Predicting psychosis in 22q11.2 by failed mitochondrial compensation - The 22q11.2 deletion syndrome (22qDS), one of the most common copy number variations at 1:4000 births, is
associated with a roughly 25% risk of developing schizophrenia-related symptoms. Since the features of
schizophrenia (SZ) in the context of 22qDS are largely shared with non-syndromic SZ in terms of typical onset
in later adolescence or early adulthood, symptoms, and brain changes, the high rate of SZ in 22qDS provides
an opportunity for longitudinal studies that identify cognitive and physiological changes that predict SZ risk. Such
identification can lead to mechanistic studies behind the variable penetrance for SZ in 22qDS, and lead to
preventative measures that may extrapolate to some instances of non-syndromic SZ.
Multiple lines of evidence, from studies of human blood, human genetics, IPSC-derived neurons, and mouse
models, suggest that aspects of the 22qDS neural phenotype involves mitochondrial dysfunction. Indeed, 6 of
the 46 genes in the deleted region encode for mitochondrial localizing proteins. We find in an ongoing study of
IPSC-derived neurons that while mitochondrial OXPHOS is reduced in the 22+SZ group relative to control (Li et
al., 2019), the 22q without SZ (22q(-)SZ) group has control-levels of OXPHOS. Remarkably, relative to both
controls and to the 22q+SZ groups, the 22q(-)SZ group has upregulated expression of multiple genes
involved in OXPHOS, and upregulation of PGC1a, a "master regulator" of mitochondrial biogenesis.
These results suggest that variable penetrance for SZ in 22qDS may be influenced by an individual's capacity
for mitochondrial compensation, a feature of some mitochondrial genetic diseases.
To test this idea with a far higher throughput approach than possible with IPSC-derived neurons, we examined
20 lymphoblastoid cell lines (LCLs) from 22qDS adults, equally split between those with 22q+SZ and those with
22q(-)SZ. By a targeted analysis of a few measures of OXPHOS activity and related gene expression, we found
that mitochondrial complex I activity is higher in the 22q(-)SZ group, as are the levels of the complex 1 gene
NDUFV1, PGC1a, and its co-factor PPARa. Statistical analyses revealed that a composite "Mito-score" based
on these 4 measures has over 90% predictability for the presence or absence of SZ-related symptoms in 22qDS.
These results raise the compelling question, can failed mitochondrial compensation identified in
lymphoblastic cell lines from teenagers with 22q11.2 deletion syndrome predict their likelihood of
developing schizophrenia?
In this R21 proposal, using existing LCL lines and existing longitudinal follow up data, we seek to determine
whether mitochondrial function/gene expression in LCLs from mid to younger teenagers predicts their risk of
developing SZ-related symptoms as later teenagers or young adults. Positive results would lead directly to a
clinical trial designed to prevent or ameliorate the development of SZ in 22qDS, and would likely have
implications for the treatment or prevention of some instances of non-syndromic schizophrenia.