Development of Cell-Type-Specific Viral Vectors to Tease Apart the Neural Circuitry that Contributes to Drug Addiction - Viral Vector Approaches in Neurobiology and Brain Diseases - page 152

of cell-specifi c promoters must be obtained or constructed. When
designing a cell-type-specifi c viral vector, there are several factors
that need to be taken into it account. First, viral vector cassettes
have a size limitation for the amount of DNA that can be packaged
that varies with each viral vector. If this limit is exceeded, packag-
ing effi ciency can be reduced and/or the plasmid can be truncated.
For example, HSV-1 plasmids permit very large inserts (in theory
up to 150 kilobases [kb] but in practice up to about 15 kb),
whereas AAV plasmids have a much smaller limit of ~5 kb and in
practice often less [ 14 , 15 ]. Thus, it is advantageous to use the
shortest length of a promoter that retains its function. Methods
such as “promoter-bashing,” which involve mutating or digesting
various promoter regions and testing the effect on gene expres-
sion, can be used to identify key regulatory regions in the pro-
moter and the nonessential sequences can be removed from the
plasmid construct if necessary. If these experiments have not been
conducted on your promoter of interest, it may be necessarily to
empirically test this yourself. Second, a promoter usually can only
drive expression of one transgene. This has led to the common use
of epitope tags for immunolabeling of transgenic protein or fl uo-
rescent fusion proteins to conveniently identify transgenic protein.
However, one must confi rm that these modifi cations do not alter
the cellular function of the transgenic protein. IRES sequences
sometimes have been used to express a second transgene (such as a
fl uorescent protein) from the same promoter, but results have
often been disappointing. Recently, “ribosomal skip” sequences
[ 16 ] have been utilized and these may offer advantages although
their broad applicability is still under investigation. Third, the pro-
moter is an important determinant of onset and duration of viral
vector expression. For example, HSV-1 viral vectors that use the
standard immediate early promoters have a typical onset of less
than 24 h, with peak expression at 3-5 days postinfusion [ 17 ],
whereas neuronal promoters have been shown to produce more
stable levels of gene expression [ 14 ]. For example, when the HSV-1
promoter is switched to a neuropeptide promoter (e.g., preproen-
kephalin or preprodynorphin), both the onset and the duration of
viral vector expression are longer (onset less than 7 days; duration
at least 1 month) [ 9 ]. For example, using the preproenkephalin
promoter to drive green fl uorescent protein (GFP) expression, we
observe strong expression at 10 and 21 days after stereotaxic infu-
sion of the viral vector into dorsal medial striatum (Fig. 1 ).
Furthermore, phenotype-specifi c promoters are usually much
weaker than popular viral promoters, so one should choose a
strong one. For example, we presumed that neuropeptide promot-
ers are generally stronger than receptor promoters, and we sus-
pected that these would more faithfully represent the phenotype of
a specifi c neuron type than receptors, which tend to be expressed
more widely. Using a moderately strong promoter may be entirely
suffi cient to express key proteins for cellular signaling and may