Anti-tumor analysis of the RIG-I agonist in and .

1
Introduction
Pattern recognition receptors (PRRs) recognize pathogen-associated molecular patterns (PAMPs) and damage-associated molecular patterns (DAMPs) to initiate innate immunity and potentiate adaptive immune responses. The major PRR families comprise Toll-like receptors (TLRs), C-type lectin receptors (CLRs), RIG-I-like receptors (RLRs), Nod-like receptors (NLRs) and its subfamily AIM2-like receptors (ALRs) [1]. Ligand engagement triggers PRR-dependent signaling cascades, inducing activating transcription factors (e.g., nuclear factor kappa-B/NF-κB, Interferon regulatory factors/IRFs) and cytokine/chemokine secretion, which collectively coordinate innate effector functions and bridge humoral/cellular adaptive immunity [2,3].
Retinoic acid-inducible gene I (RIG-I) specifically detects intracellular blunt-ended 5′-triphosphate double-stranded RNA (5′-ppp dsRNA) via its C-terminal regulatory domain (CTD) and helicase domain [4,5]. Upon RNA binding, RIG-I undergoes conformational changes and oligomerizes, exposing its N-terminal caspase activation and recruitment domain (CARD) [6]. These domains engage the mitochondrial antiviral signaling protein (MAVS) on mitochondrial membranes to trigger subsequent activation and nuclear translocation of IRF3 and NF-κB, inducing type I interferon (IFN–I) production and subsequent type I IFN-dependent apoptosis mediated by IRF3 or NF-κB [7,8].
Independent of the antiviral response, RIG-I activation drives mitochondria-dependent apoptosis by up-regulating pro-apoptotic BH3-only proteins (e.g., Noxa, Puma) and sensitizing cells to tumor necrosis factor family cytokines [9]. Also, it enhances tumor immunogenicity by promoting antigen presentation and CD8+ T cell recruitment, establishing its therapeutic potential in overcoming cancer immune evasion [10,11]. Consistently, downregulated RIG-I expression promotes tumorigenesis in both human colitis-associated colorectal cancer tissues and murine models [12]; in HCC, decreased RIG-I is associated with advanced stage, poor survival and response to IFN-α therapy [13]; RIG-I protein mutant also triggers increased susceptibility to colitis-related colon cancer [14]. These findings highlight a tumor-suppressive role for RIG-I in antitumor immunity [[12], [13], [14]].
Consequently, RIG-I agonists are emerging as promising therapeutic candidates or immune adjuvants for oncology. In addition, the cell-free synthesis of RIG-I agonists offers high synthetic tractability beyond their immunological functions. Although no RIG-I agonists have yet reached clinical approval, preclinical studies validate SLR14, MK-4621, and M8 as potent candidates, reinforcing RIG-I as a druggable target for cancer therapy [[15], [16], [17]]. However, the therapeutic efficacy of RIG-I agonists against hepatic and pulmonary carcinomas remains poorly characterized. To address this limitation, we engineered a blunt-ended 5′-ppp dsRNA agonist and analyzed its antitumor efficiency. Functionally, nCoV-L potently activated RIG-I signaling, inducing broad-spectrum tumor cell apoptosis in vitro and suppressing tumor growth in preclinical models, thereby highlighting its potential as a novel anticancer agent.

2
Materials and methods
2.1
Cells
Mouse hepatocellular carcinoma (HCC) cells Hepa1-6, Hep3B, human large cell lung cancer cell NCI–H460, and non-small cell lung cancer cell (NSCLC) NCI–H1299 were purchased from ATCC. Hepa1-6-luc cell was purchased from iCell (cat#iCell-0034a). Huh7 and Huh7.5.1 were preserved in our laboratory. Cells were cultivated in DMEM (Gibco, cat#11995065; iCell, cat#iCell-0034a-001b), RPMI 1640 (Gibco, cat#A10491-01) or MEM medium (Gibco, cat#41500034) supplemented with 10 % fetal bovine serum (Gibco, cat# F8318) and 1 % penicillin-streptomycin (10,000 U/mL) 20 ✕ 100 mL (Gibco, cat# 15140163). Cells were cultured in a 37 °C cell culture incubator containing 5 % CO2.

2.2
Mice
Female BALB/c nude mice and C57BL/6 mice (6 weeks old) were obtained from the Laboratory Animal Resource Centre of the National Institute for Food and Drug Control (NIFDC). All mice were maintained in a barrier-sustained specific pathogen-free (SPF) facility under controlled conditions: temperature 22 ± 1 °C, humidity 55 ± 5 %, 12-h light/dark cycle, with sterilized feed and water provided.

2.3
Antibodies and reagents
Antibodies against cleaved-PARP (cat#5625S), cleaved-Caspase3 (cat#9661S), PUMA (cat#98672), p-IRF3 (cat#29047), STAT1 (cat#14994), p-STAT1 (cat#9167), p–NF–κb (cat#3033), ISG56 (cat#14769) and β-actin (cat#4967) were purchased from Cell Signaling Technology. RIG-I (cat#A0550) antibody was purchased from ABclonal; β-tubulin antibody (cat#10094-1-AP) was sourced from Proteintech. The pan-caspase inhibitor Z-VAD-FMK (cat#161401-82-7) was purchased from MedChemExpress. cOmplete™. EDTA-free Protease Inhibitor Cocktail was purchased from Roche (cat#4693132001). Dithiothreitol (DTT, cat#ST041) was purchased from Beyotime and Phenylmethane-sulfonyl fluoride (PMSF, cat#BL507A) was from Biosharp.

2.4
In vitro transcription
The nCoV-L sequence was designed as:
5′-GGUUUAAUACCUAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA.
AAAAUCCCUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUAGGUAUUAAACC-3’.
The target sequence was cloned into the pUC19 plasmid vector and amplified via bacterial propagation. The purified plasmid was digested overnight at 55 °C with restriction endonuclease BsmBI-v2 (NEB, cat#R0580). Linearized nCoV-L template DNA was isolated by 1.5 % (w/v) agarose gel electrophoresis and purified with V-ELUTE Gel Mini Purification Kit (Zhuang Meng Biotechnology, cat#ZPV202) according to the manufacturer's protocol. In vitro transcription was conducted with a T7 RiboMax™ Express Large Scale RNA Production System (Promega; 20 μL system: 1 μg linear DNA template, 2 μL T7 Transcription Optimized 5 × Buffer, 1 μL T7 RNA Polymerase) at 37 °C for 2 h. RNA transcripts were then subjected to DNase digestion at 37 °C, 20 min and purified according to the manufacturer's instruction. RNA numbered 6, 19, 35 and M8 were transcribed in a similar way as nCoV-L.

2.5
RNA transfection
RNA was transfected using Lipofectamine 3000 kit (Invitrogen, cat#L3000015) with Opti-MEM Reduced-Serum Medium (Gibco, cat#51985034) when cells reached 70–90 % confluence. RNA and Lipofectamine 3000 (short for lipo) were separately diluted in Opti-MEM, gently mixed, and incubated for 10 min at room temperature at a mass ratio of 1:2 (μg RNA: μL Lipo). The control groups (Opti-MEM only, or Lipo group including Lipofectamine 3000 and Opti-MEM), or RNA-Lipo complex group were then added to the cells, respectively. For in vivo transfection, the same protocol was applied with a total transfection volume of 100 μL per injection.
5ʹ-PPP-deficient nCoV-L RNA was synthesized by Sangon Biotech Company.

2.6
Cell mortality assay (CCK8 assay)
Cell viability was assessed using the CCK8 kit (Dojindo, cat#CK04). After adding CCK-8 reagent (10 % of total volume), plates were incubated for 1–4 h at 37 °C under 5 % CO2 in the dark. Absorbance was measured at 450 nm using a microplate reader. Optical density (OD) values per group were averaged from three replicates.

2.7
Evaluation of nCoV-L anti-tumor effect in vivo
NCI–H1299 or Huh7 cells (2.5 × 106) were suspended and injected subcutaneously into the right midsternal of 6-week-old female nude mice; Hepa1-6-luc cells (3 × 106) were similarly injected into 6-week-old C56BL/6 mice. When the tumor diameters reached to 5 mm approximately, mice in the control and experimental groups received daily intratumoral injections for 5 consecutive days of equal volumes of PBS, Lipo, or 10 μg nCoV-L complexed with Lipo. For NCI–H1299 and Huh7-bearing mice, tumor length (L) and width (W) were measured daily, and volumes were calculated as V = (L × W2)/2. Hepa1-6-luc tumor growth was monitored by in vivo bioluminescent imaging assay as previously described [18], with luciferin (150 mg/kg) injected intraperitoneally 10 min prior to imaging.

2.8
Real-time quantitative PCR
Total cell RNA was extracted with PureLink RNA mini kit (Invitrogen, cat#12183018A) and subjected to RT-qPCR analysis with PrimeScript One Step RT-PCR Kit (Takara Bio, cat#RR055A), according to the manufacturers’ protocols. Gene-specific PCR primers were synthesized by Sangon Biotech Company. All the primer sequences were shown below:
PUMA-Fw:GACCTCAACGCACAGTA.
PUMA-Rev:CTAATTGGGCTCCATCT.
GADPH-Fw:GCACCGTCAAGGCTGAGAAC.
GADPH -Rev:TGGTGAAGACGCCAGTGGA.

2.9
Protein extraction and quantification
All cells were collected by centrifugation at 12,000 rpm for 2 min at 4 °C. After supernatant removal, cell pellets were resuspended in 100–200 μL of ice-cold RIPA lysis buffer (Sigma-Aldrich, cat# R0278) supplemented with 1 mM DTT, 1 mM PMSF, and 1 × protease/phosphatase inhibitor cocktail. Lysates were lysed on ice for 10 min with intermittent vertexing, followed by centrifugation at 12,000 rpm for 15 min at 4 °C to collect supernatants. The protein concentrations were quantified by the BCA Protein Quantification Kit (Thermo scientific, cat#23235) with bovine serum albumin (BSA) as a standard.

2.10
Statistical analysis
Statistical analysis of cell mortality assay data employed Student's t-test, while tumor volume data were analyzed by two-way ANOVA with GraphPad Prism 9. And P < 0.05 was defined statistically significant between sample groups.

3
Results
3.1
nCoV-L exhibited broad-spectrum cytotoxicity against tumor cells
Leveraging the 5′ UTR sequence of SARS-CoV-2, we engineered multiple stem-loop-structured RNAs bearing 5ʹ-PPP modifications. Through in vitro transcription, we synthesized double-stranded 5ʹ-PPP RNA agonists and screened four candidates for RIG-I activation potency (Fig. S1A). Primary screening across five human cancer cell lines (A549, NCI–H1299, Hepa1-6, Hep3B, Huh7) identified nCoV-L as the most potent cytotoxic agent, surpassing the reference agonist M8 in efficacy when cell death rates were quantified 24 h post-transfection with CCK8 assay. All candidate RNAs were ranked by tumor cell lethality, with nCoV-L exhibiting the highest cytotoxicity across all tested cell lines (Fig. S1B). Consequently, nCoV-L was advanced to further analysis.
nCoV-L is a short 5ʹ-PPP RNA with a stem-loop structure (Fig. 1A). To further verified the activation of RIG-I signalling in cancer cells, we detected the protein levels of RIG-I signalling components in Hep3B, NCI–H1299 and Huh7 cells. The results demonstrated that expression of RIG-I, p-IRF3, STAT1, p-STAT1, p–NF–κb and ISG56 was induced by nCoV-L treatment in those cell lines (Fig. 1B and C; Fig. S1C).
Huh7.5.1 harbored a T55I mutation which impairs MAVS interaction and consequent RIG-I signalling transduction [19]. Transfection of nCoV-L into Huh7.5.1 cells significantly attenuated nCoV-L-induced cell death compared to wild-type Huh7 cells at 24 h post-transfection (P < 0.05) (Fig. 1D). Lack of 5ʹ-PPP also abolished nCoV-L-mediated cytotoxicity in Hep3B cells (Fig. 1E), further validating the requirement of 5′-PPP for RIG-I activation. In addition, nCoV-L exhibited broad-spectrum, dose-dependent cytotoxicity: 50 ng induced ∼60 % cell death in Hep3B, NCI–H460, SW620, SW480, and CT26WT cells, wheras 100–150 ng achieved >80 % lethality across all tested lines (Fig. 1F).
In the nude mouse models implanted with NCI–H1299, daily intratumoral injections of nCoV-L (10 μg × 5 doses) significantly suppressed tumor growth versus PBS or lipo controls (P < 0.0001; Fig. 1E and F). Tumor volumes were reduced by 67 %–93 % in nCoV-L group versus control and lipo groups. Similarly, reduction in tumor volumes was observed in nCoV-L treated Huh7 or Hepa1-6 xenografts (Fig. S2). These results suggested that nCoV-L can induce potent cancer cell death and supresses tumor growth in vivo through RIG-I activation.

3.2
nCoV-L triggers tumor cell death by activating the intrinsic apoptotic pathway
Based on the established role of RIG-I/MAVS signalling in caspase-dependent apoptosis [8], we employed the pan-caspase inhibitor Z-VAD-FMK to validate the apoptotic mechanism of nCoV-L. Co-treatment with Z-VAD-FMK significantly attenuated nCoV-L-induced cytotoxicity in Hep3B cells compared to nCoV-L monotherapy (P < 0.001; Fig. 2A), proving the caspase dependency of nCoV-L-mediated tumor cell death. Concurrently, in Hep3B and Huh7, nCoV-L upregulated the mRNA level of PUMA (Fig. 2B and C), whose translation was also induced in Huh7 (Fig. S1C). Furthermore, nCoV-L activated caspase 3, evidenced by cleavage of its substrate PARP (Fig. 2D and E). Collectively, these findings establish that nCoV-L induced tumor cell death predominantly via the intrinsic apoptotic pathway.

4
Discussion
RIG-I recognizes 5ʹ-PPP dsRNA to trigger apoptosis through pro-apoptotic protein induction (e.g., PUMA/Bax) or anti-apoptotic protein (e.g., Bcl2) suppression, enabling its therapeutic exploitation against diverse cancers including melanoma, breast cancer, HCC [20,21]. Consequently, RIG-I ligands offer broad therapeutic potential as immunomodulatory agents or direct cytotoxic inducers. Although MDA5 agonists such as poly(I:C) exhibit antitumor efficacy [7,22], their large molecular size (>1 kb) complicates synthesis and escalates production costs compared to compact RIG-I ligands.
However, the therapeutic potential of RIG-I agonists in hepatic, pulmonary, and colorectal carcinomas remains largely uncharacterized. To address this gap, we engineered nCoV-L, which demonstrated potent cytotoxicity in vitro and suppressed tumor growth in xenograft models of these carcinomas. Notably, RNA delivery efficiency critically depends on nanocarrier systems. Optimizing nCoV-L encapsulation (e.g., lipid nanoparticles/LNPs) may enhance tumor targeting and cytotoxicity compared to delivery via Lipofectamine 3000. Meanwhile, further investigation is required to determine whether nCoV-L modifies the tumor microenvironment, promotes immune cell infiltration, or whether systemic administration can inhibit distant tumors through antigen-specific T cell responses.
Although conventional therapies—such as surgical resection, chemotherapy, and radiotherapy—can partially control tumor progression, their efficacy is often limited by tumor heterogeneity and immunosuppressive tumor microenvironment (TME) dynamics, leading to persistently high risks of recurrence or metastasis [[23], [24], [25], [26], [27], [28], [29]]. Recently, novel immunotherapy strategies have emerged as transformative approaches for cancer treatment, including: immune checkpoint inhibitors (e.g., PD-1/PD-L1 inhibitors), CAR-T cell therapy, oncolytic viral immunotherapy, and therapeutic cancer vaccines [[29], [30], [31], [32], [33]]. Crucially, combining therapies with orthogonal mechanisms (e.g., RIG-I agonists and ICIs) generates synergistic antitumor effects by enhancing antigenicity and adjuvanticity. Multiple clinical trials have validated the efficacy of such combinatorial strategies [17,34,35].
This study identifies nCoV-L as a novel RIG-I ligand inducing broad-spectrum cytotoxicity across diverse carcinomas, targeting the following cell lines: HCC (Hep3B, Huh7), hepatocellular carcinoma (Hepa1-6), pulmonary carcinomas (A549, H1299, NCI–H460), CRC (SW480, SW620) and colon cancer (CT26.WT). nCoV-L thus expands the therapeutic scope of RIG-I pathway and provides a pivotal tool for developing next-generation combination immunotherapies.

CRediT authorship contribution statement
Qian Wang: Writing – review & editing, Writing – original draft, Visualization, Validation, Project administration, Investigation. Ziyang Song: Writing – review & editing, Validation. Bopei Cui: Writing – review & editing. Chaoqiang An: Validation. Fan Gao: Writing – review & editing. Mingchen Liu: Writing – review & editing. Lu Li: Writing – review & editing. Xiao Ma: Writing – review & editing, Writing – original draft, Supervision. Xing Wu: Writing – review & editing, Writing – original draft, Supervision. Zhenglun Liang: Writing – review & editing, Writing – original draft, Conceptualization.

Funding
This work was supported by Chinese Academy of Medical Science Innovation Fund for Medical Sciences (2021-I2M-5-005).

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.