The brain is an intricate organ that controls various functions, such as perception, behavior, mood, and cognition, through interactions between neurons, astrocytes with neurotransmitters (NT), and neuromodulators (NM). The recent advancements in fluorescence imaging and genetic engineering have allowed researchers to study neural activity, neurochemicals, and drug-specific receptor conformations at a cellular and subcellular level.Imaging neuronal spikes is crucial in the study of neural circuits and behavior. Neuronal spikes, or action potentials, are brief electrical signals traveling along nerve fibers and are the primary means neurons communicate. Researchers can directly visualize and quantify these electrical signals in living animals with high spatiotemporal resolution using genetically encoded calcium indicators (GECIs) and genetically encoded voltage indicators (GEVIs). This ability to image neuronal spikes in real-time in behaving animals offers new possibilities for dissecting neural circuits and understanding the mechanisms behind behavior across different species. In the past ten years, GECIs and GEVIs have revolutionized systems neuroscience by allowing for the mesoscopic recording of intracellular calcium as a proxy for electrical activity and directly reporting spiking patterns and subthreshold voltage activity.
Calcium and voltage imaging are valuable tools for monitoring neurons' calcium and electrical activity, but they have limitations when studying neuromodulation. Neuromodulation involves controlling neural activity through signaling molecules like hormones, NT, NM, and growth factors. This is a crucial aspect of brain function, enabling communication between neurons and coordination of brain regions. Neuromodulation fine-tunes neural circuits and influences brain functions like learning, memory, perception, and behavior. For instance, dopamine in the striatum controls reinforcement learning and habit formation, while serotonin affects mood, anxiety, and sleep. Neuromodulation also plays a crucial role in treating neurological and psychiatric disorders, with drugs altering NT levels or neuromodulatory patterns, such as antidepressants targeting the serotonin system and antipsychotics targeting the dopamine system. However, the release of NM may not be directly coupled to neuronal activation because NM is often released from non-synaptic locations, can have diffuse effects on multiple neurons, and can be released in response to non-neuronal signals.
Many groups have pioneered the effort and developed technologies for measuring the bulk release of NM, including micro-dialysis, opto-dialysis, amperometry, and mass spectrometry. However, the traditional methods either need direct measurements of specific NM or need to be higher in spatiotemporal resolution. Therefore, to better dissect the complex dynamics of neuromodulation, it is necessary to invent new technologies for monitoring NM release with sub-cellular and sub-second spatiotemporal resolution in real-time.
My graduate thesis focused on engineering genetically encoded neurochemical indicators (GENIs) and using optical imaging to study NM release. By genetically encoding the indicators, we can achieve specific cell-type targeting of the indicators and pinpoint the release pattern of specific NM in a highly specific manner, providing sub-second temporal- and sub-cellular spatial- resolution. This enables researchers to accurately measure the real-time release of NM, including their precise location and release dynamics in behaving animals. Overall, using GENIs and optical imaging to study NM release has greatly advanced our understanding of the complex processes involved in synaptic communication and has opened new avenues for exploring the mechanisms underlying neural and physiological diseases.
In the opening chapter, I delve into the current collection of GENIs, recent advances in NM dynamic studies, and the impact of drugs and environmental stimuli on neuromodulation via fluorescence imaging with GENIs developed by our lab and others. I present the methodology of engineering GENIs that allow real-time tracking of various neural activities and specific receptor conformations affected by drugs. I also use mathematical modeling to explore the engineering and optimization methods for these indicators. Further, the GENI engineering methods can be utilized for other neurochemicals we haven’t explored and provide a comprehensive toolkit for studying neural activity and drug effects in living organisms. The text of this chapter is modified from my first-author manuscript published in the Annual Review of Neuroscience in 2022.
Despite the essential biological functions that serotonin regulates, as aforementioned, imbalances in the serotonin system have been linked to several psychiatric and neurological disorders, such as depression, anxiety, and obsessive-compulsive disorder. Thus, understanding the underlying mechanisms of serotonin signaling is crucial for developing new treatments for these conditions. In chapters two and three of this dissertation, I expand into engineering two types of GENIs for the NM, serotonin.
To understand the pharmacological mechanisms of drugs on the serotonin system, in chapter two, I present the creation and validation of a GENI, psychLight, based on the 5-HT2A G-protein coupled receptor (GPCR) and circularly permuted green fluorescent protein. After extensive screening and optimization with molecular cloning and live-cell imaging techniques, psychLight was generated. It allows for detecting changes in serotonin signaling in behaving rodents under aversive stimuli. Additionally, I use psychLight to study the mechanism of action of designer drugs on receptors. By evaluating the biased receptor conformations upon ligand binding, I investigate how to eliminate the side-effect of psychedelic-based antidepressants, i.e., hallucination. I use genetic and viral approaches to develop a cell-based drug screening platform using psychLight (patented), resulting in the synthesis and discovery of novel compounds with both short-term and long-term antidepressant potential but without the hallucination side-effect. Finally, I demonstrate the use of psychLight and other genetic tools to show that psychedelics promote plasticity through intracellular 5-HT2A receptors, and serotonin may not be the natural ligand for those intracellular receptors. This highlights the importance of considering the cellular location of 5-HT2ARs in determining their signaling properties and suggests that intracellular 5-HT2ARs may be a valuable therapeutic target.
In summary, psychLight provides high-spatiotemporal resolution and real-time monitoring of endogenous serotonin in response to behavioral stimuli. The application of psychLight combined with other methods shed light on the mechanisms behind the therapeutic effects of psychedelics and the role of serotonin in promoting brain plasticity. Furthermore, the psychLight drug screening platform demonstrates the potential for conformational indicators in discovering novel treatments for neuropsychiatric and neurodegenerative diseases with fewer side effects. The text of this chapter is modified from my first-author manuscript published in Cell in 2021 and my contributions to the manuscript published in Science in 2023.
Chapter three presents an innovative method for creating a high-dynamic range serotonin indicator, iSeroSnFR, for better understanding serotonin’s role in the brain with a high signal-to-noise ratio (SNR). I use machine learning algorithms to modify an acetylcholine-binding protein, resulting in its ability to bind serotonin selectively. I validated the iSeroSnFR’s serotonin selectivity with screenings in mammalian cells and neuronal cultures. Together with others, we use iSeroSnFR to reveal serotonin dynamics during sleep-wakefulness cycles. The text of this chapter is modified from my contributions to the manuscript published in Cell in 2020.
In chapter four, I discuss the creation of GENIs to detect neuropeptides (NP), another class of NM. NP has been linked to various brain dysfunctions, including addiction, cognitive disorders, and stress. However, the study of NP signaling is minimal due to the nature of the degradation of peptides over time. Traditional methods that need to extract NP out of the brain for analysis cannot provide detailed enough information on NP signaling. Thus, a direct measurement at the site of release/reception is the key. Of the numerous NP discovered, opioids are clinically the most significant, as they are the primary target of effective pain-relieving treatments. However, these treatments also lead to issues of abuse and overdose. To create a treatment without such adverse side effects, a better understanding of opioid signaling and opioid receptor actions on drugs is necessary.
To fill this knowledge gap, I engineered a set of GENIs based on three opioid receptors to detect endogenous opioids with sub-second temporal precision at the releasing site in real-time. I thoroughly evaluated the binding kinetics of analgesics by comparing drug screening on the indicators and traditional radio-ligand binding assay on the receptors. By incorporating these indicators with optogenetics, a technology that allows using light to activate specific neuronal projections in the brain, I observed light-induced opioid release on opioid-releasing neurons in behaving animals. Further, I observed different opioid-releasing patterns in the sub-brain regions in rewarding and aversive behaviors with optical imaging. The opioid GENI toolkit is a valuable resource that facilitates new insights into opioid signaling and drug mechanisms that were previously inaccessible. The manuscript of this chapter is finalizing for publication.
Chapter five of the thesis focuses on the future outlook, discussion, and conclusion. I begin the chapter with preliminary data on the optimization of the next generation of gastrin-releasing peptide (GRP), a critical neuropeptide for regulating fear extinction. In this section, I present an ongoing project that aims to optimize the sensitivity from a previous generation GRP sensor, grpLight1.0. Given that we have published the grpLight1.0 in 2021 and used this sensor to reveal that bombesin-like peptide plays a crucial role in enhancing fear memory by recruiting disinhibitory cortical circuits. However, grpLight1.0 is not sensitive enough to detect functionally relevant GRP levels in vivo. The first section of the chapter provides results for the sensitivity optimization of grpLight1.0 with characterizations in vitro and preliminary data for monitoring GRP release in vivo during fear conditioning. This optimized sensor variant, grpLight2.0, will be followed by more detailed ex vivo and in vivo characterizations for further dissecting GRP’s role in the brain. In the later sections, I discuss the limitations of GENIs in the field of neuroscience and present a mathematical model for potential GENI optimization directions. Finally, I summarize my thesis projects and provide a future outlook.
Together, this thesis presents the methodology and applications of GENIs to study the dynamics of selected NM signaling and drug-receptor interaction in the brain. It provided valuable insights into the role of NM in shaping circuit function, both in healthy and diseased conditions. It highlights the need for better indicators of existing and other NM for a broader understanding of brain computation. Optimizing indicators for better SNR and expanding the toolkit for multiplex imaging with indicators on different spectrums is encouraged to perform multi-NM readouts simultaneously. Methods including machine learning and fluorescence-activating cell sorting are also suggested to enhance the indicator optimization process in the future. This thesis presents the significance of GENIs in neuroscience research and their possibilities in furthering our understanding of neural circuits in complex behaviors and pharmacology with precision.
