Using Viral-Mediated Gene Transfer to Study Depressive- Like Behavior - Viral Vector Approaches in Neurobiology and Brain Diseases - page 165

from either the CMV or the HSV promoter. We have consistently
found that GFP control groups perform similarly to unoperated or
sham operated animals in most [ 8 - 10 ] but not all cases [ 11 ]. In
some cases, it could be argued that the expression of a protein
might have unintended consequences, such as altering the effi -
ciency of processing of off-target neuropeptides by endopepti-
dases. In this case, one might consider a nonfunctional protein
control, such as expressing des-tyrosine enkephalin. However,
expressing nonfunctional proteins can have unintended conse-
quences as well, such as dominant negative interactions with func-
tional endogenous proteins, so these must be used carefully. Finally,
anatomical controls can be quite powerful in demonstrating the
role of a transgene in a discrete brain region [ 9 ].

3.2 Regional
Specifi city

Traditional drug infusion strategies are a popular and powerful
method for testing the role of a target protein, such as a receptor,
in a discrete brain region. However, drug infusions have limitations
related to uncertainty about the drug concentration after infusion
into the brain region (potentially affecting the selectivity of the
injected ligand) and the potential of drug diffusion and spread
away from the intended target. Viral-mediated gene transfer has
the advantage that the surgery and drug infusion can occur long
before the behavioral experiment, so the stress of performing a
brain infusion immediately prior to behavioral testing is avoided.
Furthermore, the site of transgene expression can be precisely
determined histologically, particularly when a fl uorescent protein
or epitope tagging strategy is used. We have found that coexpres-
sion of GFP from a separate transcriptional cassette allows for con-
venient and precise identifi cation of injection sites after completion
of the behavioral experiments; it is more accurate than inferring
injection sites from cannula placements. Furthermore, we routinely
observe that about threefold more infected neurons will be revealed
by using an immunostaining procedure as compared to direct visu-
alization of GFP, for example. It is also possible to identify off-site
transgene expression, and we observe about 1-2 % as many GFP-
positive neurons in ventral tegmental area after infusion of HSV
viral particles into nucleus accumbens shell, suggesting that axon
terminals projecting to the site of infusion are occasionally infected.
We have recently started to use promoters associated with
specifi c phenotypes of neurons to further refi ne gene expression
using HSV vectors [ 12 ]. We have tested three promoters associ-
ated with specifi c subtypes of neurons (serotonin transporter, dyn-
orphin, and enkephalin), in each case using 2-3 kb of DNA
preceding a gene associated with a specifi c neuron type in addition
to the transgene of interest. Each of these promoters was previ-
ously analyzed to identify regions that were thought to be espe-
cially important in conferring cell-type-specifi c gene expression.
These hybrid vectors are interesting because they behave quite