Effects of lithium chloride on the gene expression profiles in Drosophila heads

Abstract

To gain insight into the basic neurobiological processes regulated by lithium--an effective drug for bipolar disorder--we used Affymetrix Genome Arrays to examine lithium-induced changes in genome-wide gene expression profiles of head mRNA from the genetic model organism Drosophila melanogaster. First, to identify the individual genes whose transcript levels are most significantly altered by lithium, we analyzed the microarray data with stringent criteria (fold change>2; p<0.001) and evaluated the results by RT-PCR. This analysis identified 12 genes that encode proteins with various biological functions, including an enzyme responsible for amino acid metabolism and a putative amino acid transporter. Second, to uncover the biological pathways involved in lithium's action in the nervous system, we used less stringent criteria (fold change>1.2; FDR<0.05) and assigned the identified 66 lithium-responsive genes to biological pathways using DAVID (Database for Annotation, Visualization and Integrated Discovery). The gene ontology categories most significantly affected by lithium were amino acid metabolic processes. Taken together, these data suggest that amino acid metabolism is important for lithium's actions in the nervous system, and lay a foundation for future functional studies of lithium-responsive neurobiological processes using the versatile molecular and genetic tools that are available in Drosophila.

Figure 1. Correlation of microarray data obtained from lithium-treated and untreated samples

Euclidean distance clustering analysis was performed using Ward’s method. Three biological replicates (+Lithium or −Lithium) were well clustered. Consistency of the data is visualized by the heat-map representation of expression data for the 66 genes (represented by 71 probe sets) that are differentially regulated in response to lithium treatment (fold change >1.2; FDR <0.05).

Figure 2. RT-PCR analysis of the genes whose expression is significantly affected by lithium treatment

(A) A representative result of the RT-PCR analysis (carried out in quadruplicate). The PCR products were quantitated based on the pixel intensity of the corresponding bands using Image J software. (B) Ratios of signal intensity of lithium-treated and untreated samples. Average ± SEM, N=4.

Figure 3. Biological pathways significantly affected by lithium treatment

Biological pathways identified by DAVID clustering analysis, based on the genes whose expression is affected by lithium treatment. Genes that were assigned to these pathways are also shown with values for fold change in expression and false discovery rate (FDR) in response to lithium treatment. Genes that are involved in “valine, leucine, isoleucine degradation pathway (KEGG PATHWAY)” are indicated at right.