Latest Developments in the Lab


An interactive atlas of supraspinal input! In a collaborative project with Dr. Pantelis Tsoulfas at the University of Miami, we are constructing a community resource to answer some hard questions in the SCI field. We know in general terms that many different types of neurons project axons to the spinal cord, but exactly how many exist and where exactly are they located? More importantly, to what extent does each population of axons survive any given injury? Which types of axons are responding to our pro-regenerative treatments - and which ones aren't? The starting point for all these questions is a means to identify neurons that project axons to different spinal levels, to register them to 3D atlases, and to count them in the presence or absence of graded injury. You can explore a beta version of this resource at Feedback is appreciated!

Retrograde gene delivery. A major limit to gene therapy for spinal injury is the current practice of injecting virus directly to the brain. It is fine for proof-of-principle work, but puts a low ceiling on the functional gains we can hope to see even in a rodent, let alone a patient. The problem is that there are too many supraspinal neurons in too many locations - it's impossible to hit them all. As many researchers are aware, the solution is a retrograde approach, in which vectors are delivered to spinal tracts (a small target) and then spread back to cell bodies of origin throughout the brain. Now, with a new vector developed at Janelia, it looks like this idea can become a practical reality. We have systematically tested injecting Retro-AAV to the spinal cord, both uninjured and after cervical transection, and find spectacular efficiency in supraspinal populations including the major players (corticospinal, rubrospinal, vestibulospinal, and reticulospinal). Other tracts that we don't tend to think about are also transduced, for instance hypothalamospinal tracts and cerebellar spinal tracts. To illustrate the point we collaborated with Dr. Pantelis at the Miami Project, who performed tissue clearing and 3D imaging of mouse brains after cervical injection of Retro-AAV-tdTomato. In the videos below you can appreciate how the virus reaches a broad range of supraspinal neurons. Check out our recent pre-print for more details. 

Retrograde DREADDs as rehabilitative tools? The efficiency of the new Retro-AAV opens up some interesting DREADD-based approaches. We first tested cervical injection of Retro-AAV-Gi-DREADDs from Addgene (inhibitory DREADDs). When activated with either CNO or clozapine, the result was complete paralysis of the forelimbs (!), which lasted a several hours depending on the exact ligand and dose. This quick readout made it simple to establish dose response curves (see figure to the right). It also sets up future experiments to reversibly silence regenerated axons as a means to parse out their functions. We have also tested retrograde delivery of activating DREADDs (Gq, also from Addgene). Something really interesting happened - animals started making "fictive retrieval" motions. It looked as though they were grabbing food and bringing it to their mouths to eat, over and over, except there was nothing there. To quantify this we put non-food objects on the floor their cage, which mice normally ignore after a couple exploratory nibbles. But upon Gq-DREADD activation, mice continually grabbed, nibbled and discarded the pellets over and over. It's tempting to speculate that there is some kind of CPG for this stereotyped motion, and DREADD activation of supraspinal input repeatedly triggers it. Again, that's pure speculation. But we are very interested in finding out whether this phenomenon can be used for rehabilitation after an injury, especially when combined with pro-regeneration treatments. We hypothesize that if this circuit and its supraspinal inputs are stimulated over and over - and we think that's exactly what the Retro-DREADDs are doing - then this synchronized activity may strengthen appropriate connections by regenerated supraspinal axons and ultimately improve forelimb control. We're trying to secure funding to move in this direction now.

Gi-DREADD mediated silencing of supraspinal input to cervical spinal cord. (A) Mixed AAV-Retro-Flex-Gi-DREADD-mCherry and AAV-Retro-Cre was injected to the cervical spinal cord of wildtype mice, and Retro-Flex-Gi-DREADD-mCherry alone was injected to the cervical spinal cord of CAMKII-Cre animals. Four weeks later, mCherry signal was readily detectable in the brainstem, red nucleus, and cortex of wildtype animals, but was cortically enriched in CamkII-Cre animals. (B) Clozapine produce reproducible forelimb deficits in DREADD-injected wildtype animals. (C) Dose response and timing curves for clozapine- and CNO-triggered motor deficits. Both ligands triggered reversible forelimb paralysis in Gi-DREADD injected wildtype animals, but not CamkII-Cre animals or non-injected controls. (D) Mice were tested on a horizontal ladder task atop a wheel with irregularly spaced rungs and errors in forelimb placement were scored. (E) In CAMKII-Cre animals, clozapine produced significant reduction in correctly targeted steps, consistent with selective silencing of corticospinal tract input to the spinal cord. N=5 wildtype, 4 CamkII-Cre, and 3 non-injected controls. **p<.01, t-test.