SKP2 inhibition activates tumor cell-intrinsic immunity by inducing DNA replication stress and genomic instability

Abstract

Background

S-phase kinase-associated protein 2 (SKP2) is a typical oncogene aberrantly overexpressing in a variety of cancer types, but it remains elusive whether SKP2 regulates the antitumor immunity of triple-negative breast cancer.

Methods

The efficacy of anti-PD-1 was evaluated in the orthotopic xenografts of immunocompetent mice models. The infiltration of cytotoxic T cells in tumor microenvironment(TME) were assessed by immunofluorescence staining. The levels of pro-inflammatory chemokines were analyzed by ELISA. The protein interaction was analyzed by co-immunoprecipitation and GST pull-down. The genomic instability was analyzed by fluorescent microscopy.

Results

SKP2 inhibition significantly improved the antitumor efficacy of immune checkpoint blockade (ICB). Furthermore, SKP2 inhibition activated the cGAS/STING signal pathway and induced the secretion of pro-inflammatory chemokines, thereby promoting cytotoxic T cell infiltration. Additionally, we identified CDC6, a DNA replication licensing factor as a novel substrate of SKP2 in addition to CDT1. SKP2 induced protein degradation of CDC6 and CDT1 through the ubiquitin-proteasome pathway. Conversely, SKP2 inhibition elevated CDC6 and CDT1 protein levels, which caused DNA aberrant replication, DNA damage and genomic instability, thereby resulting in the accumulation of cytosolic DNA, activating cGAS/STING signaling pathway and improving antitumor immunity.

Conclusion

SKP2 may be used as an effective therapeutic target to enable ICB antitumor immunotherapy.

Social media

Peng et al. found that SKP2 inhibition improved the antitumor immunotherapy by activating tumor cell-intrinsic immunity, thereby providing evidences that SKP2 may be used as an effective therapeutic target to enable ICB antitumor immunotherapy.

Introduction

Breast cancer is the second leading cause of cancer-related death in women [1]. Triple-negative breast cancer (TNBC) accounts for 15%–20% of breast cancer cases [2]. TNBC is a highly aggressive and heterogeneous subtype associated with a high probability of distant metastasis and poor overall survival [3]. Compared to other subtypes of breast cancer, the mortality rate of TNBC was 40% within the first 5 years after diagnosis [4].

Immune evasion is considered as the most prominent hallmark of cancer [5]. Tumor cells escape immune surveillance in the process of cancer development, progression and metastasis [6]. Multiple oncogene-associated signaling pathways have been activated during TNBC development [7], which could be vulnerable for immune surveillance. Reversely, the inhibition of these pathways might activate tumor cell intrinsic immunity [8]. TNBC patients obtained better response to immune-checkpoint blockade (ICB) due to higher PD-L1 expression levels [9]. FDA recently approved the combination of pembrolizumab and chemotherapy as the first-line treatment of TNBC based on the satisfactory data of improved progression-free survival (PFS) and overall survival (OS). However, the response rate of ICB in unclassified TNBC patients was less than 20%, thus it is urgent to improve the antitumor immunity of ICB to prolong the survival of patients [10]. The efficacy of immunotherapy is closely and positively correlated to the tumor infiltrating lymphocytes in tumor microenvironment(TME), in particular cytotoxic T cell [11, 12]. Pro-inflammatory cytokines/chemokines are critical to promote cytotoxic T cell infiltration and improve the efficacy of ICB [13, 14]. Therefore, it may be an ideal strategy to improve the therapeutic efficacy of IBC by enhancing the intrinsic infiltration capability of cytotoxic T cells to TME.

S-phase kinase-associated protein 2 (SKP2) as an E3 ligase protein of SCF complex [15, 16], has been reported to induce protein ubiquitination and proteolysis of substrates such as CDT1, p27 and E-cadherin, thereby regulating DNA replication, cell cycle and migration [17–19]. Additionally, SKP2 is a typical oncogene aberrantly overexpressed in a variety of cancer types including breast cancer, and contributes to malignant transformation and cancer progression [20, 21]. However, it remains elusive whether SKP2 regulates the antitumor immunity of TNBC.

In the present study, we first assessed the efficacy of SKP2 inhibition on anti-PD-1 immunotherapy in the TNBC xenografts of immune-competent mice, and further observed cytotoxic T cell infiltration in tumor specimens. We further tested the impact of SKP2 inhibition on cGAS/STING signaling pathway and pro-inflammatory chemokines production. Furthermore, we explored the detailed mechanisms. SKP2 induced the protein degradation of CDC6 and CDT1, while SKP2 inhibition elevated the protein levels of CDC6 and CDT1, thereby causing aberrant DNA replication, DNA damage and genomic instability. DNA damage and genomic instability caused the accumulation of cytosolic DNA fragments, thereby activating cGAS/STING signaling pathway, and enhancing the secretion of pro-inflammatory chemokines to TME, and improving the antitumor immunity of ICB.

Methods

Cell culture and cell cycle synchronization

MDA-MB-231, BT549, 4T1 and HEK293T cells were maintained in DMEM (11965092, Gibco, Shanghai, China) or 1640 medium (11875093, Gibco) supplemented with 10% fetal bovine serum, 1% penicillin and streptomycin (G4003, Servicebio, Wuhan, China). For cell synchronization in G1/S phase, the cells were treated with 2 mM thymidine (T1895, Sigma-Aldrich, Shanghai, China) for 18 h, and then washed and cultured in fresh medium for additional 9 h. The cells were repeated with thymidine treatment for additional 18 h. For cell synchronization in G2/M phase, MDA-MB-231 cells were treated with nocodazole (50 ng/ml) (S2775, Selleck, Shanghai, China) for 6 h.

Plasmids and RNAis

The cDNA fragment of SKP2 or CDC6 gene was generated by PCR, and then was subcloned to the individual vectors to construct such plasmids as pCMV-SKP2-Flag, pET28a-SKP2, pGEX-4T-1-CDC6 and pCDNA3.1-CDC6-His. The mutants were constructed by site-directed mutagenesis and verified by DNA sequencing. Cells with SKP2 knockdown were established by using lentiviral shRNA vectors. The shRNA sequences used for SKP2 knockdown were (5’-3’): SKP2(human)#1 (GCCTAAGCTAAATCGAGAGAA), SKP2(human)#2 (GATAGTGTCATGCTAAAGAAT), SKP2(mouse) (GCCTCGACTTAAGTGACAGTA). siRNA and plasmid transfection were performed by using a Mirus Transfection Kit according to the manufacturer’s instructions (TransIT-X2, Mirus Bio, Madison, WA, USA). The siRNA sequences used for CDC6 knockdown were (5’-3’):GGACAATGCTGCAGTTCAA; The siRNA sequences used for CDT1 knockdown were (5’-3’):CCTACGTCAAGCTGGACAA; The siRNA sequence used for cGAS knockdown were (5’-3’): GCCAGTAGTGCTTGGTTTCCA; The siRNA sequences used for CDK2 knockdown were (5’-3’):CDK2-1# (GCCTGATTACAAGCCAAGT), CDK2-2# (CTATGCCTGATTACAAGCC).

Immunofluorescence

Immunofluorescence assay was performed in both cells and tumor tissues from mice. For cell lines, MDA-MB-231 cells were cultured on glass bottom dishes for 48 h, and then were incubated with 10 μM EdU for 2 h before harvest. Cells were fixed with 4% paraformaldehyde solution for 15 min at room temperature, permeabilized with 0.25% Triton X-100 for 20 min, and then blocked with 5% bovine serum albumin for 1 hour. Cells were incubated with the Ser10 phosphorylated histone H3 antibodies (H3S10ph) at 4 °C for 12 h, and then incubated with the fluorescent-labeled secondary antibodies. The newly synthesized DNA was detected by using Click-iT EdU-594 kit (G1603, Servicebio). Images were captured and analyzed in fluorescence microscopy.

For tumor tissues, samples were soaked in dimethyl benzene for 15 min twice and then washed with ethyl alcohol according to concentration gradient. Antigen retrieval was conducted with EDTA (pH=8.0) and samples were blocked with 10% goat serum. The primary antibodies of were added to samples and incubated at 4 °C overnight. Images were captured and analyzed in fluorescence microscopy.

DNA Comet assays

Cells were harvested through trypsin digestion and re-suspended in PBS, then combined with molten LMAgarose (16520050, Sigma-Aldrich) at 37 °C. The molten LMAgarose was gelled on slides and then chilled for 30 min at 4 °C. The slides were sequentially immersed in lysis buffer at 4 °C for 1 h, and then in alkaline solution (pH > 13) at room temperature for another 1 h. After cell lysis and DNA denature, electrophoresis was then carried out in the alkaline buffer (pH > 13) at 20 V voltage for 30 min. Lastly, DNA was stained with Goldview dye for 15 min. Images were captured and analyzed in fluorescence microscopy.

RNA extraction and real-time quantitative PCR

The total RNA was extracted using RNAiso Plus (9108, Takara Bio, Beijing, China) and reverse transcribed into cDNA using PrimeScript RT Master Mix (RR036A, Takara Bio). The mRNA level of genes was measured by an Applied Biosystems qTower3g Real-time PCR system (Analytikjena, Germany) according to the manufacturer’s instructions. Primers for RT-PCR were listed in the supplementary table S1.

RNA-seq

Total RNA was extracted from MDA-MB-231 shCON and MDA-MB-231 shSKP2 cells using RNAiso Plus (9108, Takara Bio) according to the manufacturer’s protocol. RNA quality and integrity were determined using a NanoDrop spectrophotometer (Thermo Scientific, Massachusetts, USA). Then the libraries were constructed using NEB Next Ultra II RNA Library Prep Kit (New England Biolabs Inc; Ipswich, Massachusetts, USA) according to the manufacturer’s instructions. The transcriptome sequencing and analysis were conducted by Personal Biotechnology Co., Ltd (Shanghai, China). The data are deposited in the NCBI’s Bioproject database (BioProject ID: PRJNA1058858).

Antibodies and chemicals

The antibodies and chemicals used in the study were anti-Flag (F1804, Sigma-Aldrich, and 20543, Proteintech, Wuhan, China), anti-His (A00186, Genscript, Nanjing, China and 10001, Proteintech), anti-CDC6 (Santa sc-9964, Santa Cruz Biotechnology, Shanghai, China and CST 3387, Cell Signal Technology, Boston, USA), anti-CDT1 (sc-365305, Santa), anti-SKP2 (2652, CST and sc-74477, Santa), anti-dsDNA (sc-58749, Santa), anti-TBK1 (3504, CST), anti-p-TBK1 (5483, CST), anti-cGAS (79978, CST), anti-CD4 (96127, CST), anti-CD8 (35467, CST), anti-Granzyme B (17215, CST), anti-F4/80 (ab300421, Abcam, Shanghai, China), anti-HA (3724, CST), anti-CDK2 (18048, CST), anti-Phospho-Histone H3 (Ser 10) (66863-1, Proteintech) and anti-β-actin (AC004, Abclonal, Wuhan, China). Cycloheximide (CHX) was purchased from CST and MG132 were purchased from Selleck. SZL-P1-41 was purchased from Tocris Bioscience (Shanghai, China).

Western blot and immunoprecipitation

Cells were lysed by RIPA buffer (P0013, Beyotime, Nantong, China) containing proteinase (P1005, Beyotime) and phosphatase inhibitor cocktails (P1081, Beyotime) on ice for 30 mins. Then centrifuged at 4 °C and supernatants were collected. The protein expression was analyzed by western blot. For immunoprecipitation, cells were collected and lysed with RIPA buffer containing proteinase and phosphatase inhibitor cocktails. The cell extracts were incubated with the indicated antibodies on a rotator at 4 °C for 12 h, followed by incubating with protein A/G-magnetic beads (HY-K0202, MedChemExpress, Shanghai, China) on a rotator at 4 °C for 1 h. After incubation, beads were washed four times with lysis solution, and boiled in 1× loading buffer. The protein levels were analyzed by western blotting.

GST pull-down

SKP2-His and GST-CDC6 fusion proteins were expressed in the E.coli BL21 strain when induced with 1 mM IPTG (HY-15921, MedChemExpress). The SKP2-His was purified by Ni-NTA agarose beads (P2233, Beyotime) and GST-CDC6 was purified by glutathione-sepharose beads (C600031, Sangon Biotech, Shanghai, China). The purified proteins were characterized by Coomassie blue staining. Appropriate amounts of purified GST-CDC6 protein were mixed with the purified SKP2-His proteins in pull-down assay buffer (137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄, 2 mM KH₂PO₄, 0.02 mM EDTA, 0.1% Triton X-100) on a rotator at 4 °C for 12 h, and washed four times with the pull-down assay buffer. Resin-bound complexes were eluted by boiling and subjected to western blotting.

In vivo ubiquitination assay

MDA-MB-231 cells were transfected with the indicated plasmids for 60 h, and then treated with 50 µM MG132 (S2619, Selleck) for additional 6 h. Cells were lysed in RIPA buffer containing proteinase inhibitor cocktails on ice for 30 mins. The cell lysates were boiled at 100 °C for 10 min and then centrifuged at 12,000 g for 15 min. CDC6 proteins were immunoprecipitated with anti-His antibody on a rotator at 4 °C for 12 h, and the immune complexes were incubated with protein A/G-magnetic beads on a rotator at 4 °C for 1 h. After incubation, beads were washed four times with lysis buffer, the immunocomplex was subjected to SDS-PAGE and the ubiquitination levels were assessed by immunoblot.

Chemokines ELISA assays

The amount of CXCL9 (KE10067, Proteintech) and CXCL10 (KE00128, Proteintech) in the supernatant of cells were assessed by following the manufacturer’s instructions of ELISA kit. In brief, cell culture supernatants were collected, and then the cells and cell debris were moved by centrifugation. The ELISA reactions were quantified by reading the absorbance(OD) in a microplate spectrophotometer at 450 nm (Multiskan-GO, Thermo Fisher Scientific, China). The amount of CXCL9 or CXCL10 is proportional to the OD intensity, which can be calculated with the included formulas.

cGAMP ELISA assays

Cytosolic cGAMP was measured using Cayman Chemical 2’3’-cGAMP ELISA Kit by following the manufacturer’s instructions of ELISA kit (501700, Cayman chemical, USA). In brief, 1×10⁷ cells were collected and lysed by the lysis buffer (78501, Thermo Fisher Scientific). The supernatant then was collected after centrifugation at 20,000 g for 20 min. The ELISA reactions were quantified by reading the absorbance(OD) in a microplate spectrophotometer at 450 nm (Multiskan-GO, Thermo Fisher Scientific). The amount of cGAMP is proportional to the OD intensity, which can be calculated with the included formulas.

Animal studies

To establish orthotopic tumor xenograft models, 4T1 cells stably expressing SKP2 or control shRNA (2×10⁵ cells) were injected into the fourth mammary gland of 6-week-old female BALB/c mice. When tumors volume reached about 100 mm³, the mice were intraperitoneally injected with 100 µg isotype IgG (clone 2A3, BioXcell, Shenzhen, China) or anti-PD-1 antibody (clone RMP1-14, BioXcell) every five days for three times. In an effort to confirm the functions of CD8⁺ T cells, mice were injected 100 μg anti-CD8 antibody (clone 2.43, Selleck) every five days for three times to deplete CD8⁺ T cells when tumors volume reached about 100 mm³, and the isotype antibody (clone LTF-2, Selleck) was used as the control. Tumor sizes were measured with calipers every three days. The tumor volume was calculated by using the following formula: Volume = (width² ×length)/2. Mice were humanely sacrificed and the tumor xenografts were harvested for analysis once they reached a volume of 1500 mm³. All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) at the Institute of Laboratory Animal Science (License Number:00320412), Guangdong Pharmaceutical University (Guangzhou, China), and conformed to the relevant regulatory standards.

Statistical analysis

All data were analyzed by Origin 9.0. Results were presented as mean ± SD. Student’s two-tailed t -test was used to analyze the statistical difference between two sets of data. P < 0.05 was considered as significant. The Kaplan-Meier estimator and log-rank (Mantel-Cox) test was used for survival analysis.

Results

SKP2 inhibition improved anti-tumor immunotherapy by enhancing T cell infiltration

SKP2 as a typical oncogene aberrantly overexpressing in TNBC, and contributes to cancer progression by inducing the ubiquitination and proteolysis of substrates such as tumor suppressor proteins p27 and p21 [18–22], but has no report on antitumor immunity. To investigate the roles of SKP2 in the regulation of TNBC tumor immunity, we established a mouse-derived 4T1 cell line stably expressing SKP2 shRNA (Fig. 1a). The 4T1 cells stably expressing scrambled or SKP2 shRNA were orthotopically injected into the mammary fat pads of BALB/c mice with an intact immune system. The data showed that SKP2 knockdown dramatically inhibited tumor growth and decreased tumor volume and weight (Fig. 1b–d). Importantly, SKP2 depletion decreased tumor volume and weight more significantly in the immunocompetent BALB/c mice than immuno-deficient nude mouse (Figs. 1b–d and S1). These data indicated that SKP2 inhibition might activate antitumor immunity. The tumor nodes were harvested and the number and infiltration of CD8⁺ T cells in TME were assessed by cell immunofluorescence staining. In comparison to the tumor nodes expressing scrambled shRNA, the percentage of CD8⁺ T cells significantly increased in the SKP2 shRNA tumor nodes (Fig. 1e, f). These data indicated that SKP2 inhibition enhanced the infiltration of CD8⁺ T cells in TME. To determine whether the inhibition of tumor growth by SKP2 depletion depends on the infiltration of CD8⁺ T cells in TME, we depleted CD8⁺ T cells with a commercial anti-CD8 antibody, and monitored the growth of 4T1 tumor xenografts. Compared to isotype control, the depletion of CD8⁺ T cells using anti-CD8 antibodies significantly promoted the growth of SKP2-knockdown 4T1 tumors, and reduced animal survival (Fig. 1g, h). Compared to shCON, shSKP2 demonstrated stronger tumor suppression efficacy in the immunocompetent mice than in the immuno-deficient nude mice. The data further validated that SKP2 knockdown activated anti-tumor immunity in addition to direct tumor suppression efficacy.

Fig. 1. SKP2 depletion inhibited tumor growth and enhanced T cell infiltration.

a 4T1 cells stably expressing the scramble or SKP2 shRNA were established, and the protein levels of SKP2 were assessed by western blotting. b The growth curve of orthotopic tumor xenografts induced by 4T1-control shRNA or 4T1-SKP2 shRNA cells in BALB/c mice. c The tumor nodes were dissected at the endpoint of experiment. Scale bar, 1 cm. d Relative tumor weights of the dissected tumor nodules. e, f immunofluorescence staining and quantitative analysis the proportion of CD8⁺ T cells in the 4T1-control shRNA or 4T1-SKP2 shRNA orthotopic tumor xenografts. Scale bar, 100 μm. Data as mean ± SD. **P < 0.01. g, h The control or SKP2-knockdown 4T1 cells were orthotopically injected into the fourth mammary gland of 6-week-old female BALB/c mice (n = 24). When tumors volume reached about 100 mm³, the mice were treated with the indicated isotype control or anti-CD8 antibody. The survival of mice was analyzed by Kaplan–Meier survival curves. The P values was shown at the right row of the tables, log-rank (MantelCox) test was used to compare the survival trend of two groups.

Inhibition of SKP2 activated the cGAS/STING signaling pathway and induced the secretion of pro-inflammatory cytokines/chemokines

To reveal the mechanisms by which SKP2 modulates the infiltration of CD8⁺ T cells in TME, we generated a MDA-MB-231 cell line with SKP2 stable knockdown, and then performed transcriptomic analysis to comprehensively study how SKP2 inhibition affects immune response signaling. Compared to the scrambled control, a set of genes associated with immune response were significantly upregulated in the cells with SKP2 shRNA. The GO analysis of RNA-seq identified the innate immunity pathway among the most prominently upregulated processes (Fig. 2a). Interestingly, Th1-type cytotoxic immunity, with potently tumoricidal immune response, was also activated by SKP2 knockdown (Fig. 2b, c). It was well known that cGAS/STING is a key innate signaling pathway whose activation regulates the expression of genes associated with Th1-type cytotoxic immunity, and improved the efficacy of anti-PD-1 therapy [23]. We next sought to validate whether SKP2 knockdown activated cGAS/STING pathway. Consistently in MDA-MB-231 and BT549 cells, SKP2 knockdown significantly elevated the levels of phosphorylated TBK1, a key kinase of cGAS/STING (Fig. 2d). Furthermore, SKP2 knockdown robustly elevated the mRNA levels of several cGAS/STING downstream genes, such as CXCL9, CXCL10 and ISG15 (Fig. 2e). Furthermore, SKP2 knockdown remarkably elevated the secretion levels of CXCL10 of breast cancer cells (Fig. 2f). Consistent to the data of siRNA knockdown in MDA-MB-231 cells, a chemical inhibitor of SKP2, SZL-P1-41 significantly elevated the levels of phosphorylated TBK1 in a dose-dependent manner (Fig. 2g), and elevated the mRNA levels of CXCL9, CXCL10 and ISG15 genes (Fig. 2h). Additionally, SZL-P1-41 significantly elevated the secretion levels of CXCL10 of MDA-MB-231 cells (Fig. 2i). In addition, SKP2 knockdown dramatically elevated the mRNA levels of CXCL9, CXCL10 and ISG15 genes, as well as the secretion levels of CXCL9 in 4T1 breast cancer cells (Fig. 2j, k). Consistently, we also observed dramatically elevated mRNA levels of CXCL9 and CXCL10 genes in the SKP2-depleted 4T1 tumor xenografts (Fig. 2l). These data indicated that SKP2 inhibition enhanced CD8⁺ T cell infiltration in TME by activating cGAS/STING signal pathway and enhancing pro-inflammatory cytokines/chemokines production.

Fig. 2. SKP2 Inhibition activated the cGAS/STING signaling pathway and induced the secretion of pro-inflammatory cytokines/chemokines.

a GO analysis was performed on all upregulated genes in the shSKP2 versus shCON cells. The histogram shows the top 7 upregulated biological processes. b GSEA plot of Th1 cytotoxic genes of shSKP2 versus shCON cells. c The heatmap shows enrichment of at least 3-fold upregulated genes inside the Bosco_Th1 cytotoxic module in shSKP2 versus shCON cells. d Cells stably expressing the control and SKP2 shRNA were generated. The protein levels of p-TBK1 and TBK1 were assessed by western blotting. e The mRNA levels of IFN-stimulated genes were analyzed by RT-PCR. f The secreted CXCL10 levels of MDA-MB-231 and BT-549 cells stably expressing shCON or SKP2 shRNA were measured by ELISA. g, h MDA-MB-231 cells were treated with different doses of SZL-P1-41 for 72 h. The levels of p-TBK1 and TBK1 were assessed by western blot. The mRNA levels of IFN-stimulated genes were analyzed by RT-PCR. i MDA-MB-231 cells were treated with SZL-P1-41 for 72 h, and the secreted CXCL10 levels were assessed by ELISA. j The mRNA levels of IFN-stimulated genes in the control and SKP2 shRNA 4T1 cells were analyzed by RT-PCR. k The secreted CXCL9 levels of 4T1 cells stably expressing the control or SKP2 shRNA were measured by ELISA. l The mRNA levels of CXCL9 and CXCL10 genes inside tumors were analyzed by RT-PCR, ACTIN was used as the internal control. Data from three separate experiments were expressed as mean ± SD for each group. *P < 0.05, **P < 0.01.

SKP2 inhibition induced aberrant DNA replication and caused DNA damage and genomic instability

A growing body of evidence demonstrated that aberrant DNA replication in G2/M phase caused DNA damage and genomic instability, and then the accumulation of cytosolic DNA activated the cGAS/STING signaling pathway [24, 25]. Although SKP2 as a F-box protein with E3 ligase activity induced protein ubiquitination and degradation of CDT1, a DNA replication licensing factor [17], it is unknown whether SKP2 inhibition induced aberrant DNA replication in G2/M phase. We established a MDA-MB-231 cell line with SKP2 stable knockdown, and the cells were stained with an antibody against Serine 10 phosphorylated histone H3 (H3S10p), a typical marker of late G2 and early M phase, and then were pulse labeled with thymidine analogue 5-ethynyl-2’-deoxyuridine (EdU) labeling new synthesized DNA. The results demonstrated, almost no EdU-positve MDA-MB-231 cell was observed in G2/M phase, while SKP2 knockdown significantly increased the number of EdU-positve cells (Fig. 3a, b). The results suggested that SKP2 knockdown caused DNA aberrant replication in G2/M phase. It is well known that DNA aberrant replication would inevitably trigger genomic instability and increase cytoplasmic DNA (cytoDNA) levels, as evidenced by the formation of DNA double-strand breakage (DSB) [26–28]. We examined the effect of SKP2 knockdown on DNA damage by using a DNA comet assay. Compared to the control, SKP2 knockdown caused more serious DSB (Fig. 3c, d). Meanwhile, SKP2 knockdown significantly increased the number and percentage of cells with 6-diamidino-2-phenylindole (DAPI)-stained micronuclei and DNA bridges in the immunofluorescence images (Fig. 3e–h). Furthermore, SKP2 knockdown significantly increased cytoDNA levels (Fig. S2). These data suggested that SKP2 inhibition caused DNA replication stress and DNA damage, thereby further induced genomic instability and enhanced cytoDNA levels. The formation of cytoDNA including micronuclei activated the cytosolic DNA sensor cGAS, which in turn activated the STING-mediated type I interferons [29, 30]. We then examined whether cGAS proteins recognized cytoDNA and could be activated by SKP2 inhibition. The result showed that SKP2 inhibition induced micronuclei formation, and the majority of cGAS proteins were co-localized with micronuclei in the cytoplasm (Fig. 3i, j). Meanwhile, SKP2 inhibition significantly elevated the level of cGAMP (Fig. 3k), which activated the adaptor STING as a second messenger. Furthermore, co-depletion of cGAS and SKP2 abolished the phosphorylated level of TBK1 (Fig. 3l). Therefore, SKP2 inhibition-produced micronuclei activated innate immunity through cGAS.

Fig. 3. SKP2 knockdown induced aberrant DNA replication and caused DNA damage and genomic instability.

a, b MDA-MB-231 cells stably expressing the control or SKP2 shRNA were collected after incubating with EdU for 2 h. Cells were subjected to Click-it reaction and histone H3S10ph staining. The positive cells were analyzed by immunofluorescence microscopy. The percentage of H3S10ph and EdU double positive cells was calculated. Scale bar, 20 μm. c, d Representative images and quantification of DNA Comet assays. Scale bar, 100 μm. e, f The cells were labeled with DAPI, and DNA bridges were quantified. Red arrows indicate DNA bridges. Scale bar, 10 μm. g, h The cells were labeled with DAPI, and micronuclei (MN) were quantified. Red arrows indicate MN. Scale bar, 10 μm. i, j cGAS was stained with an immunofluorescent antibody, and the nuclei were labeled with DAPI. The MN⁺cGAS⁺ cells were quantified. Red arrows indicate the co-localization of cGAS and MN. Scale bar, 10 μm. k The level of cGAMP were assessed by ELISA. Data from three separate experiments were expressed as mean ± SD. **P < 0.01. l The SKP2-depleted MDA-MB-231 cells were transfected with the indicated siRNAs for 72 h. The protein levels were assessed by western blotting.

SKP2 regulates CDC6 protein ubiquitylation and degradation

It is well known that CDT1 and CDC6 are two licensing factors to initiate DNA replication, and ensure DNA replication once in a single cell cycle. The aberrant overexpression of CDT1 and CDC6 would inevitably cause aberrant DNA replication in G2/M phase [26]. CDT1 has been reported to be a substrate of SKP2. SKP2 promoted CDT1 protein degradation, while SKP2 inhibition sustained the protein stability of CDT1 [17]. However, the aberrant overexpression of CDT1 protein alone is not enough to induce aberrant DNA replication in G2/M phase (Fig. S3). Therefore, we speculate that CDC6 may be another ubiquitination substrate of SKP2 in addition to CDT1, since SKP2 knockdown caused aberrant DNA replication by elevating the expression of both CDT1 and CDC6. We validated our speculation by screening ubiquitin E3 ligases of CDC6 protein in a public bioinformation website UbiBrowser (http://ubibrowser.bio-it.cn/ubibrowser/) [31]. As we expected, SKP2 is indeed a E3 ligase of CDC6 (Fig. S4). Since the physical protein interaction is required for SKP2-induced substrate ubiquitination, we tested the endogenous SKP2 and CDC6 protein interaction by reciprocal co-immunoprecipitation in MDA-MB-231 cells. CDC6 protein was detected in the SKP2-immunoprecipitated protein complex and conversely SKP2 was detected in the CDC6-immunoprecipitated protein complex (Fig. 4a). SKP2 or CDC6 was individually fused with Flag or His tag, and were co-transfected to HEK-293T cells. CDC6 protein was immunoprecipitated with anti-His antibody, and SKP2 protein was detected in the immunoprecipitated protein complex (Fig. 4b). Furthermore, the physical interaction of SKP2 and CDC6 proteins was further confirmed by GST-pull down assay (Fig. 4c). To monitor the strength of SKP2 and CDC6 protein interaction in different cell cycle phase, SKP2 or CDC6 was over-expressed in MDA-MB-231 cells, and the strength of SKP2 and CDC6 protein interaction was detected by co-immunoprecipitation. The results showed that the protein interaction of CDC6 and SKP2 occurs predominantly in G2/M phase (Fig. 4d). The amino terminus of CDC6 protein contains three CDK2 phosphorylation residues S54, S74 and S106. CDK2-catalyzed CDC6 phosphorylation promoted the nuclear export of CDC6, resulting in subsequent protein degradation through ubiquitin-proteasome pathway [32]. To investigate whether CDK2-catalyzed phosphorylation of CDC6 is required for the protein interaction of SKP2 with CDC6, we co-transfected the indicated siRNAs and plasmids to MDA-MB-231 cells, and then assessed the protein interaction of SKP2 and CDC6 by co-immunoprecipitation. The results showed that CDK2 knockdown with siRNAs did not change the protein interaction of CDC6 and SKP2 (Fig. S5A). Furthermore, we mutated the CDK phosphorylation residues S54, S74 and S106 (Ser to Ala or Asp), and analyzed the protein interaction between CDC6 and SKP2 by co-immunoprecipitation. The data showed that the mutation of CDC6 serine residues did not change the protein interaction strength of CDC6 to SKP2 (Fig. S5B). These results excluded that CDK2-mediated phosphorylation was the prerequisite of protein interaction between CDC6 and SKP2.

Fig. 4. SKP2 regulates CDC6 protein ubiquitylation and degradation.

a MDA-MB-231 cell lysates were subjected to immunoprecipitation with anti-SKP2 or anti-CDC6 antibody. The immunoprecipitated protein was then blotted with the indicated antibodies. b HEK-293T cells were co-transfected with the plasmids encoding CDC6-His and SKP2-Flag. CDC6 was immunoprecipitated with anti-His antibody, and SKP2 was assessed with anti-Flag antibody by immunoblotting. c The recombinant GST, GST-CDC6 and SKP2-His proteins were purified and in vitro incubated as indicated. SKP2 was assessed by immunoblotting. d MDA-MB-231 cells were co-transfected with the plasmids encoding CDC6-His and SKP2-Flag, and then cells were synchronized at G1/S phase by double-thymidine blocking. The cells were released to cell cycle with fresh medium and were arrested at M phase with nocodazole. Cell lysates were subjected to immunoprecipitation with anti-His antibody, and the immunoprecipitated protein was then blotted with the indicated antibodies. Cell cycle phase distribution was analyzed by flow cytometry. e MDA-MB-231 and BT549 cells with stable SKP2 knockdown were treated with nocodazole for 6 hours. The protein levels of CDC6, CDT1, SKP2 and β-actin were assessed by immunoblotting. f SKP2 was depleted in MDA-MB-231 and BT549 cells. The mRNA levels of CDC6 and SKP2 genes were analyzed by RT-PCR. ACTIN was used as an internal control. g SKP2-depleted MDA-MB-231 cells were treated with cycloheximide (CHX, 50 μg/ml) for the indicated times. The endogenous CDC6 protein levels were monitored by immunoblotting. The protein levels were quantified by grayscale analysis. h Cells were transfected with pCMV-SKP2-Flag plasmid for 60 h, and then treated with 50 μM MG132 for 6 h, and the protein levels were assessed by immunoblotting. i MDA-MB-231 cells were transfected with increasing amounts of plasmids encoding SKP2(WT)-Flag or SKP2(∆F-box)-Flag. Cells were harvested 72 h post-transfection for western blotting using the indicated antibodies. j SKP2-depleted MDA-MB-231 cells were transfected with plasmids encoding HA-ubiquitin and CDC6-His for 48 h, and then treated with 50 μM MG132 for 6 h. CDC6 protein was immunoprecipitated with anti-His antibody and the polyubiquitylated CDC6 was assessed with anti-HA antibody. k Cells were transfected with plasmids as indicated for 48 h and then treated with 50 μM MG132 for 6 h. CDC6 protein was immunoprecipitated with anti-His antibody and the polyubiquitylated CDC6 was assessed with anti-HA antibody.

SKP2 as one E3 ligase protein of SKP2-SCF complex, has been reported to induce protein ubiquitination and proteolysis. We first tested CDC6 protein levels in G2/M phase when SKP2 was depleted or overexpressed in MDA-MB-231 and BT-549 cells. SKP2 depletion significantly elevated CDC6 protein level (Fig. 4e). Conversely, SKP2 overexpression significantly decreased the levels of endogenous CDC6 protein (Fig. S6A). Additionally, we excluded the influence of SKP2 on the mRNA level of CDC6 gene by RT-PCR (Figs. 4f and S6B). The results validated that SKP2 regulates CDC6 expression through a post-translational modification. To further validate these results, we assessed the effect of SKP2 on the half-life of CDC6 protein. The new protein synthesis was inhibited by cycloheximide (CHX), and the endogenous CDC6 protein stability was assessed at continuous time points. Consistently, SKP2 knockdown substantially prolonged, but overexpression reversely shortened the half-life of CDC6 protein (Figs. 4g and  S6C). Taken together, the results indicated that SKP2 promoted the protein degradation of CDC6.

To determine how SKP2 modulates CDC6 protein levels, SKP2 was overexpressed and then treated with MG132, a specific proteasome inhibitor. CDC6 protein degradation was reversed by MG132, indicating that SKP2 promoted CDC6 protein degradation through ubiquitin-proteasome pathway (Fig. 4h). We constructed a truncated mutant with F-box motif deletion (SKP2-(∆F-box)) to further validate whether SKP2-promoted CDC6 protein degradation depends on its ubiquitin ligase activity. The ectopic expression of wild-type SKP2 (SKP2-WT) significantly decreased the levels of endogenous CDC6 protein, while SKP2-(∆F-box) lost the impact on CDC6 protein level (Fig. 4i) and decreased the protein interaction with CDC6 (Fig. S7). The results were further validated by the in vivo ubiquitination assays. The plasmids encoding HA-ubiquitin or CDC6-His were co-transfected to MDA-MB-231 cells with SKP2 knockdown. CDC6 protein was immunoprecipitated with anti-His antibody, and the ubiquitinated protein was detected with HA antibody. The knockdown of SKP2 remarkably decreased the ubiquitination level of CDC6 protein (Fig. 4j). Furthermore, compared to SKP2-WT, the ubiquitin ligase-dead mutant SKP2-(∆F-box) did not affect the polyubiquitination level of CDC6 protein (Fig. 4k). These data indicated that SKP2 promoted CDC6 protein degradation by elevating the level of ubiquitination. As we knew, SKP2 induced the ubiquitination of substates through K48- or K63-linked polyubiquitination chains [33]. To identify the types of polyubiquitination chain of CDC6 protein mediated by SKP2, we co-transfected HEK-293T cells with the plasmids encoding CDC6-His and HA-tagged K48 or K63 lysine mutants. Compared to K63, SKP2 significantly induced K48-linked polyubiquitination (Fig. S8). The results suggested that SKP2 induced CDC6 protein degradation through K48-linked polyubiquitination.

SKP2 knockdown activated cGAS/STING signaling pathway through the aberrant elevation of CDC6 and CDT1

It is well known that CDT1 and CDC6 overexpression induced aberrant DNA replication. Furthermore, DNA replication stress results in genomic instability and activated cGAS/STING [26, 27]. To investigate whether SKP2 knockdown induced DNA replication stress and cGAS/STING activation through CDC6 and CDT1 overexpression, we deleted SKP2, together with CDC6 or CDT1, and tested the impacts on DNA replication stress, genomic instability and cGAS/STING signaling pathway. SKP2 knockdown significantly elevated DNA aberrant replication level in G2/M phase, DNA double-strand breakage and genomic instability, but the elevation was reversed by CDC6 or CDT1 knockdown (Fig. 5a–d). Meanwhile, compared to SKP2 knockdown alone, SKP2 knockdown, together with CDC6 or CDT1 further decreased the level of phosphorylated TBK1 protein, as well as the level of CXCL9 and CXCL10 genes (Fig. 5e–g). We further validated the impacts of CDC6 or/and CDT1 overexpression on cGAS/STING signaling pathway. Compared to SKP2 knockdown, the overexpression of CDC6 or CDT1 alone only moderately activated the cGAS/STING pathway. However, the simultaneous overexpression of CDC6 and CDT1 significantly activated cGAS/STING pathway, and the effect is close to SKP2 knockdown (Fig. S9). These results validated that SKP2 knockdown activated cGAS/STING signaling pathway by elevating CDC6 and CDT1 protein levels and inducing DNA replication stress and genomic instability.

Fig. 5. SKP2 knockdown activated cGAS/STING signaling pathway through the aberrant elevation of CDC6 and CDT1.

SKP2-depleted MDA-MB-231 cells were transfected with the indicated siRNAs for 72 h. Cells were incubated with EdU for 2 h and then subjected to Click-it reaction and histone H3S10ph staining. The percentage of H3S10ph and EdU double positive cells was calculated. Scale bar, 20 μm (a). Representative images and quantification of DNA Comet assays was calculated. Scale bar, 100 μm (b). The cells were labeled with DAPI, and DNA bridges were quantified. Red arrows indicate DNA bridges. Scale bar, 10 μm (c). The cells were labeled with DAPI, and micronuclei(MN) were quantified. Red arrows indicate MN. Scale bar, 10 μm (d). The levels of phosphorylated TBK1 were assessed by western blotting (e). The mRNA levels of CXCL9 and CXCL10 were analyzed by RT-PCR (f); the secreted CXCL10 levels was measured by ELISA (g). Data from three separate experiments were expressed as mean ± SD. **P < 0.01.

SKP2 inhibition enhanced tumor sensitivity to anti-PD-1 immunotherapy

We next assessed whether SKP2 inhibition improved the therapeutic efficacy of ICB. The orthotopic 4T1 tumor xenograft models were established by injecting 4T1 cells with SKP2 shRNA to the immunocompetent BALB/c mice. When tumor volumes reached about 100 mm³, we treated the mice with the control IgG or anti-PD-1 antibody every five days for continuous three times (Fig. 6a). Compared to the control or monotherapy, the combination of SKP2 knockdown and anti-PD-1 antibody demonstrated more significant antitumor efficacy, and prolonged the survival of mice (Fig. 6b). We further analyzed the infiltrating lymphocytes in TME by cell immunofluorescence staining. The data showed that SKP2 knockdown significantly increased the percentage of tumor infiltrating CD8⁺ T cells, CD4⁺ T cells and granzyme B⁺ cells in TME compared to the controls treated with the control IgG or anti-PD-1 antibody (Figs. 6c, d and S10). The results indicated that SKP2 knockdown enhanced the infiltration of T cells, thereby activating antitumor immunity, and improving the efficacy of anti-PD-1 immunotherapy.

Fig. 6. SKP2 inhibition enhanced tumor sensitivity to anti-PD-1 immunotherapy.

a Schematic outline of anti-PD-1 treatment for 4T1 tumor xenografts. The 4T1 cells with stable SKP2 knockdown were orthotopically injected into the fourth mammary gland of 6-week-old female BALB/c mice (n = 24). When tumors volume reached about 100mm³, the mice were treated with the indicated isotype control or anti-PD-1 antibody. b The survival of mice was analyzed by Kaplan-Meier survival curves. The P values was shown at the right row in the tables, log-rank (MantelCox) test was used to compare the survival of two groups. c, d CD8⁺ T cells in the tumor xenografts were stained with an immunofluorescent antibody and quantified. Scale bar:100 μm. Data were expressed as mean ± SD, **P < 0.01. e Molecular mechanism schematic diagram. The diagram shows the mechanisms on how SKP2 inhibition activated cGAS/STING signaling pathway, and triggered antitumor immunity in triple-negative breast cancer though CDC6 and CDT1 overexpression-induced DNA replication stress and genomic instability.

Discussion

SKP2 is a typical oncogene aberrantly overexpressing in a variety of cancer types including breast cancer, and a potential anticancer therapeutic target [20]. Recent studies indicated that SKP2 plays a critical roles in antitumor immune regulation of osteosarcoma, and SKP2 inhibition improved immune activation by increasing immune gene expression [34]. However, the detailed molecular mechanism is elusive.

As one of the members of F-box protein family, SKP2 is capable of inducing the ubiquitination of substrates such as a DNA replication licensing factor CDT1 through its E3 ligase activity, and regulates DNA replication [35]. In the present study, we identified another DNA replication licensing factor CDC6 as a novel SKP2 substrate in addition to CDT1. SKP2 induced CDC6 ubiquitination in G2/M phase, and promoted the protein degradation of CDC6 by elevating the in vivo level of CDC6 polyubiquitination (Fig. 4).

CDC6 and CDT1 are two essential initiation factors of DNA replication [36]. The overexpression of both CDC6 and CDT1, but not CDC6 or CDT1 alone induced abnormal DNA replication [26]. In this present study, we unexpectedly found that SKP2 inhibition elevated the protein levels of both CDC6 and CDT1, and induced abnormal DNA replication in G2/M phase, and thereby causing DNA damage (Fig. 3a–d). As we known, the unrepaired DNA damage and abnormal spindle architecture caused the micronuclei formation [37, 38]. We found that SKP2 inhibition didn’t change the architecture of spindle (Fig. S11), but significantly enhanced DNA damage (Figs. 3c, d and S12). Therefore, we believe that SKP2 plays a novel role in DNA replication and genomic stability, since SKP2 inhibition triggered abnormal DNA replication and DNA damage, thereby resulting in DNA shedding during mitosis, DNA fragmentation and micronuclei formation.

Emerging evidence implied that the cGAS/STING signaling pathway plays a critical role in bridging innate immunity and genomic instability [39]. In this present study, we found that SKP2 inhibition caused DNA fragmentation in the cytoplasm, such as DNA micronuclei and DNA bridges (Fig. 3e–h). cGAS as the major cytosolic DNA sensor produced the second messenger cGAMP. cGAMP then activated STING and a downstream signaling cascade to produce type I interferons and other immune mediators [40]. We unexpectedly found that SKP2 knockdown-produced micronuclei co-localized with cGAS in the cytoplasm. Furthermore, SKP2 inhibition significantly elevated the level of cGAMP, and activated cGAS/STING signaling pathway, thereby increasing the mRNA levels and promoting the secretion of pro-inflammatory chemokines (Figs. 2 and 3). Therefore, SKP2 is an ideal bridge between DNA replication and cGAS pathway, and SKP2 inhibition activated innate antitumor immunity.

Immunotherapy has showed great potential for a variety of solid cancer types. However, the response rate is very low for TNBC patients [9]. The efficiency of immunotherapy is closely related to the lymphocyte immunity of TME. The inadequate lymphocytes in TME is the major reason of immunotherapeutic failure [41]. Previous studies have shown that cGAS determined tumor immunogenicity and created an inflammatory microenvironment. The expression level of cGAS is closely correlated with poor survival of cancer patients [42]. The activation of cGAS in cancer cells effectively promoted the infiltration of lymphocytes in TME, and enhanced the efficacy of immunotherapy [43]. These studies highlighted the importance of tumor-intrinsic cGAS/STING pathway in promoting antitumor immunity and the efficacy of immunotherapy. In our present study, SKP2 inhibition produced higher secretion levels of pro-inflammatory chemokines by activating cGAS/STING signaling pathway, thereby promoting lymphocyte infiltration, and inhibited tumor growth (Figs. 2 and 6). SKP2 as an E3 ubiquitin ligase of SCF complex [15, 16], has been reported to promote tumor progression by inducing the protein ubiquitination and proteolysis of substrates such as tumor suppressor proteins p27 and p21 [18–22]. We found that SKP2 depletion inhibited tumor growth in the immuno-deficient mouse models (Fig. S1). However, compared to nude mouse, SKP2 depletion more significantly decreased tumor weight in the immunocompetent BALB/c mice. These data indicated that SKP2 depletion inhibited tumor growth by activating antitumor immunity.

Altogether, our study found that SKP2 knockdown elevated the protein levels of replication initiation factor CDC6 in addition to CDT1, resulting in aberrant overexpression of CDC6 and CDT1, and DNA replication in G2/M phase. DNA aberrant replication caused DNA damage and genomic instability, and the appearance of micronuclei and DNA bridges. The cytosolic DNA further activated cGAS/STING signaling pathway, and promoted the secretion of pro-inflammatory chemokines. Pro-inflammatory chemokines promoted the infiltration of lymphocytes in TME, and improved the therapeutic efficacy of ICB (Fig. 6e). Our study indicated SKP2 inhibitors may enhance the antitumor immunity, and potentially improved the therapeutic efficacy of PD-1 immunotherapy for TNBC in the future clinical trials.

Author contributions

YCP performed experiments, analyzed data and wrote the manuscript; XLQ performed experiments and wrote the manuscript; LYD, YHL and SZC participated in data analysis; JJH participated in animal studies and ELISA assays; RRZ participated in the real-time quantitative PCR; YMF participated in co-IP assays; LLY, TGD and YBY participated in Western Blot and ubiquitination assays; LX designed the project, supervised all experiments, analyzed the results, and wrote the manuscript. All authors read and approved the final manuscript.

Funding

This work was supported by grants from the National Natural Science Foundation of China (NSFC 81874096、NSFC 82303867 and NSFC 81572542), Research Project of Key Discipline of Guangdong Province (2019GDXK0010), Science and Technology Program of Guangzhou City (202201010161), Key team of basic and clinical research on tumor immunotherapy, Guangdong Pharmaceutical University, Project No. 2024ZZ10 and the 2024 Science and Technology Innovation Project of Guangdong Medical Products Administration “Research and Evaluation of Key Technologies for Drug Safety Risk Management” (No. 2024ZDZ10).

Data availability

The raw data of RNA sequencing are deposited in the NCBI’s Bioproject database (BioProject ID: PRJNA1058858).

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) at the Institute of Laboratory Animal Science (License Number:00320412), Guangdong Pharmaceutical University (Guangzhou, China), and conformed to the relevant regulatory standards.

Supplementary information

The online version contains supplementary material available at 10.1038/s41416-024-02909-y.