We discovered that it is possible to record and store biological information over time, such as the time course of gene expression, onto the ordered structures along intracellular linear protein self-assemblies. Analogous to how tree rings permanently store information over time as the wood grows, such protein ticker tape systems enable spatially resolved readout of live-cell physiological histories via common post-fixation immunostaining and imaging methods. Based on this concept, we designed a genetically encoded molecular recording system (Expression Recording Islands or XRIs) and applied it to record the c-fos promoter-driven expression over time in neuron populations. Because fixed cells and tissues can be scalably imaged using post-preservation processing techniques, these 'time rulers' hold great promise for physiological recordings at single-cell resolution over massive scales far beyond those permitted by the limited fields of view in conventional live-cell imaging based methods.

Living cells are intricate information processors. Unlike electronic computers that rely solely on electrical signals to process information, living cells use a complex network of interacting biochemical and biophysical signals to drive biological function and achieve biological computation. These signals include ion concentrations, molecular messenger levels, protein activities, and gene regulations. They form signal transduction networks and collectively convert cellular inputs into cellular outcomes by interacting in complex ways. Subtle defects in these processes are associated with a wide range of diseases. 

If the dynamic cellular building blocks were a collection of instruments in a symphony orchestra, listening to the sound from a single instrument, e.g. the calcium signal, is far from enough to fully appreciate and understand the ‘cellular symphony’. We discovered that ‘space’ can be used as new dimensions for multiplexed imaging and developed an entirely new way to image and record many biological signals in parallel from single cells. Using de novo designed self-assembling peptides, we engineered a toolbox of genetically encoded fluorescent reporters, where the reporters for different biological signals can spontaneously self-assemble into distinct clusters (Signaling Reporter Islands or SiRIs) randomly located throughout a given cell. The spatial separation of reporters enables accurate optical readout from individual reporters, even when the fluorescent spectra of the reporters overlap or are identical. We applied this new way of imaging to simultaneously record five different biological signals at the soma and neurites in neurons, breaking through the multiplexing limitation imposed by the number of available colors on light microscopes. 

We are developing novel ‘molecular focusing’ methods for precise neural activity imaging in large and dense cell populations in the living brain. Imaging artifacts from optical crosstalk among nearby neurons, known as the neuropil contamination, is a long-standing problem in neural activity imaging. To solve this problem, we developed soma-targeted fluorescent calcium indicators, SomaGCaMPs. The soma-targeting of the indicators was achieved by well-characterized self-assembling proteins from the protein engineering field and provided single-cell resolution optical crosstalk-free neural activity recording for the neuroscience community. We applied SomaGCaMPs to record neural activity over entire zebrafish brains and in deep and shallow brain regions in behaving mice at single-cell resolution. This soma-targeting approach demonstrated how the spatial precision of neural activity imaging can be boosted by the novel molecular focusing on top of the conventional optical focusing of light microscopes.