Ooh Ooh Ooh! They Finally Did It!
I am excited about a new set of methods articles that appeared in Nature Methods this week. My excitement stems from the fact that these papers form a relatively comprehensive collection of protocols for using the Bussey/Saksida touchscreen system. In my opinion, these protocols should raise the quality of behavioral research in rodents using touchscreen technology to a new level.
To put my excitement into context, I have never been a fan of Tim and Lisa’s work using the touchscreen systems. I saw them as an expensive solution to a problem I would solve with exploratory tasks, and said so publicly. However, over time (and to be completely honest having conversations with Tim over twitter and at patternseparation.com has elevated my perspective greatly in regards to their research) I have come to realize that, used correctly, these touchscreen systems provide the type of data that can be replicated across laboratories and used to directly compare mouse disease models. In other words, this system actually provides me the necessary toolkit to perform the types of experiments I want to pursue in the future.
The papers I am going to talk about in this post are three from the Bussey/Saksida laboratory that occupy 60 some odd pages in Nature Methods. Link 1, Link 2, Link 3. As of the time of the writing and publication of this post these protocols were freely available.
An Innovative Approach
As I stated above, I have never been a fan of operant conditioning and Skinner boxes. I had to do some serious soul searching before using fear conditioning to test hippocampus function, but finally relented because I could not think of a better paradigm I could use to test my hypotheses. I only belabor that point to make it clear just how excited I am and how important I find these methods and protocols.
Over the past few years I have been trying to convince a wide number of collaborators and the field of behavioral genetics in general to engage in what is called neurocognitive endophenotyping. This process involves identifying and subsequently evaluating subclinical or subtle behavioral deficits in mouse models? The data from these tests can then be related with genetic dosage, age, gene X age interactions, environmental stress, etc to try and provide a sort of mechanistic description of how genetic dosage and cognition are related. The main problem with this approach is that behavioral performance across labs differs, a lot. Usually these differences emerge because of methodological differences or subtle oddities in apparatus that go unexplored. Additionally, not all researchers have sufficient space and resources to perform a wide range of behavioral tasks if numerous apparatus are required. To date, these challenges have not been overcome, despite the best efforts of behavioral neuroscientists.
The fact that this problem may seemingly be solved is what I find remarkable in this work from The Bussey/Saksida lab. They have found a way to fit a very large number of tasks that span a number of cognitive domains into a single apparatus that does not take up a lot of space. Additionally, so far as I can tell from the manuscripts, the nature of these behavioral protocols make it so that a single rat or mouse could be tested for any number of their protocols in the apparatus with only a bit of extinction training and temporal breaks thrown in between tasks. An additional benefit of this touchscreen system is that these tasks give immense amounts of data that can scrutinized if one knows how. This density of data also solved a huge problem in behavioral neuroscience, the need to use lots and lots of animals to acquire sufficient data to perform meaningful statistical analyses.
What makes this work from the Bussey/Saksida lab so remarkable is that they have distilled 20 years of work of experimental design and optimization into three extremely helpful and thorough methods papers. These manuscripts give point by point and step by step how to and troubleshooting guides to the experiments to provide what I call an “idiot-proofing”. They did not have to do this. They could have sat on their in lab protocols now that they have them optimized and leave the rest of us jealous and fighting to play catch up. They didn’t. They chose the high road. They gave us all a huge boost to get to their level. And I respect them deeply for that.
So in my own future research program, what I see for my implementation of these protocols is as follows. In all likelihood I would move toward a more home cage based ArduiPod Box like that I wrote about in an earlier post rather than a Layfayette or Campden testing system (the US And UK venders of the Bussey/Saksida system they describe here). That deviation from these methods aside, I would certainly take these manuscripts as the blueprint to write the necessary experimental code. They have provided me all the necessary information I need to test executive function, general learning and memory, pattern separation (and by extension I have ideas of how to attack pattern completion using this system), paired associate learning, response learning, reaction time, and attention.
I am obviously bursting with excitement about these papers and because I see in them the future of behavioral neuroscience. Whereas in the last few decades the Morris water maze may have been the go to paradigm, I now see a touchscreen-based assay providing a grater density of data per unit experimental time and per unit research space. These paradigms may signal a quantum leap in the adoption of behavioral apparatus. I hope they do.
p>Now to my TL;DR point, go read these articles if you use behavior as a measure in you lab. Especially if you use behavior as an assay during testing. These protocols will help you and may well revolutionize your lab.