Richi Sakaguchi

Fluorescence imaging is widely used for the mesoscopic mapping of neuronal connectivity. However, neurite reconstruction is challenging, especially when neurons are densely labelled. Here, we report a strategy for the fully automated reconstruction of densely labelled neuronal circuits. Firstly, we establish stochastic super-multicolour labelling with up to seven different fluorescent proteins using the Tetbow method. With this method, each neuron is labelled with a unique combination of fluorescent proteins, which are then imaged and separated by linear unmixing. We also establish an automated neurite reconstruction pipeline based on the quantitative analysis of multiple dyes (QDyeFinder), which identifies neurite fragments with similar colour combinations. To classify colour combinations, we develop unsupervised clustering algorithm, dCrawler, in which data points in multi-dimensional space are clustered based on a given threshold distance. Our strategy allows the reconstruction of neurites for up to hundreds of neurons at the millimetre scale without using their physical continuity.

At present, in vitro phenotypic screening methods are widely used for drug discovery. In the field of epilepsy research, measurements of neuronal activities have been utilized for predicting efficacy of anti-epileptic drugs. Fluorescence measurements of calcium oscillations in neurons are commonly used for measurement of neuronal activities, and some anti-epileptic drugs have been evaluated using this assay technique. However, changes in waveforms were not quantified in previous reports. Here, we have developed a high-throughput screening system containing a new analysis method for quantifying waveforms, and our method has successfully enabled simultaneous measurement of calcium oscillations in a 96-well plate. Features of waveforms were extracted automatically and allowed the characterization of some anti-epileptic drugs using principal component analysis. Moreover, we have shown that trajectories in accordance with the concentrations of compounds in principal component analysis plots were unique to the mechanism of anti-epileptic drugs. We believe that an approach that focuses on the features of calcium oscillations will lead to better understanding of the characteristics of existing anti-epileptic drugs and allow to predict the mechanism of action of novel drug candidates.

Fluorescence imaging is widely used for the mesoscopic mapping of neuronal connectivity. However, neurite reconstruction is challenging, especially when neurons are densely labelled. Here we report a strategy for the fully automated reconstruction of densely labelled neuronal circuits. Firstly, we established stochastic “super-multicolour” labelling with up to seven different fluorescent proteins using the Tetbow method. With this method, each neuron was labelled with a unique combination of fluorescent proteins, which were then imaged and separated by linear unmixing. We also established an automated neurite reconstruction pipeline based on the quantitative analysis of multiple dyes (QDyeFinder). To classify colour combinations, we used a newly developed unsupervised clustering algorithm, dCrawler, in which data points in multi-dimensional space were clustered based on a given threshold distance. Our new strategy allows for the reconstruction of neurites for up to hundreds of neurons at a millimetre scale without manual tracing.

At present, in vitro phenotypic screening methods are widely used for drug discovery. In the field of epilepsy research, measurements of neuronal activity have been utilized for predicting efficacy of anti-epileptic drugs. Fluorescence measurements of calcium oscillations in neurons are commonly used for measurement of neuronal activities, and some anti-epileptic drugs have been evaluated using this assay technique. However, changes in waveforms were not quantified in previous reports. Here, we have developed a high-throughput screening system containing a new analysis method for quantifying waveforms, and our method has successfully enabled simultaneous measurement of calcium oscillations in a 96-well plate. Features of waveforms were extracted automatically and allowed the characterization of some anti-epileptic drugs using principal component analysis. Moreover, we have shown that trajectories in accordance with the concentrations of compounds in principal component analysis plots were unique to the mechanism of anti-epileptic drugs. We believe that an approach that focuses on the features of calcium oscillations will lead to better understanding of the characteristics of existing anti-epileptic drugs and allow prediction of the mechanism of action of novel drug candidates.

In the mouse olfactory bulb, sensory information detected by ∼1,000 types of olfactory sensory neurons (OSNs) is represented by the glomerular map. The second-order neurons, mitral and tufted cells, connect a single primary dendrite to one glomerulus. This forms discrete connectivity between the ∼1,000 types of input and output neurons. It has remained unknown how this discrete dendrite wiring is established during development. We found that genetically silencing neuronal activity in mitral cells, but not from OSNs, perturbs the dendrite pruning of mitral cells. In vivo calcium imaging of awake neonatal animals revealed two types of spontaneous neuronal activity in mitral/tufted cells, but not in OSNs. Pharmacological and knockout experiments revealed a role for glutamate and NMDARs. The genetic blockade of neurotransmission among mitral/tufted cells reduced spontaneous activity and perturbed dendrite wiring. Thus, spontaneous network activity generated within the olfactory bulb self-organizes the parallel discrete connections in the mouse olfactory system.

The stochastic multicolor labeling method ‘Brainbow’ is a powerful strategy to label multiple neurons differentially with fluorescent proteins; however, the fluorescence levels provided by the original attempts to use this strategy were inadequate. In the present study, we developed a stochastic multicolor labeling method with enhanced expression levels that uses a tetracycline-operator system (Tetbow). We optimized Tetbow for either plasmid or virus vector-mediated multicolor labeling. When combined with tissue clearing, Tetbow was powerful enough to visualize the three-dimensional architecture of individual neurons. Using Tetbow, we were able to visualize the axonal projection patterns of individual mitral/tufted cells along several millimeters in the mouse olfactory system. We also developed a Tetbow system with chemical tags, in which genetically encoded chemical tags were labeled with synthetic fluorophores. This was useful in expanding the repertoire of the fluorescence labels and the applications of the Tetbow system. Together, these new tools facilitate light-microscopy-based neuronal tracing at both a large scale and a high resolution.