Differential expression of N-linked oligosaccharides in methotrexate-resistant primary central nervous system lymphoma cells
Oligosaccharides of glycoprotein, particularly negatively-charged sialylated N-glycans, on the surface of lymphomas play important roles in cell–cell interactions and bind immunoglobulin-like lectins, causing inflammatory responses and bioregulation. However, their characterizations have largely been unknown in central nervous system (CNS) lymphoma.
Here, we investigated expression patterns of N-linked oligosaccharides of glycoproteins in cells derived from CNS lymphomas and clinical specimens.
We first generated methotrexate (MTX)-resistant cells derived from HKBML and TK as CNS lymphoma, and RAJI as non-CNS lymphoma and determined N-linked oligosaccharide structures in these cells and other non-CNS lymphoma-derived cells including A4/FUK, OYB, and HBL1. Major components of the total oligosaccharides were high-mannose type N-glycans, whose level increased in MTX-resistant HKBML and TK but decreased in MTX-resistant RAJI. We also detected sialylated biantennary galactosylated N-glycans with α1,6-fucosylation, A2G2F, and A2G2FB from HKBML, TK, and RAJI. Sialylated A4G4F was specifically isolated from RAJI. However, the ratios of these sialylated N-glycans slightly decreased against MTX-resistant compared to non-resistant cells. Interestingly, almost all complex-type oligosaccharides were α2,6-sialylated.
This is the first study for the expression profile of N-oligosaccharides on MTX-resistant primary CNS lymphoma-derived cells HKBML and TK, and tumor tissues resected from patients with CNS lymphoma,
These results propose a possibility that the differential expression of high-mannose types and sialylated A2G2F, A2G2FB, and A4G4F on the surface of CNS lymphomas may provide a hint for targets for diagnoses and treatments of the oligosaccharide type-specific lymphomas.
KeywordsPrimary central nervous system lymphoma Diffuse large B-cell lymphoma Methotrexate Sialylated N-linked oligosaccharide High performance liquid chromatography
Oligosaccharides are saccharide polymers that have various functions such as cell–cell interactions and cell recognitions, including immune responses . Glycans are normally formed by presenting sugar chains linked to lipids or amino acid side chains by N- or O-glycosidic bonds . The O-linked oligosaccharide is attached to threonine (Thr) or serine (Ser) , while the N-linked oligosaccharide is generally pentasaccharides attached to asparagine (Asn) by β-linkage to amine nitrogen of side chains . The N-oligosaccharides of glycoproteins show differential patterns of branching formations by N-acetylglucosaminyltransferase activities at N-acetylglucosamine (GlcNAc) residues in types of Pl-2, Pl-4, and B1 + 6 linkage to mannose (Man) residues of the core [5, 6]. Previous studies have also demonstrated the mechanisms of intracellular trafficking of the cell surface of glycoproteins and their subsequent returns to the cell surface [7, 8, 9].
Several studies have clarified that the structures of N-linked oligosaccharide chains on glycoproteins have been involved in tumor cell adhesion to the extracellular matrix (ECM), metastatic potentials, and cell proliferation and differentiation [8, 10, 11, 12, 13, 14, 15]. Metastatic potentials of tumor cells have been shown to correlate with the expression of highly branched tri- and tetra-antennary β-1,6-GlcNAc-bearing N-glycans [16, 17]. The increased β1,6-GlcNAc-bearing N-glycan expression is co-regulated by N-acetyl glucosaminyl transferases V (GnT-V) and the Ets-1 transcription factor, and the branching complex type N-glycans function in glioma invasivity . N-acetyl glycan structures and physicochemical properties regulate cell proliferation and differentiation in leukemia . A semi-automated systematic detection system for analyzing the N-linked oligosaccharides of glycoproteins has been developed [18, 19]. N-linked oligosaccharides are relatively easily detectable from a small amount of acetone-precipitated sample (e.g., 1–2 mg) . The oligosaccharides on the cell surface of gliomas are well-examined in T cell immune responses and sensitivity to killer lymphocytes [21, 22, 23]. Several studies clarify a strong correlation between the lectin-binding and the biological function in diffuse large B-cell lymphoma (DLBCL) [24, 25, 26, 27, 28]. However, the oligosaccharides on the primary central nervous system (CNS) lymphoma (PCNSL) surface have largely been unknown.
PCNSL is a rare subtype of DLBCL, which is an aggressive variant of extra-nodal non-Hodgkin’s lymphoma (NHL) . PCNSL only accounts for 3% of primary CNS tumors and 1% of NHLs in adults . Methotrexate (MTX) is an antifolate that inhibits the dihydrofolate reductase activity in purine and thymidine syntheses and regulates the expression of glucocorticoid receptor α and β in human blood cells in vitro [31, 32]. High-dose methotrexate (HD-MTX) is used as a first-line treatment in PCNSL . Moreover, second-line treatments are also required for 10–35% of patients with refractory diseases and for another 35–60% or more who have relapse-acquired resistances . Eventually, although treatment with HD-MTX is used as a standard treatment in PCNSL, most of the cases come to relapse-acquired resistances to MTX .
Here, we generated MTX-resistant lymphoma cell lines derived from PCNSL and non-CNS lymphoma, which were applied to the semi-automated detection system for the N-linked oligosaccharides of glycoproteins on the cell surfaces by using N-pyridylamination fluorescent labels and reverse- and normal-phase high performance liquid chromatography (HPLC), in addition to CNS lymphoma specimens derived from the patients. Consequently, we obtained the results for differential expression patterns of N-oligosaccharides among PCNSL-derived cells, non-CNS lymphoma-derived cells, and clinical specimens of CNS lymphoma and for the slightly decreased expression in MTX-resistant lymphomas, including human brain malignant lymphomas HKBML and TK, and non-CNS lymphoma RAJI, which was seemed those in CNS lymphomas. This study is the first report for the expression patterns of N-oligosaccharides on tumor tissues resected from patients with CNS lymphomas and MTX-resistant PCNSL-derived cells, including HKBML and TK. The results may be a hint for understanding the status and microenvironments of the surface glycans of lymphomas and may be useful for development of applied target therapies for cell recognitions.
Primary and secondary CNS lymphomas, which were pathologically diagnosed DLBCL tissue specimens, were obtained from Toyama Prefectural Central Hospital (Additional file 1: Table S1). All study protocols were approved by both of the Institutional Review Boards of Toyama Prefectural Central Hospital and Kyoto Prefectural University of Medicine (approval number 2011–1081), and experiments were performed in accordance with institutional guidelines. Written informed consents were obtained from all patients. Resected tumor tissues were immediately snap-frozen and fixed in 4% (v/v) paraformaldehyde (PFA) for 24 h, and then substituted with phosphate-buffered saline (PBS).
A4/FUK  and TK  were purchased from JCRB Cell Bank (NIBIOHN: National Institutes of Biomedical Innovation, Health and Nutrition) (Additional file 1: Table S1). HKBML  and RAJI  were purchased from RIKEN Cell Bank (RIKEN BRC: RIKEN BioResource Center) (Additional file 1: Table S1). HBL1  and OYB  were distributed from Kyoto University (Additional file 1: Table S1). Cells were grown according to standard protocol in Ham’s F12 medium (Nacalai Tesque) with 15% fetal bovine serum (FBS) (Thermo Fisher Scientific) for HKBML, Roswell Park Memorial Institute (RPMI) 1640 medium (Nacalai Tesque) with 20% FBS for TK, and RPMI 1640 medium with 10% FBS for A4/FUK, HBL1, OYB, and RAJI in 5% CO2 at 37 °C. MTX-resistant cells were generated as shown in Additional file 2: Figure S1A.
Purification and pyridylamination of oligosaccharides were performed as described [19, 42, 43]. Pyridylated (PA)-N-oligosaccharides were separated into neutral N-oligosaccharides and mono−/di−/tri−/tetra-sialylated N-oligosaccharides through an anion-exchange column (Mono Q5/50GL, GE Healthcare) using HPLC and DE52-packed column (Whatman, GE Healthcare) . Sialylated PA-N-glycans were treated with neuraminidase at 37 °C for 14 h in 50 mM ammonium acetate (pH 5.0) to cleave sialic acids, followed by heating at 100 °C for 5 min and filtering through 0.2 μm spin filter (Ultrafree-MC LG, Millipore). Neutral PA-N-oligosaccharides were analyzed using HPLC as described [19, 42, 43]. N-oligosaccharide structures were determined by calculating the mannose unit value from NP-HPLC (Takara Bio) and the glucose unit value from RP-HPLC (Takara Bio), as described [20, 45], also by comparison with known standards and sequential exoglycosidase digestion (see below). PA-N-glycans were quantified as described , and HPLC chromatogram data were analyzed using Unipoint (Gilson), LC station (Shimadzu), and Empower2 (Waters). The workflow of N-oligosaccharides analysis and separation by HPLC is shown in Additional file 2: Figure S1B.
Exoglycosidase digestion was performed as described . In brief, purified PA-N-oligosaccharides were digested for 3 h at 37 °C using the following enzymes: Xanthomonas manihotis β1,3-galactosidase (New England BioLabs) for β (1–3)-Gal in 50 mM sodium acetate (pH 4.5) with 100 μg/mL bovine serum albumin (BSA); Diplococcus pneumoniae β-galactosidase (Roche Diagnostics) for β1,4-Gal in 50 mM sodium acetate (pH 6.0); α1,3/4-L-fucosidase (Takara Bio) for α1,3/4Fuc in 100 mM sodium phosphate buffer (pH 6.0); bovine kidney α1,6-fucosidase (ProZyme) for α1–6 > 1–2/3/4-Fuc in 100 mM sodium phosphate buffer (pH 6.0). PA-N-oligosaccharides were also digested for 14 h at 37 °C by neuraminidase for α2,3/6/8-NeuAc in 50 mM ammonium acetate (pH 5.0) and α2,3-sialidase for α (2, 3) NeuAc in 50 mM sodium citrate (pH 6.0) with 100 mM NaCl and 100 μg/mL BSA.
Matrix assisted laser desorption ionization time-of-flight mass spectrometry (MALDI/TOF-MS)
Molecular masses of PA-sugar chains and their isobaric monosaccharide compositions were determined by MALDI/TOF-MS as described . One microliter of matrix solution of 10 mg/mL 2,5-dihydroxybenzoic acid in 30% acetonitrile was spotted on the plate, 1 μL of sample solution was added and then dried by warm air. MALDI/TOF mass spectra were acquired using REFLEX mass spectrometer (Bruker-Franzen) in the positive and reflector mode at an acceleration voltage of 20 kV and delayed ion extraction. Standard PA-oligosaccharides were used to achieve a two-point external calibration for mass assignment of ions. The mass spectra shown were the sum of at least 30 laser shots.
Heatmaps were constructed by color-coding standardized log or linear scales, indicating fold-differences of neutral and sialylated sugar chains in MTX-resistant cells compared to corresponding control cells, or by using the percent (%) area of fractionated peaks, designated by peak numbers, eluting at different retention times. Color configurations are shown in each panel. NA, not applicable.
Statistics were performed using JMPv10 built-in-modules (SAS Institute Inc., Tokyo, Japan). Principal component analysis was performed to estimate Pearson correlation. P < 0.05 was considered statistically significant.
Generation of MTX-resistant lymphoma cells
Expression and structures of N-linked oligosaccharides in lymphoma cells
DE52 diethylaminoethyl cellulose (DEAE-C) anion-exchange column chromatography was performed using HKBML, TK, RAJI, OYB, HBL1, and A4/FUK (Additional file 3: Figure S2A and Additional file 4: Figure S3A), and MTX-resistant HKBML, TK, and RAJI (Additional file 3: Figure S2A). Coupled with standard unit for sialylated sugar chains, the peaks representing non-sialylated (S0) to 4-sialylated (S4) sugar chains were detected (Additional file 3: Figure S2A and Additional file 4: Figure S3A). The left-side peaks at the S1 peak indicated the N-oligosaccharides coupling with an unknown acidic group, and the left-side peaks at the S2 peak indicated the N-mono-sulfated oligosaccharides (Additional file 2: Figure S1A and Additional file 3: Figure S2A). Fractions eluting at wide-range retention times of S0 were further analyzed using normal-phase HPLC in neutral sugar chains and sialylated sugar chains (Additional file 3: Figures S2B–G and Additional file 4: Figure S3B, C). Fractions eluting at retention times of M4-M9 were collected and analyzed, and the oligostructures included in the peaks were determined by MALDI/TOF-MS, then referred to the oligosaccharide structures in Additional file 5: Figure S4. High mannose type oligosaccharides, including M4B, M5A, M6B, M7A, M7B, M8A, and M9A were determined with mannose unit standard in all samples examined (Additional file 3: Figure S2B–G, Additional file 4: Figure S3B, C, and Additional file 5: Figure S4). Interestingly, biantennary bigalactosylated structure with α1,6-fucosylation, A2G2F, and the N-acetylglucosamine-bisecting structure, A2G2FB as complex types, were also detected in neutral and sialylated oligosaccharides in cells derived from HKBML, TK, and RAJI (Additional file 3: Figure S2B–G and Additional file 5: Figure S4). Further, tetra-antennary tetra-galactosylated structure with α1,6-fucosylation, A4G4F, was only detected in cells from RAJI (Additional file 3: Figure S2F, G and Additional file 5: Figure S4). These complex types were not detected in HBL1, A4/FUK, and OYB (Additional file 4: Figure S3B, C).
Differential expression patterns of N-linked oligosaccharides in MTX-resistant lymphoma cells
Sialylated N-linked oligosaccharides of glycoproteins on the cell surface of lymphomas
Sialylated N-linked oligosaccharides of glycoproteins in clinical specimens of CNS lymphoma
Molecular alterations characterizing PCNSL have been studied to develop potentially diagnostic and prognostic imaging and biomarkers, proposed as possible targeted therapies [46, 47, 48, 49]. There is also great interest in the Bruton’s tyrosine kinase (BTK) inhibitor against DLBCL (ibrutinib), immunomodulatory drugs (lenalidomide), and immunotherapy using anti-programmed cell death 1 (PD-1) monoclonal antibody (nivolumab) [47, 50]. Furthermore, genetic and molecular alterations activating nuclear factor-kappa B (NF-ĸB) signaling are found in most PCNSL together with MYD88 and CD79B mutations . Several studies have suggested a strong correlation between lectin-binding patterns and/or biological significances and functions in various tumors, including DLBCL [24, 25, 26, 27], lymphoma , and hematopoietic cell lines . However, few studies have been reported on the glycoproteins in CNS lymphomas. Although few clinical specimens were used, the CNS lymphoma tissue-specific oligosaccharides, including GM9, BA-2, A2G0F, BA-1, LewisXa/b-BA-2, Ga/b-BA-2, A2G2, A3G3, and A2G1 (3)(6)F, were isolated, whereas there could have been contaminating blood cells and/or macrophages from the tumor microenvironments. Thus, further studies should be performed carefully in a number of purified tumor specimens.
A2G2F was increased in glioblastoma tissues and glioma cell lines, while being less than 0.1% in normal brain tissue . In the present study, A2G2F was detected in neutral and sialylated sugar chains as 2.19–5.0% but not in neutral sugar chains in the cell lines, which had a slight change in the MTX-resistant cells compared to the control cells as 0.92–1.08-fold (Fig. 2g, h). In sialylated sugar chains, A2G2F was detected as 11.67–44.02% with decreased levels in the MTX-resistant cells compared to the control cells as 0.76–0.81-fold (Fig. 3b, c). Besides, in the analysis for the CNS lymphoma tissues, average 6.59% (range: 2.77–14.54%) of A2G2F was detected in the four clinical specimens (Fig. 4a, b). Therefore, A2G2F in CNS lymphoma tissues and cell lines is much than normal brain tissue, whereas sialylated A2G2F in the MTX-resistant PCNSL-derived cells is slightly decreased compared to the control cells, which is consistent with glioma cell lines, glioblastoma tissues, PCNSL-derived cell lines, CNS lymphoma tissues, and non-CNS lymphoma cells. In glioblastoma and/or glioma, Lens culinaris agglutinin (LCA)-lectin binding to A2G2F inhibits cell proliferation of glioma through induction of apoptosis . Therefore, A2G2F may also provide a useful marker candidate and a hint for diagnosis and development for target therapy in CNS lymphoma, as well as glioma/glioblastoma.
All of the four CNS lymphoma specimens and cell lines, including CNS lymphoma and non-CNS lymphoma, used in the study were human immunodeficiency virus (HIV)-negative, except for HBL1 and OYB as no valid information for HIV (Additional file 1: Table S1). PCNSL-derived HKBML and non-CNS lymphoma-derived Raji are Epstein-Barr virus (EBV)-positive and PCNSL-derived TK is EBV-negative (Additional file 1: Table S1). However, the expression patterns of N-oligosaccharides in HKBML were similar to that in TK than Raji, in addition to differential expression with acquired MTX-resistance in neutral and sialylated sugar chains (Fig. 2g, h), and sialylated sugar chains (Fig. 3b, c). While, the EBV infection in OYB cells was unknown (Additional file 1: Table S1). Our previous study has clarified that EBV-positive PCNSLs are counted by 20% . However, all of the four clinical specimens examined in the study were EBV-negative. Therefore, whether EBV-positive PCNSLs make a change to their oligopatterns should await future studies.
In this study, we demonstrated that the expression of specific sialylated N-linked oligosaccharides, including A2G2F and A2G2FB, in the MTX-resistant cells slightly decreased compared to the corresponding control cells. Similarly, the decreased expression of sialylated A2G2F and A2G2FB seemed to correlate with poor prognoses of the CNS lymphoma patients, despite the small sample number. Therefore, the differential expression and patterns of surface glycans on CNS lymphomas make it possible to escape the cell–cell recognition by immune cells, thereby, MTX-resistant malignant CNS lymphoma cells could re-grow. In conclusion, the above-mentioned oligosaccharides may be promising oligosaccharide marker candidates to recognize MTX-resistant cells, and primary and secondary CNS lymphomas, which may be useful for diagnosis marker development and/or applied molecular targeted therapies for CNS lymphomas.
We are grateful to Dr. Momoko Nishikori (Kyoto University) for providing cell lines.
YT, KI, and RY designed the experiments. HH diagnosed and treated patients and collected samples. YT, TY, YK, and AH performed the experiments. YT, KI, and RY analyzed data. YT and RY wrote the manuscript. All authors read and approved the final manuscript.
This work was supported by JSPS KAKENHI grant numbers 23659695, 16H05441, and 18 K09001. The funding bodies listed here do not have any roles in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.
Ethics approval and consent to participate
All study protocols were approved by both of the Institutional Review Boards of Toyama Prefectural Central Hospital and Kyoto Prefectural University of Medicine (approval number 2011–1081), and experiments were performed in accordance with institutional guidelines. Written informed consents were obtained from all patients.
Consent for publication
The authors declare that they have no competing interests.
- 6.Koenderman AH, Koppen PL, Koeleman CA, van den Eijnden DH. N-acetylglucosaminyltransferase III, IV and V activities in Novikoff ascites tumour cells, mouse lymphoma cells and hen oviduct. Application of a sensitive and specific assay by use of high-performance liquid chromatography. Eur J Biochem. 1989;181:651–5.CrossRefGoogle Scholar
- 32.Goecke IA, Alvarez C, Henríquez J, Salas K, Molina ML, Ferreira A, et al. Methotrexate regulates the expression of glucocorticoid receptor alpha and beta isoforms in normal human peripheral mononuclear cells and human lymphocyte cell lines in vitro. Mol Immunol. 2007;44:2115–23.CrossRefGoogle Scholar
- 34.Yamanaka R. Salvage therapy for primary central nervous system lymphoma. In: Yamanaka R, editor. Primary Central Nervous System Lymphoma (PCNSL): incidence, management and outcomes. New York: Nova Science Publishers; 2016. p. 175–87.Google Scholar
- 44.Torii T, Yoshimura T, Narumi M, Hitoshi S, Takaki Y, Tsuji S, et al. Determination of major sialylated N-glycans and identification of branched sialylated N-glycans that dynamically change their content during development in the mouse cerebral cortex. Glycoconj J. 2014;31:671–83.PubMedPubMedCentralGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.