Drug Target Discovery for Diseases of the Nervous
System
Existing therapeutic strategies for the treatment of neurological
disorders and mental illnesses focus almost exclusively on
symptoms rather than the underlying pathogenic mechanisms
causing disease. The available medications are thus poorly
targeted as cures and essentially incapable of preventing
or delaying the onset of neurodegenerative disorders such
as Alzheimer’s disease, Parkinson’s disease and
Amyotrophic Lateral Sclerosis (Lou Gehrig’s disease).
Acknowledging this failed strategy, most pharmaceutical companies
have now re-oriented their CNS research programs toward a
disease mechanism-based approach to drug discovery. The principal
research activity in these efforts centers primarily on characterization
of disease-associated cellular pathways and their protein
components. The next step is often the most difficult: determining
which of the implicated proteins should serve as the next-generation
drug targets. For many of these candidate molecules, it is
not easily known ahead of time whether modulating their activity
provides therapeutic benefit by affecting disease initiation
or progression or by diminishing other important aspects of
pathology. In addition, the availability and choice of animal
model systems impacts the validity of data obtained for target
assessment. More broadly, the evaluation of therapeutic strategies
for human disease in imperfect animal models is often the
road to drug development failure. Thus, successful drug discovery
lies in focusing on the right targets and in applying the
most precise animal models for any particular disease. This
is especially true for diseases of the brain.
At Neurome, research focused on genes whose protein products
demonstrably alter disease onset or pathology will enable
discovery of disease-relevant drug targets for a variety of
neurodegenerative disorders. Our strategy takes advantage
of the strengths of mouse genetics, where Neurome’s
quantitative methods have been remarkably successful at identifying
genomic regions contributing to an array of disease-associated
traits and susceptibilities that differ between inbred lines
of mice. These genomic regions are known collectively as quantitative
trait loci (or QTL), and the identity of the genes responsible
for quantitative traits offers a powerful point of departure
for the development of new pharmaceuticals. The encoded proteins
of quantitative trait (QT) genes represent genetically validated
targets, since studies that led to QTL mapping provide prima
facie evidence that differences in the activities of the proteins
directly contribute to differences in phenotype. Thus, in
a mouse model for a particular disease for which it has already
been shown by genetic studies that there are related QTLs,
each QTL-identified protein represents a validated target
for potential pharmaceutical targeting. A major goal of our
research efforts will be to systematically discover QT genes
for known QTLs related to human disease and to uncover disease-modifying
QTLs and therapeutic targets for neurodegenerative diseases.
Identifying the Best Drug Targets for CNS Disorders
The search for susceptibility genes and drug targets for neurological
and behavioral disorders has proven to be particularly difficult,
even though heritability of genetic risk factors has been
well documented for many mental illnesses. For many disorders,
genetic linkage studies have proposed one or more chromosomal
regions harboring genetic determinants, and plausible candidate
genes from these regions have been tested to determine whether
the gene co-segregates with disease within affected individuals
of a pedigree. However, due to the complexity of these disorders
and the limited statistical power of the linkage analyses,
most of these studies fail to replicate and primarily serve
to exclude hypothesized candidates within the genomic interval
for which linkage was first established. Candidate genes or
markers have also been evaluated in what are known as association
studies, a form of linkage analysis in which polymorphic alleles
or markers can be tested for association with susceptibility
after determining the genotype of variants in individuals
with disease and in unaffected controls in a large population.
Major neurological disorders, including Alzheimer’s
disease, multiple sclerosis, myasthenia gravis, Parkinson’s
disease, epilepsy, migraine, and ischemic stroke, have all
been investigated using genetic association approaches. With
the exception of Alzheimer’s disease, replication has
failed or is lacking for nearly all of these studies.
The difficulty in tracking complex trait genes for behavioral
disorders in humans is due in part to the variable roles that
environmental factors and experience play in establishing
these illnesses and to the contribution of many genes of small
effect, several of which may interact in a non-additive fashion
to influence disease traits. The variability often observed
in linkage studies is due to these inherent factors and therefore
limits the power of human molecular genetics approaches in
identifying genes for complex behavioral traits.
A Direct Route to the Targets
Fortunately, the task of identifying quantitative trait (QT)
genes is made much simpler by the use of inbred mouse strains.
As noted earlier, the observation that traits and susceptibilities
differ among different inbred lines has given rise to approaches
that identify quantitative contributions of multiple polymorphic
genes to disease-related phenotypes. Although initial QTL
mapping has proven to be rapid in mice, the molecular identification
of a QT gene and its encoded protein is often a formidable
bottleneck in QTL analysis, requiring years of intensive laboratory
effort. At Neurome, we have developed an innovative molecular
genetic strategy to overcome this obstacle. Neurome’s
strategy will employ measurements of mRNA abundance, utilizing
TOGA® technology, in a way that correlates gene expression
with genotype in order to vastly reduce the effort involved
in resolving individual loci.
TOGA® Technology
Over the past decade, advances in robotics, informatics and
molecular biology have led to the development of genomics
technologies that enable gene expression measurements to be
taken for thousands of genes in a highly parallel fashion.
The technology platform best suited to discover QT genes and
their corresponding protein therapeutic targets is Neurome’s
TOGA® (an acronym for TOtal Gene expression Analysis).
There are two principle reasons for this. First, TOGA®
is an ‘open system’ approach to gene expression
analysis. The method was designed to detect and measure the
abundance of any mRNA species present in a biological sample,
regardless of whether the gene was previously known or sequenced.
This design feature empowers TOGA® to interrogate a sample
and comprehensively survey all genes. The PCR-based detection
system provides a level of sensitivity that is sufficient
to detect genes expressed even as low as 1 part in 1,000,000.
The molecular biology steps of the procedure are outlined
in the schematic below. Second, in TOGA®, known genes
are identified based on sequence properties that provide a
unique digital address for every mRNA (see figure). For any
previously uncharacterized mRNA, a corresponding TOGA®
DNA fragment can be easily isolated and sequenced. Thus, the
TOGA® open system approach is ideal for discovery applications,
since no prior knowledge or assumptions need to be made regarding
a gene’s expressed sequence.
Full
Size
Figure 1a: The TOGA® method can be viewed
as a 3-part process involving molecular biology protocols
and robotics. The first series of steps (top box) accomplishes
the efficient capture of a 3’ fragment for every mRNA.
RNA from a biological sample serves as template for cDNA synthesis
initiated with a biotinylated anchor primer, and the cDNA
is subjected to enzymatic digest by a restriction endonuclease
with 4-base recognition specificity (i.e. MspI CCGG) followed
by fragment capture on streptavidin-coated beads. The second
part (middle box) utilizes in vitro transcription (IVT) and
reverse transcription (RT) reactions to enable highly reproducible
linear amplification and production of templates for PCR.
In the third part (bottom box), PCR is performed to systematically
parse the entire set of expressed RNAs and generate sequence-specific
subpools for analysis. PCR reactions are set up on high-throughput
robotics stations to contain a fluorescent universal 3’
primer and a unique 5’ primer taken from a 256 member
set whose 3’ termini cover all sequence possibilities
across the four bases (labeled N1N2N3N4) adjacent to the restriction
digest site. Following amplification, each of the resulting
256 PCRs is analyzed by capillary electrophoresis to measure
DNA fragment size and fluorescent intensity. Each resulting
fragment represents a single mRNA species detected by TOGA®.
The resulting gene expression data from a sample, which typically
contains expression measurements for 15,000 genes, is then
analyzed together with a logical group of samples, such as
in a comparison of normal versus diseased tissue.
Full
Size
Figure 1b: Messenger RNAs are assigned a unique
digital address by TOGA®. The addressing feature allows
a TOGA® database to be established for all detected RNAs
and enables prediction of TOGA® addresses for nearly all
transcripts contained in genomic databases. An 8-nucleotide
sequence, comprising the sequence specificity of the restriction
endonuclease together with the parsing sequence used in PCR,
is combined with the fragment length to create the TOGA®
address. In the example, the restriction sequence and parsing
sequence combine to give CCGGATCG and the fragment length
is calculated from the 5’ restriction cleavage site
to the poly A addition site to be 148 nucleotides, which together
create CCGGATCG148. A TOGA® graph for CCGGATCG148 is shown
in the bottom half of the figure.
|
