Genetic Tools for the Analysis of Neural Networks in Flies

Motion vision is of fundamental importance for moving animals from
arthropods to mammals. In this thesis I lay ground for the
functional analysis of the neural circuit underlying visual motion
detection in fruit flies by means of genetic tools. In Drosophila
melanogaster transgenic tools allow for both experimental
observation and manipulation of neural activity: genetically
encoded calcium indicators (GECIs) can be used for the
optophysiological characterization of neural activity and
transgenes for the inhibition of neural activity can be used to
determine these neurons' function. Combined, yet independent use of
both tools is a powerful approach for the functional analysis of a
neural network. However, GECI signals in vivo generally suffer from
poor signal-to-noise ratios and GECI characteristics change
dramatically and unpredictably when transfered from the cuvette
into neurons of living animals, probably due to interactions with
native cellular proteins. Here, I quantified and compared the in
vivo response properties of five new (Yellow Cameleon 3.60 &
2.60, D3cpV, TN-XL and TN-XXL) and two more established ratiometric
GECIs (Yellow Cameleon 3.3, TN-L15). In addition, I included the
single-chromophore probe GCaMP 1.6 in this comparison. The analysis
was performed under 2-photon microscopy at presynaptic boutons of
neuromuscular junctions in transgenic fly larvae. I quantified
action potential induced changes of calcium concentrations by
calibrating responses of a synthetic calcium indicator that was
microinjected under 2-photon guidance. The observed cytosolic
calcium concentration was 31 nM at rest and increased linearly with
stimulus frequency by 0.1 to 1.8 uM at sustained activity of 10 and
160 Hz, respectively. This allowed for a quantitative comparison of
the responses of GECIs in terms of their steady state response
amplitudes, signal-to-noise ratio, response kinetics, calcium
affinities and hill coefficients in vivo. The results were then
compared to in vitro properties of GECIs measured in cuvettes. The
data reveal that a new generation of GECIs retain improved
signalling characteristics in vivo. Maximum fluorescence changes
were 2-3 fold increased in new compared to former ratiometric GECI
variants. Small calcium changes in response to 10 Hz stimulation
induced fluorescence responses with signal-to-noise ratio above 2
in Yellow Cameleon 2.60 & 3.60, D3cpv and TN-XXL. Kinetics were
slowest in Yellow Cameleon 2.60 and fastest in TN-XL. The observed
changes between in vitro and in vivo performance revealed
systematic differences between GECIs of different types. GECIs in
this study employ different calcium sensing molecules:
calmodulin-M13 in Yellow Cameleons and GCaMP, a redesigned
calmodulin-M13 in D3cpv, and troponin C in TN-indicators. Those
indicators comprising calmodulin-M13 as calcium sensors displayed
reduced maximum fluorescence changes and reduced hill coefficients
in vivo, while troponin-based GECIs and D3cpv showed increased hill
coefficients and increased maximum fluorescence changes in vivo.
Calcium affinity of all GECIs was increased in vivo. The results
demonstrate that there are now suitable GECIs at hand for
experimental questions at differing expected calcium regimes.
However, in contrast to a synthetic calcium sensor, none of the
tested GECIs reported calcium concentration changes related to
single action potentials at presynaptic boutons of the
neuromuscular junction. In the visual system of Drosophila, optical
recordings from motion sensitive neurons while selectively blocking
certain classes of columnar neurons will allow for a network
analysis of the motion detection circuit. The Gal4-UAS system can
be used to express proteins that block neural activity. A similar
two-part expression system, based on bacterial protein- DNA
interaction (LexA and LexA-operator), can be used in parallel to
drive the expression of GECIs. I generated flies expressing TN-XXL
or Yellow Cameleon 3.60 under the control of the LexA-operator and
demonstrated GECI expression in olfactory receptor neurons. In
parallel, I cloned putative genomic enhancers that can be used to
drive LexA expression in motion sensitive cells of the flies visual
system. Finally, adult fixed flies expressing TN-XXL in motion
sensitive neurons were visually stimulated by large field moving
gratings. Parallel fluorescence measurements from these neurons
showed for the first time directional selective calcium responses
in Drosophila. The next step will now be the combination of calcium
imaging in these neurons and functional blocking of their
presynaptic partners.