Previous investigations showed significant anxiolytic-like activities of Gelsemium sempervirens L. (Gelsemium s.) in mice models. To provide new insights into the neural substrates of anxiety and
to identify drug targets, we decided to investigate the Gelsemium s. mechanism of action in neuronal models by assessing the genome expression changes.
The SH-SY5Y and IMR-32 human neuroblastoma cells were used since are widely employed
in neuropharmacology and well characterized. The drugs were produced by Boiron Laboratoires
(Lyon), starting from a whole-plant-hydroalcoholic extract and the cells were treated
with 6 increasing dilutions 2c, 3c, 4c, 5c, 9c, 30c. We compared the drug effects
with those of control solutions prepared by the same procedure, but with the solvent
vehicle without the plant extract. All dilution steps were followed by strong succussion.
Final ethanol concentration was 0.03% v/v. After having ruled out possible toxic effects
of any test solution on cell viability, we evaluated gene expression firstly by using
a microarray designed for the whole human transcriptome (Nimblegen, Roche). We used
the Limma statistics approach (n=4 biological replicates) to select a set of differentially
expressed genes and Friedman test followed by Wilcoxon signed-rank test to check the
null hypothesis that high dilutions have no effect in this model. The exposure to
2c dilution promoted a small (fold changes between 0.5 and 1.0) but significant down
expression of 49 genes as compared with untreated controls. With higher dilutions,
most of the genes down-regulated in the 2c-treated samples were also under-expressed
in 3c and, to a varying extent, even in higher dilutions. No changes of housekeeping
genes were recorded, confirming the specificity of drug action. The changes in the
49 selected genes of SH-SY5Y cells were in the same direction in the IMR32 cells,
showing that the expression of the same gene set was also modified in a second type
of neurocyte. Afterword we performed the RT-qPCR on a subgroup of relevant genes modulated
in 2c treatment (i.e. transcription factors, G-protein coupled receptors or neuropeptides)
and we confirmed the down-regulation for the genes DDI1, EN2, GALR2, GPR25, OR5C1,
Klkbl4 and TAC4. In the Wilcoxon analysis, applied to the 49 genes, the number of
down-regulated ones was systematically higher than the number of genes with positive
fold change over all dilutions (p<0.0001). No significant differences between treatments
and controls in a randomly chosen gene set of 49 genes were observed, suggesting that
the Gelsemium s. effects are not due to chance. In parallel we adopted, for the dilutions 2c and 9c,
an RT-PCR Array approach (SABioscience, Qiagen) containing 84 genes, including receptors
and regulators of neuronal function. In this further investigation we observed a trend
to down-regulation for DRD2, CHRN4B, CHRNG, PROKR2 and PHOX2A genes in 2c and BRS3,
GRPR genes in 9c dilution. In particular the down-regulation of DRD2 and PROKR2, effective
in the inhibition of Gabaergic neurotransmission, suggests a specific effect of Gelsemium s. in increasing the endogenous GABA activity. Overall, the genes modulated in this
experimental model outline new working hypotheses on the anxiolytic and analgesic
action of this plant. In conclusion this study provides evidence that Gelsemium s. exerts a prevalently inhibitory effect on a series of genes, in particular involved
in G-protein coupled transduction systems, in olfactory transduction, in calcium signaling
and in neurotransmission. Furthermore the whole genome expression analysis (microarray
and real-time PCR), indicates that the “omics” molecular biology is a suitable approach
to study the effects of highly diluted natural compounds.
