|
 An
estimated 100,000 genes lie on our 46 chromosomes. One third of these
are thought to be specifically expressed in our brains at different periods
over the course of our lives. Over the past few years, a small but growing
number of genes have been cloned. Mutations in a number of these genes
lead to a disruption of the normal CNS development. The ability to isolate
genes is probably one of the most important recent achievements in neuroscience.
This is a necessary first step before investigators are able to understand
the normal role of these genes and how their protein products function
in our brains. In the next several columns, we will discuss exactly how
genes are cloned and how researchers learn more about the structure and
function of the proteins they encode.
Last
month, we reviewed how cytogenetic techniques have advanced to the point
at which we can actually visualize chromosomal abnormalities such as a
deletion. These physical disruptions point to the site on the chromosome
that presumably has a mutation that contributes to the observed phenotype
in the affected individual. These types of cytogenetic analyses reduce
the search for a mutated gene from the several billion nucleotides on
our total complement of DNA down to the millions of nucleotides contained
within the deleted site. Considerable work remains, however, before the
gene of interest is actually isolated, and various methods have been developed
over the past decade to do exactly that.
Most
of the techniques to isolate genes make use of libraries. Two types
of libraries will be described. The first is called a complementary DNA
(cDNA) library. This type of library is produced from‹and therefore contains‹all
the messages expressed in a given tissue. To construct such a library,
one first isolates the pool of RNA messages from a tissue, copies these
messages into the more stable cDNA molecule, and places the cDNA molecules
into an appropriate vector. Several types of vectors exist, but in general
they serve similar functions: to carry a single cDNA molecule and to facilitate
their own replication, thereby allowing the investigator to make many
copies of the library. A single message, or ³insert,² is carried in each
vector. Each vector and its insert is termed a clone, and it is the total
collection of cDNA clones that is called the library.
cDNA
libraries consist of all the expressed sequences from a tissue of interest.
The library will contain many copies of the most common housekeeping genes
as these genes are abundantly expressed and many copies of their RNA messages
were originally present when the RNA was isolated to make the library.
The library will contain fewer copies of less commonly expressed genes.
However, if the screening methods are powerful, one should be able to
find even very rarely expressed messages. The ultimate goal in screening
a cDNA library is to isolate a clone of interest.
One then determines the nucleotide sequence for that clone, and the nucleotide
sequence defines the precise amino acid sequence for the protein the gene
encodes. It should be noted that one can construct a cDNA from any tissue
in the body. cDNAs can also be made from the same tissue at different
time periods. For example, to identify genes that are transiently expressed
during the development of the cerebral cortex, one would obtain cerebral
cortices early in development, isolate the pool of RNA messages present
in that tissue, convert these messages into cDNA molecules, and place
the cDNA molecules into an appropriate vector. The end result would be
a cerebral cortex cDNA library from early in development. Many of the
clones contained in this library would also be found in an adult cerebral
cortex cDNA library, as many genes are expressed at both times. However,
the fetal cDNA library would also contain unique messages that are likely
to be involved in the early development of the cerebral cortex.
A
second type of library is called a genomic library. In this case, the
goal of the investigator is to make a library from the total amount of
chromosomal DNA. The DNA molecules are digested into smaller, more manageable
pieces, and these smaller DNA sequences are placed into a vector. Special
vectors have been constructed for this purpose, as the amount of DNA that
must be packaged is considerably larger than for cDNA libraries. It is
DNA from human chromosomes and have these segments replicated by the yeast.
The library that is constructed using this type of vector is called a
yeast artificial chromosome (YAC) library
(Fig. 1). It allows researchers to package up to 1 million nucleotides
into a vector. This larger stretch of genomic DNA may contain several
genes, each with their respective exons, introns, and regulatory regions.
It
has been appreciated for some time that many housekeeping genes have particular
repetitive sequences near their regulatory regions. These sequences are
enriched for a particular dinucleotide repeat (CpG) and are distinguishable
from other nucleotide sequences by the extent of methylation that occurs.
This chemical modification to the DNA molecule makes the region sensitive
to digestion by rare cutting restriction endonucleases, enzymes that cleave
DNA at specific nucleotide sequences. One of the enzymes that cuts here
as a consequence of the methylation is called HpaII, and as a result these
regions have been termed HTF (HpaII tiny fragments) islands.
It
is possible to isolate some genes by taking advantage of these HTF islands.
As was mentioned, the majority of HTF islands are associated with the
regulatory regions of many genes. Isolating thedown stream nucleotide
sequence would disclose the likely coding sequence for a protein. After
subcloning genomic fragments that lie near HTF islands, the isolated genomic
clones are used as probes to screen cDNA libraries. The technique takes
advantage of the fact that complementary nucleic sequences bind very tightly
to each other. Screening a library in this fashion uses the genomic clone
to bind to the cDNA that was actually transcribed from that region. Unfortunately,
the regulatory regions of most tissue-specific genes do not have HTF islands.
Hence this technique has limited potential to identify the majority of
genes expressed in the CNS.
Additional
methods have therefore been developed. One of these techniques takes advantage
of chromosomal abnormalities that are present in a number of disorders.
It is possible to visualize these abnormalities either by routine karyotyping
methods or the more recently developed chromosomal painting technique
(see last month¹s column). One can also hone in on candidate chromosomal
regions by linkage analysis or association studies. The region immediately
surrounding such a suspected chromosomal region becomes the area that
is likely to contain a gene of interest. But how does one actually go
about isolating the gene? One method is to identify several YAC clones
that span the chromosomal region of interest. The YAC DNA is then digested
into fragments, and these fragments are immobilized on a nylon filter.
The filter is probed with members of a cDNA library. The underlying principle
is that clones that were originally transcribed from the chromosomal region
of interest will bind to the fragments of genomic DNA bound to the filter.
Once again, this is because complementary nucleic acid strands have an
enormous affinity for each other. cDNA clones that represent messages
expressed from other chromosomal regions will not bind to the YAC fragments
immobilized on the filter, and these clones will be washed away. The bound
clones are carefully removed and amplified. These clones then represent
potential genes that were transcribed from the chromosomal region of interest.
If
one were looking for genes that cause a particular disorder, then one
would sequence that gene from several individuals with the disorder and
compare the nucleotide sequence with the sequence found in unaffected
individuals. Convincing evidence for that gene being the gene responsible
for causing an illness comes when mutations are found among other affected
individuals and not found in unaffected individuals.
This
method is termed cDNA hybridization selection. Although it has been largely
successful, there are several drawbacks. For one, some of the recovered
sequences are false-positives and derive from transcribed sequences in
other regions of the genome. In addition, ribosomal and repeat sequences,
which normally represent up to 10% of the human genome, must be blocked
so as not to be selected in the experiment. Finally, the selected cDNAs
are rarely full-length and tend to be relatively short. To achieve a more
complete representation of all transcribed sequences in a given genomic
region, several cDNA libraries will need to be used, and several screenings
performed. cDNA hybridization selection performed with several cDNA sources
can typically identify 70% to 80% of the transcribed regions in a given
genomic clone.
The
method that was just describes relies on the expression of genes in order
to isolate a gene of interest. If, for example, the cDNA library used
to screen the genomic fragments was made from adult brain RNA, then genes
expressed during early development will not be found. If that gene is
the gene responsible for a disorder, it will not be isolated by screening
the genomic filters with an adult brain cDNA library.
More
recently, techniques have been developed that do not depend on the expression
of genes for those genes to be isolated. One of these methods is termed
exon trapping. The underlying principle relies on the fact that the vast
majority of genes are processed after they are transcribed from DNA into
RNA. One of the important steps in the processing of RNA messages is to
bring together the exons that contain the protein coding sequence and
to remove the large stretches of intervening, noncoding sequences. These
intervening sequences are termed introns, and the procedure that removes
the introns and joins together the exons is called splicing. A number
of small nuclear proteins are responsible for bring about the splicing
of genes into mature RNA messages. These proteins recognize conserved
nucleotide sequences at either end of an intron that needs to be removed.
When this signal is found, the small nuclear proteins bind tightly, and
the exons of a gene are spliced together while the introns are removed.
In
the expression-independent method for isolating genes, the nucleotide
signals that indicate a splicing site are used to identify the exons on
either side. Genomic DNA is digested into small fragments, and the fragments
are cloned into a specialized vector that contains the first half of the
splicing signal necessary for intron removal. If the genomic segment that
is cloned into the vector contains the remaining half of the necessary
splicing signal, a splicing event will occur. These splicing events can
be detected through a simple assay, and it is a relatively easy matter
to then isolate the clones identified. This technique is most useful when
the chromosomal region that is suspected to carry a mutated gene has been
identified. It is considerably easier to trap exons expressed in the much
smaller area surrounding a translocation, for example, then to have to
use the entire genomic sequence. The largest drawback to exon trapping
is the presence of cryptic splice sites in the genome. These sites will
lead to the isolation of false-positive clones, or clones that do not
represent exons. This necessitates the rescreening of candidate regions
in order to guarantee that isolated clones are true-positives. In addition,
this technique cannot be used to locate genes that lack introns. Despite
these problems, exon trapping is a quick and efficient way to locate coding
regions within specific regions of genomic DNA.
With
the expansion of computer databases and the sequencing of longer stretches
of DNA, computers are playing an increasingly important role in identifying
expressed sequences. Currently, only 3% of the genome has been sequenced,
and the longest continuous stretch available in public databases is less
than 1.5 million base pairs. However, over the next few years, genome
centers anticipate sequencing millions of additional nucleotides of genomic
DNA, and the completion of the sequencing of the entire human genome is
scheduled for the beginning of the next millennium. Personal computer
programs exist that can quickly search genomic DNA sequences for the presence
of HTF islands, regulatory sequences, promoters,splice sites, and open
reading frames. Large sequence databases, such as GenBank at the National
Center for Biotechnology Information, are available on the Internet. These
databases are the repositories for DNA and RNA sequences from laboratories
around the world. There are several useful tools available to search these
databases. As the human genome becomes well covered with genomic clones,
it will be increasingly important to identify the transcribed sequences
within those clones. The methods discussed here such as cDNA hybridization
selection, exon trapping, and computer-based analysis of deposited sequence
data will identify many human genes. We can then ask what their normal
function is and how they cause disease when mutated.
top of page
|
