The mission of my lab is twofold: to reveal the neural circuit mechanisms of memory storage and retrieval, and to artificially modulate memories to combat maladaptive states. We will do so in a multi-disciplinary fashion by combining virus engineering strategies, immunohistochemistry and physiology, optogenetics and functional imaging of targeted populations in vivo, and a battery of behavioral assays. Our technical repertoire, of course, will evolve as our studies evolve.
Given that any given brain region interacts with numerous targets along exquisitely precise spatial and temporal dimensions -- nothing in the brain exists in a vacuum -- we'll study the circuit-wide and behavioral manifestations of learning and memory across a variety of structurally and functionally connected areas, including, but not limited to, the hippocampus, amygdala, and prefrontal cortex. For example, I believe that a systems-level analysis of memory that perturbs genetically defined, projection-specific cell-types, while simultaneously surveying their real-time physiological dynamics, is a tractable experimental path towards understanding, and controlling, this seemingly ephemeral process.
The neuronal mechanisms of learning and memory
Previous genetic, electrophysiological, and behavioral studies have suggested that a sparse population of neurons distributed throughout the brain encodes a specific memory. These neurons can be tagged during learning for subsequent identification and manipulation. The hippocampus and its cortical inputs/ouputs in particular are pivotal for the encoding, storage, retrieval, and updating of personally experienced, or episodic, memories. Recently, our work has demonstrated that directly activating hippocampus cells that previously expressed the immediate early gene c-Fos during learning is sufficient to activate the neuronal and behavioral expression of negative, neutral, and positive memory recall. These cells also undergo plasticity-related changes during learning and are necessary for the behavioral expression of memory recall, thus corroborating their mnemonic nature.
1. Our first line of experiments focus on identifying cells in the dorsal and ventral hippocampus active during a learning experience and genetically engineering these cells, as well as their corresponding terminals, to express a variety of effectors, including light sensitive ion channels (e.g. channelrhodopsin-2, halorhodopsin), synthetic receptors (e.g. DREADDs), and calcium indicators for real-time in vivo imaging (e.g. GCaMP6f). These experiments focus on the following questions:
- Do dorsal and ventral hippocampus cells processing discrete memories differentially process mnemonic information in the spatial and emotional domains?
- What information do dorsal and ventral hippocampus axonal outputs that process positive or negative memories relay to downstream targets, such as entorhinal cortex, amygdala, prefrontal cortex, and nucleus accumbens?
- Is it possible to modulate independent features of behavioral states (e.g. contextual, emotional, or social-driven responses) by manipulating target-specific terminals in cells that carry discrete positive or negative mnemonic information?
2. Another line of experiments focuses on chronically manipulating memory bearing cells along the hippocampus-amygdala-prefrontal cortex pathway to lastingly reprogram neuronal circuits and behaviors, as well as characterizing the underlying neuronal landscape in vivo. Previous studies, for instance, have successfully utilized a variety of optogenetic, chemogenetic, and magnetogenetic stimulation protocols to enduringly modify neuronal circuits and behavior. To gain causal insight into the cellular substrates necessary and/or sufficient to induce system-wide changes related to memory, the types of questions we are currently asking include:
- If we chronically stimulate hippocampus, amygdala, or prefrontal cells processing positive or negative memories, can we enhance a given memory or extinguish the associated valence by manipulating each defined cell population or their target-specific terminals?
- What type of mnemonic information do cells in each of these brain areas process? Moreover, how does their direct modulation alter downstream neuronal activity and contribute to behavior?
- How does the structural and functional connectivity between defined sets of cells processing specific memories in these areas change in response to natural phases of memory (e.g. storage, recall, extinction) as well as in response to artificial modulation of each phase?
Thus, for these basic science projects in the lab, our conceptual scaffold consists of asking who talks to whom and when, thereby permitting us to probe the putative neuronal processes underpinning learning and memory. These projects also begin to guide ongoing and future experiments described in the subsequent section on combining the fields of memory and animal models of psychiatric disorders.
Manipulating memories to combat psychiatric disorders
Chronic stress affects numerous brain areas involved in memory, emotion, and motivation, such as the hippocampus and various cortical areas; it abnormally alters a variety of cellular events, including neuronal morphology, gene expression patterns, and neurogenesis; and, it can precipitate several maladaptive states, such as depression- and anxiety-like behaviors. Yet, the neural circuitry sufficient to mitigate or prevent such pathological phenotypes is unclear. Soberingly, most interventions rely on drug-based strategies that have not substantially improved in efficacy or specificity since their original discovery in the 1950s. Moreover, generating and utilizing animal models of psychiatric disorders is often riddled with anthropomorphic interpretations due to a lack of known and/or conserved biomarkers, a lack of objective neuroscience-based symptom criteria and, consequently, a lack of construct, face, and predictive validity.
It is not surprising, therefore that translation has often failed for half a century and that therapeutic stagnation inevitably ensued.
A tractable path forward, I believe, includes at least two key conceptual and experimental shifts:
- Deconstructing pathologies into symptom clusters with shared neurophysiological, anatomical, and biochemical substrates in humans in an attempt to provide putative nodes for interrogation in rodents. This approach helps us map objectively measured abnormalities of human neural circuit dysfunction onto rodent interventions and neurobiological readouts (e.g. rational circuit-level perturbations and genetic/physiological analyses, which collectively enable searches for candidate penetrant causes for mood and anxiety disorders).
- A shift in the view that a given brain region, transmitter, or modulator subserves a particular cognitive function to a more nuanced view focusing on the real-time activity and topology of genetically defined, projection-, target-, and synapse-specific neural networks interacting during adaptive and maladaptive states. This approach embraces the complexity of psychiatric disorders by divorcing ourselves from simplistic views of neuronal contributions to cognition (e.g. depression = too little serotonin; amygdala = fear) and, we hope, will enable far more effective forward and reverse translational outcomes.
To begin to address these caveats -- and given the extraordinary evolutionary conservation of the structural and functional properties of cortico-hippocampus memory networks -- we will commandeer the brain’s endogenous plasticity mechanisms by means of artificial memory modulation to test for and resolve its potential to regulate maladaptive circuit functioning and behavior. This proposed research aims to provide a neurobiological framework for artificially activating memories to bi-directionally modulate stress-induced states. For instance, we are currently asking:
- Are activated positive or negative memories sufficient to acutely suppress or precipitate stress-related pathologies, and what are the real-time circuit-wide responses?
- Can we deconstruct memories into their component terminals before stress, followed by physiologically correcting each terminal post-stress and test for its capacity to reverse independent features of stress-induced behaviors, including reward-related, anxiety-like, and social impairments?
- Does chronic activation of positive memories prior to stress induce neuronal and behavioral resilience? Does chronic negative memory activation mimic susceptibility? These experiments will test for memory’s potential prophylactic capacity, as well as its role in generating maladaptive states, which, if existent, will enable a brain-wide in vitro and in vivo search for key cellular and physiological loci mediating the protective or deleterious effects.
2. A second line of experiments seeks to leverage our understanding of the endogenous cellular processes supporting the various stages of memory (e.g. (re)consolidation, extinction) in the hippocampus-amygdala-prefrontal cortex pathway to intervene with disorders of fear regulation (e.g. PTSD). We are currently asking:
- By utilizing our activity-dependent and inducible strategy for labeling cells active during defined mnemonic processes, can we identify cells or physiological markers in this pathway differentially active during fear learning and fear extinction?
- When recalled, can previously learned negative associations be artificially updated and permanently attenuated by activating cells processing positive mnemonic information?
- How do extinction processes alter the physiological and structural properties of discrete sets of cells in this pathway that were previously active during fear learning?
Our species models, of course, will evolve as our experiments evolve. Collectively, the goal of these types of experiments is to provide the biomedical community with neuronal, physiological, and behavioral markers that can be future interventional and preventative targets to ameliorate a medley of maladaptive states.
"Reporting in Science, researchers write of linking a mouse's innocuous memory of a room with a more fearful memory of getting an electric shock — causing the mouse to freeze in fear upon seeing the safe room. Study author Steve Ramirez of M.I.T. and memory researcher Mark Mayford of The Scripps Research Institute discuss the implications for modifying human memories."