what-when-how
In Depth Tutorials and Information
although expensive option. However, if the missing activ-
ity is intrinsic to the cell function, then the cells need to
be engineered to produce the activity. The most common
approach is the introduction of a cDNA copy of the RNA
encoded by the missing gene whose transcriptional activ-
ity is controlled by a viral promoter. This artificial gene
construct can be delivered to the target cell in a highly
modified lentiviral vector capable of inserting the gene
sequence at random sites within the chromatin, while
still retaining a variable degree of promoter activity in
spite of cellular mechanisms that can suppress transcrip-
tional activity from a viral gene. The major concern with
this approach is the potential of the viral insertion to inter-
rupt a tumor suppressor gene or to activate an oncogene,
as was observed in the initial gene therapy trials for SCID
children.
36
Despite this concern, there are numerous ongo-
ing trials of gene therapy utilizing lentiviral vectors as the
delivery vehicle.
37,38
Other vector strategies have been developed to over-
come the objections of integrating viral vectors. Most
widely employed are adenoviral associated vectors
(AAV) that either remain as episomes, small genetic
units that replicate using host mechanisms but extrin-
sic to host chromatin, or integrate at a specific site in
the human DNA. This vector has replaced adenovi-
ral vectors, which induced major immunological and
cytopathic changes and were responsible for the col-
lapse of gene therapy in 1999.
39
Continued modifica-
tions to AAV design have reduced the immunological
side effect with the remaining difficulties being related
to the size of the gene that is carried by the vector.
40,41
Newer hybrid vectors that remain as stable episomes
and Sleepy Beauty transposons are currently being
engineered and evaluated in animal models.
42-44
Thus
for many genes that encode a relatively small protein
and do not require complex cellular regulation, a vari-
ety of delivery vectors are currently available with fur-
ther improvements on the way.
Either viral vector approach is unlikely to be effective
in expression of a major connective tissue gene such as
type I collagen. Limitations on size of the cDNA and the
inability to retain control over the activity of the gene
would be major issues that would have effectiveness and
safety implications. While other gene transfer technolo-
gies for introducing large genomic fragments are pos-
sible, the need for the approach has been superseded by
newer techniques of gene correction.
the strategy is effective it should result in a milder, type I
OI phenotype.
A variety of approaches to reduce the RNA levels that
accumulate from the mutant collagen allele using muta-
tion specific ribozymes,
45
U6RNA
46
or siRNA
47
have
not achieved the reduction that would be expected to
reduce disease severity. These results indicated that only
direct gene silencing would be effective. That goal was
achieved by the Russell laboratory using the principle of
homologous recombination in which a small gene frag-
ment is inserted into a specific site in the host DNA.
48,49
In this case, the target was the first intron of the type I
collagen gene and the fragment caused the
Col1A2
gene
to become prematurely terminated by a process called
gene trap, thus precluding production of the mutant
gene product. Because the insertion targeted indepen-
dently of the mutation, clones of cells that incorporated
the gene fragment had to be screened for the proper
event, but in those cells the abnormalities of collagen
production and secretion that were secondary to the
mutant collagen chain were completely corrected. This
was the first example of gene inactivation of a dominant
negative mutation and the concept could be universally
applied to all mutations in this category. While there are
technical challenges in the process, the ability to select for
the desired correction and to exclude off-target events
provides a margin of safety that would be appealing to
regulatory agencies. The primary disadvantage is that
the patient still has OI, albeit in a much milder form.
Gene Correction
While gene inactivation is a major achievement, it
would be preferable to correct the underlying gene
mutation and remove all aspects of the disease. Although
undirected homologous recombination is routinely
used in mouse genetics to create specific gene knock-
out lines, it has not proven to be practical in human
cells. However, when a targeting small gene fragment is
homologous to a region that also has sustained a chro-
mosomal break, the host repair process can incorporate
the fragment at the site where the mutation is located.
50
Thus it is possible to perform site-directed homologous
recombination with a repair fragment and select for cells
where the mutant sequence has been replaced with a
normal sequence (
Figure 57.1
). Key to the process is the
ability to induce a site-specific gene break.
51,50
A number
of strategies use Zinc-finger nucleases,
52-54
transcription
activator-like effector nucleases (TALENs
55,56
) and clus-
tered regularly interspaced short palindromic repeats
(CRISPR
57
). While the molecular details of each strat-
egy are beyond the scope of this chapter, these advances
indicate that gene correction is now becoming a reality
as an effective and safe means for all types of genetic
mutations. Even for a recessively inherited or haploid
Gene Inactivation
It is unlikely that enhanced expression of type I colla-
gen would be effective in the dominant negative forms
of OI because the mutant collagen chains will continue to
disrupt the processing of the normal chains. Instead, the
activity of the mutant allele has to be suppressed and if
