Mutations in the colony-stimulating factor 3 receptor (CSF3R) have been identified in the vast majority of patients with chronic neutrophilic leukemia and are present in other kinds of leukemia, such as acute myeloid leukemia. Here, we studied the function of novel germline variants in CSF3R at amino acid N610. These N610 substitutions were potently oncogenic and activated the receptor independently of its ligand GCSF. These mutations activated the JAK–STAT signaling pathway and conferred sensitivity to JAK inhibitors. Mass spectrometry revealed that the N610 residue is part of a consensus N-linked glycosylation motif in the receptor, usually linked to complex glycans. N610 was also the primary site of sialylation of the receptor. Membrane-proximal N-linked glycosylation was critical for maintaining the ligand dependence of the receptor. Mutation of the N610 site prevented membrane-proximal N-glycosylation of CSF3R, which then drove ligand-independent cellular expansion. Kinase inhibitors blocked growth of cells with an N610 mutation. This study expands the repertoire of oncogenic mutations in CSF3R that are therapeutically targetable and provides insight into the function of glycans in receptor regulation.

Significance:

This study reveals the critical importance of membrane-proximal N-linked glycosylation of CSF3R for the maintenance of ligand dependency in leukemia

Mutations in colony-stimulating factor 3 receptor (CSF3R), also known as granulocyte colony-stimulating factor receptor (GCSFR), occur in the majority of patients with chronic neutrophilic leukemia (CNL; refs. 1, 2) and also found more rarely, in patients with acute myeloid leukemia (AML; refs. 3–7). Truncation mutations that lead to a premature stop in the cytoplasmic domain are found in CNL (1) and result in increased expression of CSF3R on the cell surface (8). The most common CSF3R mutation in CNL is T618I (T595I), a point mutation in the membrane-proximal extracellular domain that causes ligand independence (Fig. 1A; ref. 9). There are two numbering conventions; the second historical one we will put in parentheses does not include the 23 amino acid signal peptide. Recently, the presence of a CSF3R T618I mutation became a part of the World Health Organization (WHO) criteria for diagnosis of CNL (10).

Figure 1.

A novel germline mutation in CSF3R. A, Schematic of the CSF3R membrane-proximal location of the N610H mutation and nearby leukemia-associated T615A and T618I mutations. B, Sanger sequencing of CSF3R exon 14 confirms the presence of the CSF3R N610H mutation in both the peripheral blood and skin biopsy from a patient with a myeloproliferative neoplasm. T.M., transmembrane domain.

Figure 1.

A novel germline mutation in CSF3R. A, Schematic of the CSF3R membrane-proximal location of the N610H mutation and nearby leukemia-associated T615A and T618I mutations. B, Sanger sequencing of CSF3R exon 14 confirms the presence of the CSF3R N610H mutation in both the peripheral blood and skin biopsy from a patient with a myeloproliferative neoplasm. T.M., transmembrane domain.

Close modal

Although CSF3R T618I is the most common mutation in CNL, there are other rarer variants that also cause profound receptor activation, such as T615A (T592A), a mutation seen in the membrane proximal domain (1). In the transmembrane domain, a point mutation at T640N (T617N), was reported in a family with congenital neutrophilia, as well as in a few patients with CNL (11, 12). This T640N mutation is predicted to cause intramolecular hydrogen bonding and has been experimentally shown to increase ligand-independent dimerization (11, 12). Although study of transmembrane and transmembrane proximal mutations has greatly improved our ability to diagnosis and treat the disease, there are still rare variants for which their significance is not clear. Herein, we study the mechanism of action of a rare germline CSF3R N610H mutation. This patient had a condition most consistent with primary myelofibrosis with mild leukocytosis. Because of N610′s proximity to other more common oncogenic CSF3R point mutations found in CNL, we were interested in understanding the functional consequences and therapeutic implications of N610 substitutions.

There is a substantive body of literature that correlates changes in glycosylation with cancer progression (13). Previously, we reported that T618I is at a site of O-glycosylation and changes in glycosylation at that site have been linked to oncogenesis (14). It is therefore critical to understand the posttranslational modifications on CSF3R. The Asn residue at 610 is part of an N-linked glycosylation consensus N-X-(S/T) motif (15) and N-glycosylated proteins typically migrate to the extracellular space. N-Glycans help to fold, traffic, and thermodynamically stabilize the protein (16).

In this study, we confirm the identity of the glycans on N610 by mass spectrometry (MS) analysis. We determined that N610 is occupied with a sialylated hybrid N-glycan. Furthermore, we identify the oncogenic pathway for these mutations. Both the N610H substitution and a second germline mutation identified (N610S) highly activate CSF3R, causing cytokine-independent growth in Ba/F3 cells. Like the common T618I mutation, these mutations render the receptor ligand–independent. Downstream, N610H and N610S activate the JAK–STAT pathway as demonstrated by an increase in the levels of phospho-STAT3. The loss of N-glycosylation in the membrane-proximal region of CSF3R promotes ligand-independent receptor activation and oncogenesis. Rare human mutations can provide significant insight into the relationship between receptor structure and function.

Sequencing

Genomic DNA from skin biopsy and peripheral blood were sequenced. Exon 14 of CSF3R was amplified using the following M13F and M13R tagged primers (forward-GTAAAACGACGGCCAGTCCACGGAGGCAGCTTTAC; reverse-CAGGAAACAGCTATGACCAAATCAGCATCCTTTGGGTG), purified on an Amicon Ultra 0.5 mL Centrifugal Filter (Millipore) followed by Sanger Sequencing (Eurofins Genomics) using M13F and M13R primers. The N610S mutation was identified through a custom RainDance Thunderbolt Sequencing panel run at Memorial Sloan Kettering on both blood and nail clippings. The genes sequenced on this panel are as follows: ASXL1, BCOR, BCORL1, BRAF, CALR, CBL, CBLB, CEBPA, CSF3R, DNMT3A, ETV6, EZH2, FLT3, GATA1, GATA2, GNAS, HRAS, IDH1, IDH2, JAK1, JAK2, JAK3, KDM6A, KIT, KRAS, MAP2K1, MPL, MYD88, NOTCH1, NPM1, NRAS, PHF6, PML, PTEN, PTPN11, RAD21, RUNX1, SETBP1, SF3B1, SMC1A, SMC3, SRSF2, STAG2, TET2, TP53, U2AF1, WT1, and ZRSR2. Patient samples were obtained with written informed consent in accordance with the Declaration of Helsinki and Institutional Review Boards of Memorial Sloan Kettering and Washington University of St. Louis (St. Louis, MO).

Plasmid construction

MSCV-IRES-GFP (MigR1) was made compatible for Gateway cloning using the Gateway Vector Conversion Kit (Invitrogen). A gateway pDONR vector for CSF3R transcript variant 1 (NM_00760.2) was purchased (GeneCopoeia). CSF3R was mutagenized as described previously for the CSF3R T618I mutation (1) or using the Quikchange II XL Site Directed Mutatgenesis Kit (Agilent Genomics). The following primers were used for site-directed mutagenesis: N610H (F-ggctggggccacccacagtacagtcct, R-aggactgtactgtgggtggccccagcc), N610Q (F-gtgaggactgtactctgggtggccccagcct, R-aggctggggccacccagagtacagtcctcac), N610S (F-gaggactgtactgctggtggccccagc, R-gctggggccaccagcagtacagtcctc). After sequence conformation of the mutagenesis, CSF3R mutants were subcloned into Gateway-MSCV-IRES-GFP using LR Clonease II, a recombination based strategy (Invitrogen).

Cell culture

Ba/F3 cells were obtained from Brian Druker at OHSU grown in RPMI1640 with 10% FBS, HyClone), penicillin/streptomycin, 15% WEHI conditioned medium (which contains IL3), and l-glutamine. BaF3 cells were not allowed to exceed passage 20. 293T17 cells (obtained from ATCC CRL-11268) were grown in DMEM containing GlutaMAX (Gibco) with 10% FBS and penicillin/streptomycin. After obtaining cells from ATCC, cells were frozen at low passage (p3–p5) and then thawed and used at passage 9 or lower. Cell line authentication was not performed for these studies. 293T17 cells were transfected with Fugene 6 (Promega) at a 5:1 ratio of lipid to DNA. Retrovirus was made by cotransfection of MigR1 constructs with pEcopac. Cell stocks were Mycoplasma tested prior to freezing or upon thaw, in addition to monthly testing of cells in culture. Mycoplasma testing was performed using the MycoAlert Mycoplasma Detection Kit (Lonza).

Ba/F3 cytokine–independent growth assays

Viral supernatants were filtered using 0.45-μm filters and then Ba/F3 cells or mouse bone marrow were spinoculated in the presence of polybrene and HEPES buffer. Cells were spun at 2,500 rpm, for 90 minutes at 30°C (brake turned off). GFP+ cells were sorted on a BD FACSAria II sorter, and then sorted cells were allowed to expand for 2 to 4 days. GFP+ Ba/F3 cells expressing CSF3R mutants or controls were washed three times, and plated at 5 × 105 cells/mL in the absence of cytokine support (RPMI1640 media, with 10% FBS, l-glutamine, and penicillin/streptomycin). Cell viability and number were monitored on a Bio-Rad TC20 cell counter. For drug treatment studies, Mig empty (control vector) and WT CSF3R were grown in the presence of IL3 (WEHI-conditioned medium) and CSF3R-mutant constructs were grown in the absence of IL3. Cells in 96-well plate format were treated with increasing doses of ruxolitinib (Selleckchem) or trametinib (Selleckchem). GCSF-independent assays were performed with GFP-sorted cells maintained in IL3-containing media. Cells were washed three times in media without IL3 and then grown in 96-well format with increasing doses of GCSF.

Mouse bone marrow colony assays

Mouse bone marrow was harvested from 6-to 10-week old C57/BL6 mice. Marrow was cultured overnight in the presence of SCF, IL6, and IL3. Virus was prepared as described in the cell culture section and then filtered using 0.45-μm filters. Mouse bone marrow (1 × 106 cells) were spinoculated with viral supernatant, HEPES buffer, and polybrene on two subsequent days. For the spinoculation, cells were spun at 2,500 rpm, for 90 minutes at 30°C (brake turned off). One day after the second sorting, GFP percentage was assessed by flow cytometry to determine that all vectors had infected the bone marrow cell. Cells were plated in triplicate with 1 × 104 cells in 1 mL mouse methylcellulose without added cytokines (MethoCult M3234, StemCell Technologies). Cells were imaged using STEMvision (StemCell Technologies) at day 14, blinded, and then manually counted. All animal work was performed in an AAALAC accredited facility with prior approval from the Oregon Health & Science University Institutional Animal Care and Use Committee, under protocol IP00000482.

Immunoblot analysis

Transfected 293T17 cells with the indicated CSF3R-MigR1, constructed as described under cell culture, were lysed 48 hours posttransfection in Cell Lysis Buffer (Cell Signaling Technology) containing Complete Protease Inhibitor Cocktail Tablets (Roche) and Phosphatase Inhibitor Cocktail II (Sigma). Lysates were centrifuged at 14,000 rpm for 10 minutes at 4°C. Supernatant was transferred to a new tube then mixed with 3× SDS sample buffer (75 mmol/L Tris pH 6.8, 3% SCS, 15% glycerol, 8% β-mercaptoethanol, 0.1% bromophenol blue) and then heated for 5 minutes at 95 °C. Lysates were run on 4% to 15% criterion TGX Precast Protein Gels (Bio-Rad). Gels were transferred using the Trans-Blot Turbo Transfer System (Bio-Rad). After blocking in TBST with 5% milk, blots were probed in anti-GCSFR (38643, R&D Systems), Rb anti-pSTAT3 (9131, Cell Signaling Technology), Rb anti-Stat3 (9132, Cell Signaling Technology), or GAPDH (#25778, Santa Cruz Biotechnology). Blots were washed with TBST and then an appropriate HRP-conjugated secondary, followed by incubation with a chemiluminescent substrate (Thermo Fisher Scientific), and imaged with a ChemiDoc Gel Imaging System (Bio-Rad).

Immunoprecipitation

Protein purification was performed as described previously (14). In brief, 293T17 cells from ATCC were maintained at 37 °C and 5% CO2 incubator in DMEM with 10% FBS and penicillin/streptomycin. 293T cells were transfected using FuGENE 6 (Promega) in Optimem. Transfected cells were lysed using cell lysis buffer (Cell Signaling Technology) containing Complete Protease Inhibitor 3 (Calbiochem), spun at 10,000 rpm for 10 minutes to pellet cell debris and supernatant collected. FLAG-tagged constructs were immunoprecipitated from cell lysates by incubation with anti-FLAG M2 affinity gel (Sigma-Aldrich) for 1 hour at 4°C on a rotator. Beads were washed with cell lysis buffer. Proteins were disassociated from beads by incubating with FLAG peptide at room temperature for 1 hour and then subjected to immunoblotting analysis.

Detailed MS protocols are provided in Supporting Information. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE [1] partner repository with the dataset identifier PXD011118.

Identification of a novel germline mutation in CSF3R

The CSF3R N610H mutation was initially identified in a patient with a myeloproliferative neoplasm through next-generation sequencing using a custom AML/MDS mutation hotspot panel for 42 genes that have biological importance in myeloid malignancies. This sequencing of the patient's bone marrow revealed a mutation at N610H in CSF3R at a 50% mutant allele frequency. This myeloproliferative neoplasm was most consistent with a JAK2, CALR, MPL mutation-negative primary myelofibrosis. This patient had a history of mild leukocytosis for several years with the most recent white blood cell counts between 13.3 and 15.3 × 103/μL with granulocytic left shift. A bone marrow biopsy revealed 90% cellularity with a mild increase in reticulin fibrosis, increased myeloid to erythroid ratio, no overt dysplasia, and less than 5% blasts. The bone marrow cells were karyotypically normal, but 59% of interphase nuclei carried a microdeletion of the 3′ end of PDGFRB (5q; identified by FISH). At the time of diagnosis, the patient had minimal symptoms with no anemia or thrombocytopenia, and was being monitored, but not receiving any specific treatment.

The near 50% allele fraction of the N610H mutation prompted us to determine whether the mutation was germline or somatic. Sanger sequencing confirmed the presence of a heterozygous N610H mutation in a sample of blood as well as in a skin biopsy (Fig. 1B), confirming the presence of a novel CSF3R N610H germline mutation in this patient with an unusual myeloproliferative neoplasm. The patient had no known family history of hematologic malignancies or neutrophilia.

Subsequent to the mutational analysis, the patient developed severe anemia and became transfusion dependent. His hemoglobin was 7.3 g/dL, with a red blood cell count of 1.9 × 103/μL, and a hematocrit of 21.9%. At this time, he continued to have moderate leukocytosis with a white blood cell count of 14.6 × 103/μL. Sequencing of a bone marrow biopsy at this time revealed the acquisition of an ASXL1 Q1448Pfs*8 mutation at 33% variant allele frequency. ASXL1 is a chromatin-modifying protein that is part of the polycomb repressor complex 2 (17). ASXL1 mutations are commonly found in myeloid malignancies and are often associated with poor prognosis (18–21). ASXL1 mutations can also occur alongside CSF3R mutations in CNL (22, 23). Because of increasing transfusion dependence, decitabine was initiated and a bone marrow transplant is being considered.

The CSF3R N610H mutation is activating and causes ligand independence

Because of N610H's proximity to the most common T618I CSF3R mutation found in CNL, we were interested in understanding whether this novel mutation might be important for the disease pathology. To test the oncogenic capacity of this novel mutation, we used a cytokine-independent growth assay. In this assay, the murine pro-B-cell line, Ba/F3, was transduced with a retrovirus expressing the WT or mutant CSF3R using a vector containing an IRES-GFP. GFP-positive cells expressing the gene of interest were sorted, allowed to recover, washed to remove cytokine support, and then cultured in the absence of exogenous cytokines. Native Ba/F3 cells and those expressing WT CSF3R depend on cytokines for growth and die when the cytokines are removed. We found that the N610H mutation, (which results in a positive charge) and also a more conservative N610Q substitution, are both highly activating in CSF3R. Both mutations promote cytokine-independent growth in the murine Ba/F3 cell line after 3 days, whereas cells expressing WT CSF3R are unable to proliferate (Fig. 2A). Glutamine (Q) was chosen as the more conservative substitution because glutamine is structurally similar asparagine (N) and also uncharged.

Figure 2.

The CSF3R N610H mutation is transforming and activates the JAK–STAT pathway in a ligand-independent manner. A, Transforming capacity of the CSF3R N610H (and a conservative N610Q) mutations in the murine Ba/F3 cytokine–independent growth assays. B, The CSF3R N610H mutation activates the JAK–STAT pathway as measured by immunoblotting of phosphorylated-STAT3 (pSTAT3) in 293T17 cells transiently transfected with WT or CSF3R mutants. N610H, the CSF3R N610H mutation; N610Q, CSF3R N610Q mutation, a conservative substitution; and T618I, the CSF3R T618I mutation found commonly in CNL. C, The CSF3R N610H, N610Q, and T618I mutation confers ligand-independence to the receptor as measured by a titration of GCSF in Ba/F3 cells expressing WT CSF3R. Cell lines were plated in decreasing concentrations of GCSF and then cell viability was measured after 72 hours using a tetrazolamine-based assay (CellTiter Aqueous One Solution Cell Proliferation MTS assay). Total protein loading was confirmed by GAPDH.

Figure 2.

The CSF3R N610H mutation is transforming and activates the JAK–STAT pathway in a ligand-independent manner. A, Transforming capacity of the CSF3R N610H (and a conservative N610Q) mutations in the murine Ba/F3 cytokine–independent growth assays. B, The CSF3R N610H mutation activates the JAK–STAT pathway as measured by immunoblotting of phosphorylated-STAT3 (pSTAT3) in 293T17 cells transiently transfected with WT or CSF3R mutants. N610H, the CSF3R N610H mutation; N610Q, CSF3R N610Q mutation, a conservative substitution; and T618I, the CSF3R T618I mutation found commonly in CNL. C, The CSF3R N610H, N610Q, and T618I mutation confers ligand-independence to the receptor as measured by a titration of GCSF in Ba/F3 cells expressing WT CSF3R. Cell lines were plated in decreasing concentrations of GCSF and then cell viability was measured after 72 hours using a tetrazolamine-based assay (CellTiter Aqueous One Solution Cell Proliferation MTS assay). Total protein loading was confirmed by GAPDH.

Close modal

One feature of the CSF3R T618I mutation is robust activation of the JAK/STAT pathway. We assessed the ability of the N610H and N610Q mutations to activate this pathway using phosphorylated STAT3 as a marker. Both of these mutations lead to robust phosphorylation of STAT3 above and beyond the increase in signaling with WT CSF3R (Fig. 2B).

The CSF3R T618I mutation is the most common in CNL and confers ligand-independent receptor activation. To test whether the N610H mutation also conferred ligand-independent growth, we grew Ba/F3 cells transduced with either WT or mutant forms of CSF3R in decreasing concentrations of GCSF, the ligand for CSF3R. Like the T618I mutant, both the N610H and N610Q mutations exhibit ligand independence, whereas WT CSF3R cell viability has a dose-dependent relationship with the GCSF concentration (Fig. 2C).

N610 is a site of N-linked glycosylation in CSF3R

We previously reported that the CSF3R T618 site is O-glycosylated. The common T618I mutation causes a loss of glycosylation, leading to ligand independence and neutrophil expansion (14). Interestingly, N610 is part of an N-X-T motif, which is a consensus sequence for N-linked glycosylation. Haniu and colleagues demonstrated that N610 is one of eight CSF3R sites that results in a lower molecular weight gel shift when treated with a cocktail of deglycosylation enzymes, indicating that N610 is likely a site of N-glycosylation (24). We set out to confirm the glycosite utilization and identifying the glycan structures at N610. We began by performing free glycan analysis of CSF3R. In brief, this was achieved by first digesting WT CSF3R-purified proteins with trypsin, and next releasing N-glycans enzymatically from the crude mixture with PNGase F. Released N-glycans are then permethylated with methyl iodide and finally analyzed by MS (Fig. 3A). Interestingly, we observed complex, biantennary, and high-mannose glycan structures (Fig. 3B). We confirmed the identity of these glycan structures by electrospray ionization mass spectrometry (ESI-MS). Although this analysis accurately assigns glycan structure, this type of MS analysis does not assign glycans to particular sites on the protein. We therefore performed additional studies to identify the glycan at the N610 site. Consistent with the previous study by Haniu and colleagues (24), MS-based glycoproteomic analysis revealed that the N610 site is occupied by a glycan (Fig. 4). In addition, we identified N-linked high mannose glycosylation at asparagine residues N389 and N474 and transient occupancy at site N51 (full mass spectrometry data have been uploaded to the EMBL-EBI PRIDE database).

Figure 3.

Annotated MALDI-TOF MS spectra of permethylated N-glycans (m/z 1,500—3,000) from wild-type CSF3R. A, Released permethylated N-glycan analysis by MALDI. Glycan structures identified were confirmed by electrospray ionization mass spectrometry. Putative structures are based on the molecular weight and N-glycan biosynthesis pathway. B, The ratio indicates a comparison of the glycan relative abundance to those identified.

Figure 3.

Annotated MALDI-TOF MS spectra of permethylated N-glycans (m/z 1,500—3,000) from wild-type CSF3R. A, Released permethylated N-glycan analysis by MALDI. Glycan structures identified were confirmed by electrospray ionization mass spectrometry. Putative structures are based on the molecular weight and N-glycan biosynthesis pathway. B, The ratio indicates a comparison of the glycan relative abundance to those identified.

Close modal
Figure 4.

Identification of the glycan at N610. Schematic of labeling of sialic acid glycoproteins with an azide-functionalized sugar (i.e., Ac4ManNAz, 1). B, Wild-type, N610H, or T618I CSF3R were transfected in 293T cells and then incubated with either Ac4ManNAz or DMSO as a control. After incubation, cells were lysed and reacted with a DIBCAC-functionalized biotin probe (Biotin-alkyne, 2). Sialoglycoproteins (biotinylated) were immunoprecipated using an avidin resin and visualized with an HRP-conjugated anti-FLAG. Equal avidin resin was loaded. CD, The MS/MS spectra of electron transfer dissociation fragmented N610 peptide “AASQAGATNSTVL.” C, The most intense peak corresponds to the peptide with a HexNAc3Hexose6NeuAc. D, Spectrum corresponds to glycan structure HexNAc3Hexose6NeuAc. Glycoworkbench (42, 43) was used for creating the glycan structure figures. Blue square, GlcNAc; yellow circle, galactose; green circle, mannose; purple diamond, N-acetylneuraminic acid (Neu5Ac).

Figure 4.

Identification of the glycan at N610. Schematic of labeling of sialic acid glycoproteins with an azide-functionalized sugar (i.e., Ac4ManNAz, 1). B, Wild-type, N610H, or T618I CSF3R were transfected in 293T cells and then incubated with either Ac4ManNAz or DMSO as a control. After incubation, cells were lysed and reacted with a DIBCAC-functionalized biotin probe (Biotin-alkyne, 2). Sialoglycoproteins (biotinylated) were immunoprecipated using an avidin resin and visualized with an HRP-conjugated anti-FLAG. Equal avidin resin was loaded. CD, The MS/MS spectra of electron transfer dissociation fragmented N610 peptide “AASQAGATNSTVL.” C, The most intense peak corresponds to the peptide with a HexNAc3Hexose6NeuAc. D, Spectrum corresponds to glycan structure HexNAc3Hexose6NeuAc. Glycoworkbench (42, 43) was used for creating the glycan structure figures. Blue square, GlcNAc; yellow circle, galactose; green circle, mannose; purple diamond, N-acetylneuraminic acid (Neu5Ac).

Close modal

Sialylation at N610

Free glycan analysis revealed several glycan structures on native CSF3R with sialic acid, also known as N-acetylneuraminic acid (Neu5Ac). Aberrant expression of sialic acid on proteins has been observed in many types of cancer (25). To confirm that WT CSF3R is indeed sialylated at N610, we performed a metabolic glycosylation assay. As previously described, sialoglycoproteins are targeted specifically by metabolically labeling cells (Fig. 4A) with a modified peracetylated N-acetylmannosamine (Ac4ManNAc) analogue, such as peracetylated N-azidoacetylmannosamine (Ac4ManNAz, 1; Fig. 4) that incorporates into sialic acid glycoproteins (26, 27). The azide-modified sugar passively diffuses through the cell and is converted by the Roseman–Warren pathway into the corresponding azidosialic acid (28), which allows for the site-specific labeling with a biotin-conjugated reagent, such as a biotin-alkyne probe (2; Fig. 4A) under copper-catalyzed azide–alkyne [3+2] cycloaddition (CuAAC) conditions (29). HEK 293 cells were transfected with WT, N610H, or T618I FLAG-tagged constructs and incubated with 50 μmol/L Ac4ManNAz or DMSO as a control. The FLAG-tagged constructs were then immunoprecipitated from whole-cell lysates and biotin-labeled. Western blot analysis revealed robust labeling of WT CSF3R and T618I, whereas we did not observe labeling in the N610H mutant, suggesting that N610 is the only site of sialylation in CSF3R (Fig. 4B). N-glycan structures are diverse and can contain a variety of different sugars, including sialic acid. Our data suggest that although there are multiple N-glycosylated sites in CSF3R, N610 is the primary site of sialylation. Thus, raising the possibility that this charged modification may have important structural or functional roles in this critical region of the receptor.

MS-based glycoproteomic analysis reveals N610 glycan structure

To confirm glycosylation at the N610 site, we performed a glycoproteomic experiment on the digested purified protein. We were not able to observe by MS the peptide containing the N610 site, in samples digested with trypsin, as the asparagine sits in a rather large 71 amino acid residue peptide, and although possible, peptides of this size are challenging to analyze by MS-based methods. However, an alternative digestion with chymotrypsin (30), which cleaves on the N-terminal side of hydrophobic amino acid residues (tyrosine, tryptophan, and phenylalanineleucine and leucine) did prove successful. FLAG-purified WT and N610H constructs allowed us to identify by MS in vitro several hybrid glycoform structures (Fig. 4C and D). We observed several diagnostic oxonium ions from HexNAc m/z at 138, 168, 186, and 204, Neu5Ac oxonium ions at 274 and 292 and the intensity of HexHexNAc oxonium ion at m/z 366. This site-directed MS analysis confirms that N610 is occupied by sialylated glycans.

Identification of a germline CSF3R N610S mutation in a patient with chronic myeloid leukemia

Subsequent to these initial studies, a CSF3R N610S mutation was detected in a 54-year-old female patient with chronic phase chronic myeloid leukemia (CML). The patient's initial white blood cell count was 13,000, with a hemoglobin of 15 g/dL and a platelet count of 428,000. Cytogenetic analysis revealed an atypical 9;22 translocation as well as t11;17(p11.2;p13). She was initially treated with imatinib and achieved complete hematologic remission and complete cytogenetic response but plateaued short of major molecular remission. She was then switched to nilotinib. This patient was in deep molecular remission after therapy with nilotinib, with minimally evident (>4 log reduction from untreated International Standard baseline) or undetectable BCR-ABL sequentially, who developed increasing thrombocytosis. Evaluation for typical myeloproliferative drivers was unrevealing and bone marrow pathology noted mild myeloid hyperplasia and megakaryocytic MPN-like changes (clustering, increased nuclear-cytoplasmic ratio). Molecular analysis by MSKC IMPACT sequencing of bone marrow and fingernail DNA revealed a germline CSF3R N610S variant and a somatic DNMT3A mutation. At this time point, the DNMT3A R882H mutation (c. 2645 G>A) was found at a variant allele frequency of 6.4% in the marrow. The CSF3R N610S mutation (c. 1829 A>G) was found at 48% in the bone marrow and 53% in fingernail DNA (Fig. 5A). The patient had no known family history of hematologic malignancies. Together these data suggest that N610 mutations represent a novel leukemia predisposition variant.

Figure 5.

A germline N610S mutation identified in a patient with CML is transforming and therapeutically targetable. A, Sequencing of bone marrow and fingernail on the MSKCC IMPACT panel reveal a germline CSF3R N610S mutation in a patient with chronic phase CML. B, The CSF3R N610S mutation is transforming in the Ba/F3 cytokine–independent growth assay. C, The CSF3R N610H and N610S mutations confer cytokine-independent growth in a mouse bone marrow colony assay. D, N610H and N610S confer serial replating capacity in the absence of cytokines. E, Sensitivity of CSF3R-expressing Ba/F3 cells to the JAK kinase inhibitor ruxolitinib. Ba/F3 cells were grown in the presence (Mig empty and WT CSF3R) and absence (N610H, N610S, and T618I) of the cytokine IL3 in 96-well format. Cell lines were plated in triplicate and were treated with a dose curve of ruxolitinib. After 72 hours, cell viability/proliferation was measured using a tetrazolium-based CellTiter Aqueous One Solution and read on a plate reader. Viability is represented as a percentage of the untreated control. F, Sensitivity of cells expressing CSF3R mutants to the MEK inhibitor trametinib as outlined in E. Parental (untransformed cells), Mig empty, empty vector control; N610H, the CSF3R N610H mutation; N610S, the CSF3R N610S mutation; T618I, the CSF3R T618I mutation. VAF, variant allele frequency.

Figure 5.

A germline N610S mutation identified in a patient with CML is transforming and therapeutically targetable. A, Sequencing of bone marrow and fingernail on the MSKCC IMPACT panel reveal a germline CSF3R N610S mutation in a patient with chronic phase CML. B, The CSF3R N610S mutation is transforming in the Ba/F3 cytokine–independent growth assay. C, The CSF3R N610H and N610S mutations confer cytokine-independent growth in a mouse bone marrow colony assay. D, N610H and N610S confer serial replating capacity in the absence of cytokines. E, Sensitivity of CSF3R-expressing Ba/F3 cells to the JAK kinase inhibitor ruxolitinib. Ba/F3 cells were grown in the presence (Mig empty and WT CSF3R) and absence (N610H, N610S, and T618I) of the cytokine IL3 in 96-well format. Cell lines were plated in triplicate and were treated with a dose curve of ruxolitinib. After 72 hours, cell viability/proliferation was measured using a tetrazolium-based CellTiter Aqueous One Solution and read on a plate reader. Viability is represented as a percentage of the untreated control. F, Sensitivity of cells expressing CSF3R mutants to the MEK inhibitor trametinib as outlined in E. Parental (untransformed cells), Mig empty, empty vector control; N610H, the CSF3R N610H mutation; N610S, the CSF3R N610S mutation; T618I, the CSF3R T618I mutation. VAF, variant allele frequency.

Close modal

To test whether the serine substitution was also transforming at the 610 position, we ran a cytokine-independent growth assay as described above. In this assay, the N610S mutation was robustly transforming and allowed for a similar growth capacity as the N610H and T618I mutations in CSF3R (Fig. 5B). To confirm the transforming potential of this mutation and assess ligand independence we performed a colony-forming unit assay. In this assay, mouse bone marrow is transduced with a retroviral vector expressing WT CSF3R, mutant CSF3R, or an empty vector control (mig empty). The cells are then plated in methylcellulose without any added cytokine support. WT CSF3R produces very few colonies but the CSF3R T618I mutant, which is able to signal in the absence of ligand produces abundant colonies. Both N610H and N610S were able to induce colony formation at similar levels as T618I, indicating that they are robustly oncogenic (Fig. 5C). Furthermore, the N610 mutations allow for enhanced replating of hematopoietic progenitors in colony-forming unit assays, similar to the T618I mutation (Fig. 5D). Together, these data indicate that CSF3R N610H and N610S mutations are oncogenic.

Therapeutic relevance of CSF3R N610 substitutions

We previously identified the JAK kinase inhibitor, ruxolitinib, as a potential therapeutic strategy for patients with CSF3R mutations (31). A recent study identified the MEK inhibitor, trametinib, as being efficacious in a mouse bone marrow transplant model of the CSF3R T618I mutation (32). We tested the ability of both ruxolitinib and trametinib to inhibit the viability of Ba/F3 cells expressing the N610H and N610S mutations and found both mutations to be as sensitive to the two inhibitors as the common T618I mutation (Fig. 5E and F). The WT CSF3R and Mig empty vector control cells are grown in medium containing IL3 for the drug studies (as they would die in its absence), which is known to exert its prosurvival effects on Ba/F3 cells through the JAK–STAT pathway. The growth inhibition of these cells by ruxolitinib is an on-target effect. Taken together, the similar properties of N610 and T618 substitutions, along with similar drug sensitivity, indicate that the CSF3R N610H and N610S variants, although rare, are clinically targetable mutations.

In the age of widespread sequencing of samples from patients with hematologic malignancies, a lack of functional annotation for less common variants represents a major challenge. In this study, we employ functional and biochemical analysis of rare germline variants of CSF3R to show that a loss of glycosylation at N610 accompanies activation of the receptor. Furthermore, this study provides evidence that these clinically-identified N610 mutations, although rare, are likely therapeutically targetable. Mechanistically, these substitutions highlight the importance of membrane-proximal N-glycosylation for regulation of CSF3R activity. CSF3R mutations are the first example of a cancer-associated mutation that alters glycosylation in a site-specific manner to cause oncogenic transformation.

CSF3R mutations are the defining genetic feature of CNL (2, 31), but can also occur in other hematologic malignancies. CSF3R mutations have been previously implicated in the development of severe congenital neutropenia associated acute myeloid leukemia (SCN-AML; refs. 33–36). CSF3R mutations are also found at lower frequency (0.5%–8%) in de novo or secondary AML (3, 5, 6, 31, 37). In CNL, SCN-AML, and AML, the main classes of CSF3R mutations are membrane-proximal point mutations (e.g., T618I and T615A) and truncation mutations in the cytoplasmic domain (e.g., Q741 and S783fs), both of which are thought to be somatic mutations. We were therefore surprised to identify a germline N610H CSF3R mutation in a patient with unusual myelofibrosis. We subsequently identified another patient with a different germline substitution (N610S) at this same site. It is essential to understand whether these mutational events in CSF3R are oncogenic and have therapeutic relevance. The N610H and N610S mutations are highly transformative in both cell line studies (Fig. 2) and in primary murine bone marrow (Fig. 5). The germline nature of the N610H/S mutations identified in this study and the familial germline T640N mutations identified in a family with neutrophilia (11) suggest that additional mutations are likely necessary for the full clinical phenotype of CNL.

Emerging evidence suggests that protein glycosylation can affect signaling, especially during oncogenesis. Takahasi and coworkers described the role of glycans in the function of EGFR (38). Proper N-glycosylation is required for ligand-binding (39); once EGFR is glycosylated, it binds specific N-glycan ligands. For example, on N420 of EGFR is involved in dimerization (40) and highlights that changes in glycosylation can have profound effects on the signaling capacity and activation of receptors. For CSF3R, WT protein treated with N-glycosidases results in a 17-kDA band shift, indicating glycosylation in a significant portion of the protein (24). When CSF3R is mutated, as with T618I, it loses O-linked glycosylation in the membrane-proximal region of CSF3R, increases ligand-independent receptor activation, and leads to oncogenesis. Although interplay is known between neighboring phosphorylation sites (41), this has not yet been studied in proximal N- and O-linked glycosylation sites. MS analysis of the CSF3R T618I–mutant protein showed that the N610 site was still occupied. One of the glycan structures was the same between the mutant and WT. Although this result does not rule out subtle differences in N-linked glycosylation at this site, it does indicate that the T618I mutation does not grossly alter the occupancy of the nearby glycosite.

In this study, we identify CSF3R N610H as a rare therapeutically relevant germline mutation in myeloid leukemia. The germline nature of these mutations suggests that they may represent leukemia predisposition variants. Furthermore, studies will be needed to determine whether CSF3R N610 mutations are recurrent in familial leukemia. CSF3R N610 mutations are highly activating and cause ligand-independence with increased phosphorylation of STAT3. We verified that both ruxolitinib and trametinib inhibit the viability of Ba/F3 cells expressing the CSF3R N610H and N610S mutations and found them to be as sensitive to both compounds as the well-characterized CSF3R T618I mutation. Furthermore, these mutations reveal the critical importance of membrane-proximal N-linked glycosylation for the maintenance of ligand dependency. This study highlights how careful investigation of cancer-associated mutations can provide critical insight into the relationship between protein structure and function.

M.J. Mauro is a consultant/advisory board member for BMS, Novartis, Pfizer, and Takeda. No potential conflicts of interest were disclosed by the other authors.

Conception and design: D.R. Spiciarich, S.L. Thompson, C.R. Bertozzi, J.E. Maxson

Development of methodology: D.R. Spiciarich, S.L. Thompson

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): D.R. Spiciarich, S.T. Oh, A.Foley, S.B. Hughes, M.J. Mauro, O. Abdel-Wahab, R. Viner, S.L. Thompson, Q.Chen, P. Azadi, J.E. Maxson

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): D.R. Spiciarich, M.J. Mauro, O. Abdel-Wahab, R.D. Press, R. Viner, S.L. Thompson, Q.Chen, P. Azadi, J.E. Maxson

Writing, review, and/or revision of the manuscript: D.R. Spiciarich, S.T. Oh, M.J. Mauro, O. Abdel-Wahab, R.D. Press, Q.Chen, C.R. Bertozzi, J.E. Maxson

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): D.R. Spiciarich

Study supervision: P. Azadi, J.E. Maxson

We would like to thank Stacy Malaker, Benjamin Schumann, and Zachary Schonrock for technical assistance and Anthony T. Iavarone at the QB3/Chemistry MS Facility at UC Berkeley for proteomic design assistance and Stephen P. Dudek for critical manuscript feedback.

This research was supported by NCI R00-CA190605 (to J.E. Maxson) and NIH R01 CA200423 and U01 CA207701 (to C.R. Bertozzi). D.R. Spiciarich was supported by a NSF predoctoral fellowship. UC Berkeley QB3/Chemistry Mass Spectrometry Facility receives NIH support (1S10OD020062-01). This research was supported in part by the NIH-funded Research Resource for Biomedical Glycomics (No. 4P41GM103490-14) and 1S10OD018530 to Parastoo Azadi at the Complex Carbohydrate Research Center.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Maxson
JE
,
Gotlib
J
,
Pollyea
DA
,
Fleischman
AG
,
Agarwal
A
,
Eide
CA
, et al
Oncogenic CSF3R mutations in chronic neutrophilic leukemia and atypical CML
.
N Engl J Med
2013
;
368
:
1781
90
.
2.
Pardanani
A
,
Lasho
TL
,
Laborde
RR
,
Elliott
M
,
Hanson
CA
,
Knudson
RA
, et al
CSF3R T618I is a highly prevalent and specific mutation in chronic neutrophilic leukemia
.
Leukemia
2013
;
27
:
1870
3
.
3.
Beekman
R
,
Valkhof
M
,
van Strien
P
,
Valk
PJ
,
Touw
IP
. 
Prevalence of a new auto-activating colony stimulating factor 3 receptor mutation (CSF3R-T595I) in acute myeloid leukemia and severe congenital neutropenia
.
Haematologica
2013
;
98
:
e62
3
.
4.
Maxson
JE
,
Ries
RE
,
Wang
YC
,
Gerbing
RB
,
Kolb
EA
,
Thompson
SL
, et al
CSF3R mutations have a high degree of overlap with CEBPA mutations in pediatric AML
.
Blood
2016
;
127
:
3094
8
.
5.
Zhang
Y
,
Wang
F
,
Chen
X
,
Zhang
Y
,
Wang
M
,
Liu
H
, et al
CSF3R Mutations are frequently associated with abnormalities of RUNX1, CBFB, CEBPA, and NPM1 genes in acute myeloid leukemia
.
Cancer
2018
;
124
:
3329
38
.
6.
Lavallee
VP
,
Krosl
J
,
Lemieux
S
,
Boucher
G
,
Gendron
P
,
Pabst
C
, et al
Chemo-genomic interrogation of CEBPA mutated AML reveals recurrent CSF3R mutations and subgroup sensitivity to JAK inhibitors
.
Blood
2016
;
127
:
3054
61
.
7.
Su
L
,
Tan
Y
,
Lin
H
,
Liu
X
,
Yu
L
,
Yang
Y
, et al
Mutational spectrum of acute myeloid leukemia patients with double CEBPA mutations based on next-generation sequencing and its prognostic significance
.
Oncotarget
2018
;
9
:
24970
9
.
8.
Ward
AC
,
van Aesch
YM
,
Schelen
AM
,
Touw
IP
. 
Defective internalization and sustained activation of truncated granulocyte colony-stimulating factor receptor found in severe congenital neutropenia/acute myeloid leukemia
.
Blood
1999
;
93
:
447
58
.
9.
Beekman
R
,
Valkhof
MG
,
Sanders
MA
,
van Strien
PM
,
Haanstra
JR
,
Broeders
L
, et al
Sequential gain of mutations in severe congenital neutropenia progressing to acute myeloid leukemia
.
Blood
2012
;
119
:
5071
7
.
10.
Arber
DA
,
Orazi
A
,
Hasserjian
R
,
Thiele
J
,
Borowitz
MJ
,
Le Beau
MM
, et al
The 2016 revision to the World Health Organization classification of myeloid neoplasms and acute leukemia
.
Blood
2016
;
127
:
2391
405
.
11.
Plo
I
,
Zhang
Y
,
Le Couedic
JP
,
Nakatake
M
,
Boulet
JM
,
Itaya
M
, et al
An activating mutation in the CSF3R gene induces a hereditary chronic neutrophilia
.
J Exp Med
2009
;
206
:
1701
7
.
12.
Maxson
JE
,
Luty
SB
,
MacManiman
JD
,
Paik
JC
,
Gotlib
J
,
Greenberg
P
, et al
The colony-stimulating factor 3 receptor T640N mutation is oncogenic, sensitive to JAK inhibition, and mimics T618I
.
Clin Cancer Res
2016
;
22
:
757
64
.
13.
Pinho
SS
,
Reis
CA
. 
Glycosylation in cancer: mechanisms and clinical implications
.
Nat Rev Cancer
2015
;
15
:
540
55
.
14.
Maxson
JE
,
Luty
SB
,
MacManiman
JD
,
Abel
ML
,
Druker
BJ
,
Tyner
JW
. 
Ligand independence of the T618I mutation in the colony-stimulating factor 3 receptor (CSF3R) protein results from loss of O-linked glycosylation and increased receptor dimerization
.
J Biol Chem
2014
;
289
:
5820
7
.
15.
Weerapana
E
,
Imperiali
B
. 
Asparagine-linked protein glycosylation: from eukaryotic to prokaryotic systems
.
Glycobiology
2006
;
16
:
91R
101R
.
16.
Imperiali
B
,
Rickert
KW
. 
Conformational implications of asparagine-linked glycosylation
.
Proc Natl Acad Sci U S A
1995
;
92
:
97
101
.
17.
Abdel-Wahab
O
,
Adli
M
,
LaFave
LM
,
Gao
J
,
Hricik
T
,
Shih
AH
, et al
ASXL1 mutations promote myeloid transformation through loss of PRC2-mediated gene repression
.
Cancer Cell
2012
;
22
:
180
93
.
18.
Gelsi-Boyer
V
,
Trouplin
V
,
Adelaide
J
,
Bonansea
J
,
Cervera
N
,
Carbuccia
N
, et al
Mutations of polycomb-associated gene ASXL1 in myelodysplastic syndromes and chronic myelomonocytic leukaemia
.
Br J Haematol
2009
;
145
:
788
800
.
19.
Gelsi-Boyer
V
,
Trouplin
V
,
Roquain
J
,
Adelaide
J
,
Carbuccia
N
,
Esterni
B
, et al
ASXL1 mutation is associated with poor prognosis and acute transformation in chronic myelomonocytic leukaemia
.
Br J Haematol
2010
;
151
:
365
75
.
20.
Bejar
R
,
Stevenson
K
,
Abdel-Wahab
O
,
Galili
N
,
Nilsson
B
,
Garcia-Manero
G
, et al
Clinical effect of point mutations in myelodysplastic syndromes
.
N Engl J Med
2011
;
364
:
2496
506
.
21.
Thol
F
,
Friesen
I
,
Damm
F
,
Yun
H
,
Weissinger
EM
,
Krauter
J
, et al
Prognostic significance of ASXL1 mutations in patients with myelodysplastic syndromes
.
J Clin Oncol
2011
;
29
:
2499
506
.
22.
Maxson
JE
,
Tyner
JW
. 
Genomics of chronic neutrophilic leukemia
.
Blood
2017
;
129
:
715
22
.
23.
Elliott
MA
,
Pardanani
A
,
Hanson
CA
,
Lasho
TL
,
Finke
CM
,
Belachew
AA
, et al
ASXL1 mutations are frequent and prognostically detrimental in CSF3R-mutated chronic neutrophilic leukemia
.
Am J Hematol
2015
;
90
:
653
6
.
24.
Haniu
M
,
Horan
T
,
Arakawa
T
,
Le
J
,
Katta
V
,
Hara
S
, et al
Disulfide structure and N-glycosylation sites of an extracellular domain of granulocyte-colony stimulating factor receptor
.
Biochemistry
1996
;
35
:
13040
6
.
25.
Bull
C
,
Stoel
MA
,
den Brok
MH
,
Adema
GJ
. 
Sialic acids sweeten a tumor's life
.
Cancer Res
2014
;
74
:
3199
204
.
26.
Saxon
E
. 
Cell surface engineering by a modified Staudinger reaction
.
Science
2000
;
287
:
2007
10
.
27.
Spiciarich
DR
,
Nolley
R
,
Maund
SL
,
Purcell
SC
,
Herschel
J
,
Iavarone
AT
, et al
Bioorthogonal labeling of human prostate cancer tissue slice cultures for glycoproteomics
.
Angew Chem Int Ed
2017
;
56
:
8992
7
.
28.
Wratil
PR
,
Horstkorte
R
,
Reutter
W
. 
Metabolic glycoengineering with N-acyl side chain modified mannosamines
.
Angew Chem Int Ed Engl
2016
;
55
:
9482
512
.
29.
Rostovtsev
VV
,
Green
LG
,
Fokin
VV
,
Sharpless
KB
. 
A stepwise Huisgen cycloaddition process: copper(I)-catalyzed regioselective "ligation" of azides and terminal alkynes
.
Angew Chem Int Ed
Engl 
2002
;
41
:
2596
9
.
30.
Giansanti
P
,
Tsiatsiani
L
,
Low
TY
,
Heck
AJR
. 
Six alternative proteases for mass spectrometry-based proteomics beyond trypsin
.
Nat Protoc
2016
;
11
:
993
1006
.
31.
Maxson
JE
,
Gotlib
J
,
Pollyea
DA
,
Fleischman
AG
,
Agarwal
A
,
Eide
CA
, et al
Oncogenic CSF3R mutations in chronic neutrophilic leukemia and atypical CML
.
N Engl J Med
2013
;
368
:
1781
90
.
32.
Rohrabaugh
S
,
Kesarwani
M
,
Kincaid
Z
,
Huber
E
,
Leddonne
J
,
Siddiqui
Z
, et al
Enhanced MAPK signaling is essential for CSF3R-induced leukemia
.
Leukemia
2017
;
31
:
1770
8
.
33.
Dong
F
,
Brynes
RK
,
Tidow
N
,
Welte
K
,
Lowenberg
B
,
Touw
IP
. 
Mutations in the gene for the granulocyte colony-stimulating-factor receptor in patients with acute myeloid leukemia preceded by severe congenital neutropenia
.
N Engl J Med
1995
;
333
:
487
93
.
34.
Dong
F
,
Dale
DC
,
Bonilla
MA
,
Freedman
M
,
Fasth
A
,
Neijens
HJ
, et al
Mutations in the granulocyte colony-stimulating factor receptor gene in patients with severe congenital neutropenia
.
Leukemia
1997
;
11
:
120
5
.
35.
Tidow
N
,
Pilz
C
,
Teichmann
B
,
Muller-Brechlin
A
,
Germeshausen
M
,
Kasper
B
, et al
Clinical relevance of point mutations in the cytoplasmic domain of the granulocyte colony-stimulating factor receptor gene in patients with severe congenital neutropenia
.
Blood
1997
;
89
:
2369
75
.
36.
Skokowa
J
,
Steinemann
D
,
Katsman-Kuipers
JE
,
Zeidler
C
,
Klimenkova
O
,
Klimiankou
M
, et al
Cooperativity of RUNX1 and CSF3R mutations in severe congenital neutropenia: a unique pathway in myeloid leukemogenesis
.
Blood
2014
;
123
:
2229
37
.
37.
Maxson
JE
,
Ries
RE
,
Wang
YC
,
Gerbing
RB
,
Kolb
EA
,
Thompson
SL
, et al
CSF3R mutations have a high degree of overlap with CEBPA mutations in pediatric AML
.
Blood
2016
;
127
:
3094
8
.
38.
Takahashi
M
,
Hasegawa
Y
,
Gao
C
,
Kuroki
Y
,
Taniguchi
N
. 
N-glycans of growth factor receptors: their role in receptor function and disease implications
.
Clin Sci (Lond)
2016
;
130
:
1781
92
.
39.
Liu
YC
,
Yen
HY
,
Chen
CY
,
Chen
CH
,
Cheng
PF
,
Juan
YH
, et al
Sialylation and fucosylation of epidermal growth factor receptor suppress its dimerization and activation in lung cancer cells
.
Proc Natl Acad Sci U S A
2011
;
108
:
11332
7
.
40.
Tsuda
T
,
Ikeda
Y
,
Taniguchi
N
. 
The Asn-420-linked sugar chain in human epidermal growth factor receptor suppresses ligand-independent spontaneous oligomerization. Possible role of a specific sugar chain in controllable receptor activation
.
J Biol Chem
2000
;
275
:
21988
94
.
41.
Schweiger
R
,
Linial
M
. 
Cooperativity within proximal phosphorylation sites is revealed from large-scale proteomics data
.
Biol Direct
2010
;
5
:
6
.
42.
Damerell
D
,
Ceroni
A
,
Maass
K
,
Ranzinger
R
,
Dell
A
,
Haslam
SM
. 
Annotation of glycomics MS and MS/MS spectra using the GlycoWorkbench software tool
.
Methods Mol Biol
2015
;
1273
:
3
15
.
43.
Ceroni
A
,
Maass
K
,
Geyer
H
,
Geyer
R
,
Dell
A
,
Haslam
SM
. 
GlycoWorkbench: a tool for the computer-assisted annotation of mass spectra of glycans
.
J Proteome Res
2008
;
7
:
1650
9
.