Kellendonk et al

In reading Kellendonk et al., I found that I was both very interested in their mouse model of schizophrenia, but also somewhat skeptical. To begin with, knowing how complex and diverse the symptoms of schizophrenia are it is expected that multiple neurological systems are dysfunctional, so I was very curious to see how these papers would approach the problem. I have read that drugs that increase dopamine activity, such as cocaine and amphetamines, can worsen many of the positive symptoms of schizophrenia, and the authors wrote that dopamine antagonists tend to reduce them, so I thought that studying the D2 receptors was a promising place to start. I think that in general the authors were able to convince me that D2R overexpression in the striatum can cause some working memory deficits, and that this system is strongly linked to PFC functioning, but there were also many areas which I felt were lacking the robust data that would have really drawn me in.

For example, the transgenic mice showed 15% increased receptor binding capacity compared to littermates, which the authors note is comparable to the 12% increase measured in schizophrenic patients. While I think this data point sounded pretty nice, I’m not sure if it was completely accidental or not, and I wondered if they had tried to increase the binding capacity even more. I understand that keeping the magnitude of the alterations small and consistent is important, but I thought that so much of their data was really barely significant that it would have been useful to investigate a more dramatic shift, for the sake of comparison. In many of the figures I found myself eyeing the error bars rather suspiciously, and the text even pointed out places where there were “no significant differences” but a “trend”. Frankly, I really wished they could have fallen clearly on one side or the other without implying that a difference may have existed, even though the numbers didn’t show it.

In addition, I appreciated the wide range of behavioral tests they did, but after a while I was beginning to think that repeating the same tests over and over (as boring as it may get) would make comparison much more straightforward, especially since they were getting so much negative data. If you’re comparing apples to apples, perhaps it is easier to spot some of the smaller changes. 

Finally, I thought that one of their more important findings was that reversing the D2R overexpression did not reverse the behavioral deficits measured. Schizophrenia has fairly well characterized roots in development, so I thought this similarity could be an important connecting point to help guide future research.