Abstract
Open access post-mortem transcriptome atlases such as the Allen Human Brain Atlas (AHBA) can inform us about mRNA expression of numerous proteins of interest across the whole brain, while in vivo protein binding in the human brain can be quantified by means of neuroreceptor positron emission tomography (PET). By combining both modalities, the association between regional gene expression and receptor distribution in the living brain can be approximated. Here, we compare the characteristics of D(2) and D(3) dopamine receptor distribution by applying the dopamine D(2/3) receptor agonist radioligand [(11)C]-(+)-PHNO and human gene expression data. Since [(11)C]-(+)-PHNO has a higher affinity for D(3) compared to D(2) receptors, we hypothesized that there is a stronger relationship between D(2/3) non-displaceable binding potentials (BP(ND)) and D(3) mRNA expression. To investigate the relationship between D(2/3) BP(ND) and mRNA expression of DRD2 and DRD3 we performed [(11)C]-(+)-PHNO PET scans in 27 healthy subjects (12 females) and extracted gene expression data from the AHBA. We also calculated D(2)/D(3) mRNA expression ratios to imitate the mixed D(2/3) signal of [(11)C]-(+)-PHNO. In accordance with our a priori hypothesis, a strong correlation between [(11)C]-(+)-PHNO and DRD3 expression was found. However, there was no significant correlation with DRD2 expression. Calculated D(2)/D(3) mRNA expression ratios also showed a positive correlation with [(11)C]-(+)-PHNO binding, reflecting the mixed D(2/3) signal of the radioligand. Our study supports the usefulness of combining gene expression data from open access brain atlases with in vivo imaging data in order to gain more detailed knowledge on neurotransmitter signaling.