The scale of the problem is enormous. Roughly 50 million people sustain traumatic brain injuries annually, and the global economic burden exceeds 400 billion US dollars. Surgeons can relieve pressure and clinicians can manage symptoms with osmotic agents, diuretics, and glucocorticoids, but none of these supports repair of the injured tissue. The real killers are the secondary cascades that unfold over hours to days: rising intracranial pressure, cerebral edema, breakdown of the blood-brain barrier, ischemia, redox imbalance, and a self-reinforcing wave of neuroinflammation. Edema and inflammation feed each other. A leaky barrier lets immune cells and plasma flood into brain tissue, swelling activates resident microglia, and the inflammatory signals they release further destabilize the endothelial tight junctions that hold the barrier together. Breaking that loop is a central therapeutic goal.

Cannabidiol, or CBD, has long been an attractive candidate for exactly this job. It is anti-inflammatory, antioxidant, and non-intoxicating, and the team’s earlier work showed that CBD restores blood-brain barrier integrity after brain injury through the PGE2-EP2-cAMP pathway while also damping microglial activation. The obstacle has always been delivery. Because CBD is highly lipophilic, intraperitoneal injection sends much of the drug into adipose tissue rather than the brain. Oral dosing loses more than 70 percent of the compound to first-pass metabolism in the liver. Intravenous injection achieves full bioavailability but suffers from a short half-life and demands professional administration, an awkward constraint in acute care. None of the conventional routes keeps therapeutic levels of CBD parked at a contused cortex for the critical first days.

The researchers’ solution was a sponge made of gelatin methacryloyl, or GelMA, a photocrosslinkable hydrogel already used in drug delivery, ophthalmic therapy, and cardiac repair. The team synthesized GelMA by reacting gelatin with methacrylic anhydride, confirmed the modification with proton nuclear magnetic resonance spectroscopy and infrared spectroscopy, and screened twelve candidate formulations by varying the gelatin-to-anhydride ratio and ultraviolet crosslinking time. They judged each variant on swelling behavior, mass loss in simulated body fluid, and microstructure imaged by scanning electron microscopy. The winning formulation swelled only modestly, between 13 and 25 percent, retained a porous interconnected architecture, and fully degraded within five days under the tested conditions, a profile suited to a temporary implant that releases drug and then disappears.

Loading the sponge with CBD changed none of those material properties. Infrared spectra of the finished product showed the hydroxyl and aliphatic carbon-hydrogen stretching peaks characteristic of cannabidiol, while swelling and degradation curves remained essentially identical to the blank scaffold. Ultraperformance liquid chromatography measured encapsulation efficiencies between roughly 50 and 63 percent, with batch-to-batch variation below 3 percent and, remarkably, center-to-periphery drug ratios of almost exactly one, meaning the CBD was distributed evenly through each sponge. In a release assay, about 22 percent of the payload emerged in the first two hours, followed by a steady trickle that reached roughly 58 percent at 24 hours and 80 percent by 72 hours, precisely the sustained profile the early post-injury phase demands.

Safety testing came next, and it was thorough. Extracts of the sponge were applied to hippocampal neurons, astrocytes, and brain endothelial cells; viability remained high up to 5 micrograms per milliliter, defining a safe exposure window, while hemolysis assays with rat red blood cells showed rates below 5 percent. The team then implanted the sponges into rats subjected to a standardized weight-drop cortical contusion and tracked CBD with high-resolution liquid chromatography mass spectrometry. The pharmacokinetic contrast with conventional dosing was striking. After intraperitoneal injection, CBD in the injured cortex peaked at about 91 nanograms per gram and faded quickly. With the sponge, cortical concentrations reached about 251 nanograms per gram and remained detectable for the entire 48-hour observation window, while blood levels stayed low, an inversion of exposure that concentrates the drug at the lesion and spares the rest of the body. Serum chemistry and organ histology showed no treatment-related abnormalities.

The therapeutic results followed. Forty-eight hours after injury, rats receiving the medium and high doses of the CBD sponge scored significantly better on a standardized neurological deficit scale, traveled farther in open-field tests, and explored the open arms of an elevated plus maze more readily than injured controls, with mannitol and injected CBD providing comparable but generally weaker benefits. Under the microscope, the treated cortex retained orderly neuronal architecture and intact Nissl bodies, showed fewer TUNEL-positive apoptotic cells, and released lower levels of neuron-specific enolase and S-100 calcium-binding protein beta, two canonical markers of neural damage. The blank sponge alone conferred little benefit, indicating that sustained CBD release, not the scaffold itself, drove the protection.

Mechanistically, the study points to the blood-brain barrier and the water channel aquaporin-4. Treated rats had less brain water content and far less Evans blue dye leaking across the barrier, alongside restored expression of the tight junction proteins ZO-1, occludin, and claudin-5. Molecular docking predicted that CBD binds aquaporin-4 with a binding energy of minus 6.7 kilocalories per mole, and surface plasmon resonance confirmed a measurable micromolar interaction with a dissociation constant of 56.6 micromolar. After injury, aquaporin-4 shifted away from vascular endothelial structures and toward reactive, GFAP-positive astrocytes; the sponge reversed that redistribution, reduced astrocytic overactivation, and preserved endothelial markers. On the inflammatory front, the treatment lowered interleukin-6, interleukin-1 beta, and tumor necrosis factor-alpha, raised the anti-inflammatory interleukin-10, and calmed Iba-1-stained microglia. RNA sequencing of the injured cortex identified nearly 800 differentially expressed genes and pointed to dampened NF-kappaB and TNF signaling alongside enhanced PI3K-Akt survival signaling, changes the authors confirmed by Western blot and quantitative PCR.

The authors are careful about the limits of their work. The study captured only the acute, two-day phase, leaving long-term cognitive and behavioral recovery untested, and the local and injected CBD regimens were not matched for dose or exposure, so some benefits may reflect the delivery platform rather than the drug alone. No formal power calculation was performed, and the transcriptomic analysis omitted the blank-sponge and injected-CBD groups. Even so, the concept is compelling: a cheap, photocrosslinkable gelatin sponge that turns a one-shot drug depot into a 48-hour local infusion, right where edema and inflammation are born. If future studies with exposure-matched controls and longer follow-up bear out these results, the humble gelatin sponge could become a serious contender in the long-delayed search for the first true neuroprotective therapy for traumatic brain injury.

Subject of Research: Local delivery of cannabidiol using a GelMA hydrogel sponge to treat cerebral edema and neuroinflammation after traumatic brain injury

Article Title: In situ delivery of cannabidiol-loaded GelMA sponge attenuates cerebral edema and neuroinflammation in traumatic brain injury

Article References: Li, H., Luo, X., Cao, Y., Jiang, H., Guo, Z., Zhu, Y., Zhang, L., Li, Z., Li, J., Wu, H., & Li, P. (2026). In situ delivery of cannabidiol-loaded GelMA sponge attenuates cerebral edema and neuroinflammation in traumatic brain injury. Materials Today Bio, 41, Article 103721. https://doi.org/10.1016/j.mtbio.2026.103721

Image Credits: AI Generated

DOI: 10.1016/j.mtbio.2026.103721

Keywords: traumatic brain injury, cannabidiol, GelMA hydrogel, drug delivery, cerebral edema, blood-brain barrier, neuroinflammation, aquaporin-4, microglia, NF-kappaB signaling, biomaterials, rat model

Writing in the journal Materials Today Bio, researchers led by Chen Zhang, Shanying Han, and Ruiyao Zhou describe a reactive oxygen species-responsive, small-molecule self-assembled nanocarrier integrated into a hydrogel, which they call HA/MPC@B. The central insight is that inflamed colonic tissue is chemically distinctive: it is bathed in abnormally high concentrations of reactive oxygen species, the volatile molecules that drive oxidative stress and destroy the protective gastrointestinal mucosal layer. Rather than fighting this harsh chemistry, the team exploits it as a trigger. The system is built around a borate ester group derived from pinacol 4-hydroxyphenylborate, a chemical bond that remains stable in ordinary physiological conditions but cleaves selectively when it encounters the elevated reactive oxygen microenvironment of a colitis lesion. This means the drug stays locked away during its journey through the stomach and healthy intestine and is released preferentially at the site of disease.

The construction of the nanocarrier itself is a feat of molecular design accomplished in three steps. First, the researchers coupled pinacol 4-hydroxyphenylborate with 1,1′-carbonyl diimidazole, producing an intermediate with non-toxic, removable imidazole as the only by-product. Second, under catalysis by 4-dimethylaminopyridine, this intermediate was covalently crosslinked with D-mannose, a sugar that serves double duty as a hydrophilic anchor and, crucially, as a targeting ligand. The resulting amphiphilic molecules self-assemble spontaneously in water into nanoparticles roughly 160 to 170 nanometers in diameter, with a hydrophilic outer shell of mannose polyhydroxy groups and amide bonds surrounding a hydrophobic core of phenyl rings and ether linkages. Fourier-transform infrared spectroscopy, ultraviolet-visible spectroscopy, and proton nuclear magnetic resonance collectively confirmed each stage of synthesis, tracking the disappearance and emergence of characteristic peaks for hydroxyl, carbonyl, boronate ester, imidazole, and sugar-ring groups.

Into these self-assembled spheres, the team loaded berberine, a natural plant-derived compound with well-documented anti-inflammatory, antibacterial, barrier-repairing, and microbiota-regulating effects, but notoriously poor solubility and bioavailability. Dynamic light scattering showed the loaded nanoparticles had a mean diameter of about 160 nanometers and a surface zeta potential of minus 33.13 millivolts. Encapsulation efficiency reached 77.93 plus or minus 2.87 percent, with a drug loading capacity of 23.76 plus or minus 0.67 percent, figures that underscore the exceptional loading power of small-molecule self-assembly. Transmission electron microscopy revealed a higher electron density in the interior of loaded particles compared with empty ones, a direct visual signature that berberine sat packed inside. Stability testing in simulated gastric, intestinal, and colonic fluids showed the particles held together through intestinal and colonic transit, yet visibly disassembled in simulated gastric fluid and, tellingly, in colonic fluid spiked with hydrogen peroxide to mimic the inflamed colon.

The final layer of the system is a hybrid hydrogel of hyaluronic acid and hyaluronate methacrylate, a biodegradable and biocompatible matrix that is injectable and shear-thinning, meaning its viscosity drops under force so it flows easily yet reforms a stable gel afterward. Scanning electron microscopy showed the composite formed an interconnected three-dimensional porous skeleton whose irregular pores facilitate diffusion of the nanoparticles and strengthen their adhesion to the intestinal wall. In drug release assays across 48 hours, the hydrogel held cumulative berberine release below 20 percent in neutral buffer, simulated gastric fluid, simulated intestinal fluid, and simulated colonic fluid. But in colonic fluid supplemented with 0.1 millimolar hydrogen peroxide, mimicking the high-reactive oxygen environment of inflamed tissue, cumulative release jumped to 52.02 plus or minus 4.11 percent. The hydrogel also delayed release compared with bare nanoparticles, acting as a depot that guards against premature leakage while the payload travels to the colon.

Safety came first in the evaluation. Cytotoxicity testing on Caco-2 intestinal epithelial cells showed no significant loss of viability at berberine-equivalent concentrations up to 1.0 milligram per milliliter, the concentration chosen for all subsequent experiments. Hemolysis assays demonstrated a hemolysis rate below 5 percent, meeting the accepted criterion for biomedical materials, and live/dead staining over 72 hours showed living cells dominating every treatment group at every time point. Targeting was verified with striking clarity in RAW264.7 macrophages. A fluorescent mannose probe was avidly internalized by the cells, but when the researchers pre-blocked the mannose receptor with free mannose or competed with unlabeled nanoparticles, fluorescence plummeted; flow cytometry quantified mean fluorescence falling from 71.9 in the probe-only group to 31.5 in the competition group. This proved the mannose-decorated nanoparticles specifically recognize macrophage mannose receptors, which are overexpressed on inflammatory immune cells at lesion sites.

The therapeutic action proved multimodal. In macrophages stimulated with lipopolysaccharide to mimic a hyperinflammatory state, the loaded nanoparticles and the full hydrogel suppressed the pro-inflammatory M1 phenotype, marked by CD86, and restored the anti-inflammatory M2 phenotype, marked by CD206. Gene expression analysis showed tumor necrosis factor-alpha, interleukin-1 beta, and interleukin-6 all falling, while the anti-inflammatory interleukin-10 rebounded above control levels. The system also scavenged reactive oxygen species broadly, neutralizing hydroxyl radicals, hydrogen peroxide, and superoxide anions in cell-based assays. In DSS-injured Caco-2 monolayers, treatments accelerated cell migration in scratch-wound assays, reduced FITC-dextran leakage across the epithelium, and restored the tight junction proteins ZO-1 and Claudin-1 along with their messenger RNA, evidence that the damaged intestinal barrier was genuinely being rebuilt, not merely soothed.

In vivo results in a mouse model of acute colitis, induced by dextran sulfate sodium in drinking water, confirmed the design logic. Mice receiving the hydrogel by oral gavage showed the best weight recovery among all treatment groups, higher survival, reduced disease activity index scores, suppressed splenomegaly, and reversal of colon shortening. Fluorescence imaging with Cy5.5-labeled formulations revealed the difference decisively: free dye vanished within hours, bare nanoparticles leaked fluorescence throughout the digestive tract within four hours after gastric acid exposure, but the hydrogel-protected formulation produced intense, sustained fluorescence in the inflamed colon from 4 to 36 hours after administration, because the crosslinked matrix shields the nanoparticles from stomach acid until they reach their target. Histology showed the lowest pathological scores in the hydrogel group, myeloperoxidase staining revealed sharply reduced neutrophil infiltration, and serum FITC-dextran levels confirmed restored barrier integrity. A healthy control group given the hydrogel showed no adverse changes, indicating good biosafety at the therapeutic dose.

Perhaps most remarkably, 16S rRNA sequencing and transcriptomic analysis traced the therapy’s effects deep into biology. Colitis had disrupted the gut microbiome, enriching pro-inflammatory genera such as Helicobacter, Escherichia-Shigella, Rothia, and Moraxella while depleting beneficial taxa like Muribaculaceae and Alloprevotella; the hydrogel treatment reversed this dysbiosis, restoring microbial profiles that clustered with healthy controls. Transcriptome sequencing of colonic tissue found 385 genes upregulated and 196 downregulated after treatment, with cell proliferation and epithelial regeneration genes such as Lgr5, Agr2, and Cldn8 rising while pyroptosis and inflammatory genes including Casp4, Nos2, Il17a, and Gsdmd fell. Gene ontology analysis showed suppression of NF-kappaB and NLRP3 inflammasome pathways, confirmed by quantitative PCR showing reduced Tlr4 and Rela and increased Nfkbia. Together, these layers of evidence, from molecular cleavage chemistry to microbiome ecology, present HA/MPC@B as an intelligent, multifunctional platform that treats ulcerative colitis through targeted delivery, on-demand drug release, antioxidant and immune modulation, barrier repair, and microbiota remodeling, offering a compelling blueprint for clinical translation.

Subject of Research: A ROS-responsive small-molecule self-assembled nanocarrier-integrated hydrogel for targeted treatment of ulcerative colitis

Article Title: ROS-responsive small-molecule self-assembled nanocarrier-integrated hydrogels for ulcerative colitis treatment

Article References: ROS-responsive small-molecule self-assembled nanocarrier-integrated hydrogels for ulcerative colitis treatment. (n.d.). https://doi.org/10.1016/j.mtbio.2026.103684

Image Credits: AI Generated

DOI: 10.1016/j.mtbio.2026.103684

Keywords: ulcerative colitis, ROS-responsive drug delivery, nanocarriers, hydrogel, berberine, hyaluronic acid, macrophage polarization, intestinal barrier, gut microbiota, NF-kappaB signaling, oxidative stress, drug targeting