Identification of RPL15 60S Ribosomal Protein as a Novel Topotecan Target Protein That Correlates with DAMP Secretion and Antitumor Immune Activation

Tumor immunity inhibits cancer progression and improves cancer immunotherapy, and its activation is affected by cancer cell–derived molecules such as cytokines and chemokines (1, 2). These molecules alter the activation of tumor-infiltrating immune cells and their composition, including CTLs and regulatory T cells (Tregs). CTLs kill tumor-associated Ag-presenting cancer cells and contribute to cancer inhibition. CTL activation is enhanced by immune checkpoint inhibitors such as anti–PD-1 and anti–CTLA-4 Abs and is repressed by CTL exhaustion and Tregs (3). Several cancer cells promote Treg expansion via secreting IL-10 and transforming growth factor-β and overexpressing PD-L1, which suppresses CTL activation by PD-1–PD-L1 interactions (4, 5). Therefore, cancer-derived molecules are closely correlated with the tumor microenvironment and antitumor immune response.

Damage-associated molecular patterns (DAMPs) are also secreted from damaged or dying cancer cells during radiotherapy and anticancer chemotherapy and alter the tumor microenvironment, including immune cell activation and angiogenesis (6). Oxidized DNA is secreted as DAMPs from irradiated cancer cells, which activates cGAS–STING signaling in macrophages and dendritic cells, resulting in increased CTL activation and increased CTL/Treg ratio in murine B16-F10 melanoma model tumors (7, 8). Previous studies reported that a few anticancer drugs (doxorubicin, gemcitabine, and cisplatin) induce the secretion of ATP and HMGB1 as DAMPs from cancer cells (9, 10). These DAMPs activate innate immune signaling, TLR signaling, and inflammasomes, which contribute to antitumor immune activation and anticancer chemotherapy efficacy (9, 11). Various endogenous molecules act as DAMPs, such as mitochondrial DNA, uric acid, and heat shock proteins, which are recognized by innate immune receptors and induce cytokine production and tumor immune activation (8, 12, 13). Thus, it is important to clarify the mechanism of DAMP secretion and DAMP-mediated immune activation for improvement of the efficacy of cancer chemotherapy and immunotherapy.

Our previous study showed that topotecan (TPT), a topoisomerase I (TOP1) inhibitor, induces the secretion of DNA-containing exosomes from cancer cells as DAMPs and activates STING-mediated tumor immune responses in tumor-bearing mice (14). cGAS recognizes pathogen- and cancer-derived DNA and activates STING, resulting in the induction of cytokine and type I IFN production, which promotes CTL activation and tumor growth inhibition. TPT activates STING-mediated antitumor immune responses via the induction of DAMP secretion from cancer cells; however, we have not elucidated how TPT induces DAMP secretion in cancer cells.

Here we show that RPL15, a 60S ribosomal protein, is a novel TPT target protein, and TPT inhibits preribosomal subunit formation via its binding to RPL15, resulting in the induction of DAMP-mediated antitumor immune activation independent of TOP1. TPT inhibits RPL15–RPL4 interactions and decreases RPL4 stability, which is recovered by CDK12 activity. CDK12 inhibition abrogates TPT-induced RPL4 degradation and DAMP secretion, indicating that CDK12 upregulates ribosomal protein expression and represses DAMP secretion during ribosomal stress. RPL15 knockdown induced DAMP secretion and increased the CTL population but decreased the Treg population in the tumors of a murine B16-F10 melanoma model. B16-F10 tumors became sensitive to PD-1 blockade by short hairpin RNA (shRNA)-mediated RPL15 knockdown, indicating that RPL15 inhibition alters the tumor microenvironment to a more immunogenic state via DAMP secretion. Our study identified a novel TPT target and indicated that ribosomal dysfunction is a trigger of DAMP secretion, which can improve the efficacy of antitumor immunotherapy.

All animals were kept under specific pathogen-free conditions. C57BL/6 mice were purchased from Sankyo Labo Service Corporation, Inc. (Tokyo, Japan). Sting^gt/gt mice were purchased from The Jackson Laboratory (Bar Harbor, ME) (15). All animal experiments were performed with the approval of the Animal Research Committee of Hokkaido University.

TPT was purchased from Cayman Chemical. THZ531 was purchased from MedChemExpress. All cell lines were cultured in DMEM (Sigma-Aldrich) supplemented with 10% FBS (Life Technologies) and 0.05 mM 2-ME (Nacalai Tesque) at 37°C in a humidified 5% CO₂/95% air atmosphere. Murine GM-CSF-induced bone marrow–derived dendritic cells (BMDCs) were prepared by culturing bone marrow cells from the femurs and tibias of 8-wk-old female C57BL/6 mice in RPMI 1640 (Sigma-Aldrich) containing 10% FBS, 10 ng/ml GM-CSF, 0.05 mM 2-ME, and penicillin/streptomycin (Nacalai Tesque) for 7 d.

cDNA of human CDK12 and ribosomal proteins was amplified from CSII-EF-hCDK12-IRES-Venus and E0771 cell cDNA by PCR and inserted into a pCIneo vector (Promega) containing a C-terminal 3xFlag tag sequence. CSII-EF-hCDK12-IRES-Venus was a kind gift from Prof. Nakajima (Osaka City University).

To generate TOP1-deficient cells using the Cas-CRISPR system, human TOP1 exon 4–targeted single-guide RNA (5′-GAGAAGACCAAACACAAAGA-3′) was inserted into pX330-U6-Chimeric_BB-CBh-hSpCas9 vector (Addgene), and DNA fragments of human TOP1 exon 4 were inserted into a pCAG-EGxxFP vector, which was a kind gift of Prof. Ikawa (Osaka University). These vectors were electroporated into MCF7 cells using the Neon transfection system (Thermo Scientific), and then EGFP-positive cells were sorted with a cell sorter (SH800, Sony) and cloned. TOP1 deficiency was confirmed by genomic sequence analysis and immunoblotting using anti-TOP1 Ab (EPR5375; Abcam).

MCF7 cells (2.0 × 10⁵ cells per well) were seeded into 96-well plates and treated with drugs at the indicated concentrations for 48 or 72 h at 37°C. After centrifugation, 2.0 × 10⁵/50 µl BMDCs were cultured with 200 µl supernatant for 48 h, and IL-6 and IFN-β production was then measured by ELISA according to the manufacturer’s instructions (R&D Systems and InvivoGen). BMDCs were prepared as described previously (14).

Total RNA was purified from MCF7 cells and B16-F10 cells using TRI reagent (Thermo Scientific) according to the manufacturer’s instructions, and then cDNA was synthesized from the RNA using ReverTra Ace (TOYOBO). mRNA levels of target genes were quantified using KAPA SYBR Green mix (KAPA Biosystems) with an Mx3005P (Agilent Technologies). Primer sequences of quantitative PCR are described in Supplemental Table I.

For small interfering RNA (siRNA)-mediated knockdown, siRNA was transfected into MCF7 cells using Lipofectamine RNAiMAX (Thermo Scientific) by reverse transfection according to the manufacturer’s instructions. shRNA-expressing MCF7 and B16-F10 cells were prepared using a pLKO.1 puro lentiviral vector. siRNA and shRNA sequences are shown in Supplemental Table I.

TPT, SN-38, or 10-hydroxycamptothecin were dissolved with water-free DMSO, and then 600 µl 20 mM TPT, SN-38, or 10-hydroxycamptothecin was incubated with 3 mg Linker FG beads (Tamagawa Seiki Co., Ltd.) and 35 mg K₂CO₃ for 16 h at 60°C with shaking. After washing, TPT, SN-38, or 10-hydroxycamptothecin beads were suspended in 120 µl Milli-Q water.

For screening of TPT-binding protein, E0771 cells were cultured in 150-mm dishes and lysed in 800 µl KCl buffer (0.5% NP-40, 100 mM KCl, 20 mM HEPES, pH 7.9, 1 mM MgCl₂, 0.2 mM CaCl₂, 0.2 mM EDTA, and 1 mM DTT) containing protease inhibitor mixture. After centrifugation, 0.75 mg TPT beads were incubated with the supernatant for 4 h at 4°C and then washed with KCl buffer. Bead-bound protein was identified by liquid chromatography/mass spectrometry analysis (MedicalProteoScope).

C-terminal 6x His– and 3x Flag tag–fused murine Rpl15 cDNA was inserted into a pFastBac1 vector (Thermo Fisher Scientific). Recombinant RPL15-His-Flag was prepared using the Bac-to-Bac baculovirus expression system according to the manufacturer’s instructions with minor modifications. Briefly, Sf9 cells were lysed in 6 M guanidine-HCl, 50 mM Tris-HCl, pH 8.0, and 0.5 M NaCl by sonication at 72 h after infection of RPL15-His-Flag expressing baculovirus. After centrifugation, the supernatant was rotated with Ni-nitrilotriacetic acid agarose beads (WAKO) for 1 h at room temperature, and then the beads were washed three times with 6 M urea, 50 mM phosphate buffer, pH 6.5, and 0.5 M NaCl. The sample was eluted with 6 M urea, 20 mM acetate buffer, pH 4.0, 0.5 M NaCl, and 0.01 N HCl, and dialyzed to 0.5 M urea, 20 mM Tris-HCl, pH 7.4, 100 mM KCl, 10% glycerol, 1 mM 2-ME, and 0.1% NP-40 by step dialysis. For preparation of recombinant RPL4, Escherichia coli Rosetta (DE3) cells possessing pColdII-Myc-mRpl4 were cultured until OD600 reached 0.6 at 37°C, and then the cells were stimulated with 0.1 mM isopropyl β-d-thiogalactoside for 24 h at 16°C. After lysing the cells, His-Myc-RPL4 was purified as above.

Two micrograms of RPL15-His-Flag was incubated with TPT in KCl buffer for 1 h at 4°C and then incubated with 2 µg His-Myc-RPL4 for 1 h at 4°C. His-Myc-RPL4 was immunoprecipitated using anti-RPL4 Ab (Bio-Rad Laboratories) and nProtein A Sepharose beads (GE Healthcare) and then blotted.

Northern blotting of rRNA was performed as previously described with minor modifications (16). Briefly, total RNA was purified from MCF7 cells using TRI reagent (Sigma-Aldrich) according to the manufacturer’s instructions, and 10 µg RNA was applied to 1.2% agarose gel electrophoresis and then transferred to a nylon membrane. After ultraviolet crosslinking, the membrane was incubated with PerfectHyb (TOYOBO) containing 50 ng/ml digoxigenin-labeled DNA probes, ITS-1 (5′-CCTCGCCCTCCGGGCTCCGGGCTCCGTTAATGATC-3′) and ITS-2 (5′-CTGCGAGGGAACCCCCAGCCGCGCA-3′) (Eurofins Genomics), for 1 h at 65°C. rRNA was then detected using a digoxigenin luminescence detection kit (Roche) according to the manufacturer’s instructions.

B16-F10 cells (1.0 × 10⁶ cells) were s.c. injected into the left flanks of 8-wk-old female C57BL/6 mice, and then the mice were i.p. administered 20 mg/kg anti–PD-1 Ab (29F.1A12; Bio X Cell) or 20 mg/kg control mouse IgG1 (MOPC-21; BioLegend) on day 11 and 10 mg/kg anti–PD-1 Ab or 10 mg/kg control mouse IgG1 on day 17. The tumors were minced on day 21 and incubated in RPMI containing 10% FBS, 0.05 mM 2-ME, 0.2 Wünsch units/ml Liberase TL (Roche), 0.1 mg/ml hyaluronidase (Nacalai), 6 U/ml DNase I (Nippon Gene), 50 ng/ml PMA (Sigma-Aldrich), 100 ng/ml ionomycin (EMD Millipore), 5 µg/ml brefeldin A (BioLegend), and 50 µg/ml kanamycin (WAKO) for 4 h at 37°C with shaking followed by incubation with 10 mM EDTA for 5 min at 37°C. After lysing the RBCs, the cells were incubated with Fc Block (BD) and then stained with Fixable Viability Dye 780 (FVD780; Thermo Scientific) and the indicated Abs. The cells were fixed and stained using the FOXP3/Transcription Factor Staining Buffer Set (Thermo Scientific) with the indicated Abs according to the manufacturer’s instructions. The samples were analyzed with an SH800 (Sony) cell sorter. The Abs for FACS are listed in Supplemental Table I.

MCF7 cells were transfected with siRNA by reverse transfection in 6-well plates and treated with TPT and cycloheximide at the indicated concentrations at 64 h after transfection. The cells were washed with PBS (−) after washing with 20 mM HEPES, pH 7.6, 10 mM KCl, and 5 mM MgCl₂, and then the cells were scraped off in 600 µl sucrose buffer (250 mM sucrose, 20 mM HEPES, pH 7.6, 10 mM KCl, 1.5 mM MgCl₂, 1 mM EDTA, and 1 mM EGTA). The lysate was rotated for 30 min at 4°C and then washed three times with 600 µl sucrose buffer for 5 min at 4°C, whereupon a second wash was performed with 0.2% NP-40/sucrose buffer. After centrifugation, the pellet was suspended in 120 µl radioimmunoprecipitation assay buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1 mM EDTA, 0.1% SDS, 0.5% deoxycholate, and 1% NP-40) containing protease inhibitor mixture (Nacalai) and shaken vigorously for 15 min at 4°C. Nuclear and membrane fractions were separated by centrifugation for 20 min at 15,000 × g, and then the supernatant was analyzed by immunoblotting.

The preribosomal subunit was analyzed as described previously with minor modifications (17). Briefly, MCF7 cells were cultured in 150-mm dishes to semiconfluency and treated with 10 µM TPT for 6 h. After the cells were washed with PBS, they were washed with 10 ml hypotonic buffer (20 mM HEPES, pH 7.6, 10 mM KCl, and 5 mM MgCl₂) and incubated with 20 ml hypotonic buffer on ice for 10 min. After centrifugation, the cells were suspended in hypotonic buffer containing protease inhibitors, 1 mM DTT, and 0.5% NP-40 and rotated for 15 min at 4°C. The lysates were centrifuged at 500 × g for 5 min, and the pellets were washed with 1 ml hypotonic buffer with rotation for 10 min at 4°C and further washed with 400 µl nuclear lysis buffer (20 mM Tris-HCl, pH 8.0, 1.5 mM KCl, 2.5 mM MgCl₂, 0.5% NP-40, and 0.5% deoxycholate) with rotation for 10 min at 4°C. The samples were centrifuged at 500 × g for 5 min and washed with 1 ml hypotonic buffer. The pellet was lysed in 300 µl nuclear lysis buffer containing protease inhibitors with gentle sonication and centrifuged at 12,000 × g for 15 min, then the supernatant was collected as the nucleolar fraction. Linear sucrose gradients (15–45%) were prepared from 60% sucrose and 10× sucrose gradient solution (200 mM HEPES, pH 7.6, 1 M KCl, and 50 mM MgCl₂). After loading the nucleolar fraction on top of the sucrose gradient, the sample was separated by ultracentrifugation for 2 h at 38,500 × g at 4°C with a SW41Ti rotor (Optima L-90K, Beckman Coulter). After ultracentrifugation, the sample was fractionated into 600-µl aliquots by pipetting. The protein was concentrated by acetone precipitation followed by methanol/chloroform precipitation and then detected by immunoblotting.

MCF7 cells were transiently transfected with siRNA by reverse transfection and then treated with TPT at the indicated concentrations and periods at 48 h after transfection. The cells were stained with propidium iodide and annexin V as described previously (14).

B16-F10 cells (5.0 × 10³ cells/well) were seeded into 96-well plates and cultured for the indicated periods at 37°C, then incubated with WST solution for 90 min at 37°C according to the manufacturer’s instructions (WAKO). OD450 of the samples was measured with an iMark plate reader (Bio-Rad Laboratories).

Our previous study showed that TPT-treated cancer cells secrete DNA-containing exosomes as DAMPs, which activate STING-mediated antitumor immune responses (14). However, we have not elucidated how TPT induces DAMP secretion from cancer cells. To clarify the mechanism of TPT-induced DAMP secretion, we investigated the contribution of TOP1, a known target of TPT, to DAMPs and tumor immune activation. First, we prepared TOP1-deficient MCF7 cells using the Cas-CRISPR system (Fig. 1A), and then BMDCs were stimulated with the conditioned medium from TPT-treated wild-type and TOP1 knockout (KO) MCF7 cells. IL-6 and IFN-β production from the BMDCs was comparable between the conditioned medium of wild-type and TOP1 KO MCF7 cells, indicating that TOP1 does not contribute to TPT-induced DAMP secretion (Fig. 1B, 1C). Supporting this, DAMP secretion was not induced in MCF7 cells by rebeccamycin, which also inhibits TOP1 but has poor structural similarity to TPT (Fig. 1D) (18). We further investigated the dependency of TPT-induced cell death on TOP1 and p53. TPT-induced cell death was decreased by TOP1 deficiency with 0.5 µM TPT treatment but comparable between wild-type and TOP1 KO MCF7 cells with 10 µM TPT treatment, which was inhibited by TP53 knockdown (Fig. 1E, 1F). However, p53 did not contribute to TPT-induced DAMP secretion (Fig. 1G). TPT-induced phosphorylation of p53 at S46 was decreased by TOP1 deficiency but not S15 phosphorylation of p53 (Fig. 1H). Expression of CDKN1A, a p53 target gene, was induced by TPT, but the expression was comparable between wild-type and TOP1 KO MCF7 cells, indicating that TPT-induced p53 activation was partially downregulated in TOP1 KO cells (Fig. 1H, 1I). These data indicated that TPT induced DAMP secretion independent of p53 and TOP1 and suggested that there is an unknown TPT target, which correlates with DAMP secretion and antitumor immune activation.

For the screening of novel TPT target proteins, we prepared TPT-conjugated beads using TPT and epoxy-linker FG beads (Fig. 2A). TPT-binding protein in E0771 cell lysate was pulled down using TPT beads and then analyzed by liquid chromatography–tandem mass spectrometry (Fig. 2B). The peptide ratio of TPT beads/Ctrl beads is shown in the horizonal axis of (Figure 2B with log₂ scale, and relative peptide amount is shown in the vertical axis (Fig. 2B, left panel). The proteomic analysis suggested that 60S ribosomal protein and its associated protein were novel TPT target candidates (Fig. 2B, right panel). To validate their interaction with TPT, ribosomal protein was overexpressed in HEK293T cells and then pulled down using TPT beads, resulting in RPL15 interacting with TPT beads (Fig. 2C). We investigated the interaction between TPT and other candidate proteins, such as LGALS3BP and NME4, by TPT bead pulldown, but we did not validate their TPT binding capability (data not shown). Because camptothecin analogs are highly hydrophobic molecules, we assumed that TPT interacts with RPL15 via its hydrophobic amino acids, such as phenylalanine, tryptophan, and tyrosine. We further narrowed down the hydrophobic amino acids of RPL15, which correlate with its interactions with other ribosomal proteins, by structural analysis (19). RPL15–TPT interactions were decreased upon amino acid substitution of Y203 of RPL15 and diminished by the deletion of eight amino acids in the C-terminal region (197–204 aa) of RPL15 (Fig. 2D). Recombinant RPL15 was pulled down by TPT beads, which was inhibited by preincubation of RPL15 with TPT, indicating that RPL15 directly interacted with TPT (Fig. 2E, 2F). RPL15 in the lysate of TOP1-deficient MCF7 cells was pulled down by TPT beads the same as wild-type cells, indicating that TOP1 does not affect RPL15–TPT interactions (Fig. 2G). RPL15 was pulled down by SN-38– or 10-hydroxycamptothecin–conjugated beads, indicating that RPL15 interacts with camptothecin analogs including TPT, SN-38, and 10-hydroxycamptothecin (Fig. 2H). Moreover, TPT–RPL15 interaction was confirmed in various cell lines (Supplemental Fig. 1). These data demonstrated that RPL15 is a novel TPT target protein and interacts with TPT via its C-terminal region.

Next, we investigated whether TPT induces DAMP secretion and cell death in an RPL15-dependent manner. TPT-induced DAMP secretion and cell death were decreased by overexpressing RPL15 ΔC, an RPL15 mutant that does not bind to TPT (Fig. 3A, 3B). siRNA-mediated RPL15 knockdown induced DAMP secretion without TPT in MCF7 cells, which activated STING-mediated signaling in BMDCs, indicating that TPT induces DAMP secretion via inhibiting RPL15 function (Fig 3C, 3D). Because the C-terminal region of RPL15 is adjacent to RPL4 in the 80S ribosome, we speculated that TPT inhibits RPL15–RPL4 interactions via its binding to the C-terminal region of RPL15 (19). Recombinant RPL15 and RPL4 form RPL15–RPL4 complexes in vitro, and this complex formation was inhibited by TPT (Fig. 3E). A previous study showed that RPL4 dissociation from the pre-60S ribosomal subunit triggered RPL4 degradation by TOM1-mediated ubiquitination (20, 21). We investigated whether TPT inhibits pre-60S ribosomal subunit formation and increases RPL4 ubiquitination in vivo. RPL4 ubiquitination was increased by TPT treatment in HEK293T cells (Fig. 3F). RPL4 knockdown induced DAMP secretion in MCF7 cells, the same as RPL15 knockdown, but not RPL7A knockdown, which is another RPL15-interacting ribosomal protein (Fig. 3G). The densities of RPL15- and RPL4-containing pre-60S ribosomal subunits were altered by TPT treatment independent of TOP1 in MCF7 cells, suggesting that TPT affects the assembly of the pre-60S ribosomal subunit in vivo (Fig. 3H, 3I). Levels of 45S and 47S pre-rRNA were decreased by TPT, but 32S pre-rRNA was accumulated after TPT treatment, which was recovered by TOP1 deficiency but not TP53 knockdown (Fig. 3J–3L). These data indicate that TPT triggers DAMP secretion via its binding to RPL15 and induces ribosomal stress, including aberrant assembly of preribosomal subunits and rRNA processing errors. Moreover, our data indicate that RPL15 and RPL4 downregulation promotes DAMP-mediated innate immune activation.

Our data showed that expression of ribosomal proteins and ribosomal stress correlate with DAMP secretion and antitumor immunity, suggesting that regulators of ribosomal protein expression also correlate with these antitumor immune responses (Fig. 3A, 3C). A previous study showed that CDK12 promoted ribosomal protein expression and that its inhibition improved anti–PD-1 therapy in a tumor-bearing mouse model (22–24). Thus, we investigated whether CDK12 promotes RPL15 and RPL4 expression and correlates with DAMP secretion from cancer cells. siRNA-mediated CDK12 knockdown enhanced TPT-induced DAMP secretion in MCF7 cells, but it was not affected by CDK13 knockdown, which is a paralogue of CDK12 (Fig. 4A–4D). Similarly, TPT-induced DAMP secretion was enhanced by THZ531 treatment, a CDK12/13 inhibitor (Fig. 4E). DAMP secretion was induced by knockdown of RPL15 and RPL4, which was augmented by THZ531 treatment, suggesting that CDK12-mediated regulation of ribosomal protein expression correlates with DAMP secretion (Fig. 4F). TPT decreased the stability of nuclear RPL4 during cycloheximide treatment, but RPL4 protein levels were not decreased by TPT treatment only (Fig. 4G). RPL4 protein expression was decreased by CDK12 knockdown together with TPT treatment, indicating that CDK12 upregulates the expression and stability of RPL4 during TPT-induced RPL4 destabilization (Fig. 4H). Consistent with this, CDK12 overexpression increased RPL4 stability in HEK293T cells, but not RPL15 (Fig. 4I). Total mRNA levels of RPL15 and RPL4 were not changed by TPT and THZ531 treatment, suggesting that CDK12 promotes the protein stability or translational efficiency of RPL4 during TPT treatment (Fig. 4J). Because CDK12 promoted the expression of ribosomal proteins during TPT treatment, we then investigated the effect of CDK12 on TPT resistance in MCF7 cells. CDK12 knockdown decreased TPT-induced cell death in both wild-type and TOP1 KO MCF7 cells (Fig. 4K, 4L). TPT-induced p53 activation was inhibited by CDK12 knockdown, indicating that TPT binds to TOP1 and RPL15 and then induces p53-mediated cell death through CDK12 activation (Fig. 4M, 4N). These data indicate that CDK12 compensates for the downregulation of ribosomal protein and represses DAMP secretion during ribosomal dysfunction.

Because our data showed that RPL15 downregulation triggered DAMP secretion and activated a STING-mediated immune response, we further investigated whether RPL15 inhibition alters tumor immune response in vivo using a tumor-bearing mouse model. Cell proliferation was slightly decreased in Rpl15 shRNA-expressing B16-F10 cells compared with control shRNA-expressing B16-F10 cells (Fig. 5A–5C). shRNA-mediated Rpl15 knockdown induced DAMP secretion from B16-F10 cells but did not affect PD-L1 expression in B16-F10 cells (Fig. 5D, 5E). Rpl15 knockdown decreased the growth of B16-F10 tumors and sensitized the tumors against anti–PD-1 treatment (Fig. 5F–5H). The population of CD8⁺ granzyme B⁺ IFN-γ⁺ T cells was increased, but Foxp3⁺ CD4⁺ T cells were decreased, in B16-F10 tumors by Rpl15 knockdown. Conversely, the myeloid-derived suppressor cell population was not changed by Rpl15 knockdown or anti–PD-1 Ab treatment (Fig. 5I–5K, Supplemental Fig. 3). These data indicate that Rpl15 knockdown induced DAMP secretion from melanoma cells and increased the CTL/Treg ratio in the tumor microenvironment, sensitizing PD-1–resistant tumors against anti–PD-1 treatment.

Many studies reported that DAMP secretion is induced by various stimuli and compounds, such as irradiation, anticancer drugs, uric acid, and pathogen infection (25). However, how these stimuli induce DAMP production intracellularly has been largely elusive. Our previous study showed that TPT induces DAMP secretion from cancer cells, indicating that there is an unknown DAMP-inducing signaling mechanism that is activated by TPT, but we have not clarified whether TOP1 is correlated with TPT-induced DAMP secretion (14). TOP1 is a known target of TPT, but it has been considered that TPT and camptothecin derivatives have unknown targets except TOP1, because TPT alters gene expression dependent on or independent of TOP1 (26). In this study, we identified RPL15 as a novel target of TPT and found that ribosomal dysfunction is a trigger of DAMP induction, which contributes to antitumor immune responses.

Ribosomal stress is induced by various cellular stresses, such as chemicals, starvation, and heat shock, resulting in the dysregulation of ribosome biogenesis and ribosomal proteins. Previous studies showed that several ribosomal proteins translocate from the nucleoli to the nucleoplasm during ribosomal stress and activate p53 via MDM2 inhibition (27–29). Although p53 activation is important for the ribosomal stress response, p53 did not correlate with TPT-induced DAMP secretion (Fig. 1G). Our data showed that TPT inhibits RPL15–RPL4 interactions and preribosomal subunit formation, indicating that ribosomal dysfunction activates the intracellular signaling that induces DAMP secretion, including DNA-containing exosomes, which are induced in several cancer cell lines but not primary cells (Fig. 3E, 3H) (14). These results suggest that p53-independent and ribosomal stress response signaling regulates DAMP secretion, and this signaling correlates with the oncogenic cellular state and thus might be an attractive target for promoting a more immunogenic state of the tumor microenvironment.

Our data showed that TPT induces DAMP secretion from MCF7 cells independently of TOP1, but TPT-induced dysregulation of rRNA processing is restored by TOP1 deficiency (Figs. 1B, 1C, 3K). TPT altered the distribution of NPM1, a nucleolar marker protein, in the nucleus in a TOP1-dependent manner, suggesting that nucleoli formation is inhibited by TPT dependent on TOP1 (Supplemental Fig. 2). Moreover, knockdown of RPL4 and RPL15 induces DAMP secretion but not RPL7A knockdown, which is an RPL15-interacting ribosomal protein (Fig. 3C, 3G). These results suggest that DAMP secretion is induced partly by ribosomal stress, including dysregulation of preribosomal subunit formation and downregulation of specific ribosomal proteins, but not all ribosomal dysfunction.

Because previous studies showed that camptothecin induces autophagy, we investigated the contribution of autophagy to TPT-induced DAMP secretion. However, TPT-induced DAMP secretion was not affected by ATG5 or BECN1 knockdown, which are critical regulators of autophagy (data not shown). ATF4 knockdown and protein kinase RNA–like endoplasmic reticulum kinase inhibitor also did not affect TPT-induced DAMP secretion, suggesting that endoplasmic reticulum stress response signaling does not correlate with DAMP secretion (data not shown). In this study, we did not elucidate DAMP-inducing signaling, which is activated by TPT treatment or downregulation of RPL15 and RPL4, and thus further studies are required to clarify this signaling.

Previous studies demonstrated that various cancer cells aberrantly express ribosomal protein, which promotes or represses cancer proliferation and metastasis (30, 31). Elevated expression of RPL15 promotes lung metastasis in breast cancer cells as well as the proliferation of gastric cancer cells (32, 33). Our data showed that RPL15 knockdown increased the CTL/Treg ratio and enhanced the antitumor effects of PD-1 blockade treatment in a murine melanoma model (Fig. 5E). These data suggest that RPL15 is a good target for cancer immunotherapy, but an RPL15-specific inhibitor has not been developed. Our study showed that TPT has immunostimulatory effects via RPL15 inhibition and independent of TOP1, and Iwai et al. showed that FM3A breast tumors were sensitized against PD-1 blockade by irinotecan treatment, which is a TPT analog, and decreased the Treg population in the tumor (34). Hibino et al. also showed that irinotecan decreased the Treg population in the murine Lewis lung tumor model (35). On the basis of these data, we speculate that camptothecin derivatives that lack TOP1 inhibitory activity act as an RPL15-specific inhibitor and enhance for cancer immunotherapy such as PD-1 and CTLA-4 blockade. Combination therapy of PD-1 and irinotecan has proceeded to clinical trials, but irinotecan has side effects, including digestive disorders and bone marrow suppression due to its high toxicity (36–38). Our data showed that TPT-induced cell death was decreased by TOP1 deficiency at its lowest concentration, suggesting that the cytotoxicity of camptothecin derivatives is largely dependent on TOP1 rather than RPL15 (Fig. 1E). Thus, an RPL15-specific inhibitor might be less cytotoxic but might have immunostimulatory effects, the same as irinotecan and TPT.

CDK12 interacts with cyclin K and upregulates phosphorylation of the C-terminal domain of the large subunit of RNA polymerase II, resulting in optimal splicing and polyadenylation of various genes, which correlates with the DNA damage response and cell differentiation (39–41). Choi et al. showed that CDK12 promotes the translation of ribosomal proteins via upregulating 4E-BP1 phosphorylation (22). Consistent with this, our data showed that CDK12 upregulated RPL4 expression during TPT-induced RPL4 destabilization and ribosomal stress (Fig. 4G). TPT-induced cell death was inhibited by CDK12 knockdown in both wild-type and TOP1-deficient MCF7 cells, indicating that ribosomal stress induces CDK12 activation, which restores ribosomal dysfunction by upregulating the translation of ribosomal proteins and promotes p53-mediated cell death (Fig. 4G, 4H, 4K). Previous studies showed that CDK12 downregulation leads to genomic instability and increased cancer immunogenicity, which sensitized prostate and breast tumors to PD-1 blockade in patients and a murine tumor model (23, 24, 42). Our data showed that CDK12 inhibition abrogated DAMP-mediated immune responses during ribosomal stress, suggesting that CDK12 inhibition increases cancer immunogenicity through both genomic instability and dysregulation of ribosomal protein expression (Fig. 4B, 4G).

Numerous studies showed that STING activation in immune cells induces type I IFN and cytokine production, which promotes the activation of CTL and its recruitment to tumor sites (43–45). Our studies showed that STING-activating DAMPs are secreted from TPT-treated cancer cells and RPL15 knockdown cells, which promotes CTL activation in E0771 and B16-F10 tumors (Figs. 3D, 5I) (14). Although previous studies showed that STING activation does not induce the reduction of Treg populations at tumor sites, Treg populations were decreased in shRPL15-expressing B16-F10 tumors (Fig. 5I) (46, 47). These data suggest that RPL15 knockdown cancer cells secrete cytokines or both STING-activating and nonactivating DAMPs, which inhibit Treg infiltration or differentiation in the tumor. Supporting this idea, BMDC responses were decreased but not diminished by STING deficiency during stimulation with conditioned medium from TPT-treated or RPL15 knockdown cancer cells (Fig 3D). Treg-suppressive DAMPs have not been reported, although several DAMPs promote Treg expansion after inflammation, such as HMGB1 and HSP90. Thus, further studies on the correlation between RPL15 downregulation and Treg suppression are required.

Finally, ribosomes and their associated proteins are essential for protein synthesis, but they also act as “transmitters” that receive cellular stress and input the signaling, which correlates with inflammation, cell differentiation, and cell death (48, 49). In addition, our work indicates that ribosome-mediated signaling is important for DAMP secretion and tumor immunity, which are induced by knockdown of RPL15 and RPL4 but not RPL7A (Fig. 3C, 3G). Previous studies showed that RPL11, RPL6, and RPS14 contribute to p53 activation, and RPS3 facilitates NF-κB–mediated transcriptional regulation (27–29, 50). On the basis of these findings, each ribosomal protein contributes to different intracellular signaling pathways, suggesting that it is important to clarify the function of individual ribosomal proteins to determine the mechanism of ribosome-mediated signaling. However, it has been largely elusive and challenging because the ribosome is a complex composed of 79 ribosomal proteins and ∼400 ribosome-associated proteins in mammals (51). Various cellular events might be further understood by clarifying the molecular mechanism of ribosome-mediated signaling, which supports the progression of studies on immunity, development, and cancer.

We thank Dr. Takumi Ito, Prof. Hiroshi Handa, Dr. Atsushi Furukawa, and Dr. Hiroshi Maita for helpful discussion. We thank H. Nikki March, PhD, Edanz Group (https://en-author-services.edanz.com/ac) for editing a draft of the manuscript.

The authors have no financial conflicts of interest.