3 October 2012. In schizophrenia, the cortex is short of transcripts for a protein critical for interneuron development, according to a postmortem study published online 13 September in American Journal of Psychiatry. Led by David Lewis of University of Pittsburgh, the study found that mRNA levels for Lhx6, a transcriptional factor involved in interneuron migration, differentiation, and maturation, was lower than those in controls. In combination with transcript deficits for other interneuron markers, the Lhx6 decreases distinguished a subset of schizophrenia cases. This suggests that some, but not all, with the disorder may have accentuated deficits in inhibitory signaling.
The study addresses the origin of disturbances in gamma-aminobutyric acid (GABA) signaling found in schizophrenia in postmortem studies
(see SRF related news story). In prefrontal cortex, Lewis and others have consistently found reduced levels of GAD67, an enzyme that makes the inhibitory GABA neurotransmitter. The GAD67 deficits are localized to parvalbumin- and somatostatin-containing interneurons, just two of many different interneuron types in the cortex. Because parvalbumin-containing interneurons have a synchronizing effect on cortical circuits, and are thought to contribute to gamma oscillations and associated working memory, they make an appealing locus of pathology for schizophrenia (Lewis et al., 2005). But why are these, and their somatostatin-containing brethren singled out for a GAD67 deficit in schizophrenia?
To answer this, the researchers looked to development, as reflected in the levels of Lhx6 and Sox6. Both genes are transcription factors involved in directing would-be interneurons to their places in the cortex, inducing them to assume their parvalbumin- or somatostatin-containing identities, and regulating their GABA production. Despite their key roles in development, Lhx6 and Sox6 remain highly expressed in adulthood, and any differences found in adult brain tissue in schizophrenia could reflect a vestige of derailed development.
Look at Lhx6
First authors David Volk and Takurou Matsubara studied postmortem brain tissue from prefrontal cortex obtained from 42 adults with schizophrenia and 42 age- and sex-matched controls. After extracting mRNA from these samples, they measured the amounts of parvalbumin, somatostatin, calretinin, Lhx6, and Sox6 transcripts with quantitative PCR. As previously reported, this revealed lower levels of parvalbumin (the difference amounting to 22 percent of control levels) and somatostatin (-36 percent) in the schizophrenia group as a whole compared to controls. An increase was also found in schizophrenia for calretinin (+9 percent), a marker of a different subtype of interneurons, ones that originate in a different area from the parvalbumin and somatostatin interneurons,.
Looking at the transcription factors, Lhx6, but not Sox6, was lower in schizophrenia compared to controls (-10 percent for Lhx6). The researchers further examined this finding with in situ hybridization in order to localize Lhx6 in the cortex of 22 subject pairs, for which they had the tissue. Again this turned up a similarly diminished level of Lhx6, particularly in layers 3, 5, and 6, which contain parvalbumin- and somatostatin-containing interneurons. The density of neurons positively labeled with their Lhx6 probe was decreased, as was the amount of labeling per cell, indicating both fewer Lhx6-expressing interneurons, and waning Lhx6 expression in those with detectable expression.
Noting that deficits in parvalbumin and somatostatin were not shared by every schizophrenia sample in their group, the researchers explored whether a certain pattern of transcript deficits distinguished a subset of their samples. Using cluster analysis to group all brains—schizophrenia and controls alike—according to the similarity in their transcript levels of GAD67, parvalbumin, somatostatin, and Lhx6, the researchers found two clusters: one containing a mixture of schizophrenia and control samples (n = 61), and one dominated by schizophrenia samples (n = 23), with only 3 controls. The 20 schizophrenia samples in the latter cluster had significantly lower mRNA levels of Lhx6, parvalbumin, somatostatin, and GAD67 compared to the schizophrenia samples not in this cluster, and to all controls; no differences emerged for calretinin and Sox6. Together the results suggest that a subset of schizophrenia cases are marked by a dearth of GABA-related transcripts, particularly those associated with Lhx6.
These transcript deficits were not explained by medication history or illness severity. Consistent with this, the researchers found that treating monkeys with the antipsychotics haloperidol or olanzapine did not curtail Lhx6 expression either, similar to their previous results for parvalbumin, somatostatin and GAD67. This, combined with the fact that Sox6 transcripts were unchanged, argues against an effect of the illness on transcriptional regulators in adulthood. This then suggests that the Lhx6 deficit observed in these adult brains may be a long-lasting mark of faulty development. Future studies will have to tease out when and how it arises, and whether it is related to the other expression deficits. If further studies uphold the idea of a subset of schizophrenia cases distinguished by GABA-related dysfunction, then this could help clear up discrepancies in the GABA-in-schizophrenia literature (see SRF related news story), and potentially predict better clinical response to drugs that boost GABA signaling.—Michele Solis.
Volk DW, Matsubara T, Li S, Sengupta EJ, Georgiev D, Minabe Y, Sampson A, Hashimoto T, Lewis DA. Deficits in Transcriptional Regulators of Cortical Parvalbumin Neurons in Schizophrenia. Am J Psychiatry. 2012 Sep 13.