Synthetic virology approaches to improve the safety and efficacy of oncolytic virus therapies

Oncolytic viruses (OVs) are promising replicating cancer therapeutics that can lyse tumor cells and deliver immunomodulatory payloads but face safety concerns and efficacy trade-offs between immune activation and intratumoral spread. The research question is whether chemogenetic gene switches can be engineered into vaccinia virus (VV) to precisely control viral replication and the timing/dose of transgene expression, thereby improving safety and therapeutic performance. The study motivates this approach by highlighting tumor heterogeneity, the need for external user-defined control, and the limited availability of robust, low-leakage regulatory systems compatible with VV transcription. The purpose is to develop and validate multiple small-molecule–responsive switches (rapamycin, doxycycline, and cumate) alone and in combination to regulate VV replication and payload delivery in vitro and in vivo, enabling safer and more effective OV therapies.

Prior work established VV as a versatile oncolytic and vaccine vector with large cargo capacity but limited controllable expression tools. Inducible systems such as TetR/TetO are widely used in mammalian cells and have been explored for poxvirus vectors, yet achieving low basal expression and robust on/off control within VV’s transcriptional context remains challenging. Rapamycin and its clinically used analogs (rapalogs) are well-characterized mTOR inhibitors and have been adapted to regulate split protein systems (e.g., split T7 RNA polymerase) in cells. The cumate (CymR/CuO) system offers another orthogonal control axis with reported low background in mammalian contexts but, to the authors’ knowledge, had not been implemented in replicating VV vectors. These backgrounds motivate engineering VV-adapted versions of these switches and testing combinatorial control to overcome leakiness and enhance safety.

- Engineered chemogenetic switches in VV: (1) Rapamycin-inducible split T7 RNA polymerase (FKBP/FRB-ST7) to drive a T7 promoter-controlled GFP–firefly luciferase (GFPLuc) reporter; (2) Doxycycline-inducible TetR/TetO regulation layered onto native and synthetic VV promoters to minimize basal expression; (3) Cumate-inducible CymR/CuO regulation fused with VV promoters, including chimeric promoter designs (e.g., PEL-CuO) to optimize dynamic range. - Construct integration: Expression cassettes inserted into the VV thymidine kinase (TK) locus; constitutive VV promoters drove regulator components (TetR, CymR, split T7 fragments). Fluorescent markers (mCherry or BFP) tracked infection. - In vitro assays: Multiple human cancer cell lines (U2OS/U205, HeLa, HT-29, A549, SKOV3, etc.) infected at defined MOIs. Inducers applied across dose ranges: rapamycin/rapalogs (temsirolimus i.v., everolimus oral), doxycycline (≤100 ng/mL in vitro; 625 mg/kg in diet in vivo), and cumate (≤100 µg/mL in vitro; 1 mg/kg i.p. or 100–6000 mg/kg via diet in vivo). Luciferase quantified as RLU; fluorescence imaging for reporter and infection markers. Multistep growth curves assessed replication kinetics versus controls. - In vivo studies: HT-29 xenograft-bearing mice received intratumoral VV and systemic inducers (single-dose rapalogs; Dox-containing diet initiated and removed at defined times; cumate via i.p. or diet). Bioluminescence imaging (IVIS) monitored inducible expression and replication. - Conditional replication safety switch: Placed essential VV gene D13L under TetO control (VV-TetR-iD13L). Evaluated replication and cytotoxicity ±Dox in cell lines and temporal control of replication in xenograft models by toggling Dox diet. - Safety and biodistribution: Compared VV-TetR-iD13L versus control VV in SCID mice for lesion formation, IVIS signals, body weight, organ titers. Assessed liver enzyme markers (alkaline phosphatase, AST) and pulmonary edema with inducible cytokine payloads. - Payload applications: Inducible expression of a fusogenic glycoprotein (TAMV-GP) to modulate intratumoral spread; tested timing (constitutive vs delayed Dox) and interaction with ST-246 (poxvirus egress inhibitor). Developed cumate-inducible delivery of cytokines (e.g., IL-2, IL-12, IL-18) and measured secreted levels by ELISA and functional splenocyte activation. - Methods details: Construct design (SnapGene; codon-optimized inserts from GenScript), cell culture (ATCC lines, standard media), bioluminescence assays (Promega luciferase reagent), viability (resazurin), virus production/titration (roller bottles, sucrose cushion, plaque assays), ELISA for RBD and cytokines, and statistical analyses (GraphPad Prism; unpaired t-tests).

- Rapamycin-inducible split T7 system in VV achieved strong on-demand expression: >10-fold luciferase induction at 24 h post infection in U205; robust induction across HeLa, HT-29, and A549 cells; rapalogs (temsirolimus, everolimus) yielded dose-dependent induction and enabled in vivo temporal control after single dosing. Viral growth kinetics were not significantly impaired by rapamycin/rapalog treatment. - Doxycycline-inducible TetR/TetO regulation of VV promoters produced low basal and high inducibility; in several cancer cell lines induction exceeded 30-fold with optimized promoter/operator fusions, achieving reporter levels comparable to strong VV promoters upon induction. - Cumate-inducible system (CymR/CuO) adapted to VV showed high dynamic range with minimal leak; a PEL–CuO promoter configuration delivered up to approximately 450-fold induction while maintaining low basal expression. In vivo, cumate rapidly and reversibly induced reporter expression within 24 h; signal dissipated within 96 h after withdrawal, demonstrating tunability. - Cumate safety profile: Dietary cumate administration (100–6000 mg/kg) for up to 25 days did not alter VV replication in vivo, body weight, or serum liver enzymes (alkaline phosphatase, AST, GLDH), supporting tolerability for control of VV payloads. - Dual-switch strategies decreased basal expression and enabled combinatorial control: In VV-ST7-TetR-iGFPLuc cells, Dox alone induced ~7-fold, rapamycin alone ~19-fold, and combined Dox+rapamycin ~28-fold, illustrating additive/synergistic control and lower leak than single-switch systems. - Conditional replication safety switch: VV-TetR-iD13L replicated poorly without Dox but matched wild-type replication with Dox; in vivo, Dox diet turned replication on and off in tumors, enabling temporal control. Compared with control VV, VV-TetR-iD13L caused drastically fewer pox lesions and reduced off-tumor bioluminescent signals in SCID mice, indicating enhanced safety. - Inducible fusogenic payloads: TAMV-GP expression enhanced syncytium formation and intratumoral spread; constitutive expression impaired viral growth, but initiating Dox 48 h post infection restored near-normal growth and improved anti-tumor efficacy in A549 xenografts. ST-246 reduced spread via extracellular virions, while TAMV-GP facilitated cell–cell fusion-based spread. - Inducible cytokine delivery: Cumate-controlled expression enabled dose-titrated secretion of IL-2, IL-12, and IL-18 from infected cells, increased IFN-γ/TNF-α in splenocyte assays, and reduced systemic toxicity markers in mice. In syngeneic models, timing and dosing of inducer improved survival compared with virus alone (e.g., reported ~40% survival benefit with induced replication/payload delivery).

The study demonstrates that chemogenetic switches can endow VV oncolytic platforms with precise, user-defined control over replication and transgene expression, addressing a central challenge of balancing safety with efficacy. By adapting rapamycin (and rapalogs), doxycycline, and cumate systems to VV transcription, the authors achieve low-leakage, high-dynamic-range control across diverse tumor cell lines and in mouse models. Dual-switch designs further suppress basal expression and enable two-input logic for safer deployment of potent payloads. Conditional control of the essential D13L gene functions as a robust safety switch, limiting off-target replication and lesions in immunodeficient mice while allowing on-demand replication in tumors. Temporal regulation of fusogenic proteins and cytokines optimizes therapeutic outcomes by decoupling initial viral amplification and spread from later payload delivery, thereby minimizing toxicity while enhancing anti-tumor activity. Collectively, these data indicate that externally controllable VV vectors can be tuned to disease context, inducer route, and timing to maximize benefit-risk profiles.

This work establishes a modular toolbox of VV-adapted chemogenetic switches—rapamycin/split T7, doxycycline/TetR-TetO, and cumate/CymR-CuO—usable alone or in combination to precisely regulate viral replication and transgene expression. The systems show robust inducibility, minimal basal activity, in vivo reversibility, and compatibility with clinically used small molecules (rapalogs, doxycycline), enabling safer delivery of otherwise toxic payloads and temporal control of essential viral functions (D13L). Dual-switch circuits further reduce leak and enhance control. The platform extends to other viral vectors (e.g., HSV), suggesting broad applicability. Future work should optimize pharmacokinetics and dosing regimens in larger animal models, evaluate immunological impacts of inducers, expand to additional payloads and logic circuits, and advance toward clinical translation for both oncology and vaccine indications.

- Some constructs (e.g., VV-adapted split T7) showed higher basal expression relative to TetO/CuO-regulated VV promoters, necessitating dual-switch designs; further leak minimization may be needed for highly toxic payloads. - Efficacy and safety were primarily evaluated in cell lines and mouse models; human pharmacokinetics, tissue distribution, and immunological effects of inducers (rapalogs, doxycycline, cumate) may differ. - Rapamycin/rapalogs possess known systemic immunomodulatory effects and pharmacologic limitations; optimal dosing to balance vector control and host immunity requires further study. - While cumate appeared safe in mice up to 6000 mg/kg diet, comprehensive toxicology, metabolism, and potential off-target effects in higher species are not yet defined. - Some data are summarized or semi-quantitative due to the focus on platform development; rigorous head-to-head comparisons across all promoter/operator permutations and payloads remain for future work.