Non-canonical olfactory pathway activation induces cell fusion of cervical cancer cells

Abstract

Multinucleation occurs in various types of advanced cancers and contributes to their malignant characteristics, including anticancer drug resistance. Therefore, inhibiting multinucleation can improve cancer prognosis; however, the molecular mechanisms underlying multinucleation remain elusive. Here, we introduced a genetic mutation in cervical cancer cells to induce cell fusion-mediated multinucleation. The olfactory receptor OR1N2 was heterozygously mutated in these fused cells; the same OR1N2 mutation was detected in multinucleated cells from clinical cervical cancer specimens. The mutation-induced structural change in the OR1N2 protein activated protein kinase A (PKA), which, in turn, mediated the non-canonical olfactory pathway. PKA phosphorylated and activated furin protease, resulting in the cleavage of the fusogenic protein syncytin-1. Because this cleaved form of syncytin-1, processed by furin, participates in cell fusion, furin inhibitors could suppress multinucleation and reduce surviving cell numbers after anticancer drug treatment. The improved anticancer drug efficacy indicates a promising therapeutic approach for advanced cervical cancers.

Keywords: Anticancer drug resistance; Cell fusion; Cervical cancer; Furin; Olfactory pathway; PKA.

Graphical abstract

Fig. 1

Establishment of cervical cancer cell lines exhibiting multinucleation. (A–E) HeLa cells were transfected with CBE4 cytosine base editor plasmid, and then single cell-derived clones were expanded. (A) Cells from clones 1 and 2 were immunostained with an anti-pan cadherin antibody (green). DAPI for nuclear staining (blue). The bottom images show magnifications of the boxed areas in the top images. Scale bar, 200 μm. (B) Box plots present the percentage of cells with two (left) or more than two (right) nuclei versus the total number of cells (≥1000 cells) (n = 5 independent experiments). **p < 0.01 by Student's t test. (C) Serial images for cell fusion were captured from clone 2 cells. Scale bar, 20 μm. Arrowheads indicate the position of cell fusion. (D and E) Serial images of cell division were captured from multinucleated (D) and mononuclear (E) CF (clone 2) cells. The number of nuclei in the cells is shown. Scale bars, 20 μm. (F and G) UCC tissues (F, stage ⅠB2; G, stage ⅡA) and NATs were excised from patients with cancer. Hematoxylin and eosin stained images (left) and magnified images of the boxed areas in the left images (right) are shown. Arrowheads indicate the multinucleated cells. Scale bars, 200 μm.

Fig. 2

The olfactory receptor OR1N2 gene is mutated in multinucleated cervical cancer cells. (A) Whole exome sequencing of NF (clone 1) and CF (clone 2) cells was performed. Venn diagram illustrates the total counts and comparison of nsSNVs in the coding sequence (CDS) region. (B) Direct sequencing of coding and template strands of genomic DNA fragments, including codon 100 of OR1N2 (underline) from NF and CF clones. (C and D) Direct sequencing of genomic DNA fragments, including codon 100 of OR1N2 from UCC tissues and NATs (C, stage ⅠB2; D, stage ⅡA). (E) Parental HeLa cells, NF cells, and CF cells were lysed and separated into soluble and insoluble fractions. The insoluble fractions were then lysed directly in 1 × Laemmli sample buffer, sonicated, and analyzed by immunoblotting to evaluate the protein expression level of OR1N2. NS points to a nonspecific band that displays equal protein loading. (F) The amino acid sequence of human OR1N2 protein surrounding the third and fourth predicted transmembrane (TM) domains, as well as the first and second extracellular loop (ECL) and the second intracellular loop (ICL) regions. Asterisks indicate cysteine residues required for the disulfide bond formation. (G) Ribbon diagrams of OR1N2 WT (blue) and C100R mutant (yellow) viewed from the extracellular space were obtained using AlphaFold2. Red spheres represent disulfide bonds. (H) The superposition between ribbon diagrams of OR1N2 WT (blue) and C100R mutant (yellow) are shown as viewpoints parallel to the membrane and from the intracellular space. (I) The model represents structural and transitional differences between the OR1N2 WT and C100R mutant. (J) Cell lysates from parental HeLa cells, NF cells, and CF cells were subjected to immunoblotting with an antibody against phosphorylated PKA substrates (p-PKA sub.). β-actin was used as a loading control. (K and L) CF cells were transfected with control (Ctrl) siRNA or siRNA against human OR1N2 for 48 h. (K) Cells were analyzed as described in (E). (L) Cell lysates were subjected to immunoblotting with antibodies indicated.

Fig. 3

Furin plays a crucial role in cell fusion of CF cells. (A) Cell lysates from parental HeLa cells, NF (clone 1) cells, and CF (clone 2) cells were subjected to immunoblotting with indicated antibodies. The solid and open arrowheads indicate the position of full-length and cleaved form syncytin-1, respectively. (B) Schematic represents the domain structure of syncytins. The positions of the disulfide bond formation motifs, the furin cleavage sites, signal peptides, fusion peptides, transmembrane domains (TMD), surface (SU) subunits, and transmembrane (TM) subunits are indicated. The immunogen ranges for syncytin antibodies are also indicated. (C–E) CF cells were treated with furin inhibitor (furin INH; 10 μM) for 96 h. The media were changed, and furin inhibitor was repeatedly added every 24 h. DMSO was used as a vehicle. (C) Cell lysates were subjected to immunoblotting with indicated antibodies. (D) Cells were immunostained with an anti-pan cadherin antibody (green). DAPI for nuclear staining (blue). The bottom images show magnifications of the boxed areas in the top images. Scale bar, 200 μm. (E) Box plots present the percentage of cells with two (left) or more than two (right) nuclei versus the total number of cells (≥1000 cells) (n = 5 independent experiments). **p < 0.01 by Student's t test.

Fig. 4

Phosphomimetic furin mutant at Ser421 induces cell fusion. (A) The domain structure of furin and sequence alignment of the consensus motif for PKA-mediated phosphorylation in furin and known PKA substrates, CREB and BAD, are shown. X, any amino acid. (B) HeLa cells were infected with retroviruses expressing PKA WT, PKA CA mutant, or control retrovirus. Cell lysates were analyzed by immunoblotting with indicated antibodies. (C) Cell lysates from parental HeLa cells, NF (clone 1) cells, and CF (clone 2) cells were subjected to immunoblotting with indicated antibodies. (D) CF cells treated with H89 (4 μM) for indicated time periods were subjected to immunoblotting with indicated antibodies. H89 was repeatedly added every 24 h. (E–H) HeLa cells were infected with retroviruses expressing furin WT, furin S421D mutant, or control retrovirus. (E) Cells were immunostained with an anti-pan cadherin antibody (green). DAPI for nuclear staining (blue). The bottom images show magnifications of the boxed areas in the top images. Scale bar, 200 μm. (F) Box plots present the percentage of cells with two (left) or more than two (right) nuclei versus the total number of cells (≥1000 cells) (n = 5 independent experiments). Data were analyzed by one-way ANOVA, followed by Tukey's multiple comparison test. *p < 0.05 and **p < 0.01. (G) Serial images for cell fusion were captured from furin S421D mutant-expressing cells. Arrowheads indicate the position of cell fusion. (H) Cell lysates were analyzed by immunoblotting. The solid and open arrowheads indicate the position of full-length and cleaved form syncytin-1, respectively. The relative intensity (RI) of full-length syncytin-1 and -2 proteins normalized to β-actin protein were evaluated. (I) The cleavage of fluorogenic furin substrates by recombinant furin proteins was evaluated using furin in vitro assay. The fluorescence intensity values of GST control protein were subtracted from those of GST-tagged furin proteins. Data are shown as the mean of three independent experiments ± SD and were analyzed by Student's t test. **p < 0.01. Recombinant proteins used in the reactions were analyzed by immunoblotting.

Fig. 5

Prevention of multinucleation improves the efficacy of anticancer drugs. (A–C) NF (clone 1) and CF (clone 2) cells were treated with cisplatin (40 µM) for 24 h (A and B) or 16 h (C). (A) Representative bright-field images are shown. Scale bar, 200 μm. (B) LDH release was measured using an LDH cytotoxicity assay. The percentage of cell death was calculated based on LDH release from cisplatin-treated and control cells. Data are shown as the mean of three independent experiments ± SD and were analyzed by Student's t test. **p < 0.01. (C) Cell lysates were subjected to immunoblotting with indicated antibodies. The solid and open arrowheads indicate the position of full-length and cleaved form PARP, respectively. (D) Serial images were captured from CF cells treated with cisplatin (40 µM) for indicated time periods. Yellow arrowheads indicate the multinucleated cells. Scale bar, 100 μm. (E and F) CF cells were treated with furin inhibitor (furin INH; 10 μM) for 96 h and then further treated with cisplatin (40 µM) for 40 h. The media were changed, and furin inhibitor was repeatedly added every 24 h. DMSO was used as a vehicle. (E) Representative bright-field images are shown. Scale bar, 100 μm. (F) The number of surviving cells after cisplatin treatment was counted using the trypan blue exclusion technique. Data are shown as the mean of three independent experiments ± SD and were analyzed by Student's t test. **p < 0.01.