Trends in oligomannosylation and α1,2-mannosidase expression in human cancers

Aberrant protein glycosylation is a prominent cancer feature. While many tumour-associated glycoepitopes have been reported, advances in glycoanalytics continue to uncover new associations between glycosylation and cancer. Guided by a comprehensive literature survey suggesting that oligomannosylation (Man5–9 GlcNAc2) is a widespread and often regulated glycosignature in human cancers, we here revisit a valuable compilation of nearly 500 porous graphitized carbon LC-MS/MS N-glycomics datasets acquired across 11 human cancer types to systematically test for oligomannose-cancer associations. Firstly, the quantitative glycomics data obtained across 34 cancerous cell lines demonstrated that oligomannosylation is a pan-cancer feature spanning in a wide abundance range. In keeping with literature, our quantitative glycomics data of tumour and matching control tissues and new MALDI-MS imaging data of tissue microarrays showed a strong cancer-associated elevation of oligomannosylation in both basal cell (p = 1.78 × 10–12) and squamous cell (p = 1.23 × 10–11) skin cancer and colorectal cancer (p = 8.0 × 10–4). The glycomics data also indicated that some cancer types including gastric and liver cancer exhibit unchanged or reduced oligomannose levels, observations also supported by literature and MALDI-MS imaging data. Finally, expression data from public cancer repositories indicated that several α1,2-mannosidases are regulated in tumour tissues suggesting that these glycan-processing enzymes may contribute to the cancer-associated modulation of oligomannosylation. This omics-centric study has compiled robust glycomics and enzyme expression data revealing interesting molecular trends that open avenues to better understand the role of oligomannosylation in human cancers.


Human cancer cell lines
Human cell lines were obtained from the American Type Culture Collection (ATCC) (Manassas, VA, USA), Cell Bank Australia (CBA) (Sydney, Australia), Cell Line Services (CLS) (Eppelheim, Germany), Garvan Institute of Medical Research (Sydney, Australia) and from multiple other sources as indicated in Supplementary Tables 1 and 2. These cell lines were collected and studied over many years by many past and present members and collaborators of our glycoanalytical laboratories as described previously [1]. The cell lines were generally stored at −140°C in 90% (v/v) FBS and 10% (v/v) dimethyl sulfoxide (DMSO) until use. All cultured cells were confirmed mycoplasma-free prior to the N-glycome profiling.

Tumour and non-tumour tissues from cancer patient cohorts
Tissue samples were obtained from multiple cancer patient cohorts or from various control individuals suffering from other non-cancer conditions at a range of hospitals and clinics as described in Supplementary Tables  3 and 4. Many of these tissue samples have been described in recent publications [2][3][4][5][6][7][8][9][10][11]. In brief, tumourigenic and non-tumourigenic tissues were obtained by surgery by trained clinicians after obtaining written and informed patient consent, see above and Supplementary Table 3 for ethics details.

Cell culture conditions
The human cancer cell lines were thawed and cultured in DMEM, RPMI 1640, MEM or MCDB 105 media in Corning cell culture flasks (see Supplementary  Table 1 for an overview of culture conditions). These culture media differed slightly in their nutritional composition, but all contained essential components including glucose, L-glutamine, sodium pyruvate, sodium bicarbonate, phenol red as pH indicator and salts. The media were supplemented with 10% (v/v) FBS, 1% (v/v) penicillin and 100 U/ml streptomycin, and, in some cases, 25 U/l insulin, 1 µM α-thioglycerol, 1 mg/ml hydrocortisone and nonessential amino acids. FBS-free media was used to obtain the secretome of some cultured cells (see below). All cells were grown at 37°C, 5% CO 2 . The cell viabilities were regularly determined using trypan blue exclusion, counted using an automated cell counter (BioRad) and visualized using an inverted light microscope (Olympus) at 10-40x magnification to monitor the morphology of the cells. The cells were sub-cultured at 90-95% confluency typically every 72-96 h [12].

Sub-culturing of non-adherent cancer cells
Suspension cells were collected by centrifugation of the culture media at 500 × g for 10 min at 20°C. The cell pellet fractions were re-suspended in culture media and transferred to new flasks. Fresh media were added every 48 h.

Sub-culturing of adherent cancer cells
The culture media were removed, and the cells were washed with sterile-filtered phosphate buffered saline (PBS) and detached by a brief treatment with 0.25% (w/v) trypsin at 37°C. Fresh media was added (1:1, v/v) to quench the activity of trypsin and the cells were transferred to new flasks. Fresh media were added every 48 h.
For all cultured cells, the cell pellets were thoroughly washed and re-suspended in cold PBS to remove traces of FBS and then centrifuged prior to storage at −30°C until being used for analysis.

Tissue sample handling
The below describes the sample handling of the tissues obtained from patients suffering from different cancers investigated in this study.

Chronic lymphocytic leukemia (CLL)
Frozen CLL samples (~20 × 10 6 mononuclear cells) were thawed in a 37°C water bath, mixed with MACS buffer in a polypropylene round-bottom tube and centrifuged at 400 × g for 5 min at 20°C. The resulting cell pellets were re-suspended in MACS buffer, magnetically isolated using CD19-conjugated microbeads and incubated at room temperature for 15 min. The samples were again centrifuged, cell pellets were re-suspended in MACS buffer, and the B cells were isolated on MACS LS columns. The column was washed with MACS buffer to remove any unbound cells and then removed from the magnet and placed on a 15 ml CellStar tube. MACS buffer was loaded into the column and flushed out immediately by firmly pushing the plunger into the column to expel the liquid containing cells with microbeads. The process was repeating several times. The tubes were centrifuged as described above. The resulting cell pellets were lysed using probe sonication (5 s cycles, 20% amplitude) on ice and proteins were extracted by chloroform-watermethanol extraction (see below).

Basal cell carcinoma (BCC) and squamous cell carcinoma (SCC)
To allow for an accurate dermato-pathological investigation, only patients exhibiting a well-circumscribed BCC or SCC of at least 5 mm diameter were included in the study. The BCC and SCC tissues were obtained by fusiform excision with micrographic control of the margins under local anesthesia. Punch biopsies were obtained from the tumourigenic and normal regions of the excised tissues, snap-frozen in liquid nitrogen and stored at −80°C. Prior to use, the samples were washed three times with 70% (v/v) ethanol and then three times with 50 mM ammonium bicarbonate. The samples were homogenized in a lysis buffer containing 50 mM ammonium bicarbonate, 1 M urea, 10% (v/v) acetonitrile (ACN) and 0.1% (w/v) sodium dodecyl sulfate (SDS) using a IKA T10 Ultraturrax homogenizer (Staufen, Germany) and then sonicated using a tabletop Branson sonifier B-12 sonicator for 30 s. To increase the protein/peptide solubility, 4 μg trypsin was added, and the samples were incubated overnight at 37°C before insoluble material/particles were removed by centrifugation at 14,000 × g for 30 min. The supernatant fractions were collected, and the protein complement fractions were isolated by chloroform-water-methanol extraction (see below).

Hepatocellular carcinoma (HCC)
Fresh liver tumour tissues were micro-dissected and prepared for hematoxylin-eosin (H&E) staining to enable a tissue sampling containing at least 70% tumour cell content [13] before being snap-frozen in liquid nitrogen and stored at −80°C. Prior to use, the frozen tissue blocks were homogenized on ice using an IKA T10 Ultraturrax homogenizer (Staufen, Germany) in a lysis buffer containing 50 mM Tris-HCl (pH 7.4), 100 mM NaCl, and 1 mM ethylenediaminetetraacetic acid (EDTA) with 5 mM EDTA-free protease inhibitor tablets. The solutions were then sonicated on ice using a tabletop Branson sonifier B-12 sonicator for 10-30 s and centrifuged at 2,000 × g for 20 min at room temperature. The supernatant fractions were collected and the proteins isolated using microsomal preparation (see below) [14].
The formalin-fixed paraffin-embedded (FFPE) tissue samples from BCC, SCC and liver were treated with 10% (v/v) formalin for 24 h followed by washing with 70%, 95% and 100% (all v/v) ethanol (2 × 1 h each), xylene (2 × 1 h at 37°C) and then liquid paraffin (3 × 1 h at 60°C). Paraffin tissue blocks were cut into 2-10 µm thick sections using a standard sliding manual microtome (Microm, Dreieich, Germany) and transferred into 1.5 ml sample tubes, washed with xylene (2x speed for 5 min) and absolute ethanol (2x speed for 5 min) and dried. The samples were then homogenized on ice in lysis buffer containing 100 mM Tris-HCl (pH 8.0) and 100 mM dithiothreitol (DTT) and sonicated using Branson sonifier B-12 sonicator for 10-30 s. SDS (4% (w/v), final concentration) was added to the samples and incubated at 99°C for 1 h under mild agitation, cooled to ambient temperature, centrifuged at 2,000 × g for 20 min, and the resulting supernatant fractions were collected [15].

Gastric cancer (GC)
The gastric cancer patients underwent a surgical procedure with a complete resection of the tumour. From each donor, tissue samples of tumour mucosa and adjacent normal mucosa were collected and snap-frozen in liquid nitrogen (~50 mg wet weight). The tissue samples were mixed on ice in a solubilization buffer containing 7 M urea, 2 M thiourea, 40 mM Tris-HCl and a protease inhibitor cocktail mixture (1:100) and homogenized using a Precellys 24 (two cycles at power 5500, 30 s/ cycle) in 2 ml reinforced homogenization tubes (Bertin Technologies, Sweden). The samples were mixed with another round of solubilization buffer with the addition of 0.5% (w/v) SDS and 10 mM DTT and left for overnight incubation on a shaker at 4°C. After overnight incubation, 25 mM iodoacetamide (IAA) was added and incubated for 40 min in the dark at room temperature. The tissue extracts were centrifuged at 19,500 × g for 20 min. The protein-containing supernatant fractions were diluted ten times with 20 mM sodium bicarbonate, applied to a preconditioned 10 kDa cut-off filter (Pall, Port Washington, USA) and washed three times in the same buffer.

Colorectal cancer (CRC)
CRC tumour samples obtained from Sydney South West Area Health Service, Australia were collected from patients undergoing surgical resection [16]. Pathological examination followed the Australian Clinico-Pathological Staging protocol for CRC, which is compatible with other staging systems e.g., the TNM (tumours/nodes/ metastases) [4]. All samples were classified as T3N0M0 stage in the TNM classification. The collected CRC tissue samples were stored in Hanks' balanced salt solution at 4°C for approximately 6 h to maintain cell viability. The samples were cut into 2 mm strips and incubated in RPMI 1640 medium containing 2% (v/v) collagenase type 4 (Worthington, USA) and 0.2% (v/v) deoxyribonuclease I from bovine pancreas for 60 min at 37°C on a shaker. The semi-digested tissue was then forced gently through a fine wire mesh strainer using the plunger of a 20 ml syringe and then washed in HBSS. The resulting cell suspension was concentrated in two rounds using 200 µm and 50 µm filters to remove cell aggregates. The cell suspensions were stored at -80°C in heat-inactivated 90% (v/v) FBS and 10% (v/v) DMSO. Thawed cell suspensions were treated with 0.1% (w/v) deoxyribonuclease I and enriched for epithelial cell fractions by using immuno-magnetic beads conjugated to an antibody against the epithelial molecular marker EpCAM (Dynabead CELLection Epithelial Enrich, Invitrogen). The beads and cells were mixed and incubated for 1 h at 4°C and washed thoroughly with PBS. The proteins were extracted by acetone precipitation (see below).
CRC primary tissue samples obtained from the Department of Pathology in Yonsei University, South Korea, were diagnosed and staged for CRC by a trained pathologist at the Severance Hospital, South Korea. All CRC primary tumours were adenocarcinomas obtained from five male patients spanning differences in age and CRC pathology i.e., varying sites (sigmoid, transverse and rectum), TNM stages (I-IV), and epidermal growth factor receptor (EGFR) expression status [5,6,17]. The microsomal protein fractions were extracted from the tissue samples as described below.

Prostate cancer (PCa)
All biopsy cores were examined for signs of PCa by a trained pathologist. Fresh tissues were collected after radical prostatectomy from a cohort of 55 patients, which included ten patients from each disease stage (PCa grade 1-5) and five individuals presenting with benign prostatic hyperplasia (BPH). The PCa grade was defined using the Modified Gleason Grading System proposed by the International Society of Urological Pathology in 2005, a system that was revised in 2014 [18,19]. Fresh prostate tissues (~60 mg), stored in RNALater (stabilization and storage solution), were washed in 80% (v/v) cold ACN and re-suspended in an extraction buffer containing 6 M urea, 10 mM DTT, 1 mM sodium fluoride, 1 mM sodium orthovanadate and a protease inhibitor cocktail (1:10). Prostate tissues were rapidly lysed (2 min, 30 Hz with a 5 mm stainless steel bead) using a TissueLyser (Qiagen, Chadstone, VIC, Australia). Protein concentrations were determined by the Qubit fluorimetric detection method. The protein samples were directly used for N-glycomics.

Protein extracts from cellular fractions
As briefly described below, protein extracts from different cellular fractions were obtained from the investigated cell lines and tissue samples.

Secretome (S)
Cells grown to high (>95%) confluency were washed at least four times with ice-cold PBS to remove FBS and incubated in FBS-free media at 37°C in 5% CO 2 for 28-48 h. The media containing the secreted proteins were collected and centrifuged at 2,000 × g to pellet any cells or cellular debris. The proteins in the supernatant fractions (hereafter the "secretome") were transferred to new vials and concentrated using a 10 kDa molecular weight cut-off Amicon Ultra Centrifugal filter device (Millipore, USA) and stored in 4°C until use.

Whole cell lysate (WCL)
Cell pellets (see above) were gently thawed on ice, re-suspended in radioimmunoprecipitation assay lysis buffer, vigorously vortexed and centrifuged at 14,000 × g, 4°C for 10-20 min. The resulting supernatant fractions (the "whole cell lysates") containing the cellular proteins were collected and stored at 4°C until use.

Microsomal fraction (MF)
This preparation follows a previously published protocol [20]. In short, cell pellets were re-suspended in 25 mM Tris-HCl (pH 7.4), 150 mM NaCl, and 1 mM EDTA with 5 mM EDTA-free protease inhibitor tablets. The cell suspension was ultra-sonicated using Sonifier 450 on ice over three cycles of 10 s bursts and then centrifuged at 2,000 × g, 4°C for 20 min. The resulting supernatant was ultra-centrifuged at 120,000 × g, 4°C for 80 min. The resulting pellet containing the total membrane complement was re-suspended in 25 mM Tris-HCl (pH 7.4), 150 mM NaCl and 1% (v/v) Triton X-114 and phase partitioned on ice for 10 min, and then at 37°C for 20 min. A subsequent gentle centrifugation step at 1,000 × g at 25°C for 10 min created two distinct phases. The dense detergent phase containing the membrane protein complement (the "microsome") was stored in 4°C until use.

Protein isolation Acetone precipitation
The proteins of all three cellular fractions (WCL, MF, and S, see above) from most cells and tissues investigated in this study were precipitated by adding ice-cold acetone in a ratio of 1:4-1:9 (v/v), vortexed thoroughly and incubated overnight at −20°C in an upright position. The samples were then centrifuged at 14,000 × g for 12-15 min and the protein pellets were stored in −30°C until further analysis.

Chloroform-water-methanol extraction
The proteins in the supernatant fractions after cell lysis of CLL tissues (B cells) and non-melanoma BCC tissues were precipitated according to the method used by Wessel and Flugge [21]. Briefly, the supernatant was mixed vigorously with methanol, chloroform and water in a 4:1:3 volume ratio. The samples were centrifuged at 14,000 × g for 5 min and the upper aqueous phase was carefully removed. Additional methanol was added, mixed and centrifuged as above. The supernatant was removed, and dried protein pellet was stored at −20°C until analysis.
The protein pellets were re-suspended in 6-8 M urea and the protein concentrations were determined using the Bradford assay or the bicinchoninic acid assay.

Protein denaturation
Proteins were reduced using 10 mM DTT, 45 min, 56°C, and carbamidomethylated using 25 mM IAA, 30 min in the dark at 20°C before the alkylation reactions were quenched using 30 mM DTT (final concentrations stated).   Tables 8 and 9 for all raw and tabulated glycome data, respectively. Statistics was performed using paired and unpaired two-tailed t tests (n, patient samples, as indicated). * p < 0.05, ** p < 0.001, *** p < 0.0001, n.s., not significant ( Statistics was performed using paired and unpaired two-tailed t tests (n, patient samples, as indicated). * p < 0.05, ** p < 0.001, *** p < 0.0001, n.s. not significant (p ≥ 0.05). The up-(black arrows) and down-regulation/unchanged levels (grey arrows) of oligomannose in tumours based on literature (Table 1) and/ or our own glycomics data (Figure 3 and Supplementary Figure 1) are indicated for each cancer type to aid the data interpretation. The number of affected cases is indicated for each of the four α1,2-mannosidases in each cohort across the investigated cancer types.