Multiple human polyomaviruses (HPyV) can infect the skin, but only Merkel cell polyomavirus (MCPyV) has been implicated in the development of a cancer, Merkel cell carcinoma (MCC). While expression of HPyV6, HPyV7, and MCPyV small T antigens (sT), all induced a senescence-associated secretory phenotype (SASP), MCPyV sT uniquely activated noncanonical NF-κB (ncNF-κB), instead of canonical NF-κB signaling, to evade p53-mediated cellular senescence. Through its large T stabilization domain, MCPyV sT activated ncNF-κB signaling both by inducing H3K4 trimethylation-mediated increases of NFKB2 and RELB transcription and also by promoting NFKB2 stabilization and activation through FBXW7 inhibition. Noncanonical NF-κB signaling was required for SASP cytokine secretion, which promoted the proliferation of MCPyV sT–expressing cells through autocrine signaling. Virus-positive MCC cell lines and tumors showed ncNF-κB pathway activation and SASP gene expression, and the inhibition of ncNF-κB signaling prevented VP-MCC cell growth in vitro and in xenografts. We identify MCPyV sT–induced ncNF-κB signaling as an essential tumorigenic pathway in MCC.

Implications:

This work is the first to identify the activation of ncNF-κB signaling by any polyomavirus and its critical role in MCC tumorigenesis.

Human polyomaviruses (HPyV) are small, double-stranded DNA viruses most intensely studied for their transforming properties. HPyVs were first noted to be clinically relevant as causes of infectious diseases in immunosuppressed patients. More recently, Merkel cell polyomavirus (MCV or MCPyV) has been identified as a critical factor in the development of approximately 80% of Merkel cell carcinomas (MCC; ref. 1), a rare but deadly skin cancer. Antibodies targeting the programmed cell death-1 (PD-1) pathway can be effective and have become the standard of care for patients with advanced disease (2–4). However, approximately 50% of patients with advanced MCC are resistant to checkpoint inhibitor therapy. Thus, a better understanding of the role of MCPyV in MCC development may help to spur the development of novel therapies.

Up to 14 species of HPyVs have been identified, and as many as 5 species have been detected on human skin, yet only MCPyV has been strongly linked to a malignancy in humans so far (5). The unique transforming properties of MCPyV appear to reside in the early region of the genome, which encodes for the small T antigen (sT) and through alternative splicing, large T antigen (LT). This region is essential for viral DNA replication (6, 7) and also for the transforming properties of murine polyomavirus and the Rhesus macaque polyomavirus SV40 (8). However, in contrast to other polyomaviruses, the cellular transforming activity of MCPyV appears to depend largely on its sT, and not on its LT, as the knockdown of sT is sufficient to cause the arrest of MCC cells in vitro (9). A more detailed analysis of how MCPyV differs from other cutaneous HPyVs is essential for understanding conserved oncogenic pathways and may yield novel therapeutic targets for patients with virus-positive MCC (VP-MCC).

One important difference between MCPyV and SV40 sTs is the presence of the large T stabilization domain (LSD domain) found in MCPyV sT, but not in the sT of SV40 or other cutaneous polyomaviruses, such as human polyomaviruses 6 and 7 (HPyV6/7). The LSD domain of MCPyV sT is essential for LT stabilization, LT-mediated viral replication, and its transforming properties. MCPyV sT can inhibit the F-box/WD repeat-containing protein 7 (FBXW7) protein (6), a tumor suppressor protein that promotes the degradation of many proto-oncogenes including c-Myc and Cyclin E (10). However, as FBXW7 knockdown alone does not rescue transformation by an LSD domain mutant MCPyV sT (6), it is likely that the MCPyV sT possesses additional transforming properties.

There have been conflicting reports on the impact of MCPyV sT on NF-κB signaling, with some reports suggesting an inhibition of canonical NF-κB pathways through an indirect interaction with the NEMO/IKKγ protein (11) while others have shown an activation of the same pathway (12). Because of NF-κB's central role in coordinating immune and inflammatory responses, its inhibition has been proposed as a mechanism to allow MCPyV to subvert the immune responses against viral infection. In addition to its role in immune signaling, NF-κB activation has been implicated in pleiotropic, and sometimes contradictory activities with respect to transformation. The activation of canonical NF-κB signaling has been linked to the development of senescence and to the upregulated secretion of a diverse set of cytokines, growth factors, and proteases (e.g., IL1β, IL6, IL8, and GM-CSF), also known as the senescence-associated secretory phenotype (SASP; refs. 13–15). In contrast, signaling from the noncanonical NF-κB (ncNF-κB) pathway, in which NF-κB2/p52 and RELB are activated, bypasses cellular senescence, at least in part through the induction of the EZH2 histone methyltransferase (16, 17). Notably, while MCPyV sT signaling has been studied in the context of canonical NF-κB activation, its impact on ncNF-κB signaling has not been directly examined.

Here, we express MCPyV sT and the sTs of two cutaneous polyomaviruses (HPyV6/7) currently not thought to be transforming. While HPyV6/7 sTs induce senescence, MCPyV sT bypasses cellular senescence and transforms primary human fibroblasts. MCPyV sT upregulated ncNF-κB signaling by increasing the levels of NFKB2 and RELB through transcriptional and posttranslational mechanisms. Using both chemical and genetic inhibition of the pathway, we find that the activation of ncNF-κB signaling was essential for the induction of SASP gene transcription and the serum independent growth of MCC. Using a specific inhibitor of ncNF-κB, we find that VP-MCC cell lines require ncNF-κB signaling for their growth in vitro and in vivo.

Cell culture and tumor tissue

BJ dermal fibroblasts were cultured in complete BJ medium (4:1 ratio of high glucose DMEM and Medium 199, 10% FBS, 2 mmol/L glutamine, and 1 mmol/L sodium pyruvate; Sigma-Aldrich). Lenti-X HEK-293T cells (Clontech) and Rat2 (ATCC) were cultured in DMEM supplemented with 10% FBS and used at early (<5 passages). All media were tested for Mycoplasma positivity by PCR-based method monthly. VP-MCC cancer cell lines (MKL-1, MKL-2, MS-1, and WaGa) were cultured in RPMI1640 supplemented with 20% FBS. VN-MCC cancer cell lines (MCC13, MCC26, UISO) were cultured in RPMI1640 supplemented with 10%FBS. MCC lines were confirmed through verifying the expression of sT (Supplementary Fig. S13A) and/or assessment of cell morphology. For chemical inhibition of RELB, DMSO vehicle control or 50 nmol/L calcitriol were added into virally infected cell culture medium. For experiments utilizing conditioned media, BJ fibroblasts pretransduced with either vector, HPyV6-sT, HPyV7-sT, or MCPyV-sT and were cultured in 2% FBS for 48 hours to produce cytokine-rich, conditioned medium. For MTT-mediated growth curve quantification, vector or LSD mutant–transduced cells were plated in 96-well plate in triplicate. Cells cultured in 2% FBS medium or MCPyV-sT WT cells cultured with protein transport inhibitor (PTI) cocktail (proprietary mixture of Brefeldin A and Monensin; Thermo Fisher Scientific) were supplemented with conditioned medium harvested from either HPyV6, 7, or MCPyV sT–expressing cell culture. Cell growth was then monitored for 6 days by MTT. Frozen MCC tumors were obtained from the UTSW Tissue Management Shared Resource. Tissue collection was performed with written informed consent from the patients and the studies were approved by the UT Southwestern institutional review board.

Cloning and lentiviral packaging and transduction

Primers used for cloning and qRT-pCR are listed in Supplementary Table S1. For constitutive expression system, HPyV6, HPyV7, and MCPyV sT antigens were amplified from the viral genome plasmid and MCPyV sT antigen was tagged with 3XFLAG at the N terminus. Amplified cDNAs were then cloned into pLEX vector with SpeI+NotI (for HPyV6 and MCPyV) and BamHI+NotI (for MCPyV). For doxycycline-inducible expression system, small T antigens were cloned into pLVX-TRE3G vector through InFusion cloning (Clontech) with single digestion by BamHI. The annealed oligos were then ligated into predigested pLKO.1-hygro vector (AgeI + EcoRI). For lentiviral packaging and transduction, transfer plasmid, pMD2.G and psPAX2 plasmids were cotransfected into Lenti-X 293T cells by Lipofectamine 3000 at molar ratio of 1:1:1. Viral supernatant was harvested at 48 and 72 hours posttransfection. A total of 8 × 104 BJ cells/well were plated into 6-well plate the day before transduction. The cells were then spin-transduced at 35°C, 2,500 rpm for 90 minutes. The cells were selected 48 hours posttransduction with 2 μg/mL puromycin (for constitutive expression plasmid), 15 μg/mL blasticidin, and 0.5 mg/mL G418 (for doxycycline-inducible expression system).

Soft agarose colony formation assay

HPyV6, HPyV7, and MCPyV small T–expressing vectors were transduced into Rat2 fibroblasts and selected with puromycin at 4 μg/mL. The stable cell lines were then resuspended and plated in 0.35% agarose DMEM on top of 0.5% agarose DMEM base in 6-well plate at 25,000 cells/well of 6-well plate. The cells were allowed to grow for 3 weeks and stained with 0.005% crystal violet dye in 6.25% ethanol for more than 1 hour at room temperature avoiding light. The colonies were then identified and counted under dissecting microscope.

MTT/XTT–mediated cell growth assay

Doxycycline-inducible HPyV6, HPyV7, and MCPyV small T antigen–expressing cells were plated at 1,000/well of 96-well plate for a total of 8 plates without addition of doxycycline. The cells were assayed for MTT per plate for every 24 hours. Doxycycline (1 μg/mL) was added to doxycycline-induced plates after 24 hours (after the first plate was assayed). MTT assay was performed according to manufacturer's instructions. Briefly, 10 μL of 12 mmol/L MTT was added to 100 μL medium and incubated for 8 hours at 37°C. MTT-containing medium was then removed and dissolved in 50 μL DMSO and absorbance read at 540 nm. All the later time points were normalized by the first plate readings. For XTT assays (Sigma Aldrich) of VP-MCC cells—MKL-1, MKL-2 and MS-1—were trypsinized into single-cell suspension and cells were plated at 5,000/well in 96-well plate. The XTT reaction was done according to manufacturer's instructions with absorbance read at 450 nm.

Senescence-associated β-galactosidase staining

Senescence-associated β-galactosidase (SA-β-gal) staining was performed according to manufacturer's instruction (Thermo Fisher Scientific). Briefly, cells were plated at 70% confluence and then fixed by formaldehyde and glutaraldehyde fixative agents for 15 minutes at room temperature. Cells were then washed and stained by X-gal at final concentration (1 mg/mL) in staining buffer at pH = 6.0 (exact) in presence of potassium ferricyanide and potassium ferrocyanide in 37°C bacteria culture incubator (free of CO2) for 16 hours.

Immunofluorescence staining

For immunofluorescence staining, cells were fixed in 4% paraformaldehyde at room temperature avoiding light for 10 minutes. Fixed cells were then permeabilized with 0.1% Triton X-100 at room temperature for 10 minutes. Cells were then blocked with 5% normal goat serum at room temperature for 1 hour. Primary antibody was then incubated at 4°C overnight, followed by 1 hour room temperature incubation of secondary antibody.

BrdU incorporation assay

BrdU incorporation assay was done as previously reported with slight modifications (18). BrdU was added to cell culture medium at a final concentration of 10 μmol/L and incubated for 4 hours at 37°C. Then BrdU-containing medium was removed and cells were washed with PBS before proceeding to immunofluorescence staining protocol. Fixed cells were treated with 1.5 N HCl at room temperature for 20 minutes for antigen retrieval before permeabilization.

Immunoblot analysis and subcellular fractionation

Cells were harvested and lysed in RIPA buffer (50 mmol/L Tris pH 8.0, 150 mmol/L NaCl, 1% NP-40, 0.1% SDS, 0.5% sodium deoxycholate) in the presence of 1× protease/phosphatase inhibitor cocktail (Thermo Fisher Scientific) on ice for 15 minutes. Patient samples and xenografted tumors were homogenized in liquid nitrogen and protein was extracted by T-PER protein extraction buffer. Chromatin and cell debris were then pelleted down by centrifuging at 12,000 rpm at 4°C. The supernatant was collected, and BCA assay was used to quantify protein concentration. Protein lysates were fully denatured by 1× Laemmli sample buffer and boiled at 95°C for 10 minutes. Protein lysates were then separated in SDS-PAGE gel and blotted onto polyvinylidene difluoride membrane. Membrane was then blocked by 5% nonfat milk in TBS/T (20 mmol/L Tris pH 7.4, 150 mmol/L NaCl, 0.1% (w/v) Tween-20) for at least 1 hour before incubation with primary antibody at 4°C for overnight. The blot was then probed with horseradish peroxidase (HRP)-conjugated secondary antibody and developed using ECL reagent. Subcellular fractionation was performed according to manufacturer's instructions. The protein concentration of each fraction was quantified by BCA assay.

Chromatin immunoprecipitation

Native ChIP

Total chromatin was purified by lysing cells with hypotonic buffer (10 mmol/L HEPES pH7.9, 10 mmol/L KCl, 1.5 mmol/L MgCl2, 0.34 mol/L sucrose, 10% glycerol, and 1 mmol/L DTT) followed by No-salt buffer (3 mmol/L EDTA, 0.2 mmol/L EGTA, 1 mmol/L DTT) in the presence of protease inhibitor cocktail and phenylmethylsulfonylfluoride. Purified chromatin was washed by hypotonic buffer for at least twice to fully remove residual EDTA and EGTA before proceeding to micrococcal nuclease (MNase) digestion. 5% of the digested chromatin was reserved as input DNA. The rest of the digested genomic DNA was precleared by rabbit IgG and then immunoprecipiated by H3K4me3 antibody bound to protein A/G dynabeads in RIPA buffer for at 4°C. The complex was then washed by RIPA buffer twice, high salt RIPA buffer (RIPA buffer + 300 mmol/L NaCl) twice, LiCl wash buffer (250 mmol/L LiCl, 0.5% NP-40, 0.5% sodium deoxycholate) twice and 1XTE buffer + 50 mmol/L NaCl. The DNA was then eluted in SDS elution buffer (50 mmol/L Tris pH 8.0, 10 mmol/L EDTA, 1% SDS) at 65°C for at least 6 hours and treated with RNase and protease K. The DNA was finally recovered by phenol chloroform extraction.

Crosslinked ChIP

The cells were crosslinked by paraformaldehyde at final concentration of 1% at room temperature for 10 minutes and quenched by 0.125 mol/L glycine. The crosslinked cells were collected with cell scraper. The chromatin was purified as native ChIPN-ChIP. The purified chromatin was then resuspended in sonication buffer (10 mmol/L Tris, pH 8, 100 mmol/L NaCl, 1 mmol/L EDTA, 0.5 mmol/L EGTA, 0.1% sodium deoxycholate, 0.5% N-lauroylsarcosine) using a Covaris M220 Focused-ultrasonicator until the chromatin was sonicated to 200–700 bp fragments. The fragmented chromatin was then precleared by mouse IgG and incubated with M2 FLAG antibody bound to protein A/G dynabeads overnight at 4°C. The magnetic beads were then washed by the same wash buffers in native ChIP. Input and ChIP DNA were reverse crosslinked by the same SDS elution buffer at 65°C for at least 6 hours. The eluted DNA was then treated with RNase and protease K and recovered by phenol–chloroform extraction.

qRT-PCR analysis

Total RNAs were isolated by Qiagen RNeasy RNA Miniprep Kit (BJ cells) or TRIzol (MCC xenografts and patient samples). RNA concentration and purity were measured and checked by Nanodrop. cDNA was synthesized using 1 μg of total RNA by iScript cDNA Synthesis Kit. The reverse transcribed cDNA was then used for real-time PCR by CFX96 Real Time Cycler (Bio-Rad) as per manufacturer's instruction using 2X SYBR green PCR Master Mix (Applied Biosystems). Target gene expression was normalized by housekeeping gene ACTB (β-actin).

c-Myc turnover assay

293T cells were cotransfected with pLEX vector, HPyV6-sT, HPyV7-sT, wild-type MCPyV-sT, or LSD-mutant MCPyV-sT and pcDNA3.1-cMyc expression vector. After 48 hours, cells were treated with cycloheximide (0.1 mg/mL) to halt translation. The cells were then harvested at 0 (immediately), 1, 2, and 4-hour intervals. c-Myc protein expression were then assessed by Western blot analysis.

Mice and xenograft of MCC lines

NOD-scid/IL2Rgamma immunodeficient mice were housed in a sterile facility and all mouse experiments were approved by UT Southwestern Medical Center's institutional Animal Care and Use Committee. For xenografts, MKL-1 and WaGa cells were cultured in RPMI + 20% FBS and allowed to grow until at least 1 × 107 viable cells were obtained. A total of 1 × 107 cells in PBS were subcutaneously injected into NSG immunodeficient mice and allowed to grow for 1 month. Five days after injection of the cancer cells, mice were treated with either SN52-mutant peptide or SN52 functional peptide twice weekly at 40 μg/injection in situ by subcutaneous injection. Tumors were then harvested for downstream analysis. Fraction of each tumor was harvested and flash frozen for protein/RNA extraction.

ELISA

Cells were typically harvested 7 days after transduction with a fresh medium changed 2 days before harvesting to allow for cytokine accumulation. Uncoated human IL8 ELISA kit was purchased from PeproTech and the experiment was performed according to manufacturer's instructions. Briefly, 1 μg/mL in PBS capture antibody was plated onto ELISA microplate for overnight incubation (at least 12 hours). Then excess capture antibody was then removed with wash buffer (0.05% Tween-20 in PBS). The plate was then blocked with 1% BSA in PBS for 1 hour at room temperature. After washing three times with wash buffer, 100 μL of standard and samples (1:10–1:10,000 dilution for all samples except for LSD mutant to fit into detection range) were added and incubated for 2 hours at room temperature. Detection antibody was then added at 0.25 μg/mL and incubated for 2 hours at room temperature. Avidin–HRP conjugate was then added and incubated for 30 minutes at room temperature. The color was developed by ABTS liquid substrate for <5 minutes and plate was read at absorbance 405 nm.

Lentiviral RNAi

The shRNAs (TP53, RELB, GFP) were cloned into pLKO.1-hygro backbone by AgeI + EcoRI. BJ cells were then lentivirally infected and selected with hygromycin B at 200 μg/mL or puromycin at 2 μg/mL for more than 7 days. The shRNA constructs (NIK, NFKB2, MAP3K14, IKKA) in pLKO.1-puro backbone were obtained from Sigma-Aldrich. Sequences of shRNAs can be found in Supplementary Table S1.

RNA sequencing and data analysis

Total RNAs were extracted by RNeasy Miniprep Kit according to manufacturer's instructions. The total RNA samples were then sent to Genewiz for RNA quality control, library preparation, and sequencing. RNA samples were quantified using Qubit 2.0 Fluorometer RNA assay (Invitrogen) and RNA integrity was checked with Agilent 4200 TapeStation (Agilent Technologies). RNA library preparations, sequencing reactions were conducted at GENEWIZ, LLC. Ribosomal RNA depletion was performed using Ribo-Zero Gold Kit (Human/Mouse/Rat probe; Illumina). RNA-sequencing library preparation used NEBNext Ultra RNA Library Prep Kit from Illumina by following the manufacturer's instructions (NEB). Briefly, enriched RNAs were fragmented for 15 minutes at 94°C. First strand and second strand cDNA were subsequently synthesized. cDNA fragments were end repaired and adenylated at 3′ends, and universal adapters were ligated to cDNA fragments, followed by index addition and library enrichment with limited cycle PCR. Sequencing libraries were validated using Agilent 4200 TapeStation (Agilent Technologies), and quantified by using Qubit 2.0 Fluorometer (Invitrogen) as well as by using KAPA qPCR assay (KAPA Biosystems). The sequencing libraries were multiplexed and clustered on two lanes of a flowcell. After clustering, the flowcell was loaded on the Illumina HiSeq instrument according to manufacturer's instructions. The samples were sequenced using a 2 × 150 Paired End (PE) configuration. Image analysis and base calling were conducted by the HiSeq Control Software (HCS) on the HiSeq instrument. Raw sequence data (.bcl files) generated from Illumina HiSeq was converted into fastq files and demultiplexed using Illumina bsl2fastq v. 2.17 program.

For data analysis, the data files of the reference genome, hg19, were downloaded from UCSC Genome Browser (https://genome.ucsc.edu/) and Illumina iGenomes (https://support.illumina.com/sequencing/sequencing_software/igenome.html). The qualities of sequencing reads were evaluated using NGS QC Toolkit (19) and high-quality reads were extracted. STAR (20) was employed to map the reads onto the reference genome and HTSeq Python package (21) was employed to count reads for genes. DESeq R Bioconductor package (22, 23) was used to normalized read counts and identify differentially-expressed (DE) genes. Gene Set Enrichment Analysis (GSEA) was performed using Broad Institute GSEA software. A custom gene set for SASP signaling was generated using genes previously identified as part of the SASP program (15).

Protein sequence alignment and protein modeling

Protein sequences were aligned in Protein BLAST program (https://blast.ncbi.nlm.nih.gov/Blast.cgi?PAGE=Proteins). MCPyV small T (wild-type and LSD mutant) structures were modeled by Expasy SWISS-MODEL program using SV40 small T antigen as template. Protein structures were then annotated in Pymol software.

Statistical analysis

Student t test (for two group comparison) and one-way ANOVA (for three group comparison) were used in this study. Levene's mean test was used to test violation of homogeneity of variance of each dataset before one-way ANOVA was used. If tested positive (assumption of homogeneity of variance was violated), the dataset was transformed to log10 fold change to correct for violation of homogeneity of variance and one-way ANOVA was then used with a Tukey post hoc analysis for comparison among each pair of groups.

HPyV6 and HPyV7 sT induce p53-dependent cellular senescence

MCPyV sT transforms rodent fibroblasts in vitro (24), but the transforming properties of other cutaneous polyomaviruses, such as HPyV6 and HPyV7, have not been explored in detail. HPyV6 and 7 have a high degree of similarity, while MCPyV is less closely related (Supplementary Fig. S1). Rat fibroblasts (Rat2)-expressing lentivirally transduced HPyV6 and 7 sTs could not generate colonies in soft agar in contrast to MCPyV sT (Fig. 1A and B). However, their expression did not impair cell proliferation (Supplementary Fig. S2A and S2B). Because dermal fibroblasts have been reported as a host cell for MCPyV (25), we next expressed cutaneous sTs in primary human foreskin fibroblasts (BJ fibroblasts). In contrast to Rat2 fibroblasts, BJ cells demonstrated a pronounced flattened and enlarged cell morphology within 2 days of lentiviral delivery of HPyV6 and 7 sT, while MCPyV sT did not induce similar changes (Fig. 1C). Similar morphologic changes occurred in BJ fibroblasts immortalized by hTERT expression (Supplementary Fig. S2B). To control the expression dynamics more precisely, we also generated lentiviral, doxycycline-inducible expression systems for each of the sT antigens. We detected expression for all PyV sTs with doxycycline induction, although the expression level varied compared with the constitutive (CMV) promoter (Supplementary Fig. S2C). Consistent with the cell morphology, both HPyV6 and 7 sT–expressing fibroblasts showed significantly slower growth rate compared with the uninduced (no doxycycline) controls as early as 2 days after induction (Fig. 1D and E; Supplementary Fig. S3A and S3B). In contrast, MCPyV sT expression did not have a strong impact on cell proliferation (Fig. 1F; Supplementary Fig. S3C and S3D).

Figure 1.

HPyV6 and HPyV7, but not MCPyV, sT induce senescence. A, A total of 2.5 × 104 Rat2 fibroblasts expressing vector control, HPyV6, HPyV7, and MCPyV small T (sT) antigens were seeded in soft agar (21 days). Representative images of soft agar colonies formed by indicated cell type. Scale bar, 1,000 μm. Only MCPyV sT–expressing cells promote anchorage-independent growth. B, Quantification of soft agar colonies in three biological replicates from each group reveals that only MCPyV sT transforms Rat2 fibroblasts. C, Representative light microscopy of BJ fibroblasts transduced with HPyV6, HPyV7, and MCPyV sT (2 days posttransduction). HPyV6/7 sT cause cells to adopt a flattened and senescent morphology. Scale bar, 200 μm. D–F, Relative growth rates of BJ fibroblasts stably transduced with doxycycline-inducible HPyV6, HPyV7, and MCPyV sT. w –dox (no expression) control. HPyV6/7, but not MCPyV, sTs induce growth arrest. G, Representative light microscopy of SA-β-gal staining of HPyV6, HPyV7, and MCPyV sT–transduced BJ cells (4 days) highlights the induction of SA-β-gal staining by HPyV6/7 sT. Scale bar, 200 μm. H, Quantification of SA-β-gal–positive cells of three biological replicates of each group confirms the significant induction of senescence by HPyV6/7 sT. I, Western blot analysis of p53, p21, p16, and EZH2 expression in HPyV6, HPyV7, and MCPyV sT–transduced BJ fibroblasts (7 days postinfection). HPyV6/7 sT–expressing cells show increased expression of p53, p21, and p16 consistent with the induction of senescence. MCPyV sT shows increased expression of H3K27 methytransferase, EZH2. 1t1.13 confirms HPyV6/7 sT expression; Flag confirms MCPyV sT expression; and HSP90 is the loading control. J, Doxycycline induction (3 days) of HPyV6/7 sT, but not MCPyV sT, results in a marked reduction of BrdU-positive BJ cells. Images show representative IF images of BrdU staining of cells without or with doxycycline induction (1 μg/mL, 24 hours before the assay). Scale bar, 100 μm. K, Expression of HPyV6/7 sT results in significantly less BrdU incorporation. Quantification of BrdU-positive cells of three biological replicates of each group. L, BJ cells were transduced with the indicated shRNA (7 days), then HPyV sT. The knockdown of p53 prevents the morphologic induction of senescence by HPyV6/7 sT (3 days). Representative light microscopy of HPyV6/7 sT–transduced shTP53-3 KD and shGFP (shRNA control) cells. See Supplementary Fig. S2. M, BJ cells were transduced with the indicated shRNA (7 days), then HPyV sT. The knockdown of p53 prevents the induction of SA-β-gal staining by HPyV6/7 sT (3 days). Representative light microscopy of transduced shGFP and shTP53-3 cells. Scale bar, 200 μm. N, Quantification of SA-β-gal–positive cells of three biological replicates of each shRNA group confirms the rescue of senescence induction by p53 knockdown. In B, D, E, F, H, K, L, and O, data represents mean ± SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Significance assessed by one-way ANOVA followed by Tukey post hoc analysis for B and H and Student t test for J.

Figure 1.

HPyV6 and HPyV7, but not MCPyV, sT induce senescence. A, A total of 2.5 × 104 Rat2 fibroblasts expressing vector control, HPyV6, HPyV7, and MCPyV small T (sT) antigens were seeded in soft agar (21 days). Representative images of soft agar colonies formed by indicated cell type. Scale bar, 1,000 μm. Only MCPyV sT–expressing cells promote anchorage-independent growth. B, Quantification of soft agar colonies in three biological replicates from each group reveals that only MCPyV sT transforms Rat2 fibroblasts. C, Representative light microscopy of BJ fibroblasts transduced with HPyV6, HPyV7, and MCPyV sT (2 days posttransduction). HPyV6/7 sT cause cells to adopt a flattened and senescent morphology. Scale bar, 200 μm. D–F, Relative growth rates of BJ fibroblasts stably transduced with doxycycline-inducible HPyV6, HPyV7, and MCPyV sT. w –dox (no expression) control. HPyV6/7, but not MCPyV, sTs induce growth arrest. G, Representative light microscopy of SA-β-gal staining of HPyV6, HPyV7, and MCPyV sT–transduced BJ cells (4 days) highlights the induction of SA-β-gal staining by HPyV6/7 sT. Scale bar, 200 μm. H, Quantification of SA-β-gal–positive cells of three biological replicates of each group confirms the significant induction of senescence by HPyV6/7 sT. I, Western blot analysis of p53, p21, p16, and EZH2 expression in HPyV6, HPyV7, and MCPyV sT–transduced BJ fibroblasts (7 days postinfection). HPyV6/7 sT–expressing cells show increased expression of p53, p21, and p16 consistent with the induction of senescence. MCPyV sT shows increased expression of H3K27 methytransferase, EZH2. 1t1.13 confirms HPyV6/7 sT expression; Flag confirms MCPyV sT expression; and HSP90 is the loading control. J, Doxycycline induction (3 days) of HPyV6/7 sT, but not MCPyV sT, results in a marked reduction of BrdU-positive BJ cells. Images show representative IF images of BrdU staining of cells without or with doxycycline induction (1 μg/mL, 24 hours before the assay). Scale bar, 100 μm. K, Expression of HPyV6/7 sT results in significantly less BrdU incorporation. Quantification of BrdU-positive cells of three biological replicates of each group. L, BJ cells were transduced with the indicated shRNA (7 days), then HPyV sT. The knockdown of p53 prevents the morphologic induction of senescence by HPyV6/7 sT (3 days). Representative light microscopy of HPyV6/7 sT–transduced shTP53-3 KD and shGFP (shRNA control) cells. See Supplementary Fig. S2. M, BJ cells were transduced with the indicated shRNA (7 days), then HPyV sT. The knockdown of p53 prevents the induction of SA-β-gal staining by HPyV6/7 sT (3 days). Representative light microscopy of transduced shGFP and shTP53-3 cells. Scale bar, 200 μm. N, Quantification of SA-β-gal–positive cells of three biological replicates of each shRNA group confirms the rescue of senescence induction by p53 knockdown. In B, D, E, F, H, K, L, and O, data represents mean ± SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Significance assessed by one-way ANOVA followed by Tukey post hoc analysis for B and H and Student t test for J.

Close modal

Because the pronounced morphologic change of the HPyV6/7 sT–expressing cells were suggestive of cellular senescence, we performed SA-β-gal staining on the cells. HPyV6 and 7 sT, but not MCPyV sT, cells showed a significantly higher percentage of SA-β-gal–positive cells compared with the vector control, with more than 80% of cells staining (Fig. 1G and H) regardless of hTERT expression (Supplementary Fig. S2B). The expression of CDK inhibitors involved in cellular senescence, including p53, p21, and p16, was strongly induced only in HPyV6/7 sT–expressing cells (Fig. 1I). The methyltransferase EZH2 represses p21 and p16 expression and is downregulated after passage or oncogene-induced senescence (26–28). Consistent with its reported functions in inhibiting senescence, EZH2 was modestly decreased by HPyV6/7 sT, but strongly induced by MCPyV sT expression (Fig. 1I). Finally, we directly assessed cell-cycle progression through the BrdU incorporation assay. Consistent with the cell proliferation assays, we found that HPyV6 and 7 sT, but not MCPyV sT, significantly inhibited BrdU incorporation after doxycycline induction (Fig. 1J and K). HPyV6 and 7 are constitutively shed from the skin, and keratinocytes, rather than fibroblasts, have been suggested as a possible host cell for these PyVs (5, 29). Therefore, we also expressed HPyV6/7 sT in keratinocytes. Consistent with its impact in fibroblasts, HPyV6 and 7 sTs also caused senescence in keratinocytes based on the pronounced change of cell morphology and increased SA-β-gal positivity (Supplementary Fig. S2D).

Because Rat2 fibroblasts show spontaneous loss or mutation of p53 (30), we tested whether HPyV6 and 7 induced senescence in primary fibroblasts, but not Rat2 cells, due to differences in p53 status. Indeed, HPyV6 and 7 sTs failed to induce senescence phenotype in Huh7 cells, which possess a p53 loss-of-function mutation (Supplementary Fig. S2E), but did induce senescence in p53 WT A549 cells (Supplementary Fig. S2F; ref. 31, 32). To confirm directly that HPyV6 and 7 sTs-induced senescence required p53, we used three shRNAs to knockdown p53 in BJ fibroblasts by at least 50% (Supplementary Fig. S2G). Knockdown of p53 rescued morphologic changes (Fig. 1L; Supplementary Fig. S2H) and significantly reduced the number of SA-β-gal–positive cells (Fig. 1M and N, Supplementary Fig. S2I). Thus, in contrast to the transforming properties of MCPyV sT, HPyV6, and 7 sTs induce p53-dependent cellular senescence.

HPyV6, HPyV7, and MCPyV sT all induce SASP gene expression

To broaden our analysis of the differences between HPyV6, HPyV7, and MCPyV sT, we performed RNA-seq on sT-expressing dermal fibroblasts. Principal component analysis (PCA) revealed that MCPyV sT possessed the most distinct transcriptomic profile compared with both EV control and HPyV6 and 7 sTs (Fig. 2A). MCPyV sT showed significantly more differentially regulated genes compared with HPyV6 and 7 (Supplementary Fig. S4C and S4D). Some of the genes most highly upregulated by MCPyV sT expression included IL6, IL1B, and IL8, cytokines which have previously been linked to the SASP, the secretory program adopted by senescent cells (Fig. 2B). To extend this finding, we quantified the expression of six classical SASP genes including CCL26, CXCL10, CSF2 (GM-CSF), IL1B, IL6, and IL8. HPyV6 and 7 sT significantly upregulated each of these well-established SASP factors as might be expected on the basis of their ability to induce senescence in fibroblasts. Despite the absence of features of senescence, MCPyV sT–expressing fibroblasts also significantly upregulated these genes, with some (i.e., CSF2, IL1B, IL6, and IL8) being upregulated 10- to 100-fold higher than HPyV6 and 7 sT (Fig. 2C). Consistent with their distinct transcriptomic profiles, HPyV6, 7, and MCPyV sT did not show uniform upregulation of all reported SASP factors (Supplementary Fig. S5A). Time course analyses revealed that expression of the SASP factors remained high up to 15 days after transduction (Supplementary Fig. S5B–S5D). ELISA of conditioned media confirmed that IL8 secretion was significantly upregulated by all sT expression, with MCPyV sT inducing IL8 secretion at levels 100-fold higher than either HPyV6 and 7 sT (Fig. 2D). In addition to highlighting Myc as being highly upregulated by MCPyV sT, gene set enrichment analysis (GSEA) confirmed SASP to be one of the most enriched signatures after MCPyV sT expression (Fig. 2E, F, and H). One reported mechanism underlying the activation of SASP program is the activation of NF-κB signaling (33). Indeed, GSEA of the RNA-seq data also demonstrated that the NF-κB pathway is highly enriched in MCPyV sT cells (Fig 2E, G, and H). To rule out any potential contributions by the lentiviral expression vectors, we knocked down MCPyV sT expression with a sT-specific shRNA, which showed approximately 90% knockdown efficiency (Supplementary Fig. S6A-B). All SASP gene expression was significantly rescued (Supplementary Fig. S6C), confirming that the observed SASP gene expression is specific to sT. Thus, HPyV6, HPyV7, and MCPyV sT–expressing cells all induce the expression and secretion of inflammatory cytokines typical of SASP.

Figure 2.

MCPyV sT induces SASP and NF-κB signaling in an LSD motif–dependent manner. A, Principal component analysis (PCA) of empty vector (EV), HPyV6 sT, HPyV7 sT, MCPyV wild-type sT, and MCPyV LSD–mutant sT transcriptomes (7 days) reveals distinct clustering of MCPyV sT compared with vector control and other PyV sTs. B, Volcano plot (single gene) of MCPyV sT compared with empty vector control highlights the upregulation of multiple SASP genes. C, qRT-PCR analysis of HPyV6, HPyV7, and MCPyV sT–transduced cells (7 days) confirms the significant upregulation of CCL26, CXCL10, GM-CSF, IL1B, IL6, and IL8 by PyV sT expression compared with vector control. For GM-CSF, IL1B, IL6, and IL8, expression induced by MCPyV sT is significantly higher than HPyV6/7. Data are normalized by empty vector control and scaled to log10 fold change. D, ELISA measurement of IL8 secreted by BJ fibroblasts after sT transduction (7 days). All PyVs induced IL8 secretion compared with the vector; MCPyV sT induced even greater IL8 secretion than HPyV6/7 sT. Data are normalized by empty vector and scaled to log10 fold change. E, Volcano plot of gene signatures induced by MCPyV sT compared with empty vector control. F, GSEA analysis of SASP signaling comparing BJ cells stably transduced with MCPyV sT to vector control. G, GSEA analysis of NF-κB signaling comparing BJ cells stably transduced with MCPyV sT to vector control signaling by MCPyV sT compared with vector control. H, Heatmap of significant genes in SASP and NF-κB signaling gene sets of MCPyV sT compared with empty vector control. Gene list (right) indicates a subset of transcripts induced by SASP induction. I, Volcano plot (single gene) of MCPyV sT LSD mutant compared with wild-type MCPyV sT reveals that many of the highly induced genes are reverted to levels comparable with the vector control (B). J, Heatmap comparing all significant genes in vector control, MCPyV sT LSD mutant, and wild-type MCPyV sT highlights the rescue of transcriptional changes by the LSD mutant. K, qRT-PCR analysis of SASP genes confirms that the MCPyV sT LSD mutant rescues changes in SASP cytokines induced by wild-type MCPyV sT (7 days). Data are normalized by vector control and scaled to log10 fold change. In C, D, and K, data are presented as mean ± SD. *¸ P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Significance assessed by one-way ANOVA followed by Tukey post hoc analysis.

Figure 2.

MCPyV sT induces SASP and NF-κB signaling in an LSD motif–dependent manner. A, Principal component analysis (PCA) of empty vector (EV), HPyV6 sT, HPyV7 sT, MCPyV wild-type sT, and MCPyV LSD–mutant sT transcriptomes (7 days) reveals distinct clustering of MCPyV sT compared with vector control and other PyV sTs. B, Volcano plot (single gene) of MCPyV sT compared with empty vector control highlights the upregulation of multiple SASP genes. C, qRT-PCR analysis of HPyV6, HPyV7, and MCPyV sT–transduced cells (7 days) confirms the significant upregulation of CCL26, CXCL10, GM-CSF, IL1B, IL6, and IL8 by PyV sT expression compared with vector control. For GM-CSF, IL1B, IL6, and IL8, expression induced by MCPyV sT is significantly higher than HPyV6/7. Data are normalized by empty vector control and scaled to log10 fold change. D, ELISA measurement of IL8 secreted by BJ fibroblasts after sT transduction (7 days). All PyVs induced IL8 secretion compared with the vector; MCPyV sT induced even greater IL8 secretion than HPyV6/7 sT. Data are normalized by empty vector and scaled to log10 fold change. E, Volcano plot of gene signatures induced by MCPyV sT compared with empty vector control. F, GSEA analysis of SASP signaling comparing BJ cells stably transduced with MCPyV sT to vector control. G, GSEA analysis of NF-κB signaling comparing BJ cells stably transduced with MCPyV sT to vector control signaling by MCPyV sT compared with vector control. H, Heatmap of significant genes in SASP and NF-κB signaling gene sets of MCPyV sT compared with empty vector control. Gene list (right) indicates a subset of transcripts induced by SASP induction. I, Volcano plot (single gene) of MCPyV sT LSD mutant compared with wild-type MCPyV sT reveals that many of the highly induced genes are reverted to levels comparable with the vector control (B). J, Heatmap comparing all significant genes in vector control, MCPyV sT LSD mutant, and wild-type MCPyV sT highlights the rescue of transcriptional changes by the LSD mutant. K, qRT-PCR analysis of SASP genes confirms that the MCPyV sT LSD mutant rescues changes in SASP cytokines induced by wild-type MCPyV sT (7 days). Data are normalized by vector control and scaled to log10 fold change. In C, D, and K, data are presented as mean ± SD. *¸ P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Significance assessed by one-way ANOVA followed by Tukey post hoc analysis.

Close modal

The LSD motif is required for sT-induced transcriptional changes, including SASP

The LSD motif of MCPyV sT stabilizes the large T antigen and is essential to the transforming properties of sT (6). The LSD motif inhibits FBXW7, which is critical to the stabilization of cellular oncogenes like c-Myc. However, the knockdown of FBXW7 alone does not promote the transforming activity of the LSD mutant MCPyV sT (6). Thus, its precise role in promoting fibroblast transformation remains an area of active investigation (34–36). The LSD motif is not conserved in HPyV6/7 sT (Supplementary Fig. S1), suggesting that it might contribute to the phenotypic differences caused by HPyV6/7 and MCPyV sT expression. To better define the role of the LSD motif, we also performed RNA-seq on a MCPyV sT with a mutation of the LSD domain (LSD; amino acids 91–95, LKDYM→AAAAA) cells (Supplementary Fig. S1, S7A). RNA-seq revealed that LSD-expressing fibroblasts were notably distinct from WT-infected cells and were more similar to vector control cells (Fig. 2A, I, and J). GSEA analyses revealed that Myc target activation by wild-type MCPyV sT was largely reversed by the LSD motif mutation (Supplementary Fig. S8A). The effect of sT expression on c-Myc expression was examined. In contrast to MCPyV sT, HPyV6/7 sT, which lack the LSD motif, did not stabilize c-Myc protein levels when expressed in BJ cells (Supplementary Fig. S8B). MCPyV sT has been reported to decrease c-Myc turnover. Using cycloheximide treatment, c-Myc turnover was assessed in 293T cells cotransfected with c-Myc and a vector control, HPyV6 sT, HPyV7 sT, wild-type MCPyV sT, or LSD-mutant MCPyV sT. Only wild-type MCPyV decreased the turnover of c-Myc in an LSD motif–dependent manner (Supplementary Fig. S8C).

We next assessed whether the LSD motif might also contribute to SASP gene expression. Indeed, compared with WT sT, the LSD mutant showed notable downregulation of the same SASP genes that were upregulated by WT (Fig. 2K; Supplementary Fig. S5A). GSEA analysis also revealed that most of the previously upregulated signaling pathways were also rescued by the introduction of LSD motif, including SASP and NF-κB pathway (Fig. 2J; Supplementary Fig. S7B–S7E). Using qRT-PCR for multiple SASP genes and ELISA for IL8, we validated the rescue of SASP activation by the LSD mutant (Fig. 2K; Supplementary Fig. S7F). Specific matrix metalloproteinase genes (e.g., MMP1, MMP3, and MMP10), previously reported to be part of the SASP program, which were significantly upregulated by MCPyV sT were largely rescued by the LSD mutant (Supplementary Fig. S7G). One notable exception to pathways rescued by the LSD mutation was IFNγ signaling. Specifically, wild-type sT, like HPyV6/7, inhibited cellular IFN signaling pathways (Supplementary Fig. S4E–I). The LSD mutant induced IFNα, but not IFNγ signaling, as assessed by GSEA analysis (Supplementary Fig. S7C). Notably, the LSD mutant did not induce cellular senescence as assessed by cell morphology (Supplementary Fig. S7H), cell proliferation as assessed by MTT (Supplementary Fig. S7I), or BrdU incorporation (Supplementary Fig. S7J–S7K). In summary, the LSD motif of MCPyV sT drives many of the transcriptional changes induced by MCPyV sT, including both NF-κB pathway activation and the induction of SASP genes.

HPyV6 and 7 and MCPyV sT activate distinct NF-κB pathways

The NF-κB pathway can be activated in two ways, referred to as canonical (classical) or noncanonical (alternative) NF-κB signaling (37, 38). While both types of NF-κB can induce inflammatory gene activation, canonical NF-κB signaling promotes senescence while noncanonical NF-κB signaling is able to evade it (13–17). To test whether HPyV6/7 and MCPyV sT might differ in their senescence response due to differential activation of NF-κB pathways, we assessed NF-κB signaling components by Western blot analysis. We first assessed the activation of the canonical NF-κB signaling pathway. HPyV6 and 7 sT–expressing cells showed increased levels of phosphor-p65 (RelA), which has been shown to be critical for the recruitment of CBP/p300 transcriptional coactivator complex and canonical NF-κB activation (39), consistent with activation of canonical NF-κB signaling. In contrast, we found that MCPyV sT decreased levels of phospho-p65 (RelA; Fig. 3A). Thus, consistent with studies showing that MCPyV sT specifically targets the IKK kinase adaptor protein NEMO (11), we find that MCPyV inhibits canonical NF-κB pathway activation. We next tested whether MCPyV sT increased levels of noncanonical NF-κB signaling components. Notably, in contrast to vector and HPyV6/7 sT–expressing cells, MCPyV sT resulted in the marked accumulation of both immature NFKB2 (p100) and its active p52 isoform and RELB (Fig. 3A). Because the LSD mutant reversed signatures of NF-κB activation and SASP, we tested whether the LSD motif was necessary for ncNF-κB pathway activation. Notably, the MCPyV sT-LSD mutant reversed immature NFKB2 (p100) accumulation, p52 processing, and RELB levels (Fig. 3B). In addition, phosphorylated IKKa, an additional marker suggestive of noncanonical NF-κB pathway activation, was elevated by WT MCPyV sT, but not the LSD mutant; total IKKα and IKKβ remained unchanged (Fig. 3B). Thus, in contrast to HPyV6/7, MCPyV activates noncanonical, and inhibits canonical, NF-κB signaling in an LSD-dependent manner.

Figure 3.

MCPyV sT activates ncNF-κB signaling through both transcriptional and posttranslational pathways. A, Western blot analyses of NF-κB components in PyV sT–transduced BJ cells (7 days) reveals induction of NF-κB by HPyV6/7 sT and induction of ncNF-κB by MCPyV sT. Arrows indicate mature p52/NFKB2 (top) and p50/NFKB1 (bottom) proteins. HSP90, loading control. B, Western blot analyses of NF-κB components in wild-type and LSD mutant MCPyV sT-transduced BJ fibroblasts (7 days) reveals induction of ncNF-κB by MCPyV sT requires the LSD domain. Arrow indicates mature p52 protein processed from immature p100 (NFKB2). HSP90, loading control. C, qRT-PCR analysis reveals significant upregulation of central ncNF-κB pathway genes, NFKB2 and RELB, in wild-type, but not LSD mutant, MCPyV sT–expressing BJ cells (5 days). Data are normalized by empty vector control and represent three biological replicates. D, qRT-PCR analysis reveals significant upregulation of H3K4me3 histone methyltransferases—MLL2, MLL4, SETD1A, and SETD1B—in wild-type, but not LSD mutant, MCPyV sT–expressing BJ cells (5 days). Data are normalized by empty vector control and represent three biological replicates. E, Western blot analyses reveals levels of of H3K4me3 are elevated in wild-type, but not LSD mutant, MCPyV sT–expressing BJ fibroblasts (5 days). HSP90, loading control. F, Subcellular fractionation of transduced BJ cells (5 days) reveals that wild-type, but not LSD mutant, MCPyV sT is present in the chromatin fraction. HSP90, cytoplasmic fraction control; HDAC2, nucleoplasmic control; H3, chromatin fraction control. G, Schematic indicating the location of putative MCPyV sT binding sequences in the NFKB2 and RELB promoter regions. H, FLAG or H3K4me3 ChIP-qPCR analyses reveal the enrichment of NFKB2 and RELB promoter regions in wild-type, but not LSD mutant, MCPyV sT–expressing BJ cells (5 days). Data are presented as percentage input and represent three biological replicates. I, Cells were first transduced with vector control or FBXW7 (5 days) and then vector or MCPyV sT (7 days) followed by Western blot analyses for indicated proteins. FBXW7 partially rescues increases in NFKB2 and RELB induced by sT. In D, E, and H, data are presented as mean ± SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Significance assessed by one-way ANOVA followed by Tukey post hoc analysis.

Figure 3.

MCPyV sT activates ncNF-κB signaling through both transcriptional and posttranslational pathways. A, Western blot analyses of NF-κB components in PyV sT–transduced BJ cells (7 days) reveals induction of NF-κB by HPyV6/7 sT and induction of ncNF-κB by MCPyV sT. Arrows indicate mature p52/NFKB2 (top) and p50/NFKB1 (bottom) proteins. HSP90, loading control. B, Western blot analyses of NF-κB components in wild-type and LSD mutant MCPyV sT-transduced BJ fibroblasts (7 days) reveals induction of ncNF-κB by MCPyV sT requires the LSD domain. Arrow indicates mature p52 protein processed from immature p100 (NFKB2). HSP90, loading control. C, qRT-PCR analysis reveals significant upregulation of central ncNF-κB pathway genes, NFKB2 and RELB, in wild-type, but not LSD mutant, MCPyV sT–expressing BJ cells (5 days). Data are normalized by empty vector control and represent three biological replicates. D, qRT-PCR analysis reveals significant upregulation of H3K4me3 histone methyltransferases—MLL2, MLL4, SETD1A, and SETD1B—in wild-type, but not LSD mutant, MCPyV sT–expressing BJ cells (5 days). Data are normalized by empty vector control and represent three biological replicates. E, Western blot analyses reveals levels of of H3K4me3 are elevated in wild-type, but not LSD mutant, MCPyV sT–expressing BJ fibroblasts (5 days). HSP90, loading control. F, Subcellular fractionation of transduced BJ cells (5 days) reveals that wild-type, but not LSD mutant, MCPyV sT is present in the chromatin fraction. HSP90, cytoplasmic fraction control; HDAC2, nucleoplasmic control; H3, chromatin fraction control. G, Schematic indicating the location of putative MCPyV sT binding sequences in the NFKB2 and RELB promoter regions. H, FLAG or H3K4me3 ChIP-qPCR analyses reveal the enrichment of NFKB2 and RELB promoter regions in wild-type, but not LSD mutant, MCPyV sT–expressing BJ cells (5 days). Data are presented as percentage input and represent three biological replicates. I, Cells were first transduced with vector control or FBXW7 (5 days) and then vector or MCPyV sT (7 days) followed by Western blot analyses for indicated proteins. FBXW7 partially rescues increases in NFKB2 and RELB induced by sT. In D, E, and H, data are presented as mean ± SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Significance assessed by one-way ANOVA followed by Tukey post hoc analysis.

Close modal

Activation of ncNF-κB signaling through H3K4me3 and FBXW7 inhibition

The RNA-seq data suggested that noncanonical NF-κB pathway members, NFKB2 (p100) and RELB, might be upregulated at the transcriptional level (Fig. 2B). qRT-PCR confirmed that that NFKB2 and RELB transcripts were upregulated by WT sT cells, but not the LSD mutant (Fig. 3C). Alterations in histone modifications have been implicated in the pathogenesis of MCC (40–42). Our GSEA confirmed chromatin and histone modification pathways to be upregulated compared with vector, HPyV6/7 sT, and LSD mutant MCPyV sT controls (Supplementary Fig. S9). Specifically, histone methyltransferases involved in H3K4 trimethylation (H3K4me3) were highly upregulated by GSEA (Fig. 2B). Upregulation of MLL2, MLL4, SETD1A, and SETD1B were confirmed by qRT-PCR (Fig. 3D). In addition, global levels of H3K4me3 were markedly upregulated by wild-type MCPyV sT, but not LSD mutant, expressing cells (Fig. 3E).

Because previous reports have demonstrated a role for sT in transcriptional regulation through chromatin remodeling (42), we used subcellular fractionation assays to verify that WT sT was found in cytoplasmic, nucleoplasmic, and chromatin-bound fractions (Fig. 3F). Notably, despite higher levels of expression in this system, we found that the LSD mutant abrogated the ability of sT to associate with chromatin (Fig. 3F). We identified multiple potential sT-binding motifs in the NFKB2 and RELB promoter regions (Fig. 3G). To determine whether sT might bind the NFKB2 and RELB promoters and promote H3K4me3 modification through direct binding, we performed ChIP-qPCR against both H3K4me3 and FLAG tagged sT antigen. In sT-expressing cells, both WT ST and the H3K4me3 mark are significantly upregulated at NFKB2 and RELB promoters. The LSD mutant, which does not activate noncanonical NF-κB, showed significantly less accumulation at these regions (Fig. 3H; Supplementary Fig. S10A).

Transcriptional upregulation of NFKB2 and RELB alone is insufficient to activate to NFKB2 ncNF-κB signaling, as the NFKB2 p100 must also be processed into its active p52 form (43). NFKB2 has been reported to be a target of FBXW7-mediated degradation and the loss of FBXW7 alone can result in the accumulation of the NFKB2 p100 precursor protein and ectopic activation of ncNF-κB pathway (44, 45). To determine whether MCPyV sTs previously reported ability to inhibit FBXW7 might contribute to ncNF-κB activation, we overexpressed FBXW7 in sT-expressing cells. The overexpression of FBXW7 decreased p100/NFKB2 accumulation and p52 processing (Fig. 3I). In addition, FBXW7 overexpression partially reversed MCPyV sT SASP gene expression (Supplementary Fig. S10B). The further expression of the LSD mutant completely blocked SASP cytokine expression, suggesting that the functions of the LSD motif were not entirely epistatic with FBXW7. In summary, we conclude that sT activates ncNF-κB signaling through promoting H3K4me3 methylation and transcriptional upregulation of NFKB2 and RELB and through inhibiting FBXW7-mediated degradation of NFKB2.

Chemical and genetic inhibition of ncNF-κB activation abrogates SASP activation

To confirm that ncNF-κB activation was required for SASP activation, we inhibited ncNF-κB activation with both chemical and genetic approaches. First, we employed calcitriol (1,25-dihydroxyvitamin D3), which inhibits ncNF-κB activation by inhibiting RELB expression (46, 47). The treatment of sT-infected cells with calcitriol for 7 days (Fig. 4A) reversed the induction of the SASP gene transcription (CCL26, CXCL10, GM-CSF, IL1B, IL6, and IL8) and secretion (IL8; Fig. 4B and C). We also confirmed the suppression of RELB protein by Western blot analysis (Fig. 4D). Unexpectedly, calcitriol treatment also increased phospho-p65 (RELA) levels, suggesting that canonical NF-κB signaling was activated by calcitriol (Fig. 4D). Because calcitriol functions as a vitamin D agonist, it has many off-target effects through its induction of vitamin D responsive (VDR) genes. Therefore, we used a complementary genetic approach and used multiple shRNAs to knockdown the ncNF-κB signaling components including IKKα, NIK, NFKB2, and RELB. Knockdowns were efficient and confirmed both qRT-PCR and Western blot analysis (Fig. 4E and F). The knockdown of ncNF-κB signaling impacted several features of MCPyV sT expression. Knockdown of NIK, NFKB2, and RELB all resulted in a significant reduction of SASP gene expression, providing additional evidence that the ncNF-κB pathway was critical for the activation of SASP by sT (Fig. 4G; Supplementary Fig. S11A). Decreased expression of EZH2 after knockdown of ncNF-κB signaling components was also noted, consistent with their reported role in EZH2 expression (refs. 16, 17; Supplementary Fig. S11B). Given a reported role for EZH2 in impacting cell proliferation, we also assessed whether the knockdown of ncNF-κB could impact the proliferation of MCPyV sT-expressing BJ cells. The knockdown of NKFB2 or RELB in these cells resulted in a significantly slower growth compared with an shGFP control, but only at late time points (>5 days) in vitro (Supplementary Fig. S11C).

Figure 4.

Chemical and genetic rescue of MCPyV sT induced ncNF-κB activation. A, Schematic of calcitriol (RELB inhibitor) rescue experiment. BJ fibroblasts were plated at approximately 50% confluence 1 day before infection. Cells were then spin-infected and puromycin selection was started 24 hours post spin infection along with 50 nmol/L calcitriol or vehicle control. Cells and supernatant were harvested 7 days postinfection for qRT-PCR, Western blot, and ELISA. B, Chemical inhibition of ncNF-κB with calcitriol reverses SASP gene induction by MCPyV sT as assessed by qRT-PCR. Data are normalized by vector + vehicle control and scaled to log10 fold change. C, ELISA of IL8 secretion by BJ cells reveals the inhibition of sT-mediated IL8 secretion by calcitriol. Data are normalized by vector + vehicle control and scaled to log10 fold change. D, Western blot analysis of NF-κB pathway components in calcitriol (50 nmol/L calcitriol) treated cells. Calcitriol induces canonical, but inhibits, noncanonical NF-κB signaling. E, BJ cells were transduced with the indicated shRNA construct (7 days) and knockdown of ncNF-κB pathway genes—IKKA, NFKB2, MAP3K14, and RELB—were assessed by qRT-PCR. Data are normalized by shGFP control and presented as three biological replicates. F, Western blot validates shRNA knockdown (7 days) of ncNF-κB pathway genes IKKA, NFKB2, MAP3K14, and RELB. G, BJ cells were first transduced with MCPyV sT (5 days), then the indicated shRNA (7 days). Induction of SASP genes by sT is partially rescued by shRNA knockdown of ncNF-κB pathway genes. Replicate shRNAs targeting the same gene were excluded for clarity and are detailed in Supplementary Fig. S9. Data are normalized by shGFP control and scaled to log10 fold change. Data are presented as mean ± SD for B–D and G. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Significance assessed by one-way ANOVA followed by Tukey post hoc analysis.

Figure 4.

Chemical and genetic rescue of MCPyV sT induced ncNF-κB activation. A, Schematic of calcitriol (RELB inhibitor) rescue experiment. BJ fibroblasts were plated at approximately 50% confluence 1 day before infection. Cells were then spin-infected and puromycin selection was started 24 hours post spin infection along with 50 nmol/L calcitriol or vehicle control. Cells and supernatant were harvested 7 days postinfection for qRT-PCR, Western blot, and ELISA. B, Chemical inhibition of ncNF-κB with calcitriol reverses SASP gene induction by MCPyV sT as assessed by qRT-PCR. Data are normalized by vector + vehicle control and scaled to log10 fold change. C, ELISA of IL8 secretion by BJ cells reveals the inhibition of sT-mediated IL8 secretion by calcitriol. Data are normalized by vector + vehicle control and scaled to log10 fold change. D, Western blot analysis of NF-κB pathway components in calcitriol (50 nmol/L calcitriol) treated cells. Calcitriol induces canonical, but inhibits, noncanonical NF-κB signaling. E, BJ cells were transduced with the indicated shRNA construct (7 days) and knockdown of ncNF-κB pathway genes—IKKA, NFKB2, MAP3K14, and RELB—were assessed by qRT-PCR. Data are normalized by shGFP control and presented as three biological replicates. F, Western blot validates shRNA knockdown (7 days) of ncNF-κB pathway genes IKKA, NFKB2, MAP3K14, and RELB. G, BJ cells were first transduced with MCPyV sT (5 days), then the indicated shRNA (7 days). Induction of SASP genes by sT is partially rescued by shRNA knockdown of ncNF-κB pathway genes. Replicate shRNAs targeting the same gene were excluded for clarity and are detailed in Supplementary Fig. S9. Data are normalized by shGFP control and scaled to log10 fold change. Data are presented as mean ± SD for B–D and G. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Significance assessed by one-way ANOVA followed by Tukey post hoc analysis.

Close modal

Activation of ncNF-κB promotes the growth of sT-expressing cells

We sought to characterize the effects of ncNF-κB on cell proliferation in greater detail. Senescent fibroblasts can promote cell proliferation in breast and prostate cancer in a paracrine manner through the secretion of SASP-associated cytokines (15, 48–50). Because many SASP cytokines have well-established, growth-promoting properties, we hypothesized that ncNF-κB might indirectly stimulate MCC proliferation in an autocrine fashion. Cells expressing MCPyV sT have the ability to grow in the absence of serum, a hallmark of cancer cells (24, 51). We confirmed that MCPyV sT fibroblasts, but not those transduced with the vector alone, could proliferate in low serum (2% FBS) conditions (Fig. 5A). The ability to proliferate in low serum was dependent on ncNF-κB activation, as shRNAs targeting RELB and NFKB2 markedly impaired proliferation in low serum, more notably than it had been impaired in normal serum (Fig. 5B; Supplementary Fig. S11C). The inhibition of ncNF-κB signaling using cell-permeable inhibitor significantly slowed the growth of MCPyV sT–expressing BJ fibroblasts in low serum (Supplementary Fig. S12A). In contrast, the chemical inhibition of canonical NF-κB had no effect (Supplementary Fig. S12B). To test whether secreted cytokines were important for cell proliferation, we blocked SASP cytokine processing and secretion using a commercial Protein Tranport Inhibitor Cocktail [PTI] (Thermo Fisher Scientific), containing a combination of Brefeldin A and Monensin. Cells treated with PTI are unable to secrete cytokines and should, therefore, be unable to promote cell proliferation through autocrine signaling. Indeed, blocking protein secretion significantly inhibited the proliferation of MCPyV sT–expressing cells in low (2%) serum, but had no effect in full (10%) serum, presumably due to the presence of excess growth factors in the serum (Fig. 5C). To directly test whether secreted factors present in the media were able to promote the proliferation of MCPyV-sT–expressing cells in low serum, we next tested whether conditioned media could rescue the growth of MCPyV sT–expressing cells in which protein secretion had been blocked. Notably, conditioned media generated by MCPyV sT–expressing BJ cells, but not vector control cells, significantly rescued the growth of MCPyV sT fibroblasts, even in the presence of PTI (Fig. 5D). Because HPyV6 and 7 sT-expressing cells also induce a SASP response, we next tested whether the conditioned media of these cells might be substituted for the conditioned media of MCPyV sT–expressing cells. Remarkably, while BJ cells expressing the LSD-mutant MCPyV sT were unable to proliferate efficiently in low serum, the addition of media harvested from wild-type MCPyV sT or HPyV6/7 sT–expressing fibroblasts all had the ability to significantly rescue the growth of the LSD MCPyV sT–expressing cells (Fig. 5E), albeit with different levels of efficacy. However, MCPyV sT fibroblasts still grew significantly better than vector control fibroblasts treated with PyV sT–conditioned media, indicating that the supplementation of SASP cytokines alone was not sufficient to promote fibroblast proliferation, consistent with the multiple transforming properties of MCPyV sT (Supplementary Fig. S12C). Thus, the induction of ncNF-κB signaling by MCPyV sT promotes the secretion of SASP-associated cytokines that promote autocrine signaling and cell proliferation.

Figure 5.

MCPyV sT induced ncNF-κB and SASP promotes proliferation through autocrine signaling. A, BJ fibroblasts were transduced with MCPyV sT or vector control (5 days). After selection, cells were plated in low serum (2% FBS) media and proliferation was assessed by MTT. B, BJ fibroblasts were first transduced with MCPyV sT (5 days) and then the indicated ncNF-κB signaling component shRNA (7 days). After selection, cells were plated in low serum (2% FBS) media and proliferation was assessed by MTT. Knockdown of NFKB2 or RELB significantly inhibited the proliferation of MCPyV sT-expressing fibroblasts compared with a GFP shRNA control. C, BJ cells transduced with MCPyV sT (5 days). After selection, cells were cultured in low (2% FBS) or normal (10% FBS) serum in the presence of a protein transport inhibitor (PTI) cocktail (Thermo Fisher Scientific) or a vehicle (DMSO) control and proliferation was assessed by MTT. Proliferation of MCPyV sT cells was significantly slower in the low (2%) serum media in the presence of the PTI cocktail. D, BJ cells were transduced with vector control or MCPyV sT (5 days). These cells were plated in low (2% FBS) serum media and this conditioned media was collected (2 days). BJ cells were separately transduced with MCPyV sT (5 days). After selection, cells were cultured in low (2% FBS) serum, a PTI cocktail, and the indicated conditioned media. Media harvested from MCPyV sT expressing, but not vector control, BJ cells rescues the proliferation of MCPyV sT-expressing BJ cells in low (2%) serum + PTI. E, BJ cells were transduced with HPyV6/7 sT or MCPyV sT (5 days). These cells were plated in low (2% FBS) serum media and this conditioned media was collected (2 days). BJ cells were separately transduced with wild-type or LSD mutant MCPyV sT (5 days). After selection, cells were cultured in low (2% FBS) serum, a PTI cocktail, and the indicated conditioned media. Media harvested from HPyV6 sT, HPyV7 sT, or MCPyV sT each significantly increases the proliferation of LSD mutant MCPyV sT-expressing BJ cells in low (2%) serum. For all growth curves, each timepoint was run as biological triplicates. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Significance assessed by one-way ANOVA followed by Tukey post hoc analysis.

Figure 5.

MCPyV sT induced ncNF-κB and SASP promotes proliferation through autocrine signaling. A, BJ fibroblasts were transduced with MCPyV sT or vector control (5 days). After selection, cells were plated in low serum (2% FBS) media and proliferation was assessed by MTT. B, BJ fibroblasts were first transduced with MCPyV sT (5 days) and then the indicated ncNF-κB signaling component shRNA (7 days). After selection, cells were plated in low serum (2% FBS) media and proliferation was assessed by MTT. Knockdown of NFKB2 or RELB significantly inhibited the proliferation of MCPyV sT-expressing fibroblasts compared with a GFP shRNA control. C, BJ cells transduced with MCPyV sT (5 days). After selection, cells were cultured in low (2% FBS) or normal (10% FBS) serum in the presence of a protein transport inhibitor (PTI) cocktail (Thermo Fisher Scientific) or a vehicle (DMSO) control and proliferation was assessed by MTT. Proliferation of MCPyV sT cells was significantly slower in the low (2%) serum media in the presence of the PTI cocktail. D, BJ cells were transduced with vector control or MCPyV sT (5 days). These cells were plated in low (2% FBS) serum media and this conditioned media was collected (2 days). BJ cells were separately transduced with MCPyV sT (5 days). After selection, cells were cultured in low (2% FBS) serum, a PTI cocktail, and the indicated conditioned media. Media harvested from MCPyV sT expressing, but not vector control, BJ cells rescues the proliferation of MCPyV sT-expressing BJ cells in low (2%) serum + PTI. E, BJ cells were transduced with HPyV6/7 sT or MCPyV sT (5 days). These cells were plated in low (2% FBS) serum media and this conditioned media was collected (2 days). BJ cells were separately transduced with wild-type or LSD mutant MCPyV sT (5 days). After selection, cells were cultured in low (2% FBS) serum, a PTI cocktail, and the indicated conditioned media. Media harvested from HPyV6 sT, HPyV7 sT, or MCPyV sT each significantly increases the proliferation of LSD mutant MCPyV sT-expressing BJ cells in low (2%) serum. For all growth curves, each timepoint was run as biological triplicates. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Significance assessed by one-way ANOVA followed by Tukey post hoc analysis.

Close modal

Noncanonical NF-κB and SASP activation in MCC cell lines and tumors

We next determined whether ncNF-κB signaling occurred in MCC cell lines. First, we assessed the activation of ncNF-κB in VP-MCC (MKL-1, MKL-2, MS-1, and WaGa) and VN-MCC (UISO, MCC13, MCC26) cell lines. VP-MCC, but not VN-MCC, lines showed increased levels of NFKB2, both activated p52 and its precursor p100, and RELB (Fig. 6A). In addition, VP-MCC lines show increased levels of global H3K4me3, consistent with its role in ncNF-κB activation (Fig. 6B). We noted the expression of sT to be lower in this isolate of MS-1, compared with other VP-MCC lines (Supplementary Fig. S13A), perhaps explaining the lower levels of H3K4me3 and ncNF-κB activation in this cell line. We next assessed whether ncNF-κB activation might be similarly upregulated in patient MCC tumors. We identified VP-MCC patient tumors and confirmed expression of MCPyV sT in the tumors, but not normal skin controls, with endpoint and qRT-PCR (Supplementary Fig. S13A–S13C). Compared with the normal skin controls, all seven patient samples were found to have increased levels of the NFKB2 precursor (p100) and RELB expression, with four samples showing strikingly higher expression of RELB. At least 6 of 7 samples showed detectable processing of NFKB2/p100 precursor protein (Fig. 6C). The variation in NFKB2/p52 and RELB levels suggests that in vivo factors, in addition to sT expression, may impact ncNF-κB activation. The expression of 6 SASP genes (CCL26, CXCL10, GM-CSF, IL1B, IL6, and IL8) were significantly elevated in VP-MCC, compared with VN-MCC cell lines (Fig. 6D). All SASP genes, except GM-CSF, were also significantly elevated in patient tumors compared with normal skin (Fig. 6E).

Figure 6.

ncNF-κB signaling is activated in VP-MCC, but not VN-MCC, cells and tumors. A, Western blot analyses of NFKB2 and RELB in VN-MCC and VP-MCC cell lines reveals activation of ncNF-κB in VP-MCC. Arrow indicates mature p52/NFKB2 processed from immature p100 precursor. HSP90, loading control. B, Western blot analyses of H3K4me3 levels of VN-MCC and VP-MCC lines. Cells were prepared as whole-cell lysates. H3, loading control. C, Western blot analyses of NFKB2 and RELB from patient VP-MCC tumor samples and normal skin controls. Arrow indicates mature p52 protein. HSP90, loading control. D, qRT-PCR analysis of relative expression of SASP genes (CCL26, CXCL10, GM-CSF, IL1B, IL6, and IL8) in VN-MCC and VP-MCC cancer cells. SASP genes are significantly upregulated in VP-MCC. Data are normalized to β-actin expression and each group run as triplicates. E, qRT-PCR analysis of SASP genes (CCL26, CXCL10, GM-CSF, IL1B, IL6, and IL8) from patient VP-MCC tumor samples and unaffected control skin. Data are normalized by β-actin expression. Each point represents one individual tumor. Data are presented as mean ± SD for (D and E). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Significance assessed by t test comparing virus-positive to virus-negative samples.

Figure 6.

ncNF-κB signaling is activated in VP-MCC, but not VN-MCC, cells and tumors. A, Western blot analyses of NFKB2 and RELB in VN-MCC and VP-MCC cell lines reveals activation of ncNF-κB in VP-MCC. Arrow indicates mature p52/NFKB2 processed from immature p100 precursor. HSP90, loading control. B, Western blot analyses of H3K4me3 levels of VN-MCC and VP-MCC lines. Cells were prepared as whole-cell lysates. H3, loading control. C, Western blot analyses of NFKB2 and RELB from patient VP-MCC tumor samples and normal skin controls. Arrow indicates mature p52 protein. HSP90, loading control. D, qRT-PCR analysis of relative expression of SASP genes (CCL26, CXCL10, GM-CSF, IL1B, IL6, and IL8) in VN-MCC and VP-MCC cancer cells. SASP genes are significantly upregulated in VP-MCC. Data are normalized to β-actin expression and each group run as triplicates. E, qRT-PCR analysis of SASP genes (CCL26, CXCL10, GM-CSF, IL1B, IL6, and IL8) from patient VP-MCC tumor samples and unaffected control skin. Data are normalized by β-actin expression. Each point represents one individual tumor. Data are presented as mean ± SD for (D and E). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Significance assessed by t test comparing virus-positive to virus-negative samples.

Close modal

Noncanonical NF-κB signaling is required in MCC xenograft growth

To test the role of ncNF-κB might in MCC in vivo, we used a specific, peptide inhibitor of ncNF-κB signaling, SN52 (52). The SN52 peptide penetrates cells and competitively inhibits NFKB2/p52 nuclear translocation and ncNF-κB activation through its consensus NFKB2/p52 NLS sequence (Supplementary Fig. S14A). It specifically inhibits ncNF-κB signaling in both cancer cells and immune cells (52–54). Consistent with the activation of ncNF-κB by MCPyV sT, the SN52 peptide slowed cell proliferation only in VP-MCC (MKL-1, MKL-2, MS-1, and WaGa), but not VN-MCC (MCC13, MCC26, and UISO) cell lines (Fig. 7A; Supplementary Fig. S14B). Next, VP-MCC lines (MKL-1 and WaGa) were xenografted onto NSG immunodeficient mice and the graft sites injected biweekly with the ncNF-κB inhibitor or a control peptide for 30 days (Fig. 7B). The inhibition of ncNF-κB significantly impaired both MKL-1 and WaGa xenograft growth, as assessed by tumor size and weight (Fig. 7CE). Consistent with the in vitro findings, the peptide inhibitor of ncNF-κB also significantly inhibited SASP gene (CCL26, CXCL10, GM-CSF, IL1B, IL6, and IL8) expression in the xenografts compared with the control peptide (Fig. 7F; Supplementary Fig. S14C). Histologic analyses revealed that control peptide–treated MKL-1 and WaGa xenografts showed sheets of tightly packed, monotonous tumor cells, with large nuclei and scant cytoplasm, consistent with the “small blue cell” histology typical of MCC. In contrast, ncNF-κB inhibited MKL-1 and WaGa tumors were smaller, possessed increased amounts of eosinophilic cytoplasm, and showed large areas with necrosis and loose stromal tissue (Fig. 7G; Supplementary Fig. S14D and S14E).

Figure 7.

ncNF-κB signaling is required for the growth of VP-MCC in vitro and in vivo. A, Proliferation of VP-MCC (MKL-1, MKL-2, MS-1, and WaGa) and VN-MCC (MCC13, MCC26) cell lines treated with a peptide inhibitor of ncNF-κB inhibitor (SN52) or a control scrambled peptide (SN52mut). One-thousand cells plated per well; peptide (35 μg/mL) added on day 0. Proliferation assessed at the indicated time point by XTT assay. Each time point was run as triplicates. VP-MCC, but not VN-MCC, is sensitive to the inhibition of ncNF-κB. B, Schematic of VP-MCC (MKL-1, WaGa) cancer cell xenograft and ncNF-κB inhibitor (SN52) treatment in NSG immunodeficient mice. Cells (1 × 107 cells) were xenografted onto the flanks of NSG mice (n = 8 per condition). C, Representative images of tumor-bearing mice and dissected MKL-1 and WaGa tumors (30 days) after xenograft and biweekly treatment with bearing mice images of MKL-1 and WaGa cell lines grown for 1 month receiving ncNF-κB inhibitor or control peptide treatment subcutaneously. Inhibition of ncNF-κB (SN52) significantly inhibits the growth of MKL-1 and WaGa xenografts. D, Weight of MKL-1 xenografted tumors receiving an ncNF-κB inhibitor or control peptide. Each point represents an individual tumor. E, Weight of WaGa xenografted tumors receiving an ncNF-κB inhibitor or control peptide. Each point represents an individual tumor. F, qRT-PCR analysis of a subset of SASP genes—IL1B, IL6, IL8—in MKL-1 and WaGa cells xenografts treated with an ncNF-κB inhibitor or control peptide. SASP gene expression is significantly inhibited by the peptide ncNF-κB inhibitor. Each point represents an individual tumor. Additional SASP genes in Supplementary Fig. S12. G, Representative images of H&E-stained tumors formed by MKL-1 and WaGa xenografts treated with an ncNF-κB inhibitor (SN52) or control peptide (SN52mut). Scale bar, 200 μm. Data are presented as mean ± SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Significance assessed by one-way ANOVA followed by Tukey post hoc analysis for A and t test (D–F).

Figure 7.

ncNF-κB signaling is required for the growth of VP-MCC in vitro and in vivo. A, Proliferation of VP-MCC (MKL-1, MKL-2, MS-1, and WaGa) and VN-MCC (MCC13, MCC26) cell lines treated with a peptide inhibitor of ncNF-κB inhibitor (SN52) or a control scrambled peptide (SN52mut). One-thousand cells plated per well; peptide (35 μg/mL) added on day 0. Proliferation assessed at the indicated time point by XTT assay. Each time point was run as triplicates. VP-MCC, but not VN-MCC, is sensitive to the inhibition of ncNF-κB. B, Schematic of VP-MCC (MKL-1, WaGa) cancer cell xenograft and ncNF-κB inhibitor (SN52) treatment in NSG immunodeficient mice. Cells (1 × 107 cells) were xenografted onto the flanks of NSG mice (n = 8 per condition). C, Representative images of tumor-bearing mice and dissected MKL-1 and WaGa tumors (30 days) after xenograft and biweekly treatment with bearing mice images of MKL-1 and WaGa cell lines grown for 1 month receiving ncNF-κB inhibitor or control peptide treatment subcutaneously. Inhibition of ncNF-κB (SN52) significantly inhibits the growth of MKL-1 and WaGa xenografts. D, Weight of MKL-1 xenografted tumors receiving an ncNF-κB inhibitor or control peptide. Each point represents an individual tumor. E, Weight of WaGa xenografted tumors receiving an ncNF-κB inhibitor or control peptide. Each point represents an individual tumor. F, qRT-PCR analysis of a subset of SASP genes—IL1B, IL6, IL8—in MKL-1 and WaGa cells xenografts treated with an ncNF-κB inhibitor or control peptide. SASP gene expression is significantly inhibited by the peptide ncNF-κB inhibitor. Each point represents an individual tumor. Additional SASP genes in Supplementary Fig. S12. G, Representative images of H&E-stained tumors formed by MKL-1 and WaGa xenografts treated with an ncNF-κB inhibitor (SN52) or control peptide (SN52mut). Scale bar, 200 μm. Data are presented as mean ± SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Significance assessed by one-way ANOVA followed by Tukey post hoc analysis for A and t test (D–F).

Close modal

In addition to promoting inflammation and regulating immune cell development, noncanonical NF-κB signaling can promote cell proliferation and survival (55). Thus, many viruses have deviously coopted the pathway to promote their own replication and dissemination. Notably, the oncoproteins of several DNA tumors viruses target ncNF-κB, including the Tax protein of human T-cell leukemia virus, the LMP1 protein of the Epstein–Barr virus, and the vFLIP protein of the Kaposi sarcoma virus (37, 43). Here, we describe the first example of a polyomavirus protein activating ncNF-κB (Supplementary Fig. S15). Clearly, not all polyomaviruses share the ability to activate ncNF-κB signaling. Specifically, we find that HPyV6/7 sT activates canonical NF-κB signaling, while MCPyV sT appears unique in its ability to inactivate canonical and activate ncNF-κB signaling. These differences are critical to the transforming properties of MCPyV sT and suggest that the activation of ncNF-κB may be a feature that is conserved among DNA viruses with transforming properties. In addition to the striking differences in NF-κB pathway choice, we also find that HPyV6/7 sT, unlike MCPyV sT, do not stabilize or activate Myc signaling. The regulation of both ncNF-κB signaling and c-Myc converge on the LSD motif present in MCPyV sT and reinforce the importance of this motif in promoting transformation.

This study provides a framework for understanding previously disparate observations about the biology of MCC tumors. Specifically, the specificity of MCPyV sT for only one arm of the NF-κB signaling pathway explains why ncNF-κB pathway components (i.e., IKKα and MAP3K14) are among the most highly upregulated genes, while canonical NF-κB (i.e., IKKβ and NEMO) are among the most significantly downregulated genes, in VP-MCC compared with VN-MCC (56). It also helps to explain why NF-κB signaling has been reported to be both inhibited and activated in different studies (11, 12). Our findings also provide mechanistic insight to previous observations that MCPyV sT could increase inflammatory cytokine expression when expressed in conjunction with MCC tumor–derived LT (57). In concert with observations that the transcriptional upregulation of EZH2 is a key mechanism through which NFKB2 is able to bypass senescence (16, 17), this study provides an explanation for EZH2 overexpression in many MCCs and suggests why it may be associated with a worse prognosis (41, 58). We further predict that EZH2-overexpressing MCCs will overwhelmingly be virus-positive. Finally, telomerase activation has also been identified as a downstream target of ncNF-κB signaling. Specifically, in cooperation with ETS1/2, NFKB2/p52 has been shown to bind and promote the overexpression of the catalytic subunit of telomerase (TERT) in cells that contain the C250T(-146C>T), but not the C228T(-124C>T), TERT promoter mutation (59). MCCs possess a higher rate of C250T promoter mutations than most other cancers (60–63). Thus, we suggest that the activation of ncNF-κB signaling early in MCC tumorigenesis selects for the C250T promoter mutation at later stages of tumor development.

Given its oncogenic potential, it is not surprising that the activation of ncNF-κB is tightly regulated. MCPyV sT induces global increases in H3K4me3 levels resulting in an increase in NFKB2 and RELB transcription. Although this transcriptional upregulation increases the pool of available NFKB2/p100 precursor and its binding partner RELB, processing of NFKB2 to its active p52 form is still necessary to activate ncNF-κB signaling. We find that MCPyV sT's ability to inhibit FBXW7 through its LSD domain and induce NFKB2 stabilization and cleavage is critical for activation of ncNF-κB signaling in fibroblasts (Fig. 3I). However, MCC tumors from patients showed marked variations in both RELB levels and NFKB2/p52 activation, suggesting that additional in vivo signals contribute to the tuning of ncNF-κB activation.

While checkpoint inhibitors are effective in many cases of advanced MCC, new therapies are still needed for progressive or recurrent disease. Inhibiting ncNF-κB signaling could be a novel therapeutic approach for VP-MCC tumors. While peptide inhibitors of NFKB2 have been used in mouse models, small-molecule inhibitors of ncNF-κB, similar to those now available for canonical NF-κB, would be more feasible in the clinical setting. These studies also revealed that the activation of vitamin D receptor (VDR) signaling with calcitriol could inhibit RELB and reduce the expression of ncNF-κB target genes. As low calcemic analogues of calcitriol have shown some benefits to patients with a wide variety of cancers in early-stage clinical trials (64, 65), it may be worth exploring whether this pathway could be exploited in VP-MCC. Notably, MCCs express the VDR, and patients with vitamin D deficiency show significantly larger tumor size and worse prognosis (66). In summary, the identification of ncNF-κB activation as a novel signaling pathway in VP-MCC presents novel opportunities for understanding and treating this deadly cancer.

No potential conflicts of interest were disclosed.

J. Zhao: Conceptualization, formal analysis, investigation, visualization, methodology, writing-original draft. Y. Jia: Resources. S. Shen: Formal analysis. J. Kim: Data curation, software, formal analysis. X. Wang: Resources. E. Lee: Resources, investigation. I. Brownell: Resources. J.H. Cho-Vega: Resources. C. Lewis: Resources. J. Homsi: Resources, writing-review and editing. R.R. Sharma: Resources. R.C. Wang: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, investigation, visualization, methodology, writing-original draft, project administration, writing-review and editing.

We thank Dr. Jerry Shay for BJ fibroblasts; Dr. Chris Buck for 1t1 antibody; Dr. Shuyuan Zhang for Tet-on doxycycline-inducible vector; Drs. Masa Shuda, Paul Nghiem, and Juergen Becker for the MCC cell lines. We thank Dr. Chris Buck for critically reviewing the manuscript. This work was supported by grants from the UT Southwestern Cary Council, ACS (RSG-18-058-01), and NIAMS (R01AR072655; to R.C. Wang).

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.
Feng
H
,
Shuda
M
,
Chang
Y
,
Moore
PS
. 
Clonal integration of a polyomavirus in human Merkel cell carcinoma
.
Science
2008
;
319
:
1096
100
2.
Mitteldorf
C
,
Berisha
A
,
Tronnier
M
,
Pfaltz
MC
,
Kempf
W
. 
PD-1 and PD-L1 in neoplastic cells and the tumor microenvironment of Merkel cell carcinoma
.
J Cutan Pathol
2017
;
44
:
740
6
.
3.
Nghiem
PT
,
Bhatia
S
,
Lipson
EJ
,
Kudchadkar
RR
,
Miller
NJ
,
Annamalai
L
, et al
PD-1 blockade with pembrolizumab in advanced merkel-cell carcinoma
.
N Engl J Med
2016
;
374
:
2542
52
.
4.
D'Angelo
SP
,
Russell
J
,
Lebbe
C
,
Chmielowski
B
,
Gambichler
T
,
Grob
JJ
, et al
Efficacy and safety of first-line avelumab treatment in patients with stage iv metastatic merkel cell carcinoma: a preplanned interim analysis of a clinical trial
.
JAMA Oncol
2018
;
4
:
e180077
.
5.
Nguyen
KD
,
Chamseddin
BH
,
Cockerell
CJ
,
Wang
RC
. 
The biology and clinical features of cutaneous polyomaviruses
.
J Invest Dermatol
2019
;
139
:
285
92
.
6.
Kwun
HJ
,
Shuda
M
,
Feng
H
,
Camacho
CJ
,
Moore
PS
,
Chang
Y
. 
Merkel cell polyomavirus small T antigen controls viral replication and oncoprotein expression by targeting the cellular ubiquitin ligase SCFFbw7
.
Cell Host Microbe
2013
;
14
:
125
35
.
7.
Kwun
HJ
,
Guastafierro
A
,
Shuda
M
,
Meinke
G
,
Bohm
A
,
Moore
PS
, et al
The minimum replication origin of merkel cell polyomavirus has a unique large T-antigen loading architecture and requires small T-antigen expression for optimal replication
.
J Virol
2009
;
83
:
12118
28
.
8.
Ali
SH
,
DeCaprio
JA
. 
Cellular transformation by SV40 large T antigen: interaction with host proteins
.
Semin Cancer Biol
2001
;
11
:
15
23
.
9.
Shuda
M
,
Chang
Y
,
Moore
PS
. 
Merkel cell polyomavirus-positive Merkel cell carcinoma requires viral small T-antigen for cell proliferation
.
J Invest Dermatol
2014
;
134
:
1479
81
.
10.
Akhoondi
S
,
Sun
D
,
von der Lehr
N
,
Apostolidou
S
,
Klotz
K
,
Maljukova
A
, et al
FBXW7/hCDC4 is a general tumor suppressor in human cancer
.
Cancer Res
2007
;
67
:
9006
12
.
11.
Griffiths
DA
,
Abdul-Sada
H
,
Knight
LM
,
Jackson
BR
,
Richards
K
,
Prescott
EL
, et al
Merkel cell polyomavirus small T antigen targets the NEMO adaptor protein to disrupt inflammatory signaling
.
J Virol
2013
;
87
:
13853
67
.
12.
Berrios
C
,
Padi
M
,
Keibler
MA
,
Park
DE
,
Molla
V
,
Cheng
J
, et al
Merkel cell polyomavirus small T antigen promotes pro-glycolytic metabolic perturbations required for transformation
.
PLoS Pathog
2016
;
12
:
e1006020
.
13.
Capece
D
,
Verzella
D
,
Tessitore
A
,
Alesse
E
,
Capalbo
C
,
Zazzeroni
F
. 
Cancer secretome and inflammation: the bright and the dark sides of NF-kappaB
.
Semin Cell Dev Biol
2018
;
78
:
51
61
.
14.
Salminen
A
,
Kauppinen
A
,
Kaarniranta
K
. 
Emerging role of NF-kappaB signaling in the induction of senescence-associated secretory phenotype (SASP)
.
Cell Signal
2012
;
24
:
835
45
.
15.
Coppe
JP
,
Desprez
PY
,
Krtolica
A
,
Campisi
J
. 
The senescence-associated secretory phenotype: the dark side of tumor suppression
.
Ann Rev Pathol
2010
;
5
:
99
118
16.
De Donatis
GM
,
Le Pape
E
,
Pierron
A
,
Cheli
Y
,
Hofman
V
,
Hofman
P
, et al
NF-kB2 induces senescence bypass in melanoma via a direct transcriptional activation of EZH2
.
Oncogene
2016
;
35
:
2813
17.
Iannetti
A
,
Ledoux
AC
,
Tudhope
SJ
,
Sellier
H
,
Zhao
B
,
Mowla
S
, et al
Regulation of p53 and Rb links the alternative NF-kappaB pathway to EZH2 expression and cell senescence
.
PLoS Genet
2014
;
10
:
e1004642
.
18.
Zhao
J
,
Lee
EE
,
Kim
J
,
Yang
R
,
Chamseddin
B
,
Ni
C
, et al
Transforming activity of an oncoprotein-encoding circular RNA from human papillomavirus
.
Nat Commun
2019
;
10
:
2300
.
19.
Patel
RK
,
Jain
M
. 
NGS QC Toolkit: a toolkit for quality control of next generation sequencing data
.
PLoS One
2012
;
7
:
e30619
.
20.
Dobin
A
,
Davis
CA
,
Schlesinger
F
,
Drenkow
J
,
Zaleski
C
,
Jha
S
, et al
STAR: ultrafast universal RNA-seq aligner
.
Bioinformatics
2013
;
29
:
15
21
.
21.
Anders
S
,
Pyl
PT
,
Huber
W
. 
HTSeq–a Python framework to work with high-throughput sequencing data
.
Bioinformatics
2015
;
31
:
166
9
.
22.
Gentleman
RC
,
Carey
VJ
,
Bates
DM
,
Bolstad
B
,
Dettling
M
,
Dudoit
S
, et al
Bioconductor: open software development for computational biology and bioinformatics
.
Genome Biol
2004
;
5
:
R80
.
23.
Anders
S
,
Huber
W
. 
Differential expression analysis for sequence count data
.
Genome Biol
2010
;
11
:
R106
.
24.
Shuda
M
,
Kwun
HJ
,
Feng
H
,
Chang
Y
,
Moore
PS
. 
Human Merkel cell polyomavirus small T antigen is an oncoprotein targeting the 4E-BP1 translation regulator
.
J Clin Invest
2011
;
121
:
3623
34
.
25.
Liu
W
,
Yang
R
,
Payne
AS
,
Schowalter
RM
,
Spurgeon
ME
,
Lambert
PF
, et al
Identifying the target cells and mechanisms of Merkel cell polyomavirus infection
.
Cell Host Microbe
2016
;19:775–87.
26.
Fan
T
,
Jiang
S
,
Chung
N
,
Alikhan
A
,
Ni
C
,
Lee
CC
, et al
EZH2-dependent suppression of a cellular senescence phenotype in melanoma cells by inhibition of p21/CDKN1A expression
.
Mol Cancer Res
2011
;
9
:
418
29
.
27.
Bracken
AP
,
Kleine-Kohlbrecher
D
,
Dietrich
N
,
Pasini
D
,
Gargiulo
G
,
Beekman
C
, et al
The Polycomb group proteins bind throughout the INK4A-ARF locus and are disassociated in senescent cells
.
Genes Dev
2007
;
21
:
525
30
.
28.
Barradas
M
,
Anderton
E
,
Acosta
JC
,
Li
S
,
Banito
A
,
Rodriguez-Niedenfuhr
M
, et al
Histone demethylase JMJD3 contributes to epigenetic control of INK4a/ARF by oncogenic RAS
.
Genes Dev
2009
;
23
:
1177
82
.
29.
Nguyen
KD
,
Lee
EE
,
Yue
Y
,
Štork
J
,
Pock
L
,
North
JP
, et al
Human polyomavirus 6 and 7 are associated with pruritic and dyskeratotic dermatoses
.
J Am Acad Dermatol
2017
;
76
:
932
40.e3
.
30.
Harvey
DM
,
Levine
AJ.
. 
p53 alteration is a common event in the spontaneous immortalization of primary BALB/c murine embryo fibroblasts
.
Genes Dev
1991
;
5
:
2375
85
.
31.
O'Connor
PM
,
Jackman
J
,
Bae
I
,
Myers
TG
,
Fan
S
,
Mutoh
M
, et al
Characterization of the p53 tumor suppressor pathway in cell lines of the National Cancer Institute anticancer drug screen and correlations with the growth-inhibitory potency of 123 anticancer agents
.
Cancer Res
1997
;
57
:
4285
300
.
32.
Hsu
IC
,
Tokiwa
T
,
Bennett
W
,
Metcalf
RA
,
Welsh
JA
,
Sun
T
, et al
p53 gene mutation and integrated hepatitis B viral DNA sequences in human liver cancer cell lines
.
Carcinogenesis
1993
;
14
:
987
92
.
33.
Chien
Y
,
Scuoppo
C
,
Wang
X
,
Fang
X
,
Balgley
B
,
Bolden
JE
, et al
Control of the senescence-associated secretory phenotype by NF-kappaB promotes senescence and enhances chemosensitivity
.
Genes Dev
2011
;
25
:
2125
36
.
34.
Kwun
HJ
,
Shuda
M
,
Camacho
CJ
,
Gamper
AM
,
Thant
M
,
Chang
Y
, et al
Restricted protein phosphatase 2A targeting by Merkel cell polyomavirus small T antigen
.
J Virol
2015
;
89
:
4191
200
.
35.
Dye
KN
,
Welcker
M
,
Clurman
BE
,
Roman
A
,
Galloway
DA
. 
Merkel cell polyomavirus Tumor antigens expressed in Merkel cell carcinoma function independently of the ubiquitin ligases Fbw7 and beta-TrCP
.
PLoS Pathog
2019
;
15
:
e1007543
.
36.
Kwun
HJ
,
Wendzicki
JA
,
Shuda
Y
,
Moore
PS
,
Chang
Y
. 
Merkel cell polyomavirus small T antigen induces genome instability by E3 ubiquitin ligase targeting
.
Oncogene
2017
;
36
:
6784
92
.
37.
Sun
SC.
. 
The non-canonical NF-kappaB pathway in immunity and inflammation
.
Nat Rev Immunol
2017
;
17
:
545
58
.
38.
Chen
LF
,
Williams
SA
,
Mu
Y
,
Nakano
H
,
Duerr
JM
,
Buckbinder
L
, et al
NF-kappaB RelA phosphorylation regulates RelA acetylation
.
Mol Cell Biol
2005
;
25
:
7966
75
.
39.
Madrid
LV
,
Mayo
MW
,
Reuther
JY
,
Baldwin
AS
 Jr.
. 
Akt stimulates the transactivation potential of the RelA/p65 subunit of NF-kappa B through utilization of the Ikappa B kinase and activation of the mitogen-activated protein kinase p38
.
J Biol Chem
2001
;
276
:
18934
40
.
40.
Busam
KJ
,
Pulitzer
MP
,
Coit
DC
,
Arcila
M
,
Leng
D
,
Jungbluth
AA
, et al
Reduced H3K27me3 expression in Merkel cell polyoma virus-positive tumors
.
Mod Pathol
2017
;
30
:
877
83
.
41.
Harms
KL
,
Chubb
H
,
Zhao
L
,
Fullen
DR
,
Bichakjian
CK
,
Johnson
TM
, et al
Increased expression of EZH2 in Merkel cell carcinoma is associated with disease progression and poorer prognosis
.
Hum Pathol
2017
;
67
:
78
84
.
42.
Cheng
J
,
Park
DE
,
Berrios
C
,
White
EA
,
Arora
R
,
Yoon
R
, et al
Merkel cell polyomavirus recruits MYCL to the EP400 complex to promote oncogenesis
.
PLoS Pathog
2017
;
13
:
e1006668
.
43.
Cildir
G
,
Low
KC
,
Tergaonkar
V
. 
Noncanonical NF-kappaB signaling in health and disease
.
Trends Mol Med
2016
;
22
:
414
29
.
44.
Arabi
A
,
Ullah
K
,
Branca
RM
,
Johansson
J
,
Bandarra
D
,
Haneklaus
M
, et al
Proteomic screen reveals Fbw7 as a modulator of the NF-kappaB pathway
.
Nat Commun
2012
;
3
:
976
.
45.
Busino
L
,
Millman
SE
,
Scotto
L
,
Kyratsous
CA
,
Basrur
V
,
O'Connor
O
, et al
Fbxw7alpha- and GSK3-mediated degradation of p100 is a pro-survival mechanism in multiple myeloma
.
Nat Cell Biol
2012
;
14
:
375
85
.
46.
Mineva
ND
,
Wang
X
,
Yang
S
,
Ying
H
,
Xiao
ZX
,
Holick
MF
, et al
Inhibition of RelB by 1,25-dihydroxyvitamin D3 promotes sensitivity of breast cancer cells to radiation
.
J Cell Physiol
2009
;
220
:
593
9
.
47.
Xu
Y
,
Fang
F
,
St Clair
DK
,
Josson
S
,
Sompol
P
,
Spasojevic
I
, et al
Suppression of RelB-mediated manganese superoxide dismutase expression reveals a primary mechanism for radiosensitization effect of 1alpha,25-dihydroxyvitamin D(3) in prostate cancer cells
.
Mol Cancer Ther
2007
;
6
:
2048
56
.
48.
Barcellos-Hoff
MH
,
Ravani
SA.
. 
Irradiated mammary gland stroma promotes the expression of tumorigenic potential by unirradiated epithelial cells
.
Cancer Res
2000
;
60
:
1254
60
.
49.
Tsai
KK
,
Chuang
EY
,
Little
JB
,
Yuan
ZM
. 
Cellular mechanisms for low-dose ionizing radiation-induced perturbation of the breast tissue microenvironment
.
Cancer Res
2005
;
65
:
6734
44
.
50.
Bavik
C
,
Coleman
I
,
Dean
JP
,
Knudsen
B
,
Plymate
S
,
Nelson
PS
. 
The gene expression program of prostate fibroblast senescence modulates neoplastic epithelial cell proliferation through paracrine mechanisms
.
Cancer Res
2006
;
66
:
794
802
.
51.
Hanahan
D
,
Weinberg
RA
. 
Hallmarks of cancer: the next generation
.
Cell
2011
;
144
:
646
74
.
52.
Xu
Y
,
Fang
F
,
St Clair
DK
,
Sompol
P
,
Josson
S
,
St Clair
WH
. 
SN52, a novel nuclear factor-kappaB inhibitor, blocks nuclear import of RelB:p52 dimer and sensitizes prostate cancer cells to ionizing radiation
.
Mol Cancer Ther
2008
;
7
:
2367
76
.
53.
Hou
Y
,
Liang
H
,
Rao
E
,
Zheng
W
,
Huang
X
,
Deng
L
, et al
Non-canonical NF-kappaB Antagonizes STING sensor-mediated DNA sensing in radiotherapy
.
Immunity
2018
;
49
:
490
503
e4
.
54.
Wang
J
,
Ferreira
R
,
Lu
W
,
Farrow
S
,
Downes
K
,
Jermutus
L
, et al
TNFR2 ligation in human T regulatory cells enhances IL2-induced cell proliferation through the non-canonical NF-kappaB pathway
.
Sci Rep
2018
;
8
:
12079
.
55.
Rahman
MM
,
McFadden
G
. 
Modulation of NF-kappaB signalling by microbial pathogens
.
Nat Rev Microbiol
2011
;
9
:
291
306
.
56.
Arora
R
,
Shuda
M
,
Guastafierro
A
,
Feng
H
,
Toptan
T
,
Tolstov
Y
, et al
Survivin is a therapeutic target in Merkel cell carcinoma
.
Sci Transl Med
2012
;
4
:
133ra56
.
57.
Richards
KF
,
Guastafierro
A
,
Shuda
M
,
Toptan
T
,
Moore
PS
,
Chang
Y
. 
Merkel cell polyomavirus T antigens promote cell proliferation and inflammatory cytokine gene expression
.
J Gen Virol
2015
;
96
:
3532
44
.
58.
Veija
T
,
Koljonen
V
,
Bohling
T
,
Kero
M
,
Knuutila
S
,
Sarhadi
VK
. 
Aberrant expression of ALK and EZH2 in Merkel cell carcinoma
.
BMC Cancer
2017
;
17
:
236
.
59.
Li
Y
,
Zhou
QL
,
Sun
W
,
Chandrasekharan
P
,
Cheng
HS
,
Ying
Z
, et al
Non-canonical NF-kappaB signalling and ETS1/2 cooperatively drive C250T mutant TERT promoter activation
.
Nat Cell Biol
2015
;
17
:
1327
38
.
60.
Maryoung
L
,
Yue
Y
,
Young
A
,
Newton
CA
,
Barba
C
,
van Oers
NS
, et al
Somatic mutations in telomerase promoter counterbalance germline loss-of-function mutations
.
J Clin Invest
2017
;
127
:
982
6
.
61.
Vinagre
J
,
Almeida
A
,
Populo
H
,
Batista
R
,
Lyra
J
,
Pinto
V
, et al
Frequency of TERT promoter mutations in human cancers
.
Nat Commun
2013
;
4
:
2185
.
62.
Xie
H
,
Liu
T
,
Wang
N
,
Bjornhagen
V
,
Hoog
A
,
Larsson
C
, et al
TERT promoter mutations and gene amplification: promoting TERT expression in Merkel cell carcinoma
.
Oncotarget
2014
;
5
:
10048
57
.
63.
Hosler
GA
,
Davoli
T
,
Mender
I
,
Litzner
B
,
Choi
J
,
Kapur
P
, et al
A primary melanoma and its asynchronous metastasis highlight the role of BRAF, CDKN2A, and TERT
.
J Cutan Pathol
2015
;
42
:
108
17
.
64.
Leyssens
C
,
Verlinden
L
,
Verstuyf
A
. 
Antineoplastic effects of 1,25(OH)2D3 and its analogs in breast, prostate and colorectal cancer
.
Endocr Relat Cancer
2013
;
20
:
R31
47
.
65.
Duffy
MJ
,
Murray
A
,
Synnott
NC
,
O'Donovan
N
,
Crown
J
. 
Vitamin D analogues: potential use in cancer treatment
.
Crit Rev Oncol Hematol
2017
;
112
:
190
7
.
66.
Samimi
M
,
Touze
A
,
Laude
H
,
Le Bidre
E
,
Arnold
F
,
Carpentier
A
, et al
Vitamin D deficiency is associated with greater tumor size and poorer outcome in Merkel cell carcinoma patients
.
J Eur Acad Dermatol Venereol
2014
;
28
:
298
308
.