University of Toronto Researchers Report Reengineered Long Noncoding RNAs Reduce Inflammation in Macrophages and Mice in Science Signaling
Researchers at the University of Toronto and The Hospital for Sick Children (SickKids) in Toronto, Canada, report that naturally occurring long noncoding RNAs can be reengineered and delivered exogenously to suppress acute inflammation in human macrophages and mouse models, according to a study published March 10, 2026, in Science Signaling. The work positions lncRNA therapeutics inflammation as a previously uncharacterized nucleic acid drug modality distinct from siRNA, antisense oligonucleotides, and mRNA.
The study, led by Pang et al. and funded through multiple Canadian research programs including the Canada First Research Excellence Fund and the Canadian Institutes of Health Research, synthesized three anti-inflammatory lncRNAs — GAPLINC, MIST, and DRAIR — via in vitro transcription, purified them by reverse-phase HPLC, and delivered them in lipid nanoparticles to LPS-challenged cells and animals. Each lncRNA modulated a distinct arm of the NF-κB inflammatory signaling cascade, and the lead candidate, GAPLINC, achieved broad multi-cytokine suppression in vivo at doses as low as 0.10 mg/kg with no observed organ toxicity.
The concept of a long noncoding RNA drug — a full-length, non-protein-coding transcript delivered from outside the cell to recapitulate its native regulatory function — has no direct precedent in clinical or late-stage preclinical development. Existing RNA therapeutic platforms from Alnylam, Ionis, and Moderna focus on short interfering RNAs, antisense oligonucleotides, or protein-encoding mRNAs. The Toronto group's approach instead co-opts the endogenous gene-regulatory machinery that lncRNAs engage, aiming to rebalance inflammatory signaling rather than block a single downstream effector.
The study screened three lncRNAs across four chemical modification states — unmodified, pseudouridine (Ψ), N1-methylpseudouridine (m1Ψ), and 5-methylcytosine (m5C) — in RAW-Dual mouse macrophage reporter cells carrying NF-κB and IRF pathway readouts. GAPLINC emerged as the lead. Its Ψ-modified form reduced LPS-induced NF-κB activation by approximately 40–50% relative to LPS-only controls (p < 0.05), while scrambled and irrelevant lncRNA controls had no effect. IRF pathway activity was largely unchanged, indicating selectivity for NF-κB. Neither a 5′ cap nor a 3′ poly(A) tail — structural features central to mRNA therapeutic design — enhanced GAPLINC's activity, and in some configurations they diminished it, underscoring that anti-inflammatory lncRNA macrophages engage cellular targets through mechanisms fundamentally different from translated mRNAs.
MIST showed modest NF-κB suppression in certain modification states but lacked the consistency of GAPLINC. DRAIR produced limited effects in the reporter system. GAPLINC was therefore advanced to in vivo testing.
In C57BL/6 mice challenged intraperitoneally with 100 μg LPS, intravenous administration of Ψ-modified GAPLINC encapsulated in lipid nanoparticles reduced serum TNF-α by approximately 50–70% and IL-6 by 40–60% compared to LPS-only controls, with statistical significance ranging from p < 0.05 to p < 0.0001 across experiments. IL-1β, MCP-1, IFN-γ, and GM-CSF were also suppressed. The anti-inflammatory cytokine IL-10 showed variable changes, consistent with selective dampening of pro-inflammatory pathways rather than blanket immunosuppression. RT-qPCR from splenic and hepatic tissue confirmed reduced Tnf, Il1b, and Il6 transcript levels.
The lipid nanoparticle RNA delivery system used custom ionizable lipids — C3-K2-E14 for in vitro work and OC2-K3-E10 for in vivo studies — formulated via microfluidics at optimized nanomaterial-to-RNA mass ratios. At a 20:1 ratio, encapsulation efficiency exceeded 85%, particle size fell within 80–120 nm, and polydispersity remained below 0.2. Biodistribution studies using luciferase mRNA as a reporter confirmed accumulation in liver and spleen, organs rich in the macrophage populations relevant to LPS-induced inflammation treatment.
A dose-response study at 0.10, 1.0, 3.0, and 5.0 mg/kg identified 3.0 mg/kg as the threshold for robust cytokine suppression, with 5.0 mg/kg offering marginal additional benefit. A time-course study showed maximal separation between treated and control groups at 6 hours post-LPS, with continued suppression at 10 hours and convergence by 24 hours as inflammation naturally resolved. An 8-day tolerability study involving two LNP administrations produced no body weight loss and no histopathological abnormalities in liver, spleen, or kidneys. Administration of GAPLINC-LNP without LPS challenge did not elevate pro-inflammatory cytokines above baseline, indicating the construct itself was not immunostimulatory.