DISCOVERING New Treatments for Brain Diseases
The world desperately needs new therapies for neurological and psychiatric diseases that work through novel mechanisms of action.
Discovering a new treatment usually starts with identifying a target protein that is involved in the biological pathways or tissues affected by a disease. Researchers then use chemistry or other methods to change that target protein’s activity while not directly affecting other proteins.
For neurological and psychiatric diseases, the brain’s complexity makes it enormously challenging to determine which protein should be targeted. Hundreds of different brain cell types are intermingled in complex circuits and pathways.
And simply targeting a protein involved in a disease may not be enough. The target protein you’ve selected may be present across different cell types with different biological roles. As a result, changing the target protein’s activity in multiple cell types may cause unwanted side effects. Many drugs acting on broadly produced proteins have caused dangerous side effects.
Therefore, an ideal target protein is produced selectively in cell types that are relevant to the disease.
To identify target proteins that are selectively produced, we need to know which proteins are made by which brain cell types. Until now, this has been extremely difficult.
Most approaches to studying human brain diseases have limitations:
Scientists try to use animals such as mice to study brain diseases, but because of the huge differences between the human brain and other animals, animal models of brain diseases affecting humans are limited and disappointing.
Scientists make induced-pluripotent stem (iPS) cells from patients’ blood or skin samples and then try to turn them into brain cells in a laboratory dish. But the brain cells produced are still immature and quite different from cells found in a mature brain. They don’t maintain age (a critical factor in neuro-degenerative disease), because they’re developmentally reprogrammed. They also lack the connectivity of a full brain (even if grown into 3D organoids).
Scientists also try to isolate single cells or nuclei from human brain tissue and then sequence their RNA to measure gene expression. While this approach typically measures only the highest expressed ~2,000 genes, proteins made by lower expressed genes often make better therapeutic targets. In addition, this approach is inefficient for profiling rare cell types.
Our PROPRIETARY Platform: NETSseq
Our proprietary “NETSseq” approach, invented by Nat Heintz and Xiao Xu at Rockefeller University, enables us to comprehensively profile specific brain cell types – including both neurons and glial cells – in mature human brain tissue. The approach involves using antibodies against nuclear proteins (e.g., transcription factors), endoplasmic reticulum proteins and membrane proteins, as well as RNA probes against any cell-type-specific transcripts. These probes tag specific cell types in brain tissue so that we can sort their nuclei. Read More.
Antibody labeling of protein by immuno-histochemistry or immuno-fluorescence
Probe labeling of RNA by in situ hybridization
We then capture the nuclei from one cell type by using florescence-activated sorting and read the sequence of the nuclear messages, allowing us to quantify the expression levels of genes in that cell type. Our RNAseq data from specific cell populations are robust and highly reproducible, even with human brain tissue samples that were frozen as long as 48 hours after death.
Nuclear Enriched Transcript Sort sequencing: NETSseq
Our approach enables us to measure the expression of significantly more genes, including genes expressed at lower levels, than single-cell or single-nuclei analysis. This more sensitive approach is critical for the discovery of molecular signatures of late onset degenerative disease and for the identification of potential therapeutic targets (e.g., cell surface receptors, enzymes, etc.) that may be expressed at low levels in the mature brain.
We employ network and pathway analysis, machine-learning and other informatics approaches to analyze these large data sets, revealing potential targets that are selectively expressed in circuits and cell types affected by disease.
Some of the Exciting Things We’re Starting to See
We’re starting to reveal cell-type-specific differences between:
- Different brain cell types
- Brain donors from age 17 to 97
- The same cell type across different brain regions (e.g., glial cells)
- The same cell type in different sub-regions of the same brain structure (e.g., dorsal versus ventral striatum)
- Healthy and diseased donors
- Groups of healthy individuals
We can also look at the cell-type-specific expression levels of genes/loci in which variants are known to cause familial forms of brain diseases or have been associated with brain diseases in genome-wide association studies. And, we can look at differential chromatin accessibility (epigenetics) in specific types of neuronal and glial cells.