Biomedical Engineering Reference
In-Depth Information
7.2
Gene-Expression Systems in
Drosophila
Specifi c behaviors of animals are often generated and organized by activity patterns
across ensembles of neurons mediating sensory processing, decision-making, and
motor control. Analyzing the neural processes that underlie a specifi c behavior thus
requires measuring neural activity in vivo. In
D. melanogaster
, transgenes for moni-
toring neural activity can be targeted to almost any population of neurons using
binary expression systems (Venken et al.
2011
).
A binary expression system consists of a transcription activator and its target
sequence to which the transcription activator binds. These two elements are sepa-
rated on two different transgenic
Drosophila
strains, a “driver strain” that deter-
mines where and when the gene of interest is expressed and a “responder strain” that
determines which gene of interest is actually expressed. The GAL4/UAS system
was the fi rst and is still the most widely used binary system developed for
D. mela-
nogaster
(Brand and Perrimon
1993
) (Fig.
7.1
). The key for this was the fi nding that
the transcription activator GAL4 identifi ed in the yeast
Saccharomyces cerevisiae
functions also well in the fruit fl y when a responsive DNA element for it, called the
upstream activation sequence (UAS) element, is present (Fischer et al.
1988
). In the
GAL4/UAS system, the two elements are now separated on a GAL4 strain (“driver
strain”) and a UAS strain (“responder strain”). If both strains are crossed, the F1
generation carries both transgenic DNA insertions, and the transcription factor
GAL4 can bind to the UAS sequence that induces the expression of the desired
gene, e.g., a Ca
2+
sensor (Fig.
7.1a, b
).
This principle is not limited to
Drosophila,
and the GAL4/UAS system is also
used in other genetically tractable organisms such as zebrafi sh or mice (Ornitz et al.
1991
; Scheer and Campos-Ortega
1999
). The beauty of the GAL4/UAS system in
Drosophila
, however, relies on a very large collection of fl y strains; thousands of
GAL4 strains have been created by random
P
element insertion, and each strain
expresses GAL4 in a designated spatiotemporal pattern (Hayashi et al.
2002
)
(Fig.
7.1c
). Moreover, a variety of defi ned promoters, such as the pan-neuronal
ELAV promoter (Yao and White
1994
) and the eye-selective GMR promoter (Hay
et al.
1994
), have been used to generate tissue-specifi c GAL4 strains. Recently,
GAL4 strains that drive gene expression in more restricted patterns were systemati-
cally generated by inserting small fragments from the fl anking noncoding and
intronic regions of genes (Pfeiffer et al.
2008
). This can lead to a transgene expres-
sion in small subsets of cells, e.g., very few selected neurons within the brain.
Furthermore, the effi cacy of the GAL4 system was improved by optimizing factors
that affect the pattern and strength of GAL4-driven expression such as codon usage,
mRNA stabilization, transcription activation domain, and the number of UAS sites
(Pfeiffer et al.
2010
). A variety of additional techniques have been developed to
further narrow down the expression of the transgene of interest in time and space.
Genetic mosaics can be created using the MARCM system (Wu and Luo
2006
) that
