The long-term goal of our laboratory is to develop, scale up, and broadly disseminate molecular technologies for mapping cells and functional circuits. At the sub-cellular scale, maps document the spatial organization of proteins, RNA, DNA, and metabolites with nanometer precision and temporal acuity on the order of seconds. Maps also chart the connectivity between these molecules, elucidating the circuits and signaling processes that give rise to function. 

Beyond the single cell, we also strive to map cellular ensembles, such as brain tissue. Can we create tools that contribute to the construction of cell and tissue atlases, and can we map the cellular circuits that give rise to function and behavior? To achieve these ambitious goals, our laboratory has focused on the development of scalable technologies to detect, measure, and manipulate molecules and circuits, both at the sub-cellular level, and at the level of cell populations.


One of the most commonly asked questions in biomedical research is: what is interacting with or in the vicinity of my molecule or organelle of interest? Existing technologies to address this question are plagued with false negatives (e.g., immunoprecipitation) or false positives (e.g., biochemical fractionation and crosslinking), or not even applicable to certain cellular structures (e.g., RNA granules). More generally, there is a pressing need for a simple, generalizable technology to map the protein interaction network or proteomic composition of specific target molecules or organelles of interest, including unpurifiable ones. To address this need, our laboratory developed enzyme-mediated proximity biotinylation. A peroxidase that we engineered by directed evolution, APEX2, is genetically targeted to a cellular region of interest. Addition of a small molecule substrate for APEX2, biotin-phenol, results in the covalent biotinylation of endogenous proteins within 1-10 nm of APEX over a 1 minute time window in living cells. After labeling, cells are lysed, and biotinylated proteins are harvested and identified by mass spectrometry.

Using this approach, our laboratory produced the most specific proteomic maps of two human mitochondrial subcompartments (the matrix and intermembrane space) to date. We also mapped the proteomes of the excitatory and inhibitory synaptic clefts in living neurons for the first time. These structures are impossible to purify or enrich, but crucially important for neuroscience. Our proteomes resulted in the discovery of 76 novel mitochondrial proteins, 33 novel synaptic proteins, and a new mechanism for inhibitory synapse specification. Other laboratories have applied APEX proteomic technology to the primary cilium, ER-plasma membrane contact sites, C. elegans specialized cell types, Drosophila mitochondria, microsporidia infection, and GPCR signaling.


We also strive to develop technologies that can be used to map cell populations. One of the greatest technological challenges in modern neuroscience is how to identify the ensemble of neurons that encode or control a specific memory, behavior, or emotional state of interest – the so-called “memory trace”. To address this challenge, we are building molecular tools that can permanently “tag” neurons active during a specific time window defined by light. A separate effort focuses on the visualization and manipulation of specific synaptic contacts, useful for dissecting the circuitry of neuronal networks.


Our laboratory has used protein engineering and chemical synthesis to create a suite of tools that push the boundaries of protein imaging in cells. For example, we introduced APEX peroxidase as a genetically-encoded tag for electron microscopy. Analogous to GFP for fluorescence microscopy, APEX reveals the nanometer-scale localization of specific tagged proteins or organelles in electron micrographs. 

We have also developed a suite of fluorophore ligases and accompanying chemical probes that permit site-specific tagging of cellular proteins modified only by a 13-amino acid recognition peptide, rather than GFP, SNAP tag, or HaloTag which are ~20 times larger and much more perturbative. We developed small, monovalent, targetable quantum dots to empower single molecule biophysicists interested in much brighter, photostable, protein-specific, and live cell-compatible probes of manageable size and without induction of protein crosslinking.

Finally, we developed several technologies aimed at detection and measurement of cellular protein-protein interactions. These include ID-PRIME and BirA-mediated proximity biotinylation, as well as our recent split HRP which was used for highly sensitive visualization of trans-synaptic protein-protein interactions in the in vivo mammalian brain. 


We are applying our technologies to address several fundamental questions: How does the mitochondrion replicate and divide? How does the mitochondrion coordinate its activities with those of other cellular organelles, such as the endoplasmic reticulum? How does the mitochondrion build itself and replenish its proteins, the vast majority of which are encoded by the nuclear genome?