Samsung Science & Technology Center

The objective of research in the field of neuroscience is to delve into a more profound comprehension of the brain's functioning, encompassing its role in shaping behavior, cognition, and emotions. This pursuit of knowledge is pivotal as it serves as the foundation for the development of novel treatments and therapies targeting brain related disorders and diseases. Significantly, considerable progress has been made in refining techniques and tools, particularly in the realms of molecular, cellular, circuit, and system-level investigations. However, there has been relatively little emphasis on the measurement and quantification of the functioning (behavioural phenotype and other), despite the overarching aim of brain science being to grasp the intricacies of brain-related phenomenon and, subsequently, applying this understanding to clinical assessments, particularly in the context of drug trials. Moreover, it is well-established that the brain is significantly shaped by environmental factors and interpersonal interactions. Yet, it's noteworthy that there has been a scarcity of focus on these aspects within experimental settings. In this presentation, I will advocate for the importance of creating methods to assess the phenotype and designing experimental frameworks aimed at exploring the impact of environmental and social interactions.

My lab is interested in how sensory information regulates behaviors that maintain homeostasis. We focus on ingestive and thermoregulatory behaviors, the two interrelated behaviors that can be parametrically controlled by mechanical and thermal stimuli. Specifically, we employ optogenetic and pharmacogenetic manipulations, two-photon Ca2+ imaging and fiber photometry, tracing and molecular tools, surgical approaches, and internal organ manipulations to identify and characterize the neural circuits underlying these behaviors in both central and peripheral nervous systems, throughout the brain-body axis. We have recently found that Pdyn+ neurons in the parabrachial nucleus monitor the intake of both fluids and solids using mechanosensory signals arising from the upper digestive tract and mediate negative feedback control of ingestion (Kim et al., Nature 2020; also see Kim et al., Nat Rev Neurosci 2022). In parallel, we also identified Vgat+ neurons in the lateral hypothalamus as a neural substrate necessary for diverse thermoregulatory behaviors (Jung et al., Neuron 2022). Using these as entry points, we are now concentrating on defining the additional circuit components for ingestive and thermoregulatory behaviors and understanding the precise encoding of information by these neurons at the single-cell level.

In parallel, leveraging our accumulated expertise in surgical targeting and manipulation of various vagus nerve branches, we are currently seeking to achieve organ and tissue-specific vagus nerve stimulation (VNS). The vagus nerve plays a vital role in the body-brain axis, facilitating the two-way communication between various internal organs and the brain. VNS, a form of neuromodulation that involves the stimulation of the vagus nerve, has emerged as a potentially effective therapeutic approach for the treatment of many neurological, psychiatric, and metabolic disorders. Notwithstanding its purported benefits, the present methodologies for VNS rely on electrical stimulation that non-selectively activates all vagal afferent and efferent fibers innervating various organs, resulting in unintended adverse effects.

The first part of this talk will discuss our recent progress in the study of homeostasis maintaining behaviors and its neural circuit underpinnings focusing on body brain communication. Then, I will overview the history and current of vagus nerve stimulation strategies and discuss the vision for the selective stimulation of vagal afferents with specific target organs, as well as related ongoing works in my laboratory.

3D Neural interfaces are emerging as platforms for brain-computer interfaces (BCI) that not only enable new forms of interaction with technology, but also act as therapeutic prostheses for tetraplegic patients. Neural interfaces are currently facing potential widespread application in the public with ventures such as Neuralink. However, the invasiveness and limited lifetime due to chronic immune response of neural interfaces has acted as the major bottleneck in their widespread usage. In order to address the issues, we propose a fully-bio-integrated neural probe via a two-part strategy: [1] stealth neuron-like fiber and [2] a brain-ECM mimetic hydrogel for active regeneration for next-generation clinical neural interfaces. Firstly, the stealth fibers are three-dimensionally distributed to enable tissue regeneration to occur between the fibers with tissue-like bending stiffness (from neuron-scale size and tissue-like modulus). These stealth fibers are embedded in a regenerative gel, which enables neuron migration between the fibers and have a multistage drug release mechanism to minimize critical stages in chronic gliosis formation. Additionally, this project aims to extend the application of multifunctional neural probes for clinical treatment by integrating key functional aspects that have acted as translational barriers (wireless connectivity, large animal testing) in human usage. Through this fully-biointegrated neural interface, we aim to drastically reduce the traumatic healing process of neural probes and ensure the functional integration of regenerated neuronal tissue for minimal disabilities or surgical complications after implantation.

Non-invasive technologies for neuromodulations and brain drug deliveries are essential to daily maintenance of brain health. Although the uses of currently available approaches of stimulating brain for neuromodulations or opening blood-brain-barrier for drug delivery show decent feasibilities in brain therapeutics, the reliability of the non-invasive neurostimulation needs to be further investigated or the dimension of operational systems is too large for patients to access the facility everyday. Therefore, omni-neurotheragnosis are defined as a group of technologies to provide brain therapy in any place and time, and the technology implies a combination of controllable neuromodulations, real-time image guidance and stimulation and portability or miniaturizations. Among many stimulation and imaging methods, ultrasound technologies should be appropriate for omni-neurotheragnosis because transcranial ultrasonic stimulations have benefits in delivering spatially precise stimulations to arbitrary lesions in brain parenchyma with a reduced stimulatory and imaging system form factor. In this presentation, the current technologies for ultrasonic neurostimulations are reviewed, and the essential research topics on ultrasonic omni neurothegnosis are proposed.

For decades, Alzheimer’s disease (AD) researchers have proposed that brain amyloid beta (Aβ) deposition is the primary pathophysiological event that somehow triggers tau neurofibrillary tangles to spread beyond the medial temporal lobe and out into the heteromodal neocortex. This hypothesis remains in doubt, however, because: (1) Aβ and tau deposition begin within spatially disparate brain regions and (2) Aβ lowering agents have shown modest and inconsistent benefits, at best, in AD clinical trials. In our recent research work, we identified two pivotal, region-specific Aβ-tau interactionns that explain the topographical dissimilarity between Aβ and tau and provided a roadmap for enhanced patient selection in AD clinical trials. Although this work focuses on the fundamental molecular anatomy of AD, it also has major potential clinical implications. Here, in this talk, I seek to identify who is most likely to benefit from Aβ-lowering drugs by applying the connectivity-based Aβ-tau interaction mapping algorithm, and to improve the clinical efficacy of Aβ-lowering drugs at the individual level in the treatment arm. This will increase the probability of success in future AD clinical trials.

Mapping the entire neuronal circuitry throughout the body at single-neuron resolution has been one of the most desired goals in neuroscience. However, multiple hurdles have prevented the achievement of this goal in vertebrates, especially in large vertebrates such as mice. The biggest hurdle in achieving this goal is resolution. As neurites form bundles in the peripheral nervous system, resolving individual neurites in such dense bundles requires sub-100-nm resolution over the entire body. Such high resolution can be achieved through various imaging methods, including electron microscopy and super-resolution optical microscopy. However, imaging the entire vertebrate bodies with such high resolution has been impossible until now. In this talk, I will introduce our latest work on expansion microscopy imaging of whole vertebrate bodies, which enables sub-100-nm resolution imaging of the entire vertebrate bodies within a reasonable timeframe. Our technique allows for the 4-fold expansion of mouse embryos. When imaged with high-NA long-working distance objectives, it enables sub-100-nm resolution imaging of the entire vertebrate bodies. I will demonstrate that this technique is compatible with various molecular labelling techniques, showing the potential for tracing individual neurons throughout the entire bodies. During the talk, I will discuss various efforts made in this field and the challenges and hurdles that need to be overcome in scaling up this technique. More specifically, I will discuss the optical microscopy system needed for accelerating the imaging of large specimens and automated data analysis enabling fast automated detection of neurites and tracing throughout the entire bodies. I will also discuss the state-of-the-art molecular labelling techniques that can be combined with our technique and new labelling techniques that need to be developed to maximize the potential of our technique. Finally, I will discuss how this technique could be used in the future once it is further developed.

Current brain organoid technology fails to provide adequate patterning cues to induce a mature structure that represent the complexity of the human brain. Recently, we developed a module-based cellular reconstitution technology to sequentially build uniform forebrain assembloids with mature cortical structures and functional connectivity. The uniformity and maturity of the newly-conceived forebrain assembloids were achieved by creating single-rosette-based organoids at the early stage, whose sizes were big and consistent with the treatment of Wnt and Hedgehog agonists, followed by spatial reconstitution with the Reelin-expressing neuronal layer and non-neuronal glial cells. The resulting single-rosette-based forebrain assembloids were highly uniform and reproducible without significant batch effects, solving major heterogeneity issues caused by difficulties in controlling the number and size of rosettes in conventional multi-rosette organoids. Furthermore, these forebrain assembloids structurally and functionally recapitulated the physiology of the human brain, including the six-layered cortical structure, functional connectivity, and dynamic cellular interplay between neurons and glial cells. Using patient-derived forebrain assembloids, our study further elucidated the nature of the cellular interplay between neurons and glial cells, mediated by two signalling axes, in human schizophrenia brains, and suggested an innovative model to study human neurodevelopmental disorders, understanding the pathogenesis of which requires an organoid system capable of representing the dynamic cellular interactions of the human brain.

The development of technology to generate brain organoids from human induced pluripotent stem cells (hiPSCs) allows the utilization of these in vitro models for human brain research, which has led to the advancement in knowledge of neurological disorders. However, there are still many limitations in these models that hamper precisely elucidating mechanisms of pathogenesis and developing efficient therapeutics for diseases: such as difficulty in implementing a vascular structure to support the long-term survival of cells in organoids, incompleteness in mimicking the complexity of the brain and diversity in different regions/cell types, and impossibility of behavioral assessment. In this talk, I will share current works using hiPSCs-derived brain organoids to study neurological disorders and discuss the strategy to overcome their shortcomings.

Human induced pluripotent stem cell (iPSC)-derived brain organoids have garnered substantial attention as a promising model for drug screening, given their close resemblance to the human brain. Particularly noteworthy is the recent advancement in the development of disease model organoids sourced from patients, amplifying expectations for applications in drug screening and personalized medicine. Various platforms incorporating recording electrodes have been introduced to evaluate the functionality of brain organoids, revealing promising outcomes in recording electrophysiological signals[1-3]. In contrast to surface-type electrodes, which are affixed to the organoid's surface for signal recording, penetrating electrodes offer the capability to analyze the internal functional connectivity of brain organoids. Moreover, electrodes with stimulation capabilities enable the modulation of specific regions within a brain organoid. Our innovative organoid-based one-step drug screening system facilitates the exploration of functional changes induced by different drug doses across multiple brain organoids. The system comprises a culture chamber housing a microfluidic chip capable of accommodating 10 organoids, delivering distinct doses of two drugs to each organoid. Additionally, we integrate a 3D multi-electrode array (MEA) with 10 shanks, enabling the simultaneous evaluation of the functionality of 10 brain organoids. In this presentation, I will delve into the evolving landscape of systems designed for the functional assessment of brain organoids and discuss the prospective directions for this pioneering system.

The elderly population aged over 65 years gradually constitutes a major proportion of the global population and life expectancy has also continued to rise. As life expectancy continues to rise, ageing and age-related diseases are issues to address for maintaining a healthy lifestyle. Brain ageing, which is characterized by a progressive decline in memory and cognition, is recognized as the greatest risk factor for neurodegenerative diseases. For last few decades, animal models have extensively used for aged models; however, aging of animal is different from that of human in terms of genetics and metabolism. In addition, two-dimensionally (2D) cultured cells are also used for studying intracellular changes upon the aging-related stimuli; however, 2D cultured cells could not reflect the cell-cell and cell-microenvironment interactions observed in vivo. Recently, the brain-on-a-chip platform has risen as one of the human brain avatar model for recapitulating human brain physiology and pathology. In this presentation, we present a concept to utilize a brain-on-a-chip model for studying aging and further, reverse brain aging. To this end, preliminary aged brain models based on brain-on a-chip platform are presented with a particular focus on the mimicry of aged brain microenvironment and pathological features. Next, possibility of studying reverse brain aging by using brain-on-a-chip will be discussed.

[Part 1] Often known as the glymphatic system, the brain clearing system is vital for sustaining brain health, chiefly by expediting the removal of waste elements. It is pivotal for the swift removal of soluble proteins and metabolites, facilitating the distribution of compounds including glucose, lipids, amino acids, and neurotransmitters within the brain. However, as individuals age, this system's efficiency notably declines, resulting in the accumulation of waste in the brain, potentially leading to cognitive dysfunction and heightened risk of neurodegenerative ailments. In the following discussion, the spotlight will be on the formulation of innovative therapeutics aimed at rejuvenating the aging brain clearing system. When used in tandem with traditional Alzheimer’s antibody medications, these novel drugs may offer promising outcomes. Effective drug screening for clinical success necessitates the employment of sophisticated humanized models that accurately reflect the distinctive functions of the brain clearing system and the aging process. In this regard, I propose the development of microphysiological systems specific to brain clearing—also known as 'organ-on-chips', which may hold breakthroughs in both fundamental and applied scientific research.

[Part 2] The failure of most central nervous system (CNS) drugs to penetrate the market can be attributed to their inability to traverse the blood-brain barrier (BBB), a challenge that complicates the delivery of most antibody drugs and therapeutic stem cells to the CNS. This situation presents a considerable obstacle for pharmaceutical entities endeavoring to formulate drugs for the CNS. In this segment, I will introduce a strategy for identifying effective brain drug delivery systems. This approach leverages microphysiological systems and employs screening technologies such as CRISPR-Cas screening, SELEX, and phage display. Furthermore, it incorporates predictive modeling through artificial intelligence (AI), providing a comprehensive strategy for advancing drug delivery to the brain.

Modeling brain development and neuronal disorders remains a great challenge due to the complexity of the structure, physiology, and function of human brain. Several in vitro models have been established for brain disease modeling and drug discovery. However, they are often limited in recapitulating complex cellular/extracellular interactions and physiological aspects in human brain. To overcome such limitations of current models, we have developed a brain-on-a-chip with organoids and microfluidics. Human brain organoids generated with brain extracellular matrix and microfluidic device could reconstitute structural and physiological features of neuroepithelium development and corticogenesis, which can be used for neurodegenerative disease modeling. We have also established neurovascular unit chip with functional blood brain-barrier for modeling infectious brain diseases. Our strategy using brain-on a-chip with cerebral organoids and neurovascular units would be able to facilitate elucidation of brain disease mechanisms and development of novel therapeutics. In this presentation, recent advances and future perspectives of human brain models will be discussed.

Brain-computer interfaces (BCIs) have been actively developed as a non-verbal way of communication, particularly for the disabled or the paralyzed. Recently, with the rapid development of artificial intelligence (AI) and wearable technology, increasing interest has been drawn toward technologies that allow users to monitor their own brain states with wearable EEG recording devices. However, the currently available commercial EEG devices do not recognize the surrounding environment or activities of the users. In this talk, I would like to propose a new brain-AI interface (BAI) to implement a “super-intelligent personal assistant.” The proposed BAI system recognizes the user’s ambient environment or activities based on computer vision technologies while simultaneously collecting their EEG data to estimate their current brain or mental states. For example, if a virtual assistant system were to learn that the user is currently studying mathematics and their concentration is gradually decreasing, the assistant could suggest to the user whether studying Korean language might be a good idea. More use cases as well as detailed implementation methods would be presented at the conference. To implement a more practical personal assistant, the development of wearable EEG devices with convenient and stylish designs is necessary, an interesting topic that I would like to discuss at the conference.

Traditionally, in order to analyze biosignals, electrodes are attached to a predetermined area, and diseases can be diagnosed or health status monitored through biosignals measured only at that location. However, biosignals can also be measured at other points, in which case meaningful interpretation is difficult. The reason for this is that diseases have been diagnosed through statistical analysis methods and a lot of data has been accumulated. For example, in the case of electrocardiograms, in 1911, before the development of electrodes, both hands and feet were immersed in a water tank and used as electrodes. Because diagnostic data has been accumulated through measured signals, signals at new points become meaningless. On the other hand, now, over 120 years later, new measuring devices based on MEMS and ICT technology have been developed, and even the form of electronic skin has been developed, making various services using biological signals possible. In order to give clinical meaning to this, a new standard was needed, and to propose this standard, we developed an approach that broke away from measuring the potential difference at two points and measured it only at a single point, and explored its usefulness. I would like to discuss this. If measurement at a single point is successful, it is expected that differences in specifications will be reduced, and the simplicity of measurement will be dramatically improved, suggesting a new paradigm for bioelectrical signal measurement devices, especially brain wave signal analysis. It is judged to be very useful in this regard.

The glymphatic system is a recently discovered waste clearance system in the brain, and it plays a crucial role in removing metabolic waste products and other toxins from the central nervous system. Dysfunction in the glymphatic system has been linked to various neurological conditions, including Alzheimer's disease. However, therapeutic methods specifically designed to target the glymphatic system are still an area of ongoing research and development. In this presentation, we propose a novel approach involving electroceuticals to enhance the glymphatic system by facilitating the flow of cerebrospinal fluid (CSF) through the brain's spaces. Additionally, I will briefly discuss the potential application of this approach in the treatment of neurological disorders.

Neuromodulation technologies, including optogenetics, have been essential in investigating the roles of neurons and their connectivity and brain functions. However, due to light scattering, optogenetic approaches typically require the need for invasively implanted optical fibers to control the deep brain neural activity. Such wired systems significantly interfere with animals’ natural movements and social interactions, posing challenges for studying brain circuits underlying behaviors and neurological disorders. Magnetogenetic (MG) and sonogenetic (SG) neuromodulation, in principle, have the merit for wireless and deep tissue stimulation with both magnetic fields and ultrasound can penetrate deep tissues with negligible perturbations. However, they have been less widely accepted by neuroscience community due to the poor understanding of the working principles and limited in vivo applications.

Recently developed MG systems showed new possibilities by using nanoscale magnetic actuators for activation of genetically encoded ion channels and remote stimulation of deep brain neurons in intact, freely moving animals. Here, I will introduce our on-going development of MG towards versatile neurotechnology: (1) cell-type specific MG for neural circuit research, (2) molecular toolbox for MG neural silencing, and (3) implant-free, anatomically selective deep brain stimulation for effective treatment of neurodegenerative diseases. Then, I will discuss the possible mechanisms and the current progress of sonogenetic control of genetically encoded neurons towards potential clinical translation. With the advantages of magnetism and ultrasound, these emerging neurotechnologies provide non-invasive control of the targeted neurons with spatiotemporal precision, allowing the interrogation of brain mechanisms underlying cognition and sociality. The development of these techniques towards larger animals (i.e., primates) and human has the potentials to revolutionize our understanding and treatment of various neurological disorders, including Parkinson’s disease, epilepsy, addiction, and depression.

The realm of implantable devices is garnering significant attention for its capacity to execute functions related to patient diagnosis and treatment over an extended period after a single implantation. However, a fundamental challenge plagues many of these implantable devices due to the unique environmental conditions within the body, rendering long-term reproducible use difficult. Consequently, it has often proven challenging to establish a clear rationale for invasive implantation procedures. If we were to compare the unique environment inside the body to a new celestial body, most implanted devices can be likened to astronauts wearing spacesuits that protect the astronauts from the outer environment, surviving on air and sustenance self contained in the spacesuits or unreasonably supplied from outside the star. Indeed, many implantable devices are unable to sustain themselves independently within the body, which is why a permanent use remains impossible for a significant number of them. In nature, there are organisms like parasites that actually survive and thrive in new and challenging environments. They adapt to survive by concealing themselves on surfaces, neutralizing surrounding threats, and consuming substances in their vicinity as sustenance, essentially leading a form of "lifestyle" for their survival. Certainly, many of these parasites are harmful, as they often kill their hosts once their reproductive cycle is complete. Therefore, not everything that implantable devices can learn from them is necessary. However, if we can develop implantable devices that mimic the positive aspects of their survival strategies, such as camouflage, immune evasion techniques, and highly efficient energy utilization, we may usher in an era where every individual can have at least one implantable device for lifelong health management. In this presentation, I aim to discuss the technologies that may need to be developed for implantable devices to mimic beneficial aspects of parasites and share perspectives on this approach.

Neuromodulation refers to the therapeutic alteration of neural activity using various techniques including electrical, optogenetic, magnetic, chemical stimulation, and other means. As evidenced by the success of deep brain stimulation, neuromodulation holds remarkable promise for revolutionizing contemporary healthcare by offering innovative solutions to a wide range of incurable neurological and neuropsychiatric conditions such as chronic pain, Parkinson's disease, epilepsy, and depression. Recent advances of neuromodulation techniques have opened new doors for treating neurodegenerative diseases and augmenting brain-computer interfaces for enhancing human cognition or enabling individuals with disabilities to regain lost functionality such as hearing or vision.

Despite the tremendous potential and growing significance of neuromodulation as a dynamic and transformative field in modern medicine, neuromodulation faces various challenges including the need for personalized treatments, long-term safety assessments, and ethical considerations. Particularly for sophisticated and complex high-level cognitive and sensory functions such as emotion and vision, researchers and clinicians must work collaboratively to precisely optimize stimulation parameters of each neuromodulation modality.

In this talk, two stimulus optimization approaches will be discussed: stimulus optimization using machine learning vs. neural plasticity maximization by creating natural neural signals. By comparing engineering and neurophysiological aspects of neuromodulation, it will be explored which one will achieve greater success for millions of patients seeking improved treatments and enhanced quality of life.