis and recycling in the endoplasmic reticulum. Glycobiology 2001, 11, 61R–70R. 10.1093/glycob/11.5.61R. [DOI] [PubMed] [Google Scholar]
  8. Aebi M. N-linked protein glycosylation in the ER. Biochim. Biophys. Acta 2013, 1833, 2430–2437. 10.1016/j.bbamcr.2013.04.001. [DOI] [PubMed] [Google Scholar]
  9. Shrimal S.; Cherepanova N. A.; Gilmore R. Cotranslational and posttranslocational N-glycosylation of proteins in the endoplasmic reticulum. Semin Cell Dev Biol. 2015, 41, 71–78. 10.1016/j.semcdb.2014.11.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Lundborg M.; Widmalm G. Structural analysis of glycans by NMR chemical shift prediction. Anal. Chem. 2011, 83, 1514–1517. 10.1021/ac1032534. [DOI] [PubMed] [Google Scholar]
  11. Wang N.; Seko A.; Takeda Y.; Ito Y. Preparation of asparagine-linked monoglucosylated high-mannosetype oligosaccharide from egg yolk. Carbohydr. Res. 2015, 411, 37–41. 10.1016/j.carres.2015.04.011. [DOI] [PubMed] [Google Scholar]
  12. Dell A.; Morris H. R. Glycoprotein structure determination by mass spectrometry. Science 2001, 291, 2351–2356. 10.1126/science.1058890. [DOI] [PubMed] [Google Scholar]
  13. Ashline D.; Singh S.; Hanneman A.; Reinhold V. Congruent strategies for carbohydrate sequencing. 1. Mining structural details by MSn. Anal. Chem. 2005, 77, 6250–6262. 10.1021/ac050724z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Kailemia M. J.; Ruhaak L. R.; Lebrilla C. B.; Amster I. J. Oligosaccharide analysis by mass spectrometry: a review of recent developments. Anal. Chem. 2014, 86, 196–212. 10.1021/ac403969n. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Zaia J. Mass Spectrometry and Glycomics. OMICS 2010, 14 (4), 401–418. 10.1089/omi.2009.0146. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Nwosu C. C.; Aldredge D. L.; Lee H.; Lerno L. A.; Zivkovic A. M.; German J. B.; Lebriaal C. B. Comparison of the human and bovine milk N-Glycome via high-performance microfluidic chip liquid chromatography and tandem mass spectrometry. J. Proteome Res. 2012, 11, 2912–2924. 10.1021/pr300008u. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Dong X.; Zhou S.; Mechref Y. LC-MS/MS analysis of permethylated free oligosaccharides and N-glycans derived from human, bovine, and goat milk samples. Electrophoresis 2016, 37, 1532–1548. 10.1002/elps.201500561. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Houel S.; Hilliard M.; Yu Y. Q.; McLouglin N.; Martin S. M.; Rudd P. M.; Williams J. P.; Chen W. N- and O-glycosylation analysis of etanercept using liquid chromatography and quadrupole time-of-flight mass spectrometry equipped with electron-transfer dissociation functionality. Anal. Chem. 2014, 86 (1), 576–584. 10.1021/ac402726h. [DOI] [PubMed] [Google Scholar]
  19. Song T.; Aldredge D.; Lebrilla C. B. A method for in-depth structural annotation of human serum glycans that yields biological variations. Anal. Chem. 2015, 87, 7754–7762. 10.1021/acs.analchem.5b01340. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Harvey D. J.; Abrahams J. L. Fragmentation and ion mobility properties of negative ions from N-linked carbohydrates: part 7. Reduced glycans. Rapid Commun. Mass Spectrom. 2016, 30, 627–634. 10.1002/rcm.7467. [DOI] [PubMed] [Google Scholar]
  21. Ashwood C.; Lin C. H.; Thaysen-Andersen M.; Packer N. H. Discrimination of isomers of released N- and O-glycans using diagnostic product ions in negative ion PGC-LC-ESI-MS/MS. J. Am. Soc. Mass Spectrom. 2018, 29, 1194–1209. 10.1007/s13361-018-1932-z. [DOI] [PubMed] [Google Scholar]
  22. Morimoto K.; Nishikaze T.; Yoshizawa A. C.; Kajihara S.; Aoshima K.; Oda Y.; Tanaka K. GlycanAnalysis Plug-in: a database search tool for N-glycan structures using mass spectrometry. Bioinformatics 2015, 31, 2217–2219. 10.1093/bioinformatics/btv110. [DOI] [PubMed] [Google Scholar]
  23. Klein J.; Zaia J. Glypy: an open source glycoinformatics library. J. Proteome Res. 2019, 18, 3532–3537. 10.1021/acs.jproteome.9b00367. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Rojas-Macias M. A.; Mariethoz J.; Andersson P.; Jin C.; Venkatakrishnan V.; Aoki N. P.; Shinmachi D.; Ashwood C.; Madunic K.; Zhang T.; Miller R. L. Towards a standardized bioinformatics infrastructure for N-and O-glycomics. Nat. Commun. 2019, 10, 3275. 10.1038/s41467-019-11131-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Prien J. M.; Ashline D. J.; Lapadula A. J.; Zhang H.; Reinhold V. N. The high mannose glycans from bovine ribonuclease B isomer characterization by ion trap MS. J. Am. Soc. Mass Spectrom. 2009, 20, 539–556. 10.1016/j.jasms.2008.11.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Wei J.; Tang Y.; Bai Y.; Zaia J.; Costello C. E.; Hong P.; Lin C. Toward automatic and comprehensive glycan characterization by online PGC-LC-EED MS/MS. Anal. Chem. 2020, 92, 782–791. 10.1021/acs.analchem.9b03183. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Hsu H. C.; Liew C. Y.; Huang S. P.; Tsai S. T.; Ni C. K. Simple approach for De novo structural identification of mannose Trisaccharides. J. Am. Soc. Mass Spectrom. 2018, 29, 470–480. 10.1007/s13361-017-1850-5. [DOI] [PubMed] [Google Scholar]
  28. Hsu H. C.; Liew C. Y.; Huang S. P.; Tsai S. T.; Ni C. K. Simple method for De novo structural determination of underivatised glucose oligosaccharides. Sci. Rep. 2018, 8, 5562. 10.1038/s41598-018-23903-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Tsai S. T.; Liew C. Y.; Hsu C.; Huang S. P.; Weng W. C.; Kuo Y. H.; Ni C. K. Automatically full glycan structural determination with logically derived sequence tandem mass spectrometry. ChemBioChem. 2019, 20, 2351–2359. 10.1002/cbic.201900228. [DOI] [PubMed] [Google Scholar]
  30. Hsu H. C.; Huang S. P.; Liew C. Y.; Tsai S. T.; Ni C. K. De novo structural determination of mannose oligosaccharides by using a logically derived sequence for tandem mass spectrometry. Anal. Bioanal. Chem. 2019, 411, 3241–3255. 10.1007/s00216-019-01817-y. [DOI] [PubMed] [Google Scholar]
  31. Ni C. K.; Hsu H. C.; Liew C. Y.; Huang S. P.; Tsai S. T.. Modern mass spectrometry techniques for oligosaccharide structure determination: logically derived sequence tandem mass spectrometry for automatic oligosaccharide structural determination. In Comprehensive glycoscience from chemistry to systems biology, 2nd ed.; Barchi J., Ed.; Elsevier Science Ltd: Oxford, U.K., 2021; Vol. 1, pp 309–339. 10.1016/B978-0-12-409547-2.14903-0. [DOI] [Google Scholar]
  32. Liew C. Y.; Yen C. C.; Chen J. L.; Tsai S. T.; Pawar S.; Wu C. Y.; Ni C. K. Structural identification of N-glycan isomers using logically derived sequence tandem mass spectrometry. Commun. Chem. 2021, 4, 92. 10.1038/s42004-021-00532-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Liew C. Y.; Chan C. K.; Huang S. P.; Cheng Y. T.; Tsai S. T.; Hsu H. C.; Wang C. C.; Ni C. K. De novo structural determination of oligosaccharide isomers in glycosphingolipids using logically derived sequence tandem mass spectrometry. Analyst 2021, 146, 7345–7357. 10.1039/D1AN01448J. [DOI] [PubMed] [Google Scholar]
  34. Weng W. C.; Liao H. E.; Huang S. P.; Tsai S. T.; Hsu H. C.; Liew C. Y.; Gannedi V.; Hung S. C.; Ni C. K. Unusual free oligosaccharides in human bovine and caprine milk. Sci. Rep 2022, 12, 10790. 10.1038/s41598-022-15140-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Ni C. K.; Liew C. Y.; Luo H. S.; Yang T. Y.; Hung A. T.; Magoling B. J. A.; Lai C. P. K.. Identification of novel high mannose N-glycan isomers undescribed by current multicellular eukaryotic biosynthetic pathways. Chemrxiv 2022/8/3. 10.26434/chemrxiv-2022-rhv80. [DOI] [PMC free article] [PubMed]
  36. Li Q.; Xie Y.; Wong M.; Barboza M.; Lebrilla C. B. Comprehensive structural glycomic characterization of the glycocalyxes of cells and tissues. Nat. Protoc. 2020, 15 (8), 2668–2704. 10.1038/s41596-020-0350-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Yang M.; Wei M.; Wang C.; Lu Y.; Jin W.; Gao X.; Li C.; Wang L.; Huang L.; Wang Z. Separation and preparation of N-glycans based on ammonia-catalyzed release method. Glycoconjugate J. 2020, 37, 165–174. 10.1007/s10719-020-09909-z. [DOI] [PubMed] [Google Scholar]
  38. Liew C. Y.; Chen J. L.; Ni C. K. Electrospray ionization in-source decay of N-glycans and the effects on N-glycan structural identification. Rapid Commun. Mass Spectrom. 2022, 36, e9352. 10.1002/rcm.9352. [DOI] [PubMed] [Google Scholar]
  39. Tsai S. T.; Chen J. L.; Ni C. K. Does low-energy collision-induced dissociation of lithiated and sodiated carbohydrates always occur at anomeric carbon of the reducing end?. Rapid Commun. Mass Spectrom. 2017, 31, 1835–1844. 10.1002/rcm.7961. [DOI] [PubMed] [Google Scholar]
  40. Chen J. L.; Nguan H. S.; Hsu P. J.; Tsai S. T.; Liew C. Y.; Kuo J. L.; Hu W. P.; Ni C. K. Collision-induced dissociation of sodiated glucose and identification of anomeric configuration. Phys. Chem. Chem. Phys. 2017, 19, 15454–15462. 10.1039/C7CP02393F. [DOI] [PubMed] [Google Scholar]
  41. Huynh H. T.; Phan H. T.; Hsu P. J.; Chen J. L.; Nguan H. S.; Tsai S. T.; Roongcharoen T.; Liew C. Y.; Ni C. K.; Kuo J. L. Collision induced dissociation of sodiated glucose, galactose, and mannose, and the identification of anomeric configurations. Phys. Chem. Chem. Phys. 2018, 20, 19614–19624. 10.1039/C8CP03753A. [DOI] [PubMed] [Google Scholar]
  42. Chiu C. C.; Tsai S. T.; Hsu P. J.; Huynh H. T.; Chen J. L.; Phan H. T.; Huang S. P.; Lin H. Y.; Kuo J. L.; Ni C. K. Unexpected dissociation mechanism of Sodiated N-Acetylglucosamine and N-Acetylgalactosamine. J. Phys. Chem. A 2019, 123, 3441–3453. 10.1021/acs.jpca.9b00934. [DOI] [PubMed] [Google Scholar]
  43. Tomiya N.; Lee Y. C.; Yoshida T.; Wada Y.; Awaya J.; Kurono M.; Takahashi N. Calculated two-dimensional sugar map of pyridylaminated oligosaccharides: Elucidation of the jack bean α-mannosidase digestion pathway of Man9GlcNAc2. Anal. Biochem. 1991, 193, 90–100. 10.1016/0003-2697(91)90047-W. [DOI] [PubMed] [Google Scholar]
  44. Pabst M.; Bondili J. S.; Stadlmann J.; Mach L.; Altmann F. Mass+retention time = structure: a strategy for the analysis of N-glycans by carbon LC-ESI-MS and its application to fibrin N-glycans. Anal. Chem. 2007, 79, 5051–5057. 10.1021/ac070363i. [DOI] [PubMed] [Google Scholar]
  45. Aldredge D.; An H. J.; Tang N.; Waddell K.; Lebrilla C. B. Annotation of a serum N-glycan library for rapid identification of structures. J. Proteome Res. 2012, 11 (3), 1958–1968. 10.1021/pr2011439. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Pabst M.; Grass J.; Toegel S.; Liebminger E.; Strasser R.; Altmann F. Isomeric analysis of oligomannosidic N-glycans and their dolichol-linked precursors. Glycobiology 2012, 22 (3), 389–399. 10.1093/glycob/cwr138. [DOI] [PubMed] [Google Scholar]
  47. Abrahams J. L.; Campbell M. P.; Packer N. H. Building a PGC-LC-MS N-glycan retention library and elution mapping resource. Glycoconjugate J. 2018, 35 (1), 15–29. 10.1007/s10719-017-9793-4. [DOI] [PubMed] [Google Scholar]
  48. Hwa K.; Khoo K. Structural analysis of the asparagine-linked glycans from the procyclic Trypanosoma brucei and its glycosylation mutants resistant to Concanavalin A killing. Mol. Biochem. Parasitol. 2000, 111, 173–184. 10.1016/S0166-6851(00)00310-8. [DOI] [PubMed] [Google Scholar]
  49. Ohta M.; Ohnishi T.; Ioannou Y.; Hodgson M.; Matsuura F.; Desnick R. Human alpha-N-acetylgalactosaminidase: site occupancy and structure of N-linked oligosaccharides. Glycobiology 2000, 10, 251–261. 10.1093/glycob/10.3.251. [DOI] [PubMed] [Google Scholar]
  50. Kimura Y.; Miyagi C.; Kimura M.; Nitoda T.; Kawai N.; Sugimoto H. Structural features of N-glycans linked to royal jelly glycoproteins: structures of high-mannose type, hybrid type, and biantennary type glycans. Biosci. Biotechnol. Biochem. 2000, 64, 2109–2120. 10.1271/bbb.64.2109. [DOI] [PubMed] [Google Scholar]
  51. Tseneklidou-Stoeter D.; Gerwig G. J.; Kamerling J. P.; Spindler K. D. Characterization of N-linked carbohydrate chains of the crayfish, Astacus leptodactylus hemocyanin. Biol. Chem. Hoppe. Seyler. 1995, 376, 531–537. 10.1515/bchm3.1995.376.9.531. [DOI] [PubMed] [Google Scholar]
  52. Olafson R.; Thomas J.; Ferguson M.; Dwek R.; Chaudhuri M.; Chang K.; Rademacher T. Structures of the N-linked oligosaccharides of Gp63, the major surface glycoprotein, from Leishmania mexicana amazonensis. J. Biol. Chem. 1990, 265, 12240–12247. 10.1016/S0021-9258(19)38336-X. [DOI] [PubMed] [Google Scholar]
  53. Amatayakul-Chantler S.; Dwek R.; Tennent G.; Pepys M.; Rademacher T. Molecular characterization of Limulus polyphemus C-reactive protein. II. Asparagine-linked oligosaccharides. Eur. J. Biochem. 1993, 214, 99–110. 10.1111/j.1432-1033.1993.tb17901.x. [DOI] [PubMed] [Google Scholar]
  54. Prien J. M.; Prater B. D.; Cockrill S. L. A multi-method approach toward de novo glycan characterization: A Man-5 case study. Glycobiology 2010, 20 (5), 629–647. 10.1093/glycob/cwq012. [DOI] [PubMed] [Google Scholar]
  55. Chantret I.; Dupre T.; Delenda C.; Bucher S.; Dancourt J.; Barnier A.; Charollais A.; Heron D.; Bader-Meunier B.; Danos O.; Seta N.; Durand G.; Oriol R.; Codogno P.; Moore S. E. H. Congenital disorders of glycosylation type Ig is defined by a deficiency in dolichyl-P-mannose: Man7GlcNAc2-PP-dolichyl mannosyltransferase. J. Bio. Chem. 2002, 277 (28), 25815–25822. 10.1074/jbc.M203285200. [DOI] [PubMed] [Google Scholar]
  56. Romero P. A.; Herscovics A. Transfer of nonglucosylated oligosaccharide from lipid to protein in a mammalian cell. J. Bio. Chem. 1986, 261 (34), 15936–15940. 10.1016/S0021-9258(18)66656-6. [DOI] [PubMed] [Google Scholar]
  57. Ren W. W.; Jin Z. C.; Dong W.; Kitajima T.; Gao X. D.; Fujita M. Glycoengineering of HEK293 cells to produce high-mannose type N-glycan structures. J. Biochem. 2019, 166 (3), 245–258. 10.1093/jb/mvz032. [DOI] [PubMed] [Google Scholar]
  58. Fukuda M. N.; Akama T. O. In vivo role of α-mannosidase IIx: ineffective spermatogenesis resulting from targeted disruption of the Man2a2 in the mouse. Biochim. Biophys. Acta 2002, 1573, 382–387. 10.1016/S0304-4165(02)00407-5. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ac2c05599_si_001.pdf (4.7MB, pdf)

Articles from Analytical Chemistry are provided here courtesy of American Chemical Society

RESOURCES