Genetic system converts scar-forming cells into neurons to restore movement after spinal cord injury

What if cells that form scars after a spinal cord injury could become part of the repair process? Researchers have developed a genetic system that converts certain scar-forming cells into neurons, improving movement in mice and rats with spinal cord injuries.

A joint research team led by Director C. Justin LEE of the Center for Memory and Glioscience at the Institute for Basic Science and Professor HA Yoon of Yonsei University College of Medicine developed TRANsCre-DIONE, a system designed to selectively convert reactive astrocytes at an injury site into neurons. The findings offer a potential new direction for spinal cord repair, although the approach has not been tested as a treatment in humans.

Using scar-forming cells as a source of new neurons

Spinal cord injuries damage neurons that transmit signals between the brain and the body. Because these neurons have limited capacity to regenerate, an injury can cause lasting loss of movement and sensation.

After injury, support cells called astrocytes become reactive and help form a glial scar around the damaged tissue. This scar can protect nearby tissue and contain the injury, but some reactive astrocytes may also contribute to an environment that makes nerve repair more difficult. The challenge is to target cells that could be repurposed without indiscriminately affecting astrocytes that perform essential or protective functions.

Earlier approaches to converting astrocytes into neurons commonly relied on a genetic signal associated with the protein GFAP. But GFAP is also found in normal astrocytes and some neural progenitor cells. Its activity can also decline as an astrocyte changes into a neuron, potentially interrupting the conversion process.

TRANsCre-DIONE addresses these problems with a two-part molecular switch. In mice, the system uses genetic regulatory elements associated with GFAP and Lcn2, an inflammation-related protein. Both signals must be active in the same cell to switch on Neurog2, a gene that promotes neuronal development. A separate regulatory element then sustains Neurog2 expression as the cell changes identity.

This design aims to improve targeting while allowing the reprogramming process to continue even after the cell begins to lose its original astrocyte characteristics.

Converted neurons help animals regain movement

The researchers delivered TRANsCre-DIONE near spinal cord injury sites in mice and rats two weeks after injury. In the treated mouse spinal cords, 87% of the labeled cells examined expressed a neuronal marker. The converted cells also expressed markers associated with motor neurons, including Isl1 and ChAT.

Electrical recordings showed that the new neurons could generate nerve impulses and receive signals from other neurons—evidence that they had acquired functional neuronal properties.

The treated animals showed significant improvements in movement. Eight weeks after injury, mice receiving the reprogramming treatment reached an average score of 3.88 on a nine-point locomotor scale, compared with average scores below 1 in the control groups. Treated rats also showed significantly better motor performance than controls.

The team then tested whether the new neurons were contributing to that improvement. They used a chemogenetic method to temporarily suppress the activity of converted neurons in treated mice. When the neurons were silenced, the animals lost some of their regained motor function. This provided evidence that the newly generated neurons played an active role in recovery.

Professor HA Yoon said, "TRANsCre-DIONE shows the possibility of rebuilding neural circuits at a fundamental level and offers a way to move beyond the limitations of current spinal cord injury treatments. If successfully translated to the clinic, this approach could help shift spinal cord injury treatment from preserving remaining function toward actively promoting neural regeneration."

The researchers also observed reduced markers of astrocyte reactivity and improvements in the structure of injured spinal cord tissue. Changes in nearby cells suggested that reprogramming might help create conditions more favorable to repair, although the mechanisms behind those effects remain to be established.

Neuronal identity depends on the surrounding tissue

The study also found that the same reprogramming factor, Neurog2, produced different neuronal characteristics depending on where it was delivered.

In the mouse striatum, a brain region involved in movement, 62% of labeled cells expressed a neuronal marker. Most of these newly generated neurons expressed GABA, a signaling molecule used by the region's predominant neuronal type. In the injured spinal cord, by contrast, converted cells expressed markers associated with motor neurons.

The researchers also tested an adapted version of TRANsCre-DIONE in the brains of cynomolgus monkeys. Because the injury-associated signals differed between species, the primate system combined GFAP with iNOS rather than Lcn2. In the control group, 93.13% of labeled cells expressed the astrocyte marker GFAP, indicating preferential targeting of astrocytes under the experimental conditions. Following Neurog2 expression, labeled cells showed increased neuronal-marker expression and reduced astrocyte-marker expression, supporting the feasibility of neuronal reprogramming in a non-human primate.

Note: The primate experiment was conducted in the brain; it did not test spinal cord repair or motor recovery.

TRANsCre-DIONE could potentially be applied not only to spinal cord injury, but also to a wide range of neurological disorders, including amyotrophic lateral sclerosis (ALS), Parkinson's disease, and stroke. Just as astrocytes can seemingly transform into neurons, we hope this technology can move toward practical use and one day help create a moment when people with spinal cord injury can stand again."

C. Justin LEE, Director, Center for Memory and Glioscience, Institute for Basic Science

Because TRANsCre-DIONE works with cells already present in the tissue, it does not require transplantation of externally produced cells. Nevertheless, further research is needed to establish its long-term safety, targeting accuracy, and effects on neural circuits before clinical applications can be considered.

The study, "TRANsCre-DIONE transdifferentiates scar-forming reactive astrocytes into functional motor neurons," was published online in Experimental & Molecular Medicine on September 11, 2026.