Laminectomy and Spinal Cord Window Implantation in the Mouse

The method described here allows for stable imaging of the spinal cord in mice through a glass window. This method has been applied to assess BBB remodeling in transgenic eGFP:Claudin5+/- mice that express a fluorescent BBB tight junction protein, but it could be applied equally well for studies of any fluorescent proteins or cells in the spinal cord.

Multiple methods for laminectomy and spinal cord stabilization have been developed. All protocols address stabilizing the spinal cord during imaging and window implementation for visual access to the structure of interest. The number of vertebra removed and the degree of invasiveness of the available protocols vary (e.g., components glued on the surface of superficial bone, as in the present protocol, versus embedded more deeply). Davalos and Akassoglou² developed a laminectomy method using removable clamps on each side of the spinal column and one clamp at the base of the tail of the mouse in order to stabilize the spinal cord. This innovative strategy to suspend the animal relieved some of the thoracic displacement caused by the movement of the lungs expanding against the surgical table. To create a well for containing immersion fluid for a water-immersion microscopic lens, a rim of gelatin seal (e.g. Gelseal) was made around the spinal cord and filled with aCSF. The seal rim could be disrupted during imaging, but could also be easily wiped away at the end of the session to enable wound closure and subsequent reimaging. This method has been widely adopted^12,13. Other groups have developed alternative stabilization methods. Fenrich et al.⁴ hand-prepared modified paper clips as a way to secure the spinal column. These modified paper clips were secured in the lateral vertebral pedicles with cyanoacrylate glue and maintained as permanently implanted handles for a removable external clip and external holding fork for excellent motion stability. Cupido and colleagues have presented variations on the aforementioned methods with incorporation of agarose overlaid onto the cord^2,4,12. Farrar and Schaffer³ developed a quadrilateral metal stabilizer that could allow for the implementation of a glass window onto three vertebrae rather than only one. This method also allowed the spinal cord to be attached with screws to a larger bridge stabilizer during imaging to reduce potential motion.A miniaturized one-photon microscope implanted directly onto the laminectomy imaging chamber has also been developed for in vivo recording in freely moving mice at the one-photon level, but is not yet readily available to most laboratories¹³. In a different approach, Weinger et al.¹ dissected an entire spinal cord and embedded it in agarose for ex vivo imaging, which allows for unsurpassable motion stability and access to the ventral spinal cord, but abrogates blood flow. Some limitations of these developments include lengthy surgical time⁴, possible disruption of the gelatin seal rim², the need to customize the cover-glass dimensions to fit the desired area of spinal cord⁴, manual modification of paper clips^4,12, relatively invasive surgical techniques^12,14, and air bubbles forming when using the silicon elastomer^3,4.

We have developed an alternative method that offers several advantages. This protocol has been optimized to reduce the amount of time spent during surgery. Whereas some surgical protocols require longer procedure times ranging from an hour and a half⁴ to an hour³; once mastered, this laminectomy method can be performed in approximately 30 min. Reducing the time spent in surgery can decrease physiologic stress to the mouse, and facilitate higher-throughput experimentation. This protocol removes a single vertebra, and incorporates superficial adhesion of the stabilization device making it less invasive than some comparable protocols^4,5,12,14. Like the method of Figley et al., by utilizing a plastic implant this protocol offers compatibility with acoustic imaging⁵.

To avoid light scattering (during intravital microscopy) that could be caused by the differences between the refractive indices of air, water, and tissue, most protocols overlay an optically transparent substrate over the exposed spinal cord. Common substrates include high purity, low-melting temperature agarose^10,12 or silicone polymers^3,4,5. Agarose offers the advantage of ease of use, with minimal bubble formation, and is appropriate for acute imaging sessions. To protect the tissue from heat damage, it is convenient to heat the agarose to beyond the melting point and then allow it to cool to ~39 °C in a water bath during the laminectomy, so it can be ready at the appropriate time for application to the exposed spinal cord. For chronic imaging, silicone polymers are more resistant to dehydration. Pilot trials during the development of the current protocol omitted either the agarose layer or the overlying cover-glass, and found that the consequent light scatter reduced available depth of imaging.

A differentiating feature of this protocol is the incorporation of a 3D printed backplate and supportive backplate fork holder. After the laminectomy and window implantation, the preparation is stabilized by the addition of an 3D-printed oval backplate that is fixed in place with dental cement. The backplate serves two functions: first, it provides structural support and stabilization of the spinal cord, and second, it creates a lip to hold fluid for immersion objectives for microscopy. In prototypes of this setup, commercially available pillar posts, adaptors, and holding forks were used; we have recently switched to custom machined parts as depicted here. In either case, the essential feature is to provide structural rigor to stabilize the imaging field against the perturbations in space and time induced by heartbeat and respiration. Although the body of the animal is resting loosely on the heating pad, the spinal cord and its imaging field are slightly suspended from the holding fork, which also decreases respiratory displacement. The plastic substrate offers slight flexibility to accommodate tension from screwing into the plate holder. The black plastic color used for printing reflects minimal light into the fluorescence field. Through these methods, we successfully generate image stacks that can be used without post hoc alignment adjustment. Furthermore, the 3D backplate described herein is inexpensive to produce, costing only pennies in material for each print, once the printer is purchased. Furthermore, the costs of 3D printers have fallen in the recent years. The 3D printed backplate structural files (See Supplementary Files 1 and 2) published with this protocol can be readily modified to accommodate individual laboratory needs. We designed the long dimension of the backplate to accommodate the intervertebral space created by removal of thoracic 12 vertebral spine, which overlies the lumbar 2/3 spinal cord segments¹¹. To apply this technique to a different vertebral section, the accompanying CAD files can be modified.

This protocol utilizes a commercially available low-flow anesthesia system that deploys a digital integrated direct injection vaporizer as an alternative to the traditional passive vaporizer. The major feature of the low-flow unit is the reduced operator exposure to isoflurane, a substantial health benefit. The low-flow anesthesia unit also offers cost savings due to the reduced consumption of isoflurane and utilization of room air instead of compressed gas. In the present study, 2% isoflurane delivered by integrated digital vaporizer at 150 mL/min, together with feedback-controlled thermal support, achieved a stable plane of anesthesia and appropriate maintenance of core body temperature. Consistent with this, published comparisons of digital integrated vaporizers and traditional vaporizers have also concluded that the digital integrated vaporizer yields a stable plane of anesthesia and good preservation of core body temperature, heart rate, respiratory rate, and recovery while utilizing less isoflurane^15,16.

A non-steroidal anti-inflammatory drug (NSAID) such as carprofen can be administered Pre-operatively as a supplementary analgesic. Over the course of few hours, NSAIDs inhibit inflammatory cytokine transcription and interstitial edema; multi-day administration attenuates severity of neuroinflammatory diseases including experimental autoimmune encephalomyelitis, an animal model of multiple sclerosis^17,18. Particularly in the study of neuroinflammatory disease, the beneficial effects of carprofen analgesia must be carefully weighed against disease modifying effects when determining analgesia and anesthesia for an experiment in close coordination with appropriate regulatory boards.

A limitation of this method is that it is not readily amenable to repeated imaging sessions across multiple days. The main reason is that the backplate structure is too large to close the skin over. Therefore, there is a risk that a mouse would dislodge the backplate upon waking from anesthesia. If repeated imaging was essential, there are several strategies that could be deployed, including reducing the size of the backplate, or changing the mount. As with any surgical procedure, there is a learning curve for operators. Close coordination with institutional animal care offices and review boards is required.