An error occurred while loading the PDF.
Can Mouse Models be Trusted? Uncovering Differences in Human and Mouse Serotonin Receptor Expression
Voiceover
Presenter(s)(s)
Ali Cramer
Presentation Number
1907
Abstract or Description
Rodent models are used to test the efficacies of most candidate drugs intended for human use. In particular, the widely prescribed SSRI family of drugs were first validated on rodent models of depression. However, there has always been controversy about whether it is appropriate to model human psychiatric illness in rodents.
The SSRI drugs act on the serotonin reuptake transporter, directly influencing the effects of serotonin on many neurons. So far there has been no comprehensive attempt to catalog the varied effects of serotonin on different neuron types or brain regions or to compare these effects between human and rodent brain.
In this study we used single cell RNA sequence data and regional expression from the Allen Brain Institute to compare the expression levels of serotonin receptor genes in the brains of mice and humans across cell types and brain regions. We find much more contrast in 5-HT receptor (HTR) expression between humans and mice than previously reported in literature. Perhaps the most striking difference is the inhibitory receptor HTR1E: this gene is absent from the mouse genome but abundant across the human brain. Through our examination of regional expression of HTRs we have also found several mouse HTR genes which are widely regarded as homologs of human HTR genes yet have markedly different expression profiles in both inhibitory and excitatory neurons.
In conclusion, we demonstrate a surprising discrepancy between human and mouse serotonin systems, calling into question the widespread use of rodent models for testing SSRIs for human use.
Mentor
Mark Reimers
Comments
Madison Janey4 years ago
Hi Ali! I loved your voiceover, you went into great detail. I do have one question though, would it be correct to assume that the receptor expressions can be the same in the different types of mice? If not, how are you able to decipher between all kinds of mice to be able to identify the difference of receptor cell types between humans and all kinds of rodents?
• • 1 comment
Ali Cramer4 years ago
Hello Madison! It would be safe to assume all types of mice would have roughly the same pattern of expression as seen in the figures since we used trimmed means data instead of individual cell points. We also verified that the genes present in humans and not in mice were not found in other rodents such as rats using the UCSC Genome Browser and NCBIs BLAST Algorithm.
Aaron Reifler4 years ago
Hi Ali, great work on your poster. I was wondering whether you consider the study of 5HT receptor types across species to be more or less significant than neuronal connectivity across brain regions for 5HTRs.
• • 1 comment
Ali Cramer4 years ago
Hi Aaron! Both the receptors that each cell expresses and where those cells are located/project to are important. While we only showed the significance of the cell type expression differences between organisms in this poster, our next step is to identify regional 5HT receptor expression differences across both organisms to get a clearer picture of what circuits each cell type is participating in and therefore the larger affects of widespread differences on psychiatry and brain function.
Denise Zavala4 years ago
Gret presentation! I had one quick question. I am not sure if you mentioned any limitations in your project. Do you have any limitations that you found and if so in the future how could you potentially fix them?
• • 1 comment
Ali Cramer4 years ago
Perhaps the biggest limitation is that the human single cell RNA-seq data contain only cortical brain cells, while the mouse data contain both cortex and hippocampal cells. Human hippocampal data should be coming out soon, so that should allow us to make more accurate comparisons between organisms. We also used trimmed means in this study instead of the reads of each cell, sacrificing a bit of granularity for a more space and time efficient workflow since we are using R Studio, which will not load the entire mouse single cell file, instead forcing us to save small cuts of the full file using unix.