Light-inducible Spinal Cord Injury

Reproducible, tract-specific SCI using chimeras between rhodopsin and perforin

To edit collaborative proposals, please register by clicking 'Join this Space' on the left,

then click the 'Edit This Page' button on the right!!!

To make comments without becoming a member, use the 'discussion' tab.

Posted 11 April 2008 by Noam Y. Harel



Current animal models of spinal cord injury (SCI) suffer from a lack of tract-specificity and lesion reproducibility. We need improved SCI models that allow reproducible, tract-specific lesions along with the following additional advantages: 1) Simpler surgical procedure; 2) less inflammatory response and scar formation; 3) the ability to specifically correlate contributions of individual spinal tracts with specific behavioral outcomes such as locomotion; and 4) the ability to measure post-injury plasticity of individual spinal tracts.
These goals may be accomplished using transgenic expression of protein complexes that mediate light-inducible membrane lysis of selected spinal tracts at targeted locations.


Two of the most commonly-used animal models of spinal cord injury (SCI) are the weight-drop contusion and the dorsal hemisection models. Both procedures require placing the rodent under general anesthesia, then exposing the spinal cord (usually at the thoracic level) by removing part of the bony vertebra (laminectomy).

Weight-drop contusion injury

The weight-drop contusion involves dropping a weighted rod (usually 25g) a defined height onto the exposed cord, resulting in a traumatic compression injury that somewhat resembles the mechanism of most human SCIs. The weight is dropped from different standardized heights to result in mild, moderate, or severe contusion injury. Though the contusion model is more pathologically representative of human injury, it is harder to produce a consistent lesion from animal to animal, and results in a variable amount of inflammation and tissue sparing that can make it difficult to interpret histology at later time points.

Dorsal hemisection injury

The dorsal hemisection model involves transecting the dorsal half of the exposed spinal cord with a scalpel or scissors. It ensures complete disruption of the targeted tracts (especially the dorsal corticospinal tract, involved in fine motor control), and results in less inflammation than the contusion model. However, the hemisection model also suffers from lack of perfect reproducibility, and results in somewhat variable lesioning of tracts that lie near the border of the blade edge.

Need for better SCI models

The variability and non-specificity of the contusion and hemisection models of SCI do not just make rodent SCI experiments more difficult to interpret. They make rodent SCI results somewhat less reliable. This partly underlies the failure of many promising therapies to succeed in more rigorous human clinical trials (see also *).
We need specific, reproducible methods for producing SCI that can be targeted to individual spinal tracts. A potential approach to achieve this specificity would employ several steps: 1) Design of chimeric 'Perfodopsin' molecules incorporating a light-sensitive rhodopsin domain into the membrane-lysing perforin protein. 2) Transgenic expression of Perfodopsin in individual spinal tracts using tract-specific promoters. 3) Illumination of the exposed spinal cord with the appropriate Rhodopsin-activating wavelength, triggering perforin domain oligomerization and membrane lysis of Perfodopsin-expressing tracts at the selected spinal level.
While more aspects of the experimental approach are described below, many important details remain to be worked out. Interested readers are urged to contribute their ideas and expertise to this collaborative project!


  1. Engineer chimeras between rhodopsin and perforin ('Perfodopsin') such that light-excitation of the rhodopsin domain induces oligomerization of the perforin domain within cell membranes. Confirm functional osmotic pore formation and membrane lysis in cell culture.
  2. Create transgenic mice expressing Perfodopsin with a frameshifted GFP cassette flanked by loxP recognition sites. Crossing these mouse lines with spinal tract-specific Cre expressors will result in tract-specific Perfodopsin expression. In most cases, a combination of promoters unique to the given tract will be used to express a chimeric transcription factor that would then activate Cre expression (ref).
  3. Confirm light-activated Perfodopsin membrane lysis in vivo. This will be shown first in principle by exposing a ubiquitously-expressing animal to the appropriate wavelength light and measuring epidermal cell membrane lysis. Then, tract-specific Perfodopsin mice will be tested by surgically exposing the cord at desired levels via laminectomy. Subsequent experimental paradigms will be further discussed.


Aim 1: Engineer chimeras between rhodopsin and perforin ('Perfodopsin'). Confirm functional osmotic pore formation and membrane lysis in cell culture.

Aim 2: Create transgenic mice expressing Perfodopsin in selected spinal tracts.

Aim 3: Confirm light-activated Perfodopsin membrane lysis in vivo.


Some potential questions, hurdles and possible solutions are listed below:
How does tract-specific SCI mimic natural human SCI?
The Perfodopsin model will differ from natural human SCI not only in its tract specificity, but also due to the relative lack of inflammation and scarring expected. However, the expected benefit of reproducibility and the ability to systematically study the function and regenerative capacity of individual spinal tracts will provide enormously useful information that can be applied to treatment of human SCI.

Are there other potential uses for Perfodopsin-mediated membrane lysis?
Crossing the floxed Perfodopsin transgenic mouse with any existing tissue-specific Cre mouse lines would allow study of light-inducible lesions in any tissue accessible with the desired wavelength.


T Knopfel, Nat Methods 4/08