Neurocognitive Endophenotypes in CGG KI and Fmr1 KO Mouse Models

What I Have Been Up To!

So in this post I want to talk about a theoretical review paper I published this last month. This paper is important to me for a couple of reasons: 1) It serves as a logical extension to my theoretical work wherein I first claimed to demonstrate a behavioral endophenotype in CGG KI mice Link, Link. I suggested in this paper that I would be able to predict where the Fmr1 KO would compare based on the molecular genetics. And 2) The data I used in this review came from a series of studies from another laboratory using behavioral tasks I developed See Earlier Post.

The new paper of mine that I am writing about in this post is “Neurocognitive endophenotypes in CGG KI and Fmr1 KO mouse models of Fragile X-Associated Disorders: an analysis of the state of the field”. It is in F1000Research and is open access-thus free to download.

What Did I Study and Why?

My primary focus in this review was to bolster a theory that I developed with a colleague at UC Davis, that of behavioral endophenotyping of mouse disease models. I extended my findings in the CGG KI mouse model of the Fragile X Premutation to include the Fmr1 mouse model of Fragile X Syndrome (Note, the paper I refer to defines these terms, so I will defer you there rather than explain the genetics here).

What I mean by behavioral endophenotyping is the process by which a researcher uses a set of tasks to test specific disease-related hypotheses in a mouse model. These hypotheses include neuropathological data, cognitive/psychiatric/behavioral data, as well as any medical co-morbidities present in the population. Often times, we see these features in a human clinical population, but we are not able to specifically design experiments to test what underlie these pathology and what cellular mechanisms are involved.

For Fragile X-Associated Disorders, a colleague at UC Davis, Tony Simon, proposed that the fragile X population had fundamental impairments in spatial and temporal attention, such that they needed larger spatial or temporal differences between items in memory to distinguish them. As his lab pursued this question in human Fragile X Premutation carriers, I used a mouse model of the Premutation, the CGG KI mouse.

My Innovative Approach

While at UC Davis, I developed a battery of tests for the CGG KI mouse that evaluated the resolution of spatial and temporal processing/memory/attention in mice. Using these tasks, I demonstrated that there were clear spatiotemporal processing deficits in the CGG KI mice, as well as some low level motor deficits. I also demonstrated that the brain and organ pathology associated with the Fragile X Premutation was also present in the CGG KI mouse model. Unfortunately, I did not have access to the Fmr1 KO mouse model of fragile X syndrome, so I was unable to fully evaluate the neurocognitive endophenotyping theory across the range of Fragile X-Associated disorders. So instead, I proposed what I thought the results would be and left it to someone that had access to this mouse to go ahead and do the experiments.

Well, they did! I was so giddy I blogged about it!

Richard Jope’s lab at the University of Miami did the research I was hoping someone would do. They specifically used the experiments I developed for the CGG KI mice without modifying the protocols, and tested the Fmr1 KO mouse. Based on the graphs in these papers, it appeared that the Fmr1 mouse’s behavioral performance fell precisely where I had predicted it would Link, Link.

I emailed Dr. Jope to ask for the data so I could verify I was correct in this assumption. They sent the data to me without delay. What astounded me was that these data fell where one would predict in the human population, along a roughly linear trend toward increasingly poor performance with wildtype mice performing better than CGG KI mice with small mutations better than those with increasingly large mutations, and Fmr1 KO mice performing worse still (see figures below).


Endophenotype of Mouse Models of Fragile X-Associated Disorders. A. Spatial Processing Task 1. B. Spatial Processing Task 2. C. Temporal Processing Task. D. Object recognition task. Note that the wildtype (normal) mice are red boxplots, CGG KI mice with small mutations are green, CGG KI mice with large mutations are orange, and Fmr1 KO mice are in blue. The trend line is only for the CGG KI mice (wildtype and Frm1 mice not included in making the rend line). The Fmr1 mice fit on or just below the line, which is intriguing.


Heatmap comparing CGG KI and Fmr1 KO mice Notice that almost all of the mice are placed into the correct groups based solely on behavioral performance. In fact, only 1 Fmr1 KO mouse and 1 CGG KI mouse with a large mutation were misclassified. This heatmap was made by feeding all the behavioral data into a classification algorithm that did not have access to what animal was in which group.

What further made all this cool was the fact that in the papers from Dr. Jope’s lab, when they treated the Fmr1 KO mice, they performed almost as well as the CGG KI mice with small mutations, never as well as wildtype mice. And when treatment was discontinued, the performance returned to the more profound impairments as soon as serum levels fell to baseline. These data suggest that the treatments being used DO actually target cellular mechanisms specifically disrupted by the mutations, and thus these tasks are valid as outcome measures for drug studies in these models!

All the manuscripts of mine that I mention above but do not link to are available in the “Behavioral Endophenotyping” and “Histological Analysis of Neuropathology” sections of my Publications List.

Take Home Message

TL;DR, I think my new review paper may be useful for researchers looking to develop novel behavioral assays to test mouse models of human disease. It provides an example that can be followed by researchers modeling any genetic disorder in mice (or rats). Researchers simply need to focus on what gaps there are in knowledge of the human clinical population and move from there.

To date there has been a tendency for clinical researchers to diminish the utility of mouse research because mouse results often fail to translate back to the population of interest. I tend to think this is not the mouse’s fault, but rather ours for not being creative enough to make sure the mouse is being given a fair test compared to the human patient population being modeled.

As a teaser to the utility of the behavioral endophenotyping approaach, I have just finished collecting data to make a mouse variant of the Arizona Cognitive Task Battery for Down Syndrome. Similar to the case of fragile X described above, the mouse seems to be an appropriate analogue for the human Down Syndrome related cognitive deficits. In my opinion, it is possible for any clinical population to be modeled in rodents if we take the time to do so conscientiously! …so long as the primary deficits are not verbal, I still haven’t learned to speak mouse.


I would love to hear your thoughts on this!

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s