Changes in Protein Glycosylation as a Result of Aptamer Interactions with Cancer Cells

2.1 Cell Culture

Breast cancer cells MCF7, purchased form Sigma–Aldrich ECACC (The European Collection of Authenticated Cell Cultures), were cultured in RPMI 1640 (Sigma–Aldrich) supplemented with 10% Fetal Bovine Serum FBS (Life) and 1% antibiotic/antifungal solution (Sigma–Aldrich) at 37 °C in 5% CO₂ humidified atmosphere.

2.2 Aptamers Binding Assay

Aptamers (Microsynth AG) stored at –20 °C were thawed to the room temperature. Stock solutions (25 mm in DPBS buffer) of the oligonucleotide sequences were incubated at 90 °C for 5 min, and then cooled down to room temperature by the natural cooling process to the 37 °C facilitating the proper 3D structures of aptamers during renaturation. Oligonucleotides were diluted in binding buffer (BB; 5 mm MgCl₂, 1 mm CaCl₂, 0.1% ssDNA in DPBS [Sigma–Aldrich]) to a final concentration of 25 nm. Cells were incubated at 37 °C for 45 min with the aptamer sequences A26 and A33 that were found to be highly specific toward MCF7 cells,8 and with A33sc (scrambled sequence) characterized by the lack of specificity to the studied cells (Table 1). Additional control sample was prepared after incubation with BB only. After binding, the excess of aptamers was removed for 2 min with 5 mL of washing buffer (5 mm MgCl₂, 1 mm CaCl₂, 1 mg mL^–1 BSA in DPBS) on rotator at 37 °C and the process was repeated three times. Washing step did not affect viability of the cells, nor binding kinetics between cells and aptamers. The cells were carefully scraped off the flask surface, transferred into a 15 mL Falcon tube, and centrifuged at 200 × g for 3 min. All experiments were repeated three times as three independent biological replicates. After the cells have formed a monolayer on the surface of the T75 dish (Eppendorf), the passage was performed in proportion 1:3. Cells were grown until the proper amount of biological material necessary to complete the experiments was obtained.

2.3 Homogenization

MCF7 cells were homogenized three times by sonication on ice (5 s each) in DPBS buffer with the proteinase inhibitors cocktail (Roche). Homogenates were centrifuged for 45 min at 25 000 × g at 4 °C. Protein concentration was measured using Bradford method (Sigma–Aldrich), according to the producer's protocol.

2.4 Glycoprotein Enrichment

For the identification of N-glycosylation profiles, PHA-L (Vector Laboratories) was selected. PHA-L specifically binds to the variety of highly branched-type N-glycans that are characteristic for many tumor cells.7 Therefore, MCF7 homogenates were incubated with 50 µL of PHA-L lectin linked to agarose for 1 h at room temperature in the presence of 300 µL of 10 mm HEPES buffer, pH 7.5, with 0.15 m NaCl, 0.1 mm CaCl₂, and 0.01 mm MnCl₂ (all from Sigma–Aldrich), followed by 16 h incubation at 4 °C. Successively, suspensions were centrifuged at 25 000 × g for 5 min at 4 °C. Precipitates were washed three times with 1 mL of PBS (Sigma–Aldrich) followed by 5 min incubation at 95 °C with 40 µL of the Laemmli sample buffer (Bio-Rad; protein concentration 250 µg).

2.5 Electrophoretic Separation

1D electrophoresis was performed as follows. Proteins were separated with the aid of the 10% SDS-PAGE in a Mini Protean system (Bio-Rad). The gels were stained with the Sypro Ruby (Thermo Fisher Scientific) for total protein visualization, and the ProQ Diamond (Thermo Fisher Scientific) prior to glycoprotein identification, followed by the nanoLC–MS/MS analysis. Electrophoretic separations were run individually for each biological replicate. Gel scans were performed using the XR+ GelDoc (Bio-Rad).

2.6 Isolation of Peptides

The spots corresponding to glycoproteins visualized by the ProQ Diamond were excised from the gel with a scalpel, chopped into cubes, rinsed with water, and transferred into the LoBind tubes (Eppendorf). Dye was removed with 100 mm NH₄HCO₃ (Sigma–Aldrich), and an equal volume of 100% acetonitrile (ACN; JT Baker) was added after 10 min. Then, the gel pieces were treated with 100% ACN and re-swollen in 5 ng mL^–1 trypsin (Promega) in 50 mm NH₄HCO₃ for 45 min on ice. The supernatants, which were not absorbed by gel particles, were removed and gel pieces were immersed in 50 mm NH₄HCO₃ and incubated overnight at 37 °C. After completion of digestion, the supernatants were transferred into fresh tubes, followed by the addition 100 µL of 50 mm NH₄HCO₃, and after 10 min, an equal volume of 100% ACN was added. The samples were incubated under shaking for 15 min at 37 °C. Extraction of peptides was repeated twice with 5% formic acid (FA; Sigma–Aldrich) (v/v) in 100% ACN, and the combined extracts were evaporated to dryness in a vacuum centrifuge. The resulting peptides were dissolved in 20 µL of 0.1% FA.9

2.7 NanoLC–MS/MS Analysis

The nano liquid chromatography – tandem mass spectrometry (nanoLC–MS/MS) analysis was performed with the aid of the Proxeon capillary chromatography system combined with the AmaZon ETD mass spectrometer (Bruker Daltonics).10 The separation was performed in a capillary column filled with the PepMap reversed-phase material (C18, 15 cm long, 75 µC;m ID, 3 µm particle size, 100 Å pore size [Thermo Fisher Scientific]). The gradient was formed using 0.1% FA in water (solvent A) and 0.1% FA in ACN (solvent B), and it was delivered at a flow rate of 300 nL min^–1. The system was controlled by the Hystar software (Bruker Daltonics). A gradient was produced from 2% to 50% B in 30 min and up to 90% B at 35 min, then kept until 45 min, and again reduced to 2% until 55 min. The AmaZon ETD instrument operated in a positive-ion mode. During analysis, two most intense peaks (threshold above 100 000) in the range 450–1800 m/z were automatically fragmented using CID in the data-dependent acquisition mode. The acquired spectra were analyzed using the Bruker Data Analysis 4.0 software and were identified using the Mascot 2.4.1 algorithm against the Swiss-Prot sequence database 2018_03 (557012 sequences; 199714119 residues). Search parameters were set as follows: taxonomy, human; modification, carbamidomethyl (fixed); methionine dioxidation (variable), up to 1 missed cleavage; peptide charges, +1, +2, and +3; mass tolerance, 0.8 Da for precursor mass and 0.6 Da for fragment mass. Proteins with more than two fragmented peptides detected were considered, and an additional criterion was an ion score above 40. Accumulation time to achieve the number of ions entering the trap was less than 1 ms. Analysis of glycopeptides was performed using the GlycoQuest software (Bruker Daltonics). Search parameters were set as follows: minimal precursor m/z, 600 Da; occurrence of oxonium ions (IonTrapOxoniumIons List) with absolute mass tolerance, 0.3 Da; minimum intensity coverage, 50%; distance-based glycopeptides pattern established with the minimal number of consecutive m/z distances required, 2; MS tolerance, 0.3 Da; MS/MS tolerance, 0.35 Da; peptide mass offset, 204.09 Da.

2.8 Western Blotting

Verification of the MS-identified protein sequences was performed with standard anti-KI13B and anti-PCNA protein antibodies. Moreover, affirmation of the N-glycan attachment to studied proteins was achieved by using the biotinylated PHA-L form, followed by detection with alkaline phosphatase-conjugated streptavidin.

To confirm aptamers binding to their protein targets, we have also used alkaline phosphatase-conjugated streptavidin to detect biotinylated forms of aptamers on the surface of identified proteins. Standard proteins could not be used, because of the lack of information on their glycosylation state, therefore MCF7 cell homogenates after biotin-conjugated aptamers incubation were applied, followed by lectin affinity enrichment.

Blotted membranes, along with the Bio-Rad standard protocol, were treated with casein solution (Vector Laboratories) for 1 h at room temperature on rotator, washed three times with Tris buffered saline TBS (Sigma–Aldrich), and subsequently incubated with anti-KI13B (Thermo Fisher Scientific), anti-PCNA (Thermo Fisher Scientific), and biotinylated PHA-L lectin for 2 h, according to the manufacturer's recommendations. After washing with TBS buffer, the blots were incubated with (Vector Laboratories) for 2 h with TBS supplemented with 0.1% Tween 20. Finally, the blots were visualized by alkaline phosphatase substrate kit II (Vector Laboratories). Blots were documented with the XR+ GelDoc (Bio-Rad).

3.1 Protein Profile Changes

As a result of SDS-PAGE separation, three gels were obtained that were first stained with a ProQ Diamond dye characteristic for glycoproteins, followed by staining with SyproRuby for total protein visualization (Figure 2).

Confirmation of the aptamer–protein target complexes formation was achieved by the Western blotting method, using biotinylated forms of aptamers, followed by the detection with alkaline phosphatase-conjugated streptavidin (Figure 3).

Seven fractions of N-glycoproteins in PHA-L enriched pull-outs were excised and analyzed using capillary LC–MS/MS. It was possible to identify 522 proteins in all types of samples A26, A33, A33sc, and BB (Table S1, Supporting Information).

Fourteen proteins were prevalent in the samples incubated with both aptamers A26 and A33 that were not present in control samples A33sc and BB (Figure 4). Functional characteristics of these proteins were presented in Figure 5.

In the group of identified proteins, there are molecules of TRAP1 Heat shock protein 75 kDa responsible for nucleotide binding, protein folding, response to stress, and apoptosis signaling pathway,11 likewise a membrane component responsible for protein localization and trafficking VP13A (Vacuolar protein sorting-associated protein 13A).12 Also, a co-chaperone, which acts as a regulator of the Hsp70 chaperone machinery that may be involved in the processing of ataxia-linked proteins SACS (Sacsin), has been identified in samples interacting with aptamers A26 and A33.13 Furthermore, a protein that plays an important role in the regulation of stress-induced apoptosis UACA (Uveal autoantigen with coiled-coil domains and ankyrin repeats) has been found in the samples incubated with A26 and A33 aptamers.14 Essentially, all proteins present in the MCF7 protein extracts after incubation with potential therapeutic aptamers sequences and not identified in control samples were described in the literature as cancer-related agents.15, 16

3.2 Glycoproteome Changes

Finally, using MS approach, it has been recognized, for the first time, the presence and localization of the specific N-glycan in the amino acid sequences among identified proteins (Table 2). Four proteins with altered N-oligosaccharides profiles were identified; two of them characteristic for the A26 sequence KI13B and PCNA, and one for the A33 oligonucleotide ANXA2. Also, one protein HS90B was found to be deglycosylated as a result of A26 and A33 interactions.

Table 2. Unique glycosylation localizations in the sequences of identified proteins

Protein name	Peptide sequence	Peptide spectrum	Glycan spectrum	Glycan structure/sample
Kinesin-like protein KI13B	R.TFVNGSSVSSPIQLHHGDR.I 566 AA
Proliferating cell nuclear antigen PCNA	R.NLAMGVNLTSMSK.I 121 AA
Annexin A2 ANXA2	K.GVDEVTIVNILTNR.S 130 AA
Heat shock protein HSP 90-beta HS90B	R.ELISNASDALDK.I 96 AA

The protein identified in our study as uniquely glycosylated is KI13B (Kinesin-like protein KIF13B), which is a phosphoprotein responsible for the reorganization of cytoskeleton, intracellular trafficking, microtubule-based movement, protein targeting, and nucleotide binding.17 Mass of this molecule is more than 200 kDa, and so far, it was localized in the cell interior; however, the presence of the N-glycan on the 566 amino acid residue might bring a light on the new locus of KI13B on the membrane surface.

The presence of N-glycosylated KI13B in A26 sample was confirmed using Western blot analysis in two-step immunodetection; first using antibodies characteristic for their polypeptide sequences, and second occurrence of highly complex antennary N-oligosaccharides by the biotinylated anti-PHAL antibody (Vector Laboratories; Figure 6).

Another protein found to be N-glycosylated is PCNA (Proliferating cell nuclear antigen) that plays a key role in DNA damage response, controls eukaryotic replication, cell proliferation, and epithelial cell differentiation, and was identified in A26 samples. The presence of PCNA proteins with attached N-glycan was confirmed using Western blot approach (Figure 7). Glycosylation of this protein at the 121 AA position may influence its localization and cause cancerous transformation by changing cell differentiation and proliferation mechanisms.18

Annexin 2 ANXA2 was described as an aptamer-binding target in MCF7 cells.8 This membrane-bound protein that is involved in heat stress response, angiogenesis, cell proliferation, and adhesion may cross the membrane as a constituent of the lipid raft assembly; however, it has not been described to occur in the glycosylated form. Using MS-based approach, we have identified, for the first time, the structure of N-glycan at the 130 asparagine residue of Annexin 2 in A33 samples.

HS90B (Heat shock protein HSP 90-beta) is a molecular chaperone that promotes, along with its co-chaperones, the maturation, folding, and proper regulation of specific target proteins involved in cell cycle control, differentiation, gene expression, transcription factors, and signal transduction.19 It was found to be N-glycosylated at the 96 arginine residue in control samples, while the lack of glycan was observed in MCF7 samples incubated with aptamers A26 and A33.

In summary, the strategy described in this work enabled efficient discovery of N-glycosylation patterns and their precise localization in cancerous cells as a result of aptamers interactions.

The presented approach represents an effective identification method to reveal aptamer–protein target interactions that may influence their glycosylation profile. We believe that after formation of aptamer–protein conjugates, the 3D structure is changed, and steric effect might therefore alter the action of glycotransferases, which in turn have an impact on the protein glycosylation state. Changes of the glycosylation profile occur at different stages in diseased cells. The detection of these changes might be critical for understanding the molecular mechanisms underlying pathogenesis, as well as for diagnosing disease states and monitoring therapeutic effects. Aptamers have been shown as candidates for molecular theranostics agents simultaneously serving as therapeutic and diagnostic tools.20