PROJECT SUMMARY
Brain activities involve neurons generating fast-propagating signals to encode and relay information within
dynamic neural networks. Neuroscientists aspire to obtain access to such networks in unconstrained animal
models (e.g., rodents) with high spatiotemporal resolution, which will shed light on the fundamental working
mechanisms of the brain. Optical imaging, particularly multiphoton microscopy, has played a significant role in
this endeavor. The past decade has seen impressive progresses, from head-restrained benchtop microscopy
with virtual navigation to large FOV microscopy for neuron population imaging, three-photon microscopy for deep
brain imaging, and two-photon (2P) miniscopy for in vivo imaging in freely-walking (but limited rotation) mice.
Despite these exciting technological advances, tools for simultaneous, large-scale, and high-resolution
imaging over multiple brain regions in freely-behaving rodents are still lacking. Successful development of
such tools can accelerate the process of uncovering general principles of neural networks in a working brain
under nearly natural conditions. The free-moving style for imaging would minimize the differences between
experimentally controlled actions and natural spontaneous behaviors, thus allowing for precise examination of
neural network functions. The capability of simultaneous imaging over two interconnected neural populations
would provide a comprehensive and precise timeline of the neural circuit dynamics associated with various
behaviors at both cellular and population levels.
Our proposed research is motivated by the need for such imaging tools with the above-mentioned features. The
main objective is to develop a 3D-scanning, ultrathin and light 2P fiberscope technology for enabling high-
resolution, simultaneous imaging of dynamic neural activities over a large FOV at two brain regions in freely-
moving rodents. To achieve our objective, we propose the following aims:
(1) To develop a fast scanning 2P fiberscope of a large FOV (Ø500 um) using a cascaded magnification
strategy while maintaining a compact probe size (Ø2.5 mm). The larger FOV will be achieved by using an
innovative micro-optics design. In addition, a modular scanner head design will be implemented in the 2P
fiberscope to improve the probe robustness for in vivo imaging at a high scanning frequency (e.g., ~2.8 kHz);
(2) To develop a miniature (Ø2mm) tunable lens that can be integrated into our 2D scanning fiberscope
for enabling depth (focus) scanning/selection over 150 um. Focus scanning allows for convenient selection
of a proper layer or population of neurons. The tunable lens can create a curved refractive index profile when
applied with a low-voltage (<10 V, safe) electrical drive. Compared with other tunable lenses, the tunable lens
will be extremely compact and light, critical for imaging freely-moving rodents. A fiberscope integrated with a
tunable lens will be developed and tested using phantoms, fluorescent tissue slides, and a mouse model in vivo.
(3) To develop a dual-probe system, enabling simultaneous 2P imaging of two brain regions in freely-
walking/rotating mice. The ultracompact size and lightweight of the fiberscope permit two fiberscopes to be
mounted a mouse head, allowing for simultaneous imaging of two brain regions (cortex or deep brain). A novel,
proactive, dual-probe optoelectrical commutator (dpOEC) will be developed for the first time to sense and
compensate the torque built up in the fiberscopes, allowing the mouse to walk/rotate freely during imaging;
(4) To assess the feasibility of the dual-probe 2P technology for exploring neural network dynamics in
two different brain regions simultaneously during social decision making. Social behavior involves
sensory, cognitive, and motor functions and thus depends on the interactions of many neurons, but until now no
technology is available to record from a large population of neurons with subcellular resolution over multiple
interconnected regions in freely-behaving mice. Here we choose to study the dynamic neural connectivity
between the primary motor cortex (M1) and a critical sensory information routing node, periaqueductal gray
(PAG). Both areas are critically involved in social behavior, but how these interconnected regions synergize to
process information remains almost completely unknown. In addition to testing the performance of the 2P
fiberscopy technology, this aim could also shed light on how social preference is encoded. As a control, we will
monitor these regions during a locomotion (but nonsocial) activity (Rotarod running), for which the information
on M1 that is independent of PAG is already available.
In summary, successful completion of the proposed study will establish a new two-photon fiberscope imaging
platform for the neuroscience community to enable simultaneous high-resolution imaging of neural network
dynamics of different cell types over different brain regions in freely-behaving rodents. In addition, focus/depth
scanning will be made possible. The fiberscope can be easily attached to and detached from the mouse head,
permitting repeated use. Although beyond the scope of current proposal, the technology can also have many
translational applications, including internal luminal organ imaging for diagnosis or guidance of intervention.
