A fundamental question in the study of memory is: Why do we remember some events but forget others? It has been proposed that people preferentially remember motivationally relevant information, as these memories may be useful in guiding choices in the future, a framework called adaptive memory. This dissertation examined the brain mechanisms that support adaptive memory, specifically focusing on how memory is shaped by rewards and dopamine, using a combination of pharmacological manipulations and behavioral assays. First, we found that rewards retroactively prioritize memory for preceding neutral events, and consistent with models of hippocampal replay, two periods of consolidation are necessary for this effect: a period of rest immediately following encoding and overnight consolidation. Second, motivated by research showing that neurotransmitters, such as dopamine, potentiate motivationally relevant memories to endure over long durations, we administered d-amphetamine (a dopamine agonist) before encoding. We found that when hippocampus dependent memory is tested after a short delay, working memory best accounts for memory performance, but when tested after a long delay, d-amphetamine level directly predicts memory performance. And third, we tested how d-amphetamine modulates different memory systems after a delay, using two different behavioral paradigms in which participants learned about overlapping associations using either stimulus-response learning or deliberate associative encoding. In both experiments, we found that d-amphetamine during encoding enhanced test performance on the trained items a week later; however, we did not detect any evidence that d-amphetamine modulates the integration of the overlapping pairs. Together, the work reported in this dissertation suggests that memory for motivationally relevant information is prioritized, dopamine enhances performance across different memory and learning systems, the effect of both reward and dopamine on memory and learning emerge after consolidation, and dopamine does not bias the hippocampus to encode memories in an integrated manner.

The mammalian olfactory system is capable of decoding complex mixtures of volatile chemical odorants into identifiable percepts. While the general mode of peripheral signal transduction is largely known, the mechanism relies on a rather complicated combinatorial “olfactory code”, where each of the hundreds of expressed odorant receptors (ORs) detects multiple odorants, and a given odorant in turn activates multiple ORs (Malnic, Hirono et al. 1999). Since the first identification of mammalian ORs in 1991, the deorphanization, i.e. solving of the substrate, of ORs has proven to be a challenge. Many attempts at systematic monitoring of the olfactory code have seen marginal successes for a number of reasons. First of all, there are still no solved structures of mammalian ORs to be used for high throughput computational modeling. Second, experimental validation methods such as heterologous expression still face considerable challenges. Lastly, primary chemical features of odors that allow for OR tuning are not yet defined. The traditional organic chemistry-based classification of odorants fails to predict biological activity, while percept-based computational analyses isolate esoteric descriptors that are difficult to chemically manipulate. Receptor level structure-activity analysis can provide a missing context to the odorant discrimination in the peripheral olfactory system. A critical finding by Manic et al (1999) indicates that each mature olfactory sensory neuron (OSN) only expresses one type of OR, allowing for high throughput screening of carefully crafted odorant panels using dissociated OSN calcium imaging. A few bioisosteric substitutions widely utilized in medicinal chemistry were used to construct odorant panels, showing greater success in defining odorant-OR interaction than previously used organic chemistry-based clustering methods. Among classical substitutions used by medicinal chemists, heteroaromatic ring exchanges are especially well tolerated when heteroatoms with a similar topological polar surface area (TPSA) are used as replacements. Among odorants with differing TPSA, it is likely that an OR activated by analogous odorants at two extremes of the TPSA spectrum will be activated by an odorant with an intermediate TPSA. Flipping of a polar functional group, which is often used with amides in drug target replacements, is well tolerated by the ORs in esters. Furthermore, there is a predictable activation pattern relative to number of carbons in a hydrophobic chain uninterrupted by polar epitopes. Using binary mixtures, the OR activity can be further surveyed through enhancement or inhibition of OSN activation signals. Odorants activating a smaller subset of an OR population may also be binding to a larger subset of ORs, resulting in mixture inhibition. Specifically, this work indicates that extracted odorant fragments may be binding but not activating some of the OR repertoire of the original odorant. The concept of non-classical bioisosteres is applied to the OR repertoire using aliphatic and aromatic aldehydes. It appears that the specialized electronics of a fully conjugated benzene ring can in fact be dispensable, only acting as conformational restrictor of the odorant in most cases. Not only do analogous non-conjugated systems substitute well for benzaldehyde, but so do non-cyclic odorants possessing tiglic moieties. Conformationally restricted extractions act as more faithful replacements for larger molecules in a subset of ORs. While the dissociated OSN results alone have broad implications for binding patterns of GPCRs in general, simple behavioral tests in mice using the same odorant panels indicate concrete perceptual links to medicinal chemistry-based odorant discrimination. The results from the behavioral data suggest that there may be a maximum constraint for percent OSN activation for two sequentially presented odors to be interpreted as the “same”. The results open a window to exploring other medicinal chemistry-based substitutions. Furthermore, many methodological improvements have been made over the past decade to allow for increased efficiency of deorphanization and validation of ORs.

The study of cerebellum-like circuits allows many points of entry. These circuits are often involved in very specific systems not found in all animals (for example electrolocation in weakly electric fish) and thus can be studied with a neuroethological approach in mind. There are many cerebellum-like circuits found across the animal kingdom, and so studies of these systems allow us to make interesting comparative observations. Cerebellum-like circuits are involved in computations that touch many domains of theoretical interest - the formation of internal predictions, adaptive filtering, cancellation of self-generated sensory inputs. This latter is linked both conceptually and historically to philosophical questions about the nature of perception and the distinction between the self and the outside world. The computation thought to be performed in cerebellum-like structures is further related, especially through studies of the cerebellum, to theories of motor control and cognition. The cerebellum itself is known to be involved in much more than motor learning, its traditionally assumed function, with particularly interesting links to schizophrenia and to autism. The particular advantage of studying cerbellum-like structures is that they sit at such a rich confluence of interests while being involved in well-defined computations and being accessible at the synaptic, cellular, and circuit levels. In this thesis we present work on two cerebellum-like structures: the electrosensory lobe (ELL) of mormyrid fish and the dorsal cochlear nucleus (DCN) of mice. Recent work in ELL has shown that a temporal basis of granule cells allows the formation of predictions of the sensory consequences of a simple motor act - the electric organ discharge (EOD). Here we demonstrate that such predictions generalize between electric organ discharge rates - an ability crucial to the ethological relevance of such predictions. We develop a model of how such generalization is made possible at the circuit level. In a second section we show that the DCN is able to adaptively cancel self-generated sounds. In the conclusion we discuss some differences between DCN and ELL and suggest future studies of both structures motivated by a reading of different aspects of the machine learning literature.

Amyotrophic Lateral Sclerosis (ALS) is a debilitating and fatal neurodegenerative disease affecting upper and lower motor neurons. Though studied for over two decades since the first ALS-associated genetic mutation was discovered, researchers have yet to uncover the pathological processes that lead to progressive degeneration of motor neurons in ALS, or to develop effective treatments. One prominent hypothesis proposes that excitotoxicity caused by increased motor neuron firing plays a role in ALS pathogenesis. While prior studies reported increased action potential firing in early postnatal ALS-model motor neurons in vivo, it remains unknown whether the increased activity stems from increased intrinsic excitability of ALS motor neurons or from increased excitatory drive, and whether these changes are transient or persist into adulthood, when ALS symptoms emerge. In this thesis, I circumvented the difficulties in standard measurement of electrophysiological properties of adult spinal motor neurons in vivo by relying on the visualization of the axon initial segment, a subcellular structure known to undergo compensatory structural changes in response to perturbations in excitatory input. I discovered that cultured motor neurons derived from stem cells of the SOD1G93A mouse model of ALS display shortened axon initial segments and hypoexcitable electrophysiological properties. The shortening of the axon initial segment is compensatory, as ALS motor neurons receive increased numbers of excitatory inputs and manifest increased spontaneous activity. Remarkably, similar shortening of the axon initial segment was detected in early presymptomatic spinal motor neurons in vivo. The shortened axon initial segment persists into the symptomatic stages and is particularly pronounced in motor neurons containing p62 immunoreactive aggregates and neurons exhibiting swollen mitochondria, two signs of stress and neurodegeneration in the disease. Based on these observations I propose that early in the presymptomatic stages of the disease, spinal motor neurons recruit excessive excitatory inputs, resulting in their increased activity that is in part compensated by shortening of the axon initial segment. This state persists and becomes even more pronounced in motor neurons exhibiting biochemical changes preceding neurodegeneration. While these observations support the potential role for excitotoxic stress in spinal ALS motor neurons, I paradoxically observed the opposite phenotype in ALS-vulnerable cranial motor neurons in the brainstem of the SOD1G93A animals, raising the possibility that the cellular stress that drives the neurodegeneration in ALS is motor neuron subtype specific.

Locomotion, including running, walking, and swimming, is a complex behavior enabling animals to interact with the environment. Vertebrate locomotion depends upon sets of interneurons in the spinal cord, known as the central pattern generator (CPG). The CPG performs multiple roles: pattern formation (left-right alternation and flexor-extensor alternation) and rhythm generation (the onset and frequency of locomotion). Many studies have begun to unravel the organization of the neuronal circuits underlying left-right and flexor-extensor alternation. However, despite pharmacologic, lesion, and optogenetic studies suggesting that the rhythm generating neurons are ispilaterally-projecting glutamatergic neurons, the precise cellular identification of rhythm generating neurons remains largely unknown. Traditionally, CPG networks (both pattern formation and rhythm generation) are thought to reside upstream of motor neurons, which serve as the output of the spinal cord. Recently however, it has been discovered that direct stimulation of lumbar motor neurons using the intact ex vivo neonate mouse spinal cord preparation can activate CPG networks to produce locomotor-like behavior. Furthermore, depressing motor neuron discharge decreases locomotor frequency, whereas increasing motor neuron discharge accelerates locomotor frequency, suggesting that motor neurons provide ongoing feedback to the CPG. However, the circuit mechanisms through which motor neurons can influence activity in the CPG in mammals remain unknown. Here, I used motor neurons as a means of accessing CPG interneurons by asking how motor neuron activation might induce locomotor-like activity. Through intracellular recording and morphological assays, I discovered that ventral spinocerebellar tract (VSCT) neurons are activated monosynaptically following motor neuron axon stimulation through chemical and electrical synapses. A subset of VSCT neurons were located close to or within the motor neuron nucleus. VSCT neurons were found to be excitatory, have descending spinal axon collaterals, and influence motor neuron output, suggesting that VSCT neurons are positioned advantageously to initiate and maintain locomotor-like rhythmogenesis. Intracellular recording from VSCT neurons revealed that they exhibit rhythmic activity during locomotor-like activity. VSCT neurons were found to contain the rhythmogenic pacemaker Ih current and to be connected to other VSCT neurons, at least through gap junctions. Optogenetic and chemogenetic manipulation of VSCT neuron activity provided evidence that VSCT neurons are both necessary and sufficient for the production of locomotor-like activity. Silencing VSCT neurons prevented the induction of such activity, whereas activation of VSCT neurons was capable of inducing locomotor-like activity. The production of locomotor-like activity by VSCT neuron photoactivation was dependent upon both electrical communication through gap junctions as well as the pacemaker Ih current. The evidence presented in this thesis suggests that VSCT neurons are critical components for rhythm generation in the mammalian CPG and are key mediators of locomotor activity.

Walking is an essential behavior across the animal kingdom. To navigate complex environments, animals must have highly robust, yet flexible locomotor behaviors. One crucial aspect of this process is the selection of an appropriate walking speed. Speed shifts entail not only the scaling of behavioral parameters (such as faster steps) but also changes in coordination to produce different gaits, and the details of how this switch occurs are currently unknown. Modulatory substances, particularly small biogenic amine neurotransmitters, can alter the output and even the connectivity of motor circuits. This work addresses the hypothesis that one such neuromodulator – serotonin (5HT) – is a key regulator of walking speed at the level of motor circuitry. To explore this question, I use the model organism Drosophila melanogaster which, like vertebrates, displays complex coordinated locomotion at a wide range of speeds. In Chapter 2, I will describe our efforts to characterize the anatomy of the serotonergic cell populations that provide direct input to motor circuitry. I find that innervation of the neuropil of the ventral nerve cord - a structure roughly analagous to the mammalian spinal cord - is provided primarily by local modulatory interneurons. Using stochastic single cell labeling techniques, I will detail the specific anatomy of individual neuromodulatory cells, and also the distribution of synapses across their processes. In Chapter 3, I will show that optogenetic activation or tonic inhibition of VNC serotonergic neurons produces opposing shifts in walking speed. To analyze behavior, I will use two complementary approaches. On the one hand, I will use an arena assay to holistically assess walking velocity and frequency. On the other, I will use a behavioral assay developed in the lab - the Flywalker - to assess walking kinematics at high resolution. The combination of these technique will give us a broad and specific picture of how the VNC serotonergic system modulates walking. In Chapter 4, I will identify natural behavioral contexts under which serotonin is used to shift walking behavior. I will use a variety of paradigms that induce animals to shift their speed, from changes in orientation and nutrition state, to pulses of light, odor, and a vibration. I will assess the requirement for the VNC serotonergic system under all of these conditions, to build a clearer picture of its role in modulating behavioral adaptation. In Chapter 5, I will describe our efforts, in collaboration with Pavan Ramdya's lab at EPFL, to functionally image VNC serotonergic cells while the animal is walking, to understand how activity is endogenously regulated in this population. Finally, in Chapter 6 I will characterize the circuit elements which might be responsible for serotonin's effect on walking. I will use recently developed mutant lines to identify the particular serotonergic receptors responsible for enacting shifts in walking behavior. Using genetic labeling tools, I will identify potential targets of serotonergic signaling in the VNC, and formulate a model by which action on these targets could adjust locomotor output. Altogether, this work seeks to characterize the anatomy and behavioral role of the VNC serotonergic system in Drosophila. I hope that through this work, I will gain a deeper understanding of not only this particular modulatory system in this particular behavioral context, but also of how static circuits are conferred with essential flexibility in behaving animals.

Touch sensation is initiated by diverse mechanosensory neurons that innervate distinct skin structures; however, little is known about how touch receptors are patterned during mammalian skin development. During the course of my PhD training, I analyzed embryonic and neonatal development of mouse touch domes, which contain Merkel cell-neurite complexes that encode pressure and object features. I found that developing touch domes share three key features with canonical sensory placodes: discrete patches of specialized epithelial, co-clustered mesenchymal cells capable of engaging in molecular crosstalk with the epithelium, and selective recruitment of sensory neurons. During embryogenesis, molecularly distinct patches of epithelial Merkel cells and keratinocytes clustered with a previously unsuspected population of BMP4-expressing dermal fibroblasts in nascent touch domes. Concurrently, two populations of sensory neurons preferentially targeted touch domes compared with other skin regions. Surprisingly, only one neuronal population persisted in mature touch domes. Overexpression of Noggin, a BMP antagonist, in epidermis at embryonic age 14.5 resulted in fewer touch domes, a loss of Merkel cells, and decreased innervation density in skin areas where touch domes are typically found. Thus, touch domes bear hallmarks of placode-derived sensory epithelia that require BMP signaling for proper specification.

The peripheral nervous system (PNS) is composed of a diverse array of neurons that mediate sensation. This includes sensory circuits that encode external stimuli, as well as circuits that provide information flow from our internal organs. My PhD training has focused on addressing two questions: 1) what molecular mechanisms underlie this functional diversity, and 2) can we engineer non-invasive therapeutics to modulate PNS activity? To study the molecular mechanisms of sensory function, I employed the Merkel-cell neurite complex as a model system. Merkel cells are mechanosensory epidermal cells that have long been proposed to activate neuronal afferents through chemical synaptic transmission. RNA sequencing of adult mouse Merkel cells demonstrated that they express presynaptic molecules and biosynthetic machinery for adrenergic transmission. Moreover, live-cell imaging showed that Merkel cells mediate activity- and VMAT- dependent release of fluorescent catecholamine neurotransmitter analogues. Touch-evoked firing in Merkel-cell afferents was inhibited either by silencing of SNARE-mediated vesicle fusion from Merkel cells or by neuronal deletion of b2-adrenergic receptors. Next, to develop non-invasive technologies for peripheral nerve modulation, I employed targeted focused ultrasound (FUS) stimulation and electrophysiology to record activity of individual mechanosensory neurons. Parameter space exploration showed that stimulating neuronal receptive fields with high-intensity, millisecond FUS sonication reliably and repeatedly evoked action potentials in peripheral neurons. FUS elicited action potentials with latencies comparable to electrical stimulation, demonstrating both speed and reliability of the technique. Lastly, I show that peripheral neurons can be both excited by FUS stimulation targeted to either skin receptive fields or peripheral nerve trunks, a key finding that increases the therapeutic range of FUS-based peripheral neuromodulation.

Signals for each of the five tastes have previously been shown to be processed by distinct labeled lines from taste receptor cells (TRCs) on the tongue to the ganglion neurons that innervate them. Furthermore, different tastes have been shown to be represented by distinct neurons in the taste cortex. We recorded calcium activity using fiber photometry from genetically defined populations in the mouse rostral nucleus of the solitary tract (rNST), the first brain station receiving taste signals from the tongue. We found that Somatostatin- (Sst) expressing cells respond exclusively to bitter chemicals while Calretinin- (Calb2) expressing cells respond exclusively to sweet chemicals. Immunostaining and viral strategies demonstrated that Sst and Calb2 mark distinct neuronal populations in the rNST. We then showed that optogenetic activation of Sst and Calb2 cells elicits prototypical bitter and sweet behaviors, respectively and demonstrate that ablation of these cells strongly impairs aversion to bitter tastants and attraction to sweet tastants, respectively. These findings reveal how taste information is propagated into the brain.

Synaesthesia is a cognitive trait in which stimuli of one sensory modality are automatically and consistently experienced in conjunction with perceptions in a separate modality or processing stream. Investigations of synaesthesia may help determine the neural processing required in the generation of a conscious experience. In order to gain the most complete understanding of synaesthesia, we have applied an integrated neurophenomenological approach. In Chapter 2 we present an extended case study of spatial-form synaesthesia (SFS) phenomenology. This investigation goes significantly beyond the rudimentary accounts of provided elsewhere, and provides novel observations on inducer-concurrent relationships, suggesting that guided introspection techniques can provide neurobehaviourally relevant information. In Chapters 3-5 we investigate neural activity in grapheme-colour synaesthesia (GCS). In Chapter 3 we demonstrate that activation in colour selective areas during synaesthetic colour processing is dependent on individual differences in phenomenology, thereby reconciling previous attempts to replicate this key finding in the GCS literature. In Chapter 4 we find no evidence for trait level differences in context specific functional connectivity in GCS, however, we demonstrate that localisation of the synaesthetic concurrents modulate connectivity between colour and low-level visual areas. In 5 we replicate findings of trait level differences in resting state fronto-parietal networks, suggesting that the RFPN may be a significant network in aspects of the synaesthetic experience common to all participants. We demonstrate that localisation of concurrents also modulates resting state visual networks, whilst automaticity of concurrents modulates parietal networks. Both Chapters 4 and 5 support a model of synaesthesia in which localisation of concurrents is modulated by bottom-up connectivity, between colour and early visual areas. This thesis demonstrates that individual differences in synaesthetic phenomenology significantly impact neural activity. We propose that future investigations place emphasis on the phenomenological experience of the participant in the interpretation of neural effects.