Cell line and culture conditions

Cancer cells and others, CP70 (Culture media: DMEM 10% FBS), A2780 (Culture media DMEM: 10% FBS), SKOV3 (Culture media: McCoy5a 10% FBS), TOV112D (Culture media: MCDB 15% FBS), OVCAR8 (Culture media: DMEM 10% FBS), OV81 (Culture media: DMEM 10% FBS), HEK293T (Culture media: DMEM 10% FBS), Jurkat (Culture media: RPMI 10% FBS) cells were grown in recommended growth media in humidified incubator containing 5% CO2. For each experiment, cells were grown up to a confluence of ~ 70% and treated or analyzed as indicated.

Chemicals and reagents

Cell culture medias were purchased from the Lerner Research Institute Media Core at the Cleveland Clinic. Cisplatin was purchased from the Cleveland Clinic Pharmacy. Cell titer glow reagent (Promega), FBS (Atlas Biologicals), anti-CD55 antibody from Proteintech and EMD Millipore, anti-ZMYND antibody (Proteintech), anti-GAPDH antibody (Proteintech), anti-Lamin A/C antibody (Proteintech), anti-α Tubulin antibody (Proteintech), Goat anti-Rabbit IgG Alexa Fluor 488 (Thermo Scientific), and Goat anti-mouse IgG Alexa Fluor 568 (Thermo Scientific) were used in different studies. Details of chemicals/reagents described in supplementary Table 1.

Generation of wild type CD55 and domain deletion mutant lentiviral plasmids

Wild type CD55 and deletion mutants (Δ1, Δ2, Δ3, Δ4, Δ1234, ΔS/T, Δ34, and Δ34S/T) were generated using D-TOPO cloning method. Briefly, the gene block of CD55 cDNA was purchased from Integrated DNA Technologies (IDT). CD55 cDNA was cloned into pENTR™/D-TOPO® Vector followed by LR recombination reaction to clone CD55 into the destination vector pLenti CMV Puro DEST. pENTR™/D-TOPO® CD55 mutants were generated using primers for site-directed mutagenesis (Supplementary Table 5). Mutants were cloned into the destination vector pLenti CMV Puro DEST for mammalian expression. Details of the plasmid sequence of wild-type CD55 and domain deletion CD55 mutants described in supplementary Table 4.

Generation of CRISPR/Cas9 knockout cells

Gene knock outs were performed from indicated cancer cells using a CD55 CRISPR/Cas9 knockout plasmid (Santa Cruz Biotechnology) or ZMYND8 CRISPR/Cas9 knockout plasmid (Santa Cruz Biotechnology). Briefly, cancer cells were seeded in six well plates and transfected with GFP labelled CRISPR/CAS9 knockout plasmids using Lipofectamine 3000. After 24 h, the transfection media was replaced with fresh serum-enriched medium. Cells were sorted by flow cytometry and GFP positive cells were collected as single cells in 96 well plates and cultured for 10 days. Each clone was screened to check for knockout efficiency. After confirming sufficient knockout, cells were grown and used in the indicated experiments.

Limiting dilution assays to determine cancer stem cell frequency

Ovarian cancer cells were counted and plated as single cells in nonadherent 96-well plates in 200 µL of stem cell media. Cells were plated at a density of 1, 5, 10, and 20 cells per well in triplicate rows. Stem cell media contained serum free DMEM/F-12, Fibroblast growth factor (bFGF) 20 ng/mL, Epidermal growth factor (EGF) 10ng/mL, B27 supplement 2%, and Insulin 10 µg/mL. After two weeks, each well was examined under a phase contrast microscope to detect tumorsphere-formation. Stem cell frequency or sphere-forming frequency was estimated using ELDA (Extreme limiting dilution algorithm) software (http://bioinf.wehi.edu.au/software/elda/) [10].

Cell proliferation assay using IncuCyte

Cancer cell proliferation was measured using an earlier reported method [11]. Briefly, CP70 cells (CD55 OE, KO, Δ1, Δ2, Δ3, Δ4, Δ1234, ΔS/T, Δ34, and Δ34S/T) were collected from cells during growing phase. Cells were seeded (1000 cells/well) in Geltrex-coated 96-well plates and incubated in the IncuCyte Live Cell Analysis System (Sartorius). Cell proliferation was measured up to 96 h as cell count normalized to day zero.

Immunoblot analysis

Immunoblots were performed to check protein expression as previously reported [12]. In brief, at the end of each study, plates were washed 3X with ice-cold PBS and placed on ice. Cells were harvested with NP-40 lysis buffer containing 250 mM NaCl, 50 mM Tris, pH 7.4, 1 mM Na3VO4, 5 mM EDTA, 1% Nonidet™ P40 (NP40), 50 mM NaF, 0.02% NaN3, 2 µg/ml protease cocktail inhibitor and 1mM PMSF by dropwise addition to the plates and kept on ice for few minutes. Cells were scrapped and collected into 1.5 ml centrifuge tube and kept on ice for 1.5 h. Lysates were vortexed vigorously every 10 min. Lysates were then centrifuged for 10 min at 12,000 rpm at 4 °C. The supernatants were collected and placed in a new 1.5 centrifuge tube and kept on ice. Protein concentration was measured by BCA (Thermo Scientific). 6× Laemmli buffer containing β-mercaptoethanol was added to the protein lysates and protein samples were boiled for 6 min. Protein samples were then resolved on SDS-PAGE electrophoresis using precast gel (4–20% Gradient gel, Biorad) and transferred to PVDF membrane using a wet transfer method at 4 °C for overnight. After transfer, PVDF membranes were blocked in 5% BSA for one hour at room temperature followed by the addition of primary antibodies at 4 °C overnight with gentle rocking. The next day, membranes were washed three times with 1×TBST (Tris-Buffered Saline, 0.1% Tween® 20) on a platform shaker followed by incubation with HRP-conjugated secondary antibodies for one hour at room temperature. Membranes were then washed three times with 1×TBST. Immunodetection was carried out using a chemiluminescence reagent (PerkinElmer) in a Chemidoc imaging system (GE Healthcare) and band densitometry was quantified by Image J software.

Cytoplasmic and nuclear protein fractionation

Cytoplamic and nuclear fractions were isolated using a commercially available NE-PER™ Nuclear and Cytoplasmic Extraction Kit (Thermo Scientific). Briefly, cells were cultured in 100 mm Petri Dish and grown to. ~70% confluency, cells were harvested, and cytoplasmic and nuclear fractions were isolated. After protein isolation, immunoblot study was performed according to the above-mentioned methods. For cytoplasmic protein, α-Tubulin antibody was used as loading control and for nuclear protein, Lamin A/C antibody was used as loading control.

Subcellular fractionation

Subcellular fractionation assay was performed according to the manufacturer’s instructions (Thermo Scientific). Briefly, ovarian cancer cells were grown in 100 mm Petri dish and grown to 80% confluence. Cells were harvested and washed two times with chilled PBS. The cell pellets were stepwise lysed according to the manufacturer’s protocol. In this experiment, cytoplasmic, soluble nuclear, chromatin-bound, and membrane fraction proteins were isolated to detect. Subcellular fractions were validated using tubulin for cytoplasmic faction, Na+/K + ATPase for cell membrane, Lamin A/C for soluble nuclear, and Histone H3 for chromatin-bound fraction.

CellTiter-Glo® assay

Cancer cells (CP70 cells 2000/well and SKOV3 4000/well) were seeded in 96 well plates (White non-transparent). The next day, cells were treated with increasing doses of cisplatin (0, 0.1, 0.3, 1, 3, 10, 30, and 100µM) for 48 h. Subsequently, CellTiter-Glo reagent mix (100 µl) was added to each well by replacing 100 µl cisplatin containing media according to the manufacturer’s protocol (Promega). Plates were placed on a platform shaker for 5 min with gentle rocking. The luminescence reading of each well was measured in a luminometer and viability was calculated as a percentage normalized to untreated control. Percent viability for each concentration was plotted using GraphPad Prism software (GraphPad Software, Inc.).

Cycloheximide treatment and protein stability

Ovarian cancer cells (Parental cells and CD55 WT/mutants transduced cells) were grown in 100 mm Petri dish until 60% confluence. Cells were then treated with cycloheximide for indicated times. Cycloheximide containing media was discarded and washed with chilled PBS two times. Cells were collected by scraping in PBS and centrifuged to obtain cell pellets. Cytoplasmic and nuclear fractions were enriched using commercially available NE-PER™ Nuclear and Cytoplasmic Extraction Reagents (Thermo Scientific). Protein samples were prepared, resolved by SDS-PAGE, and subjected to immunoblot analysis.

PIPLC treatment in cancer cells

To release CD55 from lipid rafts, Phosphatidylinositol-Specific Phospholipase C (PIPLC, Thermo) was used to cleave the GPI anchor [13]. Briefly, cancer cells were grown in 100 mm Petri dishes to 70% confluence. Fresh media was added to the cells followed by the addition of PIPLC to the media at concentrations of 10, 25, or 35 Unit/ml. Cells were then incubated at 4 °C for 30 min or 37 °C for indicated time with gentle rocking. Cells were then washed with ice-cold D-PBS three times and cytoplasmic and nuclear fractions were isolated. Immunoblots of the enriched fraction were performed to check CD55 expression.

Analysis of N-linked and O-linked glycosylation of CD55 protein

Cell surface CD55 is heavily O-glycosylated at the serine/threonine rich domain whereas it is lightly N-glycosylated at the SCR1/2 domain [14]. To elucidate whether nuclear CD55 is glycosylated we used a deglycosylation mix II enzyme (NEB) to remove O-linked and N-linked glycosylation. Briefly, cancer cells were grown in 100 mm Petri Dishes to a confluence of 70%, cells were harvested, and cytoplasmic/nuclear fractions prepared. Deglycosylation mix II enzymes were incubated with the nuclear and cytoplasmic fractions according to the manufacturer’s protocol. Protein samples were processed for immunoblot analysis.

For the analysis of N-linked glycosylation of CD55 protein, cancer cells were grown in 100 mm Petri Dishes. We used tunicamycin, a previously reported N-linked glycosylation inhibitor [15]. Cells were grown to 70% confluence and treated with tunicamycin for 24 h. Cells were harvested, and cytoplasmic/nuclear proteins were fractionated. Protein samples were resolved on SDS-PAGE and immunoblotted for CD55.

Immunofluorescence analyses

Immunofluorescence study was performed to visualize the distribution of CD55 protein in cancer and non-cancerous cells. Briefly, Cells were plated on coverslips in cell media and incubated for 48 h. For processing, cells were washed with D-PBS two times and fixed in a 4% paraformaldehyde solution for 10 min, washed with D-PBS two times. Coverslips were either incubated in 0.01%Triton-X-100 solution for 2–3 min or untreated followed by incubation in blocking buffer (3% BSA + 2% Goat serum in TBST) for 1 h at room temperature. Primary antibodies with indicated dilution were added to the cells and coverslips were incubated in a humidified chamber at 4 °C overnight. The next day, coverslips were washed three times with TBST on a platform shaker. Alexa Flour conjugated secondary antibodies were added to the coverslips and cells were incubated for 1 h at room temperature then washed three times with 1×TBST on a platform shaker. Coverslips were mounted on glass slides using VECTASHIELD mounting media (Vector Lab) containing the DAPI and visualized on Confocal microscope.

Lentivirus generation and transduction

For the generation of stable overexpression lines, lentivirus particles were used to infect the indicated cells. In brief, 4 × 106 HEK293T cells were seeded in 100 mm Petri Dish. Next, day, HEK293T cells were transfected with pMD2.G (Viral envelope expressing plasmid), pMDLg/pRRE (Packaging plasmid), pRSV-Rev (Packaging plasmid) and pLenti-puro Dest vectors (pLenti CMV Puro DEST, CD55 WT pLenti CMV Puro DEST, Δ1 pLenti CMV Puro DEST, Δ2 pLenti CMV Puro DEST, Δ3 pLenti CMV Puro DEST, Δ4 pLenti CMV Puro DEST, Δ1234 pLenti CMV Puro DEST, ΔS/T pLenti CMV Puro DEST, Δ34S/T pLenti CMV Puro DEST). After 24 h of incubation, fresh DMEM media was added to replace the transfection media and incubated for 24 h. Viral particle-containing media was filtered to remove cell debris and floating cells. In parallel, 0.5 × 106 ovarian cancer cells were seeded in six well plates and viral particle containing condition media was added to the cancer cells. After 24 h, a second batch of fresh viral particles from HEK293T cells was used to replace the first batch of viral-containing media and to infect the cancer cells. Virus transduced cells were kept in the incubator for 48 h and then treated with puromycin to select transduced cells. Once cells were ready, protein expression was evaluated to confirm the efficiency of knockdown or overexpression.

Immunoprecipitation and co-immunoprecipitation

To determine protein-protein interactions, immunoprecipitation and co-immunoprecipitation studies were performed according to the earlier reported method with few modifications [16]. Briefly, OC cells were lysed with Immunoprecipitation (IP) lysis buffer (Thermo Scientific) supplemented with protease cocktail inhibitor. Alternatively, cytoplasmic or nuclear fractions were isolated and IP lysis buffer was added. For immunoprecipitation, 3 µg of CD55 or ZMYND8 antibodies were added to the protein lysate and incubated for overnight at 4 °C. In parallel, a control antibody (Cell Signaling) was added to the protein lysate. Whole cell, cytoplasmic and nuclear lysates were incubated overnight at 4 °C. The next day, pre-cleaned magnetic A/G bead (Thermo Scientific) was added to the protein lysates containing the antibodies and incubated for 4 h at 4 C with constant rotation. Magnetic beads were collected and washed three times with lysis buffer. Laemmli buffer was added to the beads and boiled for 6 min. Beads were removed to obtain the protein samples and resolved on SDS-PAGE followed by immunoblot analysis.

CD55 immunoprecipitation and LC-MS/MS

Ovarian cancer cells were grown in 150 mm Petri dishes to 70% confluence. Cells were washed with D-PBS two times and harvested by scrapping in D-PBS. Samples were centrifuged at 2000 rpm for 5 min to collect the cell pellet. Cytoplasmic and nuclear fractions were isolated using NE-PER™ Nuclear and Cytoplasmic Extraction Kit (Thermo). Protein concentration of each fraction was estimated by BCA. CD55 antibody (3 µg) was added to the protein lysates and incubated at 4 °C overnight with continuous and mild rotation. The next day, protein A/G Agarose beads were added to the samples and incubated for 3 h at 4 °C. Agarose beads were washed three times with lysis buffer containing protease cocktail inhibitor. 2x Laemmle buffer was added to the beads and boiled for 5 min. Protein samples were collected and beads were discarded, followed by SDS-PAGE. Intact gels were transferred to the mass-spec core at the Lerner Research Institute (Cleveland Clinic) for LC-MS/MS (Liquid chromatography-tandem mass spectrometry) analysis to identify binding partners of cytoplasmic and nuclear CD55. The gel lanes were analyzed using a GeLC method, where large areas of the gel lane were cut, bands were washed/destained in 50% ethanol, 5% acetic acid, and dehydrated in acetonitrile. The bands were then reduced with DTT and alkylated with iodoacetamide prior to the in-gel digestion. All bands were digested in-gel using trypsin, by adding 5 µL of 10 ng/µL chymotrypsin in 50 mM ammonium bicarbonate and incubating overnight digestion at room temperature to achieve complete digestion. The peptides that were formed were extracted from the polyacrylamide in two aliquots of 30µL 50% acetonitrile with 5% formic acid. These extracts were combined, and half of the protein extracts were evaporated in a Speedvac and resuspended in 30 µL 0.1% formic acid for LCMS analysis.

Samples were analyzed by LC-MS using a Fusion Lumos Tribrid MS (ThermoScientific) equipped with a Dionex Ultimate 3000 nano UHPLC system, and a Dionex (25 cm x 75 μm id) Acclaim Pepmap C18, 2-µm, 100-Å reversed-phase capillary chromatography column. Peptide digests (5 µl) were injected into the reverse phase column and eluted at a flow rate of 0.3 µl/min using mobile phase A (0.1% formic acid in H2O) and B (0.1% formic acid in acetonitrile). The gradient was held at 2%B for 5 min, %B was increased linearly to 35% in 80 min, increased linearly to 90% B in 10 min, and maintained at 90% B for 5 min. The mass spectrometer was operated in a data-dependent manner which involved full scan MS1 (375–1700 Da) acquisition in the Orbitrap MS at a resolution of 120,000. This was followed by CID (1.6 Da isolation window) at 35% CE and ion trap detection. MS/MS spectra were acquired for 3 s. The second method was used for glycopeptide identification and involved full scan MS1 7 (350–1700 Da) acquisition in the Orbitrap MS at a resolution of 120,000. Dynamic exclusion was enabled where ions within 10 ppm were excluded for 60 s.

Raw data were analyzed by using all CID spectra collected in the experiment to search the human SwissProtKB database (downloaded on 4-29-2021, 26,594 entries) and more specifically against the sequence CD55 with the program Sequest which is integrated into Thermo Proteome Discoverer (V2.3) software package. Peptide and protein validation was performed using the percolator node with protein, peptide, and PSM thresholds at < 1% FDR. For the differential enrichment analysis, the protein abundance was estimated using the total number of spectra identified for each protein [17].

H&E and immunohistochemistry of A2780 CSC tumors

Hematoxylin and eosin (H&E) and immunohistochemistry were utilized in tumor tissues derived from the A2780 CSC mice ovarian tumor model, as described in a previously [3].We analyzed tumors that were enriched for CD55 (A2780 CSC) and two tumors with CD55 silenced [A2780 CSCs CD55 KD1(TRCN0000057167), A2780 CSCs CD55 KD2 (TRCN0000255377)] from our previous publication. H&E and IHC methods were described below.

Quantitative real-time polymerase chain reaction

Quantitative real-time Polymerase chain reaction (qRT-PCR) was performed according to the earlier reported method [18]. Briefly, cancer cells were grown in 100 mm Petri dishes to 70% confluence. Cells were harvested and total mRNA was separated using an RNA isolation kit (Takara Bio). RNA concentration as well as quality were evaluated using Nanodrop. First strand cDNA synthesis (PrimeScript 1st strand cDNA Synthesis Kit, Takara) was carried out from the isolated RNA. Primers for ZMYND8 and GAPDH were designed using Primer-BLAST (supplementary Table 6). Then we mixed primers sets, and SYBR Green master mix (Applied Biosystems). The gene expression was evaluated as 2−ΔΔCT and plotted in the graph using GraphPad Prism software.

RNA extraction/sample preparation and bulk-RNA sequencing and bioinformatics

Ovarian cancer cells (CP70 CD55 OE and KO) were grown on 100 mm Petri dishes to 70% confluence. Cells were washed two times with D-PBS and cells and harvested. Total RNA was extracted using RNeasy kit (Qiagen Inc). RNA concentration and RNA integrity number (RIN ≥ 7) were measured using Bioanalyzer. High-quality RNA was processed for library preparation and sequencing using methods published by Sangwan [19]. Raw sequencing reads were quality-trimmed using trimmomatic pipeline [20]. Quality filtered reads were mapped to the reference genome (GRCm39) [21] using STAR aligner [22], and gene expression levels were quantified using the count module in RNA-Seq by Expectation-Maximization (RSEM) v.1.3.3. Raw gene count matrices were prefiltered and processed for downstream analysis using methods published by Sangwan [19, 23]. Briefly, differential gene expression and pathway enrichment analysis were performed using edge [24] and topGO package [25]. Gene set enrichment analysis was performed using the command line pre-ranked GSEA application downloaded from the Broad Institute’s website with Mouse MSigDB (v2022.1.Mm) as a reference database [26].

In vivo experiments in NOD-scid IL2Rgammanull(NSG) mice

Pre-clinical studies were performed in NSG mice under a protocol reviewed and approved (Protocol number# 2986) by the Cleveland Clinic Institutional Animal Care and Use Committee (IACUC). Briefly, CP70 CD55OE, KO, ∆1234 and ∆ST cells were transduced with pCDH-EF1a-eFFly-mCherry lentivirus. FFly luciferase labeled (0.20 × 106 cells/mice) cells were injected in NSG mice intraperitoneally. After 10 days mice were divided into eight cohorts. (1) OE Vehicle (N = 9), (2) OE Cisplatin (N = 9), (3) KO Vehicle (N = 9), (4) KO Cisplatin (N = 9), (5) ∆1234 Vehicle (N = 9), (6) ∆1234 Cisplatin (N = 9), (7) ∆S/T Vehicle (N = 9), (8) ∆S/T Cisplatin (N = 9). The dose of Cisplatin was 2 mg/kg twice a week. During the study, mice were kept in an isoflurane inhalation chamber to induce anesthesia. D-luciferin solution was injected and bioluminescence images of the tumor in each mouse were captured by IVIS Lumina (PerkinElmer). IVIS images were analyzed by Living Image Software (Caliper Life Sciences). The fold change in tumor growth was measured and plotted in the graph. At necropsy, tumors were harvested and fixed in paraformaldehyde solution for immunohistochemical analysis according to the earlier reported method [27]. Additionally, TUNEL assay was performed to visualize the extent of DNA fragmentation in mice tumor specimens according to the manufacturer protocol.

Human ovarian tumor specimen collection and immunohistochemical analysis

Human ovarian tumor samples (FFPE tissue sections) were collected from patients in the Cleveland Clinic Foundation (IRB 19–185) and immunohistochemistry of the FFPE tissue sections was performed. Briefly, tissue slides were put into Histo-Clear (National Diagnostics) to remove the paraffin. Tissue sections were rehydrated in graded ethyl alcohol (100%, 95%, 80%, and 60% ethanol). Antigen retrieval was performed by boiling the tissue sections in Tris–EDTA buffer (pH9). After cooling, sections were incubated in 0.01% triton X-100 for 10 min for permeabilization. Tissue sections were then blocked in 5% goat serum for one hour and primary antibodies (CD55 1:500 dilution, Proteintech, and EMD Millipore) were added to the sections and incubated overnight at 4ºC in a humidified chamber. Following antibody incubation, tissue sections were washed three times with PBS. Peroxidase labeled polymer antibody (Vector Labs) was added to the tissue sections and incubated for 30 min. After washing, DAB chromogen was added for 5 min. Slides were washed with PBS three times and sections were then counterstained with hematoxylin to visualize the nuclei.

Software and statistical analysis

Image J software and Graph pad prism were utilized for image processing and graphical representation. Every experiment was performed at least three times. For multiple group analyses, One-way Analysis of Variance (ANOVA) was performed with Tukey’s post hoc comparison to determine p values (* p < 0.05, ** p < 0.01, *** p < 0.00, **** p < 0.0001). In other indicated assays, unpaired t test (* p < 0.05) was performed to determine the statistical significance. ELDA software was used to measure the stem cell frequency of ovarian cancer cells.