Intraspinal injections - To Laminectomy or Not?


Intraspinal viral injections are increasingly used to study spinal circuits. Traditionally, the approach would involve the removal of the vertebrae overlying the region of interest (usually L3-L5), providing an open territory for injection.

Here is an exemplar of that approach:

Recently, Kohro et al. reported a method that circumvents laminectomy and accesses the spinal cord through the interspace between the spinous processes of T13 and L1.

Each method has its advantages. The full laminectomy version allows greater territory to be injected, with multiple injections (e.g. Foster et al 2015), but it also is very invasive and I fear that the area of exposed spinal cord is prone to infection and scarring.

The Kohro method avoids the damage of the laminectomy but really only allows a single injection, which could limit the spread of the virus to smaller area (L4 in their paper). Peirs et al (@rpseal lab) used an approach on very P9-P21 animals using the non-laminectomy approach to good effect.

I’ve tried both approaches this summer here at the MBL, and I’ve settled on the non-laminectomy method for now. It’s actually not that hard. I thought it was going to be challenging to find that space, but with a little tweezing and dissection, and the right stereotactic setup (Nashige spinal one here), it’s not hard. More challenging is filling the glass pipette properly with mineral oil and not clogging the tip, and also breaking through the dura. If anyone has tips on that, please share.

Clamping the cord and finding the interspace is not that hard

Dye injection demonstrating proper location
What are your thoughts and experiences? Which do you prefer and why? With the non-laminectomy method, would you do another injection, say at the T12/T13 interspace or would this be too high? It looks like it could be L2-L3, depending on strain.

@zhzhj131421 @zhisrich @liz @fmoehring @tberta @sshiers @SamineniV @runDRG


I don’t have experience doing spinal cord injections, but the possibility of injection without laminectomy is very exciting! I’ve always been concerned with the effect of the laminectomy itself–in terms of causing additional pain and impacting the biomechanics of the spine. It’s not quite the same, but in orthopaedics, we have groups modeling arthritic pain in the spine by damaging the facets of the spinous processes, which both induces pain and significantly changes the way the spine is loaded, so I would not be surprised to see somewhat similar effects (although likely on a smaller scale) with laminectomy. So to have a model without that potential effect is pretty neat. For trying to go more rostral with this method, it may be difficult, as the ribs may impair your ability to manipulate the spine in the way necessary for exposing the cord. Have you seen anyone try this in larger models, such as rats or rabbits?


@runDRG Thank you for your insights. Avoiding bone damage is why I prefer the non-lami method. I can definitely see the interspace between T12 and T13. I think I could do it. I’ll let you know.