Design a site like this with
Get started


Spatial Navigation, Memory & Computing Mechanisms of Neurons

Ching-Lung Hsu Lab

Academia Sinica

Institute of Biomedical Sciences (IBMS)

Neuroscience Program of Academia Sinica (NPAS)

徐 經 倫 實 驗 室
中央研究院    空間認知、神經訊息整合與可塑性研究室

We are interested to bring cellular physiology to the big picture of systems neuroscience
for a better understanding of the power and the physical implementation of neuron computations that support memory and spatial cognition.

...we could say that creating a new theory is not like destroying an old barn and erecting a skyscraper in its place. It is rather like climbing a mountain, gaining new and wider views, discovering unexpected connections...

But the point from which we started out still exists and can be seen, although it appears smaller and forms a tiny part of our broad view gained by the mastery of the obstacles on our adventures way up.

 The Evolution of Physics (1938), Albert Einstein & Leopold Infeld

Background image credit: Julia Kuhl

About Us

General Background

Synaptic Integration for Context-dependent Spatial Coding in The Brain

How does the brain respond differently to small changes in the environment so quickly and robustly? Consider this scenario: you are in a hallway that looks identical to another hallway in the same building; simply relying on a tiny sign on the wall, however, you never have difficulty understanding right away where you are and how to get into a correct room from that hallway.

Mammalian brains are deep neural networks; the cognitive areas are typically many synapses (or ‘stages’) away from peripheral sensory organs. It is very challenging, in principle, for a small difference in the input to rapidly produce distinct effects in a higher-order region. However, there is some ‘magic’ existing in each individual neuron.

Pyramidal neurons, as a canonical principle cell type in all cortical areas, can be modeled as sophisticated combinations of small active and passive electronic devices. They could exhibit amazing nonlinear computational properties we have yet to fully understand. Therefore, one hypothesis is that the complexity of single-neuron integration and biophysics is a crucial basis for the rapidly flexible and robust computations described above.

I use place cells (in the hippocampus) and “relay” cells (in the thalamus)in rodents as model systems to address these problems. Moreover, the proposed cellular and circuit mechanisms for memory might also play important roles in the context-dependent neural coding, as suggested by some of my earlier and preliminary work. If it sounds exciting and interesting to you, welcome to join my lab and together explore this amazing field.




我的初步研究、和近年來許多研究顯示,神經元的非線性系統特性可能在輸入差異很小的情況下,產生極為不同的運算結果;與記憶相關的細胞與網路機制也可能在此扮演一個重要角色。我使用小鼠/大鼠的海馬迴place cells作為主要的研究模式系統。如果你/妳覺得很有趣,歡迎加入我的實驗室,一起探索這個美妙的世界。

Our Approach

Who are we? What matters to us?

The mission of Hsu lab is to do rigorous experiments, ask insightful questions, cultivate technical dexterity and intellectual independence, and communicate sciences through networking.

In short, we want to maximize each lab member’s potential.

We want to think more deeply in light of what an understanding of how the brain works means.

We promote a systems neuroscience view, in which an understanding involves explanations of the nervous system from the molecular, cellular, circuital and behavioral levels altogether, and both from a bottom-up, experimental/modeling angle and a top-down, theoretical angle.

To facilitate this, we hold a regular journal club in the lab. Different from common ones, it aims to provide a systematic review of literature in the historical context, covering from the very basic to more advanced findings.

Patch Clamp

Patch-clamp intracellular recordings of soma and dendrites in acute brain slices and in awake, behaving mice

Computational Modeling

Computational techniques & mathematical framework to understand how cells work and support more complex functions

Virtual Reality

Mouse behavior and navigation through VR environments

Expansion Microscopy

Highly adoptable technique, which overcomes diffraction limit, to combine with patch-clamp recording. An idea started at Janelia to continue

Virus-Assisted Circuit Analysis

Cell-type-based neural tracing, histology and immunohistochemistry

Live Cell Imaging

Establishing synaptic resolution live-cell imaging by kilohertz frame-rate two-photon microscopy. Find the methods here

Research Projects

Principles of synaptic integration

in the whole-neuron and behavioral context

Place cells in the hippocampus, which fire according to the animal’s position, can be created by nonlinear activity (called Ca2+ plateau potentials) and synaptic plasticity in the dendrites. Recently, we observed the strengthened synaptic response in awake, goal-directed spatial behavior is not always stable within the timescale of long-term synaptic potentiation (LTP). We are interested in the following questions: (1) what is the circuit and synaptic basis supporting goal-directed spatial navigation? (2) what are the cellular and biophysical mechanisms underlying place-cell emergence following the plateau potentials? (3) what governs the stability and reversibility of the strengthened response? (4) the profile of synaptic weights in a given cell is likely redistributed by a second plasticity induction event in induced place cells (Milstein et al., 2020); similar phenomenon was also observed for native place cells (Zhao, Hsu et al., in prep). What are the mechanistic principles and the computational implication? Our strategy is to use a combination of complementary approaches, including somatic/dendritic patch-clamp recording in acute brain slices, and in awake mice in virtual reality, assisted with optogenetic/chemogenetic methods, as well as computational modeling.

Circuit architecture

of intra- & extra-hippocampal synaptic inputs

Inhibitory interneurons serve important functions like gain modulation, homeostatic regulation, noise suppression, and possibly shape cellular feature selectivity. Despite the description of many interneuron subtypes, their exact role, especially in the context of multiple synaptic inputs from within and outside the hippocampus, is not clear. An important focus of the lab is the input from the prefrontal cortex, an executive and planning center, through the ventral midline thalamus called nucleus reuniens. How is the inhibition involved in this information flow in comparison with other inputs from entorhinal cortex or CA3? We will use anterograde and retrograde neural tracing techniques, with Cre mouse lines and new variants of AAVs and rabies viruses, to derive cell-subtype-based transsynaptic mapping. Together with what has been known about the interneuron connectivity to subcellular compartments of CA1 pyramidal (place) cell, it will depict the microcircuit architecture supporting different information streams.

Neural basis underlying memory-dependent,

goal-directed navigation in real-world and VR settings

The hippocampus is important for working and episodic memories, and place cells constitute a mapping system for guiding navigation behavior with specific goals. However, how the neurons respond to the animal’s position and particular task situations remains a big problem in the field. We want to start by using behavioral assays for freely moving mice in real mazes and head-restrained mice in virtual-reality mazes, combined with pharmacological and chemogenetic perturbation of neural circuits, to address this problem. There have been several different aspects of information considered to support the place code, and we are particularly interested in the precise role of proprioception, the sense of limb position and movement, in the brain navigational system and the encoding of space. The contribution of proprioception is very much understudied because it is hard to accurately manipulate. We will collaborate with Dr. Chih-Cheng Chen at IBMS, taking advantage of the transgenic methods from his lab to investigate how proprioception influences the performance of memory-dependent spatial tasks. We are generally interested in the role of the ventral midline thalamus in goal-directed navigation with memory demands, given the projection of reuniens to dorsal CA1 place-cell dendrites. For the proprioception project, in particular, the ventral thalamus is also a candidate target because of its relation to central proprioceptive pathways. Eventually, electrophysiological or imaging can be applied to the VR setting to reveal the underlying physiological basis.

Context-dependent neuron computation

with extensive mapping at synaptic resolution

Pyramidal neurons, a canonical cell type in essentially all cortical areas, have complex morphological and biophysical properties and can be modeled as a computing device of multiple subunits. It might be important to some characteristics of intelligent behavior. In solving the input-output transformation of neural computations, the critical aspects are not only the dendritic integration problem, which could be addressed by the approaches mentioned above, but also the detailed structure of synaptic inputs in space and in time, and the information they transfer—which could be studied only by mapping inputs to individual neurons, with sufficient power for synaptic resolution and large, 3D field-of-view, in more physiological situations. Dr. Kaspar Podgorski at Janelia will assist us with establishing a new imaging method, called SLAP, to achieve these goals (see Collaborators). There have been several great sensors developed, with different colors, suitable for the laser wavelength used by this microscope. They are used to measure calcium, glutamate, acetylcholine etc. in fine spatio-temporal scales (such as this and this). In particular, a crucial aspect of this project is to combine the imaging with patch-clamp recording in brain slices, the gold standard of investigating synaptic integration, in conditions approximating the situations when multiple, complex inputs or synaptic plasticity occurs.

We are interested in the problems of how inputs carrying different information to distinct locations of the dendritic trees collectively determine the emergence of context-dependent spatial coding during goal-directed navigation behavior. In addition, how specific and prevalent are the effects of the place-cell inducing dendritic plasticity in the re-distribution of synaptic strengths in an entire neuron’s context? And what are the implications of the answers to these questions on the computing capabilities of neurons? Eventually, we plan to ask these questions by using the imaging method together with currently developing 2-photon genetically encoded voltage indicators in awake, navigating mice.

Virtual reality for a head-restrained mouse

Our new VR system for head-restrained animals. The mouse can turn in Y-shaped virtual space

Anterograde one-step transsynaptic circuit tracing, using recombinant viral tools. Only granule cells of the dentate gyrus that received specific inputs were labeled.

Super-resolution microscopy
for molecular analysis of synaptic connectivity assisted with patch-clamp recording
(red: Homer; green, Bassoon;

expansion by Boaz Mohar and Paul Tillburg)

Electrophysiological recording in the hippocampus of a mouse running in auditory virtual reality to find location-dependent rewards (auditory frequency changed with virtual position)

Our new VR system for head-restrained animals. The mouse can run in virtual corridor of infinite length


Ching-Lung Hsu

徐經倫 博士



Ching-Lung had a B.Sc. of Zoology (major) and Electric Engineering (minor) and received his Ph.D. from National Taiwan University. After that, he moved to Janelia Research Campus of HHMI to work as a Postdoc and then a Research Scientist.

His lifelong interests include to understand how the algorithmic properties required for dynamic cognitive processes can be supported by the computations of individual neurons. He is enthusiastic at the idea of pursuing neurosciences from both experimental and theoretical perspectives.

Assistant Research Fellow

Academia Sinica

Institute of Biomedical Sciences (IBMS)

Neuroscience Program of Academia Sinica (NPAS)

Adjunct Assistant Professor

National Taiwan University

Department of Life Science, College of Life Science

Wan-Ting Teresa Liao


Research Associate

M.Sc. in Neuroscience
National Taiwan University, School of Veterinary Medicine

B.Sc. in Medical Biotechnology and Laboratory Science
Chang Gung University

Teresa’s research will focus on characterizing the neuronal circuitry between the hippocampus and thalamus by viral neuronal tracing. Her further work is to link targeted neuronal subtypes with their circuit connectivity and function in spatial memory using patch-clamp techniques.

Ching-Tsuey Jessica Chen


Research Associate

M.Sc. in Life Science
National Taiwan University

B.Sc. in Pharmacy with minor in Neurobiology and Cognitive Science
National Taiwan University, School of Pharmacy

Jessica’s research is focusing on detailed mechanisms of synaptic plasticity capable of supporting new place cells and episodic memory, using patch clamp in acute brain slices. She also plans to combine such technique with new super-resolution approach to correlate electrophysiology with protein organization following synaptic plasticity induction.


Kaspar Podgorski | Allen Institute for Neural Dynamics

Kaspar is helping us set up novel random-access two-photon microscopy with micrometers (spatial) and milliseconds (temporal) resolution, SLAP2. Find the SLAP1 paper. Here is the detector to use. Kaspar and Janelia GENIE team also created many new biological sensors.

Paul Tillberg | Janelia Research Campus, HHMI

Paul is the creator of expansion microscopy that overcomes the diffraction limit of light microscopy. See the new method.

Boaz Mohar | Janelia Research Campus, HHMI

Boaz is an expert of neurophysiology and imaging techniques. We are interested to apply expansion microscopy with protein labeling to well-defined synapses undergoing robust LTP in brain slices.

William Kath | Northwestern University

I’ve worked with Aushra Abouzeid in Bill’s lab for developing genetic algorithms to perform parameter fitting of neuron models to electrophysiological data. This work started when I was in Nelson Spruston’s group.

Past Collaborators

Dezhe Jin | Penn State University

Jianglai Wu, Ji Na | Janelia Research Campus; UC Berkeley

Francesco Piccolo (Nathaniel Heintz) | Rockefeller University


Benjamin Liu (劉品喆)

B.A. in Integrative Biology with minor in Data Science
UC Berkeley


Recent Publications

Kilohertz two-photon fluorescence microscopy imaging of neural activity in vivo

Mar 2, 2020

Understanding information processing in the brain requires monitoring neuronal activity at high spatiotemporal resolution. Using an ultrafast two-photon fluorescence microscope empowered by all-optical laser scanning, we imaged neuronal activity in vivo at up to 3,000 frames per second and submicrometer spatial resolution. This imaging method enabled monitoring of both supra- and subthreshold electrical activity down to 345 μm below the brain surface in head-fixed awake mice.

Persistent Sodium Current Mediates the Steep Voltage Dependence of Spatial Coding in Hippocampal Pyramidal Neurons

Jul 11, 2018

The mammalian hippocampus forms a cognitive map using neurons that fire according to an animal’s position (“place cells”) and many other behavioral and cognitive variables. The responses of these neurons are shaped by their presynaptic inputs and the nature of their postsynaptic integration. In CA1 pyramidal neurons, spatial responses in vivo exhibit a strikingly supralinear dependence on baseline membrane potential. The biophysical mechanisms underlying this nonlinear cellular computation are unknown. Here, through a combination of in vitroin vivo, and in silico approaches, we show that persistent sodium current mediates the strong membrane potential dependence of place cell activity. This current operates at membrane potentials below the action potential threshold and over seconds-long timescales, mediating a powerful and rapidly reversible amplification of synaptic responses, which drives place cell firing. Thus, we identify a biophysical mechanism that shapes the coding properties of neurons composing the hippocampal cognitive map.

Dendritic sodium spikes are required for long-term potentiation at distal synapses on hippocampal pyramidal neurons

Aug 6, 2015

Dendritic integration of synaptic inputs mediates rapid neural computation as well as longer-lasting plasticity. Several channel types can mediate dendritically initiated spikes (dSpikes), which may impact information processing and storage across multiple timescales; however, the roles of different channels in the rapid vs long-term effects of dSpikes are unknown. We show here that dSpikes mediated by Nav channels (blocked by a low concentration of TTX) are required for long-term potentiation (LTP) in the distal apical dendrites of hippocampal pyramidal neurons. Furthermore, imaging, simulations, and buffering experiments all support a model whereby fast Nav channel-mediated dSpikes (Na-dSpikes) contribute to LTP induction by promoting large, transient, localized increases in intracellular calcium concentration near the calcium-conducting pores of NMDAR and L-type Cav channels. Thus, in addition to contributing to rapid neural processing, Na-dSpikes are likely to contribute to memory formation via their role in long-lasting synaptic plasticity.




Interested in joining our lab?


Website structure: Benjamin Liu