Chemically programmed STING-activating nano-liposomal vesicles improve anticancer immunity

Cancer immunotherapy using immune checkpoint blockade (ICB) inhibitors (e.g., anti-PD-1/PD-L1 and anti-CTLA-4 antibodies) has achieved remarkable clinical success, yielding durable and long-term therapeutic responses in multiple cancer types. However, only a small subset of patients benefit from this treatment. Many inaccessible tumors are immunologically cold, characterized by abundant infiltration of immune suppressors and paucity of tumor-infiltrating lymphocytes, enabling escape from immune surveillance. Consequently, most cancers exhibit de novo refractoriness to FDA-approved ICB antibodies. Therefore, effective immunotherapeutic approaches beyond those directly targeting the adaptive immune response are needed. STING is an intracellular receptor regulating innate immune pathways critical for initiation of antitumor immunity. Activation by endogenous cyclic dinucleotides (e.g., cGAMP) induces type-I interferons and proinflammatory cytokines, stimulating dendritic cell activation and cross-presentation, reversing tumor immune desertification. However, CDN-based agonists suffer metabolic instability, rapid clearance, delivery challenges due to hydrophilicity/negative charge, and risk of systemic inflammation upon IV dosing, limiting clinical trials to intratumoral injection for accessible tumors. Systemic intravenous regimens could better address unresectable and metastatic disease. MSA-2 is a non-nucleotide STING agonist with oral administration in animals, but low oral bioavailability and inadequate cytosolic entry may limit efficacy. The authors hypothesized that the carboxyl moiety (10-OH) on MSA-2 impairs compatibility with drug carriers and that rational chemical derivatization could improve delivery and therapeutic outcomes. They develop esterase-activatable MSA-2 prodrugs formulated into liposomal vesicles (SAProsomes) to enhance pharmacokinetics, tumor/lymphoid delivery, and antitumor immunity via systemic administration.

Study design: The authors designed esterase-responsive MSA-2 prodrugs to improve liposomal formulation and systemic delivery of STING agonists. Four morpholine (MP)-type alkanol ester MSA-2 prodrugs (compounds 1–4) with varying linker hydrocarbon lengths were synthesized, characterized, and formulated into liposomal vesicles (SAProsomes) for in vitro and in vivo evaluation in multiple murine tumor models, alongside pharmacokinetic and biodistribution studies, immune profiling, and safety assessments. Synthesis and characterization: Four MSA-2 prodrugs (1–4) were synthesized by esterifying the carboxyl group of MSA-2 with MP-typed alkanols of different lengths. Structures were confirmed by 1H NMR and high-resolution mass spectrometry. An alternative tocopherol-derived prodrug (compound 5) was also synthesized and evaluated in preliminary studies. Prodrug activation kinetics: Hydrolysis was assessed in PBS at 37 °C with or without porcine liver esterase (PLE, 50 U/mL). Conversion to free MSA-2 was quantified; pseudo-first-order rate constants were derived. SAProsome preparation and characterization: Liposomes were prepared via ethanol dilution. Lipids (egg-phosphatidylcholine:cholesterol:DSPE-PEG2000 at 7:2:1 w/w) were dissolved in ethanol (0.9 mL). MSA-2 prodrug in DMSO (0.1 mL, 20 mg/mL; lipid:MSA-2 equivalence 14:1 w/w) was added, then rapidly injected into PBS (9 mL, pH 7.4) to form liposomes (0.2 mg/mL MSA-2 equivalence). Liposomes were ultracentrifuged (100,000 × g, 30 min) and washed thrice with PBS. Loading efficiency (>90%) and particle size (~120 nm), PDI, morphology (TEM, Cryo-TEM) and stability in serum-containing buffer (20% w/v) were evaluated. Cell lines and culture: THP1, LLC, B16F10-OVA cell lines were used. Primary BMDCs were derived from C57BL/6 mice (GM-CSF 20 ng/mL, IL-4 5 ng/mL; days 6–8). Cellular uptake and STING activation in vitro: BMDCs and THP1 cells were incubated with free MSA-2 or SAProsomes (40 µM MSA-2 equivalence). Intracellular drug fluorescence (MSA-2 intrinsic signal) was quantified by flow cytometry (405 nm excitation, 450/45 nm emission). STING pathway activation was assessed by qRT-PCR (IFN-β, TNFα, CXCL10 mRNA), ELISA for IFN-β (dose-response, EC50), and Western blot in THP1 for p-STING, p-TBK1, p-IRF3. DC maturation markers (MHC II, CD80/CD86) were measured by flow cytometry. Antigen cross-presentation was evaluated using SIINFEKL-H-2Kb staining after OVA exposure. CTL cytotoxicity was measured via LDH release in co-cultures of BMDCs pretreated with SAProsomes+OVA, splenic T cells, and B16F10-OVA or B16F10 target cells (ratio 10:1:5). Animal models and treatments: Ethical approval obtained; animals housed SPF. Tumor volume = 0.5 × length × width^2; endpoint 2000 mm^3. - MC38 colorectal model (C57BL/6J): 5×10^6 cells s.c.; when ~100 mm^3, mice randomized (n=10/group) to saline (IV), free MSA-2 (PO), or SAProsomes (IV) at 35 mg/kg MSA-2-equivalent on days 0, 4, 8. Tumor growth, body weight monitored every 2 days. Histology (H&E), TUNEL, and immunofluorescence (CD8, granzyme B) on excised tumors; survival (Kaplan–Meier). Long-term survivors (day 150) were rechallenged with MC38 cells to assess immunological memory. - Immune profiling (MC38): On day 10 post-treatment, flow cytometry quantified tumor-infiltrating lymphocytes (CD3+CD8+, CD3+CD4+), intracellular IFN-γ/TNFα, CD107a, NK cells, DC activation (CD80/CD86), macrophage polarization (M1/M2; CD206), and MDSC subsets (gMDSC, mMDSC). Tumor-draining lymph nodes (TDLNs) and spleens were analyzed for DC maturation, T cell subsets, and macrophage polarization. - Orthotopic 4T1-luc breast cancer (BALB/c): 2×10^5 4T1-luc cells inoculated in mammary fat pad. When tumors ~100 mm^3, mice (n=5/group) received saline (IV), free MSA-2 (PO), prodrug 3 (PO), or SAProsome-3 (IV) on days 0 and 4. On day 12–13, primary tumors resected; bioluminescence imaging tracked metastases over time; organ weights (liver, lung, spleen, TDLN), lung macrometastases counted, and histology (H&E) performed. Survival monitored to 120 days post-surgery. - LLC lung carcinoma (C57BL/6): Dose–route comparison of free MSA-2 (PO 80–160 mg/kg; IV 40 mg/kg) vs SAProsome-3 (IV 30–40 mg/kg). Tumor growth, complete response rates, survival, body weight, and tumor histology (H&E, TUNEL) and CD8/granzyme B staining assessed (n=6/group). - B16F10 melanoma (C57BL/6): ~150 mm^3 tumors treated with single dose saline (IV), free MSA-2 (PO, 35 mg/kg), SAProsome-3 (IV, 35 mg/kg), with/without anti-PD-L1 (5 mg/kg IP ×3). Tumor growth, survival (Kaplan–Meier), and body weight monitored (n=6/group). Pharmacokinetics (rats): Sprague-Dawley rats (~200 g; n=3/group) received single doses of free MSA-2 (PO or IV) or SAProsomes (IV) at 17.5 mg/kg MSA-2 equivalent. Plasma collected at predefined time points; analytes quantified by reverse-phase HPLC after complete hydrolysis. PK parameters (t1/2, C0, Cmax, AUC0-∞, MRT) calculated. Biodistribution: MC38, 4T1, and B16F10 tumor-bearing mice (n=6/group) received single doses of MSA-2 (PO, IV) or SAProsome-3 (IV). Tumors and lymph nodes harvested at 8 and 24 h; MSA-2 quantified by HPLC to assess tissue accumulation. Additional organ distribution assessed (Supplementary). Safety/toxicology: C57BL/6 mice (n=4/group) were treated every 3 days ×3 with free MSA-2 PO 240 mg/kg or IV 60 mg/kg, or SAProsome-3 IV 45 mg/kg. Serum chemistry (AST, ALT, UA, BUN) measured on day 8; histology (liver, kidney) examined. Cytokines/chemokines including IL-6 were measured by ELISA at 6 h post-dose in serum and tissues (n=5). Statistics: Data presented as mean ± s.d. (unless stated). One-way ANOVA or unpaired two-tailed Student’s t-test for group comparisons; Kaplan–Meier survival with log-rank test. GraphPad Prism 8.0 used.

- Prodrug design and activation: Four MSA-2 ester prodrugs (1–4) exhibited esterase-triggered release of active MSA-2 with tunable kinetics; ~90% conversion within 30 min in presence of PLE. Pseudo-first-order rate constants (×10^-3 min^-1): prodrug 1: 31.9 ± 0.3; 2: 23.9 ± 3.5; 3: 64.1 ± 0.2; 4: 199.1 ± 1.2. No hydrolysis in PBS without esterase. - Liposomal formulation: All prodrugs (not free MSA-2) stably formulated into liposomes (SAProsomes) with >90% loading, ~120 nm diameter, narrow PDI; stable in serum-containing buffer except SAProsome-4. - In vitro delivery and immunostimulation: SAProsomes markedly increased intracellular MSA-2 uptake in BMDCs and THP1 versus free MSA-2. SAProsomes upregulated IFN-β, TNFα, and CXCL10 mRNA; elicited dose-dependent IFN-β secretion with lower EC50s for SAProsome-3 (22.21 µM) and -4 (19.93 µM) versus free MSA-2 (30.11 µM). SAProsome-3 activated STING pathway (p-STING, p-TBK1, p-IRF3) in THP1. DC maturation (MHC II, CD80/CD86) increased across SAProsomes. Antigen cross-presentation (SIINFEKL-H-2Kb) improved 2.9–3.4× vs saline and 1.8–2.1× vs free MSA-2. SAProsomes+OVA primed CTLs with enhanced B16F10-OVA killing. - MC38 efficacy (IV dosing 35 mg/kg MSA-2 eq., days 0,4,8): Free oral MSA-2 (35 mg/kg) failed to impede growth; SAProsomes shrank tumors with good tolerability. SAProsome-3 achieved 100% complete regression (CR 10/10) and extended median survival time (MST) >150 days versus MSA-2 MST 15 days; other SAProsomes MST 93–99 days. Tumors from SAProsome-3 showed increased necrosis (H&E), apoptosis (TUNEL), dense CD8+ infiltration with granzyme B versus sparse CD8+ with free MSA-2. Rechallenge at day 150 yielded no tumor regrowth over 60 days, indicating durable tumor-specific memory. - Immune remodeling (MC38): SAProsome-3 increased tumor CD3+CD8+ T cells and CD8+/CD4+ ratio (2.82 vs 0.26 untreated, P=4×10^-7); elevated IFN-γ+ CD8+ T cells (37.2% vs 15.7% untreated and 18.0% MSA-2; P=2×10^-6 and 9×10^-6). Decreased CD206+ M2 macrophages; increased NK cells and activated DCs (CD80/CD86); reduced intratumoral gMDSC and mMDSC. In TDLNs and spleen, DC maturation increased (CD80/CD86; TDLN 3.0× vs saline, 2.1× vs MSA-2), CD8+ T cells increased, and macrophage polarization shifted toward M1. - Postsurgical metastasis (orthotopic 4T1-luc): SAProsome-3 (IV, days 0 and 4) induced sustained primary tumor regression and reduced bioluminescence; after surgical resection (day ~13), SAProsome-3 minimized metastatic burden (no detectable 4T1-derived signals), decreased liver, lung, spleen, and TDLN weights; reduced lung macrometastases; normal organ histology; 100% survival at 120 days (n=5). Free MSA-2 or oral prodrug 3 were inferior. - LLC model dose–route comparison: SAProsome-3 IV produced dose-dependent tumor control with CR rates of 50% (35 mg/kg) and 66.7% (40 mg/kg), shrinking tumors to ~30 mm^3. Free MSA-2 (PO or IV) showed limited efficacy. Comparable efficacy observed for oral MSA-2 160 mg/kg, IV MSA-2 40 mg/kg, and IV SAProsome-3 30 mg/kg. MSTs: saline 12 d; MSA-2 PO 160 mg/kg 18 d; MSA-2 IV 40 mg/kg 19 d; SAProsome-3 IV 30 mg/kg 23 d; 35 mg/kg 46 d; 40 mg/kg >60 d. - Combination therapy (B16F10): SAProsome-3 (IV 35 mg/kg) suppressed tumors (92.3% inhibition at day 10) vs free MSA-2 (24.4%). Combining SAProsome-3 with anti-PD-L1 eradicated large established tumors by day 10 and improved survival (MST 27 d) versus saline (12 d), free MSA-2 (12 d), anti-PD-L1 (14 d), or SAProsome-3 alone (22 d), with stable body weights. - Pharmacokinetics (rats, 17.5 mg/kg eq.): SAProsomes prolonged t1/2 to 4.61–5.95 h vs IV MSA-2 0.82 h (5.6–7.3× longer). SAProsome-3 had highest AUC0-∞, 10.6× greater than oral MSA-2. - Biodistribution: SAProsomes increased tumor MSA-2 concentrations 19.0–45.1× vs IV MSA-2, and 22.0–449.1× vs oral MSA-2 across MC38, 4T1, and B16F10 models; lymph node accumulation also increased. SAProsome-3 showed lower distribution to normal organs compared to IV MSA-2 (which accumulated in heart and kidneys). - Safety: At 1.5× therapeutic doses, SAProsome-3 maintained normal serum AST/ALT/UA/BUN and normal liver/kidney histology, with negligible systemic cytokine changes. Free MSA-2 (IV) elevated AST/ALT (liver toxicity); oral MSA-2 elevated AST, ALT, UA, BUN; histology showed organ toxicity. IL-6 levels (cytokine storm marker) were 12.5× (PO) and 18.6× (IV) higher with free MSA-2 than with SAProsome-3.

The study addresses the limitation of systemic STING agonist therapy—poor pharmacokinetics, suboptimal cytosolic delivery, and systemic inflammatory toxicities—by chemically re-engineering the MSA-2 scaffold into esterase-activatable prodrugs compatible with liposomal encapsulation. The MP-typed ester linkers reduced polarity and enabled efficient liposomal loading and systemic administration. A clear relationship emerged between prodrug hydrolysis kinetics and biological activity: faster esterase-mediated release correlated with stronger STING activation (type-I IFN and cytokine induction) and superior antitumor efficacy. Among candidates, SAProsome-3 balanced rapid activation, formulation stability, and pharmacokinetics to maximize tumor and lymphoid accumulation while minimizing off-target exposure. Consequently, SAProsome-3 transformed immunologically cold tumors into hot microenvironments, increasing activated CD8+ T cells, NK cells, and mature DCs, while reducing M2 macrophages and MDSCs. This immune reprogramming produced complete regression and durable memory in MC38, robust control in LLC, effective postsurgical metastasis suppression in orthotopic 4T1 breast cancer, and synergy with PD-L1 blockade in B16F10 melanoma. Pharmacokinetic enhancement (5.6–7.3× longer t1/2; highest AUC) and 19–449× higher tumor accumulation versus free MSA-2 explain the efficacy and safety gains. Importantly, SAProsome-3 mitigated systemic cytokine surges (notably IL-6) and hepatonephrotoxicity observed with free MSA-2, widening the therapeutic window for intravenous STING agonist therapy.

This work establishes a proof of principle for chemically programmed, esterase-activatable MSA-2 prodrugs formulated as liposomal vesicles (SAProsomes) to safely and effectively activate STING via systemic administration. SAProsome-3 emerged as a lead candidate, delivering superior pharmacokinetics and tumor/lymph node targeting, potent innate and adaptive immune activation, complete tumor regression and durable immunological memory, reduced metastasis post-surgery, and synergy with PD-L1 blockade, all with minimal systemic toxicity. The straightforward, scalable synthesis of non-nucleotide MSA-2 prodrugs and clinically translatable liposomes support potential clinical translation. Future studies should deepen structure–activity relationship understanding, further optimize linker chemistry, evaluate long-term safety and immunogenicity, explore oral administration of optimized prodrugs (e.g., prodrug 3), and test combinations with additional immunotherapies across diverse tumor settings and patient-relevant models.

- Preclinical scope: Efficacy and safety were evaluated in murine tumor models and rats; human translation requires further validation, including species-specific STING allelic responses and clinical safety studies. - Structure–activity exploration: Although a correlation between hydrolysis rate and activity was observed, a comprehensive SAR remains incomplete; only a limited set of MP-type linkers (1–4) and one tocopherol prodrug were tested. - Administration routes: While IV SAProsome-3 was optimized, oral efficacy of optimized prodrug 3 in vivo was not fully explored here despite indications of improved activity. - Tumor model diversity: Strong effects were shown in MC38, LLC, 4T1, and B16F10; broader assessment across additional tumor types and microenvironments is needed. - Long-term toxicity and immunogenicity: Extended safety beyond the acute and short-term assessments, including repeated dosing and potential anti-PEG or anti-lipid immune responses, were not fully characterized. - Mechanistic granularity: While immune cell reprogramming was profiled, detailed mechanisms of uptake, endosomal escape, and cell-type specific delivery in vivo warrant further study.