Engineered stem cells help paralyzed mice walk again

Paralysis from spinal cord injury is a devastating, life-altering condition. For centuries, doctors and scientists have considered it permanent because of one simple fact: unlike neural cells elsewhere in the body, injured neurons in the spinal cord cannot regenerate.

Jessica Aijia Liu smiles and wears a white lab coat over a dark brown turtleneck sweater.

Jessica Aijia Liu studies neuropathology treatments at the University of Hong Kong.

credit: Jessica Aijia Liu

University of Hong Kong neurologists Martin Cheung and Jessica Aija Liu have both dedicated their careers to studying the Sry-related HMG box (SOX) gene family, which plays an important role in regulating neural cell regeneration in the peripheral nervous system (1). Cheung has been studying SOX9’s role in neurodevelopment since he was a PhD student, and Liu was originally working on SOX10 before learning that SOX9 might have a bigger and more unexplored role in neuropathology. Eventually, their mutual interest in SOX9 brought Liu to join Cheung’s lab as a postdoctoral fellow. “SOX9 is one of my favorite genes,” said Cheung.

Now in a new study, Cheung and Liu modulated SOX9 expression to recover movement in mouse models of spinal cord injury (2). Their work brings scientists a few steps closer to healing what is currently a life-altering and irreversible event.

As Cheung explained, researchers had previously shown that SOX9 hampered neuro-regeneration in mice (3). “No one has ever studied whether it has a similar role in humans,” said Cheung. “We got interested in whether we [could] play around with the level of SOX9, especially in neural stem cells in humans and whether [decreasing SOX9 levels] promotes neuronal regeneration.”

The team first used a mix of growth factors to differentiate human pluripotent stem cells (hPSCs) to neural stem cells in vitro. They found that SOX9 levels decreased as hPSCs further differentiated. “We tried the total knockdown of SOX9,” said Liu. “It turns out all the cells undergo apoptosis. They cannot differentiate.” Instead, the researchers injected hPSCs with plasmids in which a drug-inducible promoter inhibited SOX9 protein expression in the presence of Dox. After confirming that SOX9 expression decreased proportionally with increased Dox, they found that a 50 percent reduction caused the hPSCs to form spherical aggregates called neurospheres. “Only neural stem cells have this ability [to create neurospheres],” said Cheung, who further explained that the presence of neurospheres confirmed that hPSCs had successfully differentiated into neural stem cells, which would eventually lead to neuronal cells.

We got interested in whether we [could] play around with the level of SOX9, especially in neural stem cells in humans, and whether it promotes neuronal regeneration.
- Martin Cheung, University of Hong Kong

To test what happened when SOX9 expression was reduced more permanently, they next treated cells with a constitutively expressed short hairpin RNA (shRNA) that reduced SOX9 expression by 50 percent. Quantitative PCR analysis showed that across the board, treated cells expressed higher levels of genes associated with neuronal differentiation than untreated cells. Cheung and Liu concluded that they could potentially use the shRNA-treated hSPCs for longer-term treatment.

They grafted the shRNA-treated hPSCs with 50 percent-reduced SOX9 expression into a mouse model of spinal cord injury, and these cells generated new neuronal cells in the injured area. The resulting neuronal cells grew healthy axons and connected to host neurons. The team confirmed that the new neurons were communicating by injecting an mCherry-labeled virus into the spinal cord below the injury. In untreated mice, the spinal cord stayed severed, and the team saw no mCherry fluorescence in the brain. However, in the grafted mice, the brain lit up. “That means the neural circuit is reconnected,” said Liu.

“Neural stem cells are picky,” said Stephanie Willerth, a biomedical engineer at the University of Victoria who was not involved in the study. “When you normally put [neural stem cells] into the environment of the spinal cord, you get a lot of cell death, and so the fact that they were able to find a way to program the cells to overcome that was really exciting.” Willerth explained that the more common approach is to transplant a biomaterial scaffold with growth factors that cause host stem cells to differentiate, rather than to directly reprogram and transplant new stem cells (4). She anticipates that Liu and Cheung’s next challenge will be safely grafting genetically modified cells into human patients.

Martin Cheung stands by a floor-to-ceiling window wearing an all black suit.

Martin Cheung studies the molecular mechanisms of neuroregeneration at the University of Hong Kong.

credit: Martin Cheung

Liu and Cheung’s team then tested their animal models for motor function. Two months after the transplant, grafted mice had regained muscle mass in and control of their hind legs, leaving prints with their hind paws that showed coordination with their front paws. When placed on top of a cage, they used their hind feet to grip the wires, demonstrating fine motor control. Grafted mice moved around more freely than untreated mice, whose hind legs stayed paralyzed.

“We spent almost seven years to complete this work,” said Liu, citing both technical challenges within the experiments and restrictions caused by the COVID-19 pandemic. Still, when asked how she felt about successfully getting mice to walk again, Liu immediately said that she is looking forward to clinical trials testing whether SOX9 reduction can lead to a similar recovery of motor control in humans with paraplegia. The team is working to acquire funding and the necessary approvals, and they already have a line of patients eager to participate. “They’re asking us whether we're going to do the clinical trials,” said Liu. Cheung added, “It gave them hope, but we also told them it will take some time.”