Kellendonk et al make measured statements throughout the paper as to not suggest that they are overinterpreting their data and the implications. Both the title and the introduction make it clear that, though the findings show similarities with schizophrenia and it is possible that D2 receptors are involved in the cognitive impairments characteristic of the disease, the most they can say conclusively is that the overexpression of D2 causes cognitive impairment phenotypes in their transgenic animals. The experiments themselves are simple and straightforward but I found the experiment in Figure 6F difficult to interpret. I am not sure what to make of this experiment, in which adult animals that were fed doxycycline for 14 days showed a significant decrease in the number of c-fos positive neurons. Here’s a little bit of background to explain my confusion:
First, they state that the D2 overexpression decreased DA turnover and therefore increased DA levels in the previous section. Then, they state that depletion of DA increases D1 ligand binding in rat models, which makes me think that D2 overexpression should cause decrease D1 activation/ligand binding. Next, they show that animals with D2 overexpression increases c-fos expression with the addition of D1/5 agonist, suggesting that D2 overexpression is actually causing more D1 activation, which is the opposite of what I expected based on the previous information. Finally, they show the experiment in Figure 6F, which then suggests that maybe something more complicated than D2 dependent alteration in D1 ligand binding is at work. I know they say that the results in Figure 6F suggest, “that [the animals] developed compensatory processes for D1 receptor function that are independent of concurrent D2 receptor function,” but I cannot for the life of me understand what the implications of this are in combination with the previous experiment and how this fits in with the rest of their hypothesized/potential model for cognitive impairment in schizophrenia.
The MAM experiment in the Moore et al paper seems incredibly heavy handed, but as Kellendock et al point out in the first paper, they are limited by the little knowledge they have on the biological basis of mental illness so you have to start somewhere. That said, administering MAM on E17 to prevent inducing microcephaly was well reasoned in the introduction, but the treatment still causes widespread changes, and likely dysfunction, throughout most of the cortex. I would be surprised if they didn’t find evidence to suggest phenotypes characteristic of most mental illnesses are likely to result from this, not just schizophrenia, at which point the conclusions of this experiments lose some impact. I’m less interested in what was wrong with the animals, because it seems like it would be and is a lot, and more interested in what cortical functions managed to remain intact. The decreased number of trials required for MAM-E17 animals to learn the novel discrimination task would be an example.
The part of the paper that struck me was the last line, which states: “Finally, perhaps the most innovative use of such a model may be to identify early behavioral physiological abnormalities that may serve as biomarkers for preventative treatments.” I agree with the preceding statement which states that these model can be used to identify genes that regulate development of cortical circuits, in fact I think this is its best use as a developmental model, but the final line seemed far-fetched. The proposal seemed like a catch-22 because, if this is a good simulation of the start of the disease, then their experiments suggest you are doomed from the get-go and there is no preventative measure you can take. Or, if this is a sufficient way to induce the symptoms of the illness even if it is not the natural progression of the disease, then any kind of early identifiers in this model are not liable to be indicative of the ones we would see in the progression of the actual disease. Not to mention there is a high likelihood that these animals probably don’t go through a “progression” of any sort and instead are phenotypically altered from birth, which is distinct from schizophrenic individuals who don’t show cognitive symptoms until late adolescence.