After the gene was removed, the animals, which had been trained to use external cues to look for chocolate treats buried in sand, couldn't learn a similar task, the researchers report in a paper appearing in The Journal of Neuroscience.
The researchers deleted the gene, which codes for a part of a protein involved in passing signals between nerve cells needed for learning and memory. When a similar protein is blocked by drugs in humans, it leads to a psychotic state similar to schizophrenia.
"We think that both our genetic rodent model as well as a new learning and memory test we developed may provide valuable tools in the investigation of schizophrenia," said Dr. Robert Greene, professor of psychiatry and senior author of the study. Dr. Greene holds the Sherry Knopf Crasilneck Distinguished Chair in Psychiatry, in Honor of Albert Knopf.
The researchers developed the training method to test the animals' memories. Chocolate was buried in a cup containing scented sand, which hid the treat's odor. A second cup contained sand with a different scent but no treat. The researchers could change the cage's environment by affixing colored cutouts to the transparent cage walls, adding a textured floor and making other modifications.
The normal mice learned that in the first environment, the chocolate was linked to the first scent. When the researchers changed to a second environment, the mice learned to find the chocolate using the second scent.
Once the mice were trained, an area of the brain called the hippocampus was injected with a genetically engineered virus that selectively cut out the NR1 gene. NR1 produces a protein that is critical for molding nerve messages in an area of the hippocampus called the CA3, which is associated with distinguishing complex patterns. It is this molding that underlies the hippocampal-dependent learning and memory that is needed to distinguish the complex patterns.
The researchers then attempted to train the mice in memory tasks with new scents and new environments, but the animals lacking the gene couldn't learn. The control group, which received an injection that doesn't cut out NR1, learned as quickly as before.
This shows that the treated animals couldn't react properly to situational cues, which also happens in people with schizophrenia, Dr. Greene said.
The researchers hope to see in future studies if similar small changes to nearby brain regions involved in learning and memory result in the same kind of problems.
"In addition, we want to use a similar task in humans to that used in this study to see if patients with schizophrenia have similar deficits in cognition as we observed in our experimental mice," Dr. Greene said. "This will help determine whether our genetically altered animals provide a good model of the psychosis associated with schizophrenia."
Former UT Southwestern researchers involved in the study were Dr. Tarek Rajji, a psychiatry resident now at the University of Pittsburgh, and Dr. David Chapman, a postdoctoral fellow now at UCB Pharma. Dr. Howard Eichenbaum of Boston University also participated in the study.
swmed/