How Can Gene Expression Studies Help Design Treatments for Adult Degenerative Brain Diseases?

James P. Bennett, Jr. M.D., Ph.D.
President and Chief Scientific Officer
Neurodegeneration Therapeutics, Inc.
April 29, 2018


Genes are stretches of DNA in our chromosomes that define the primary structure of proteins. Our cells use these thousands of proteins to carry out their many different tasks. Current estimates are that ~5% of our DNA in chromosomes is used for genes and the remaining ~95% used to be called “junk DNA”. We now recognize this to be a misnomer, for our “junk DNA” codes for other molecules that in turn regulate how many copies of each gene (“gene expression”) are made. This is our “epigenetic” (above genetics) system that has been recognized as increasingly important in our growth and development and occurrence of diseases.

In our cells’ nuclei, DNA in genes is first made into a string of nucleotides called “messenger RNA” (technically, pre-messenger RNA, or pre-mRNA). This pre-mRNA string, which faithfully reflects the DNA sequence of the gene it came from, is chemically modified at both ends, sliced-and-diced into various “isoforms” by a process called “alternative splicing”, and exported as “mature mRNA” from the nucleus to the cell interior (cytoplasm). The mature mRNA makes its way to protein synthesizing machinery, called ribosomes that turn the mRNA sequence into protein. The resulting proteins are folded into 3-dimensional shapes and then used by the cells.

Humans have ~20,000 genes in our DNA genomes (chromosomes), but we make ~100,000 different proteins. This means that on average, each gene yields ~5 different proteins through the “alternative splicing” mechanism. The number of genes we possess is nothing special in the animal kingdom, but the diversity of proteins we make is above average.

Best current estimates are that only ~30% of newly made proteins properly fold; the remainder are tagged for removal and shredded into their amino acid components and recycled. mRNA is degraded after it is turned into protein, and some mRNA never makes it to the ribosomes. It is degraded by “microRNA’s”, aka miRNA’s, that originate from the “junk DNA” referred to above. miRNA’s (of which ~4000 are currently known in humans) can also block the conversion of mRNA into protein, by inhibiting the mRNA-ribosome machinery. miRNA’s appear to exist as an evolutionarily old system to protect against invasion by RNA viruses.

What is gene expression?

Gene expression is the estimation of the numbers of copies of each mRNA inside a cell or tissue. Modern gene expression studies use “next-generation” sequencing techniques to determine the ordering of letters for each RNA molecule, followed by complex mathematical algorithms to align these sequences to known genes (in this case, in humans) and to estimate their abundances. Older approaches to gene expression include quantitative polymerase chain reaction (qPCR) and microarray slides. These older approaches still appear in the scientific literature but are being replaced by newer RNA-sequencing technologies.

At best, all gene expression approaches are estimates of the actual numbers of RNA molecules derived from each gene. There are assumptions about sequencing accuracy, alignment and normalization that influence the final answers, so one of the important aspects of any gene expression study is to have appropriate “controls”.

Does gene expression add to knowledge of what causes degenerative brain diseases?

The answer to this question is important but is still not clear. The levels of genes in a cell or tissue reflect the adaptive responses to stresses present at death (if one is studying postmortem brain tissues). The genes may or may not reflect stresses that caused the disease process, and instead may reflect strategies of “survivors”, cells that have withstood the disease process. Thus, one must be very careful in with interpretations of findings.

That being said, genes inside cells or brain tissues may represent deficiencies (which are amenable to supplementation) or excesses (which are amenable to repression) of one or more processes. Gene expression studies may define areas of cell function that are overactive or underactive in a disease and which could be altered as a therapeutic strategy.

Are degenerative brain diseases heterogeneous?

In contrast to diseases from infection or toxins, which arise in response to invasion from the external world, degenerative brain diseases represent changes in our own cells, similar to what happens in cancer or diabetes. These “complex” diseases arise from within, not as a result of invasion from without.

We are very close but not identical to each other genetically. Thus, there is no reason to assume common etiologies of diseases or responses to treatment at the molecular levels. This approach helps to understand why there are many different cancers of breast, lung or pancreas tissues- they frequently have different molecular genetic abnormalities that have been revealed by modern DNA sequencing techniques. And drugs that work for one type of DNA mutation in cancer may not work for a different DNA mutation.

The goal of “personalized” or “targeted” therapy is to define which therapies work (or don’t) for specific genetic abnormalities. A similar approach is needed for degenerative brain diseases. “One size fits all” has not been successful for drug development in these conditions, beyond reducing certain symptoms.

An effective therapy may work for only a small fraction of the population with a complex condition. As an example, only ~20% of breast cancer patients are help by antibody immune therapy directed against a specific protein receptor family (HER2) on the breast cancer cell surface. If all patients with breast cancers were treated with this antibody, the positive signal would have been lost in the statistical noise. Only by pre-selecting breast cancer patients carrying the HER2 mutations was the specific antibody therapy shown to be effective against this otherwise lethal breast cancer of young women.

This experience, among others, informs us that we need reliable biomarkers of a specific molecular abnormality, in order to offer therapy for that specific deficit. For complex diseases like neurodegenerative conditions, it is highly unlikely that a single agent will be effective for all patients, or even a majority of patients. Yet, outside of cancers, that single agent approach dominates current drug development trials (and notably their failures).

Where should research efforts be directed?

I believe that we should now focus on identifying ways to separate groups of degenerative diseases into subgroups based on molecular and biochemical criteria. In this manner we hopefully will identify biomarkers that could be used prospectively to separate a large heterogeneous group into smaller, more homogeneous groups that receive specific treatments (again based on what is deficient or what needs to be repressed). “One size fits all” is not helpful as a therapeutic philosophy anymore, putting aside successes in symptom control which are not the same as altering underlying disease.

Gene expression is one approach to defining subgroups. It may not be the best approach, but that is a question that will be answered in time.