Search

Stressful experiences can produce a variety of emotional states, including elevated and prolonged levels of anxiety that can lead to anxiety disorders. Anxiety disorders are the most prevalent class of mental disorders worldwide; however, the neurobiological mechanisms that regulate anxiety and its disorders are still not well understood. In this context, the current project delves into downstream effects of the cleavage of EphA4 receptor by tissue plasminogen activator (tPA)/plasmin proteases; an event recently discovered by former members of our laboratory. The initial hypothesis of this project was, therefore, that tPA/plasmin-induced proteolysis of the murine EphA4 receptor is present in the mouse brain and can regulate stress-related phenomena. An initial necessary first step was to confirm the presence of the proteolytic cascade in areas relevant for the study of anxiety. In agreement with this hypothesis, I demonstrated that tPA and plasminogen co-localise with EphA4 in the GABAergic neuronal synapses of the central amygdala (CeA) through immunhistochemical techniques. In line with this discovery, the relevant literature sufficiently proves that inhibitory interneurons in the central amygdala of the mouse brain regulate anxiety-related processes by controlling the activity of downstream output cells. Specifically, those of the lateral subdivision of the central amygdala (CeL) expressing protein kinase C delta (PKCδ+) are important for aversive stimuli processing and memory. In the present work, I show that all tPA-expressing cells in CeA are also PKCδ+, which establishes a strong link between PKCδ+ cell-types and the location of an assumptive tPA/ plasmin/EphA4 cascade in areas relevant for stress-related events and anxiety- like behaviours. Conceivably, PKCδ+ (tPA) cells can regulate the properties of their downstream GABAergic synapses during stress through a cleavage of EphA4 associated with the tPA/plasmin proteolytic cascade. It is known that stressful stimuli produce the tPA-mediated conversion of plasminogen into the active enzyme, plasmin; and, as demonstrated here, plasmin would subsequently be able to cleave the tyrosine kinase receptor, EphA4. Cleavage of EphA4 has multiple neurobiological consequences. At the molecular level, I examined how shedding of EphA4 affects postsynaptic GABAergic protein-protein interactions. Here, I show that cleavage induces the dissociation of EphA4 from the GABA-receptor anchoring protein, gephyrin. The repercussions of this event are still unknown. Furthermore, this shedding can regulate the dendritic spine shape as evidenced by spine morphology experiments. Spine morphology is thought to reflect the strength and activity of a synapse whereby the excitatory or inhibitory tone of a neuron can be tuned. Moreover, consistent with a crucial role of the tPA/plasmin/EphA4 signalling cascade in anxiogenesis, EphA4 main cleaved form of EphA4 is increased after restraint stress. Accordingly, increased protein levels in the central amygdala of a plasmin-resistant variant of EphA4 (crEphA4) prevents the expression of stress-induced anxiety-like behaviours in mice, whereas the expression of a truncated EphA4 variant that mimics the cleavage by plasmin (tEphA4) increases this expression. This indicates that the cleavage of EphA4 potentially helps to modulate the expression of anxiety-like behaviours. Therefore, the present work identified a central molecular cascade that potentially controls the structure and function of GABAergic synapses downstream of CeL-PKCδ+ interneurons in the CeA and has the ability to modify the expression of anxiety-like behaviours. Additional pieces of data presented in this work indicate that the cleavage of EphA4 is affected in other brain conditions in which tPA/ plasmin cascade is involved, such as rodent models of stroke or epilepsy. Therefore, this work opens future possibilities for the study of other mechanisms regulated by tPA/plasmin/EphA4 cascade.

Rho GTPases and their regulators such as guanosine exchange factors (GEFs) and GTPase activating proteins (GAPs) represent an important class of molecules controlling dendritic spine plasticity. Although they are typically described as cytoskeletal modulators, roles for the GTPases in endocytosis and cell polarity establishment have also been defined. The neuronal proteins a1- and a2-chimaerins belong to a group of Rac and Cdc42 GAPs that inactivate these GTPases; in addition to a GAP domain, the a-chimaerins share a phosphokinase C (PKC)-like C1 domain but have distinct N-terminal domains (NTDs). My project has explored the importance of specific domains of a1-chimaerin both in induction of a morphological cellular protrusion collapse phenotype ('circularisation') and in interactions with partner proteins that may help to explain the phenotype. The results described in my thesis show that a1-chimaerin possesses a previously undescribed C-terminal domain (CTD) that is indispensable for the ability of the protein to induce collapse of protrusions, and consequent circularisation, in various cell types; moreover, an intact CTD is also important for association of a1-chimaerin with its known effector EphA4, and potentially with other undefined membrane proteins, in a C1-domain- dependent manner. In addition, my results show that a1-chimaerin associates via its NTD with the Src kinase Fyn, and via its C1 domain with the NR2A subunit of the NMDA receptor. Further experiments explored a1-chimaerin effects on EphA4 and NMDA receptor cell surface expression, as well as binding to other putative partners - including the adaptor protein p35 and the polarity protein PAR6. Finally, I have shown that inhibition of a pathway involving the Rho-associated coiled-coil containing protein kinase (ROCK) reverts circularisation induced by a1- chimaerin, and that a blocking peptide based on the CTD may be employed to partially counteract the phenotype. These results uncover a novel domain in a1-chimaerin that may have a crucial importance for the induction of cellular process collapse by a1-chimaerin with a potential relevance to the EphA4-induced dendritic spine retraction, EphA4 receptor endocytosis, and cell surface expression of NR2A-containing NMDA receptors. This suggests a model of a multi-protein signalling complex involving a1-chimaerin that coordinates cellular process remodelling, and that is likely to be important both for adult neuronal circuit plasticity and for neurodegenerative diseases.

Learning is crucial for survival. Learning involves the remodeling of synaptic connections in the brain which alters the strength of the synapses in response to activity. These changes in synaptic strength are known as synaptic plasticity. Synaptic plasticity has been a heavily studied topic since the discovery of long-term potentiation (LTP) by Bliss and Lomo in 1973. Much of this research focuses on the synaptic plasticity of connections in the hippocampus, a brain region which is critical for learning. It has been observed that learning can be more effective when using the spaced learning technique, where learning sessions are broken up by short breaks. Interestingly, it has been shown that spaced learning paradigms improve learning in animal models of Down’s and fragile X, which exhibit impaired learning. It has been proposed that this efficacy of spaced learning is due to saturation of plasticity, a state in which synapses become unresponsive to plasticity stimuli following the initial induction of plasticity. Saturation of plasticity has been observed in the hippocampus and artificial induction of saturation leads to learning deficits in rodents. It has also been shown that following the induction of plasticity at hippocampal circuits, these become unresponsive to further plasticity-inducing stimulation. After some time, these circuits can undergo plasticity again. Yet, how saturation of plasticity is occurring and what molecular mechanisms are responsible for the onset and offset of saturation have not been identified. I hypothesized that saturation of plasticity can occur at individual dendritic spines—membranous protrusions on dendrites which are the postsynaptic site of excitatory neurotransmission in the cortex. In Chapter 1 of this dissertation, I review the current literature on the molecular makeup of dendritic spines, their plasticity mechanisms and saturation of plasticity. Chapter 2 of this work contains experiments which I conducted to elucidate mechanisms which are involved in saturation of plasticity at individual spines. I show that plasticity can be saturated at individual dendritic spines and that after a period of 60 minutes, these same spines are able to undergo further plasticity. This time period of saturation can be described as a refractory period for plasticity. I also show that CaMKII signaling is impaired in saturated spines and that increasing the strength of the plasticity-inducing stimulation is not sufficient to recover their plasticity during the refractory period. Additionally, I show that the refractory period can be shortened by the overexpression of the synaptic scaffolding protein, PSD95. In Chapter 3, I explore mechanisms that may be related to saturation of plasticity, including a potential role for an NMDA receptor subunit switch, metabotropic glutamate receptor and the timing of the arrival of PSD95 at potentiated synapses. Chapter 4 of this work is a reflection of my contributions to the field of plasticity and my thoughts on the future directions on this line of research. Finally, I have included an Appendix in which I tested the role of mGluR5 in mGluR-mediated long-term depression (LTD) in spines of different sizes.

Activation of centrally expressed luteinizing hormone receptor (LHCGR) rescues spatial memory function and dendritic spine density losses associated with gonadectomy and high circulating gonadotropin levels in females. However, whether this extends to the AD brain, and the mechanisms that underlie these benefits remain unknown. To address this question, we delivered the LHCGR agonist human chorionic gonadotropin (hCG) intracerebroventricularly (ICV), under reproductively intact and ovariectomized conditions to mimic the post-menopausal state in AD model mice. Then, to delve deeper into the fundamental role of this receptor in the CNS, we developed a CRISPR/cas9 conditional knockout system to knockdown LHCGR in the hippocampus and hypothalamus. Central LHCGR activation restored spatial ability, hippocampal dendritic spine density, and BDNF expression in ovariectomized AD model female mice to wild-type levels without altering Aβ. Oppositely, hippocampal LHCGR knockdown in gonadally intact wild-type adult animals lead to spatial memory deficit and reduced BDNF transcription. Additionally, hypothalamic LHCGR knockdown uncovered novel physiological roles of the receptor on HPG and HPA axis function. Together, this work highlights the relevance of the LHCGR on cognition and its therapeutic potential in the “menopausal” AD brain, as well as its functional relevance in the CNS.

The synaptic tagging and capture (STC) hypothesis not only explains the integration and association of synaptic activities, but also the formation of learning and memory. The synaptic pathways involve in the synaptic tagging and capture phenomenon are called STC pathways. It is challenging to associate the physical attributes of STC pathways with structural changes along with synaptic strength. Mathematical modelling and computational analysis provide a way to explain the complexity of STC mechanisms along with their structural impact on a single dendritic spine. In this study, we develop mathematical models based on significant reported networks involved in synaptic tagging. We use this model to explore events associated with synaptic tagging candidates and evaluate them against the assumption based on STC hypothesis. Our model explains the integrated activities of kinases (CaMKII, PKA, MAPK/ERK) in response to the initial stimulation and setting of tags, the effect of this on overall synaptic tag strength, and the induction of L-LTP with protein synthesis. We use this model to investigate the behaviour of different published synaptic tagging candidates and examine it against the criteria laid out in the synaptic tagging hypothesis. Our model reveals that CaMKII activation (Tag1) is critical for the setting of the initial tag; however, coordinated activity with other kinases and the biochemical pathway is necessary for the tag to be stable. Similarly, PKA modulates NMDAR-mediated Ca2+ signalling. PKA and CaMKII likely act in concert in the process of tagging. We extend our model to investigate structural aspects of synaptic plasticity; we add the effects of actin remodelling and AMPARs anchoring on spine dynamics. The Spine model verifies that AMPARs are anchored in the PSD via scaffolding proteins and cytoskeletal elements to ensure reliable synaptic transmission. We introduce critical variables in the Extended-spine model to analyse the contribution of significant events leading to dendritic spine structural change. Our model associates the physical and structural aspects of synaptic plasticity link with spine dynamics. We find that a simple activation-inhibition loop can be used as paradoxical signalling to investigate the dynamics of kinases like CaMKII, the RhoGTPases-Rho and Cdc42, and actin remodelling. Hence, the interface between actin barbed end generation and signalling is a source of natural robustness, regardless of model sensitivity to kinetic parameters.

Diabetes is one of the most devastating and prevalent health conditions worldwide, affecting over 30 million Americans (Centers for Disease Control and Prevention 2017). Of recent interest is the study of insulin resistance and diabetes’ contribution to cognitive decline and impairment. A unique aspect this relationship is the study of Rho GTPases, a member of the Ras superfamily of small GTPases, and one of their regulatory proteins, the Rho GTPase Activating Proteins, or RhoGAPs. These Rho GTPases and regulatory RhoGAPs have been shown to alter hippocampal dendritic spine morphology. Dendritic spines, which branch off of a neurons’ dendrites, are critical in proper synapse formation, influencing spatial learning and memory. In addition, improper dendritic spine formation has been shown to lead to cognitive disabilities and cognitive decline. One of these RhoGAP proteins, Arhgap39, has shown to be regulated by insulin and insulin-like growth factor-1 (IGF-1). This suggests a possible explanation into how insulin resistance may hasten cognitive decline. However, the mechanism behind insulin and IGF-1’s effect on Arhgap39 expression is still unclear. As a result, investigation into the process by which insulin regulates Arhgap39 and its downstream effectors presents a promising avenue in the field of diabetes and cognitive dysfunction research.This project focused on investigating the effects of increasing insulin concentration treatments on wild type and insulin receptor knockdown cells. The aim of this study was to determine how insulin regulates Arhgap39 expression, and through which receptor it exhibits its effects. It was hypothesized that insulin would downregulate the suppressive effect of insulin on Arhgap39, confirming that insulin acts through its own receptor in its regulation of Arhgap39. Results from this study show that treating cells with increasing insulin concentrations results in increased Arhgap39 expression, which is predominantly meditated through the insulin receptor. This work provides a new perspective into how insulin regulates Arhgap39, offering greater insight into how Arhgap39 may be a key mediator for how insulin resistance impacts cognitive decline and impairment.

The group I metabotropic glutamate receptors, mGluR1 and mGluR5, are important modulators of neuronal signaling and plasticity. One specific way that group I mGluRs appear to influence excitatory neurotransmission is through the remodeling of neuronal structure by inducing changes to dendritic spines. However, group I mGluR spine remodeling has only been studied in an extremely limited number of regions and cell types, leaving the contribution of this mechanism to plasticity in many systems unknown. Group I mGluRs, especially mGluR5, have been associated with the synaptic plasticity in the reward circuitry of the brain that is believed to underlie addiction. Structural changes in this circuitry, particularly in the nucleus accumbens (NAc) are strongly correlated with the development and maintenance of addiction. Yet a potential relationship between mGluR5 signaling and spine plasticity in the NAc has not been directly studied. Here, the effects of mGluR5 signaling on spine plasticity in medium spiny neurons of the NAc are characterized, with particular attention on the sex differences and hormonal regulation of these effects. Activation of mGluR5 signaling is found to decrease spine density in the NAc with sex differences in subregion specificity. Additionally, primary gonadal hormones are found to trigger mGluR5 signaling to produce structural modulation in the NAc, with previous evidence implicating mGluR5 in estradiol-induced spine changes in this region in females, and research here finding a similar, novel role for androgen signaling in males. The mechanisms of mGluR5-mediated spine plasticity are also explored. Endocannabinoid signaling was found to be required for mGluR5-induced spine decreases in the male NAc, and spine changes were found to be correlated with a change in NAc F-actin content. Collectively, these results indicate that mGluR5 signaling results in structural plasticity in a region that is critical to reward in a sex-dependent manner, suggesting that the activity of this receptor might contribute differently to both natural and pathological motivated behavior in males and females.

Actin cytoskeleton is important for neuronal morphology and function. Many accessory proteins, including brain-specific drebrin A and diaphanous family of formins (mDia), regulate actin dynamics in neuronal substructures, such as dendritic spines. In order to probe for possible interrelation of drebrin and mDia2 formin, we examined their combined effects on actin polymerization, bundling, and binding. We identified drebrin A as a new interacting partner of mDia2 formin. Our results suggest two sites of interaction between drebrin A and mDia2, with the C-terminal region of drebrin binding to formin’s FH2 domain and the “actin binding core” of drebrin binding to formin’s C-terminal “tails.” We found that drebrin A inhibits mDia2-driven actin nucleation without affecting the rate of its processive elongation. Drebrin also abolishes actin bundling mediated by FH2 domain of mDia2. Thus, our data contribute to the mechanistic understanding of actin regulation and its role in neuronal function and development.

Abstract of the DissertationRational Design, Synthesis and Evaluation of Benzothiazole Amphiphiles with Applications in Neurodegenerative Diseasesby Jessica Lynn CifelliDoctor of Philosophy in ChemistryUniversity of California, San Diego, 2016Professor Jerry Yang, ChairNeurodegenerative diseases (NDDs), or disorders with progressive neuronal or nervous system dysfunction, affect millions of people worldwide each year. With a steadily increasing aging population, the amount of people affected is projected to exponentially increase unless there are significant advances made in research. As such, this dissertation will explore small molecule therapeutics that can potentially alleviate common pathological features of NDDs, with a focus on Alzheimer’s disease (AD). First, this dissertation will describe the rational design strategy towards improving the biocompatibility, of a class of amyloid-binding benzothiazoles particularly, with focus on decreasing toxicity. Here, novel benzothiazole amphiphiles or BAMs were designed, synthesized and evaluated to have a decreased toxicity over the parent compound when observed in various cell lines. In the subsequent chapters, these improved derivatives were further evaluated and modified for use in NDDs diagnostics or therapeutics. Specifically, BAMs capability to protect cells against the pathological hallmark of AD, β-amyloid (Aβ), its related toxicity and oxidative stress, was examined. Next, charged derivatives of BAM1-EG6 were synthesized for use in improving the binding to several NDD associated amyloids, including Aβ and α-synuclein (αS). This non-covalent strategy has potential future applications to differentiate between amyloids for diagnostics, or in therapeutics if metal chelation was utilized instead of charge-charge interactions. The last two chapters of this dissertation, analyzes the ability of BAMs to increase dendritic spine density. Since decreased synaptic density and function is associated with many NDDs, particularly as one of the initial pathologies, small molecules that could alleviate or even reverse these effects would be quite valuable. Here, the BAMs spinogenic properties were evaluated in both primary hippocampal neurons as well as human induced pluripotent stem cell (hiPSC)-derived neurons. Importantly, the ability of BAMs to counteract Aβ-induced dendritic spine loss was also examined in primary culture. Lastly, the possibility of these effects translating into humans was evaluated utilizing hiPSC-derived neurons. Collectively, this dissertation presents on small molecules with a dual-modality (i.e amyloid binding capabilities and spinogenic activity), which may find use as part of a therapeutic regiment for AD and other NDDs.

Drug addiction is a major global health problem, persisting due to a lack of effective treatments. This is largely due to insufficient understanding of the pathophysiology and neural circuitry underlying the progression of a healthy individual into drug addiction. Rodent models of cocaine addiction implicate the activity of medium spiny neurons (MSNs) in the nucleus accumbens (NAc) in the progression to addiction. MSNs of the NAc can be classified as dopamine receptor D1 or D2 expressing neurons, and cocaine exposure has previously been shown to induce many changes in these two cells types, including differential modifications of dendritic spine densities. While the neural substrates driving these changes remains unknown, these changes are likely due to synaptic changes driven by the neural adaptation in the regions upstream of the NAc in response to cocaine administration. Therefore, we seek to determine whether major excitatory projections to the NAc MSNs exhibit changes during the progression to addiction. In particular, previous work has shown the basolateral amygdala (BLA), prefrontal cortex (PFC), and the ventral hippocampus (VH) projections to the NAc as facilitating behavioral aspects of addiction. We utilized a method of monosynaptic retrograde tracing using an EnvA pseudotyped rabies virus that would allow us to label neurons in these areas that directly project to MSNs. The effect of cocaine on spine density was measured in multiple stages: initial use (5 days of cocaine exposure), withdrawal (5 days of cocaine followed by 2 week withdrawal period without cocaine), and relapse (5 days of cocaine followed by a single cocaine reinstatement shot at the end of 2 week withdrawal). Here, we characterize the distinct effects of cocaine exposure on dendritic modifications in a projection-, cell type-, and stage-specific manner. Our findings suggest cocaine differentially alters spine densities on neurons in brain areas that send excitatory projections to NAc MSNs. These neuroanatomical results establish a framework to further understand the circuit-basis of progression to drug addiction.

Fragile X syndrome (FXS) is the most common form of genetically inherited mental retardation and although it has been investigated for over 20 years, how cognitive function is specifically disrupted remains unknown. One of the most consistent neuronal phenotypes observed in FXS is increased density of dendritic spines, the protrusions that form postsynaptic components of excitatory synapses. This study further investigates spine abnormalities of the mouse model of FXS in different neuronal compartments, cell types, and cortical regions of wild type and mutant mice, as well as heterozygotes. Increased spine density was identified in mutant mice specifically on apical dendrites of layer 5 neurons. This phenotype only manifests during adulthood and the effect is consistent across multiple sensory regions of the cortex. However, effects on spine density differed between neuronal cell types. These findings identify previously unobserved region, cell type, and cell compartment-specific effects in the mouse model of FXS.

Studying spatial arrangement and relationships in full tissue samples can improve our understanding of the various developmental/pathological processes that underlie proper organ or organism function. In particular, it has been found that neuronal or vascular structures are pervasive in many tissues, and oftentimes are spatially correlated with other cells. This work aims to discover those relationships, by extracting biological knowledge from cellular and sub-cellular imaging using spatial point process methods.In this dissertation, we present discoveries on spatial distributions and attributes of dendritic spines and retinal astrocytes, two crucial elements in the mammalian nervous system. Although little is known about the spatial distributions of either respective to their surroundings and attributes, this thesis attempts to pose some possible biological hypotheses based on strong statistical evidence, as well as further extend the tools used for spatial analysis. In particular, we develop a multitype version of the linear network K-function, a summary function used for measuring clustering or repulsion of point features existing on a linear network.

Structures with linear features are very common in images, for example, the neurites, the vessels and the leaf veins. The image analysis techniques for linear structures can be used in clinical and industrial applications. In this thesis, the development of novel algorithms in linear structures, which include junction detection in linear structures, linear feature enhancement and small object detection (dendritic spines) along linear structures, is developed. Junctions have been demonstrated to be important features in many visual tasks such as image registration, matching, and segmentation, as they can provide reliable local information. In this thesis, a method for detecting junctions in 2D images with linear structures as well as providing the number of branches and branch orientations is presented. Feature enhancement is an important preprocessing step in many image processing tasks. It is the process of adjusting image intensities so that the enhanced results are more suitable for analysis. Good enhancement results for linear structures such as vessels or neurites can be used as inputs for segmentation and other operations. We develop a morpho-Gabor filter for enhancing the linear features. Dendritic spines are tiny membranous protrusions from neuron's dendrites, which are structures with typical linear feature. They play a very important role in the nervous system. It is very challenging to detect spines in neuron images. We present a novel pipeline to detect dendritic spines in 2D images and a new dendrite backbone extraction method is developed in the pipeline. In this thesis, firstly, we will introduce the definitions of linear structure. Then we will give a brief review of the existing algorithms. Next, the junction detection algorithm will be described. At last, We will introduce the spines detection method and give the conclusions for the whole thesis.

The regulation of female sexual receptivity by estradiol is complex and requires both the classical nuclear receptors and membrane estrogen receptors. It is only through the combined actions of both of these types of receptors in the brain and peripheral tissues that the female is ready for sexual behavior. The nuclear receptors dimerize, bind to DNA and regulate transcription, and thereby the translation of new proteins. The membrane estrogen receptors exert their effects through another method. In the arcuate nucleus of the hypothalamus (ARH), hippocampus and striatum, estrogen receptor-α (ERα), the only estrogen receptor shown to be required for sexual receptivity, transactivates metabotropic glutamate receptors (mGluRs) in order to initiate G protein signaling. This leads to the phosphorylation of a diverse array of signaling molecules depending upon which mGluR is activated. In a membrane to nucleus signaling schema, membrane-initiated estradiol signaling can lead to the activation of cAMP response element binding protein (CREB), which also in turn affects transcription. For the activation of sexual receptivity, the association of ERα with mGluR1a, leading to the phosphorylation of PKCθ, is required in the ARH. In the present set of experiments, I show that three other events within the ARH are necessary for sexual receptivity: caveolin-1 (CAV1) mediated ERα trafficking, spinogenesis, and modulation of activity regulated cytoskeleton associated protein (Arc).Estradiol regulates levels of membrane ERα modulating its own signaling, and CAV1 is a scaffold protein that traffics receptors to the membrane. In vitro, it has been observed to move ERα and the ERα-mGluR1 complex to the membrane. Without this protein, signaling in these neurons was significantly attenuated. Until now, no work had been done to examine whether CAV1-mediated ERα trafficking is involved in the activation of sexual receptivity in vivo. siRNA directed against CAV1 was used to knock down CAV1 protein in the ARH. This led to a significant reduction in membrane ERα, circuit activation and sexual receptivity.Estradiol-mediated spinogenesis has been seen in several areas of the brain related to sexual receptivity including the ventromedial hypothalamus (VMH), which is considered the final common output from the limbic-hypothalamic lordosis behavior regulating module. Increases in dendritic spine density in the VMH and ARH have been proposed to regulate this behavior, but there has been no formal test of the idea. I induced the formation of spines in the ARH with estradiol and showed that this was mediated by membrane-initiated signaling requiring mGluR1a and the phosphorylation of cofilin. Deactivation of this natural actin depolymerizing agent allows the formation of new spines. I then used the β-actin polymerization inhibitor, cytochalasin D, to inhibit spinogenesis in the ARH. The loss of estradiol-induced dendritic spines markedly reduced lordosis behavior underlying the importance of morphological changes in the ARH for the estradiol-induction of sexual receptivity. Mating experience has been shown to upregulate the immediate early genes in areas relevant to lordosis behavior like the VMH. Mating is a rich sensory experience involving tactile, olfactory, and auditory stimulation, which seems likely to modify the circuits that regulate it. I hypothesized that the activation of immediate early genes may modify future sexual behavior. To this end, I examined differences between sexual experienced females that were allowed a mating test after each injection of estradiol and sexually naïve females that were behaviorally tested once, but were given a similar number of estradiol injections as the experienced females. Unexpectedly, experienced rats were less receptive than the naïve females, even though they had more dendritic spines. The immediate early gene Arc was upregulated by the first sexual encounter and may play a role in the reduction of levels of membrane ERα in the ARH in experienced animals. When the Arc upregulation was inhibited with antisense oligodeoxynucleotides, receptivity in experienced females resembled that in naïve rats. These experiments indicate a vital role for Arc in the estradiol-regulation of sexual behavior.In summary, the regulation of sexual behavior in females is complex and requires the coordination of a multitude of molecular targets. The trafficking of the ERα to the membrane by CAV1 allows for the initiation of estradiol-mediated spinogenesis. During sexual behavior, the circuit that regulates receptivity is further modified by Arc. When all of these molecular and morphological events are positively coordinated, a female will be sexual receptive.

The ability to distinguish between similar experiences is a critical feature of episodic memory and is primarily regulated by the dentate gyrus (DG) region of the hippocampus. However, the molecular mechanisms underlying such pattern separation tasks are poorly understood. We report a novel role for the small GTPase ADP ribosylation factor 4 (Arf4) in controlling pattern separation by regulating dendritic spine development. Arf4+/- mice at 4-5 months of age display severe impairments in a pattern separation task, as well as significant dendritic spine loss and smaller miniature excitatory post-synaptic currents (mEPSCs) in granule cells of the DG. Arf4 knockdown also decreases spine density in primary neurons, whereas Arf4 overexpression promotes spine development. A constitutively active form of Arf4, Arf4-Q71L, promotes spine density to an even greater extent than wildtype Arf4, whereas the inactive Arf4-T31N mutant does not increase spine density relative to controls. Arf4's effects on spine development are regulated by ASAP1, a GTPase-activating protein that modulates Arf4 GTPase activity. ASAP1 overexpression decreases spine density, and this effect is partially rescued by concomitant overexpression of wildtype Arf4 or Arf4-Q71L. In addition, Arf4 overexpression rescues spine loss in primary neurons from an Alzheimer's disease-related apolipoprotein (apo) E4 mouse model. Our findings reveal that Arf4 is a critical modulator of pattern separation and regulates dendritic spine development both in vitro and in vivo.

Fragile X syndrome (FXS) is the most common single-gene inherited form of intellectual disability, with behaviors characteristic of autism. FXS is caused by silencing of the Fragile X mental retardation gene (FMR1). Individuals with FXS display childhood seizures, hyperactivity, anxiety, developmental delay, attention deficits, and visual-spatial memory impairment, as well as repetitive and perseverative behavior. Many of these behavioral deficits are also associated with mice that lack the Fmr1 gene (Fmr1-/-). Furthermore, Fmr1-/- neurons, similar to the human condition, exhibit defects in dendritic spine maturation which may underlie the cognitive and behavioral abnormalities in FXS. Dendritic spines are small protrusions on the surface of the dendrite that receive the majority of excitatory synapses in the brain and changes in their morphology affect synaptic efficacy. Much research has focused on treatment options to improve the deficits associated with FXS to ultimately improve the quality of life for the individuals afflicted with this syndrome. The basis of my research examined how minocycline improved behavioral deficits, such as a reduction in hyperactivity, anxiety, obsessive compulsive tendencies and susceptibility to audiogenic seizures as well as promoted dendritic spine maturation in the Fmr1-/-. Minocycline is a semi-synthetic antibiotic that also has many non-microbial targets, such as inhibiting the enzyme matrix metalloproteinase-9 (MMP-9). MMP-9 has been implicated in many neurological disorders, including multiple sclerosis, cerebral ischemia and mild cognitive impairment. The concurrent focus of my project was on the involvement of MMP-9 in FXS. We discovered that active MMP-9 promoted an immature dendritic spine profile in cultured hippocampal neurons, and that MMP-9 expression and activity were upregulated within the hippocampus of the Fmr1-/-. To further expound upon the role of MMP-9, we also analyzed MMP-9's expression levels within the blood and brain of individuals with FXS and found a decrease in relative activity in the blood, but an increase in total expression within postmortem hippocampal and neocortical tissue. Overall, this study elucidated a novel mechanism for MMP-9 that appears to play a causative role in the pathology of FXS as well as provided a new therapeutic option for FXS treatment.

Several lines of evidence suggest that Arp2/3 may play a role in regulating dendritic spine morphology. First, inhibition of the Arp2/3 activators N-WASP, Cortactin, and Wave alters spine morphology and density (Racz and Weinberg 2004, Pipel and Segal 2005, Soderling et al. 2007b, Wegner et al. 2008). Second, electron microscopy of dendritic spines revealed that the actin filaments within the spine head appear to be organized in "Y" shaped branches (Fifkova and Delay 1982). Since, Arp2/3 is the only complex known to make "Y" branched actin filaments, the presence of such branches in spine heads strongly suggests that Arp2/3 is involved in actin polymerization in spines. Consequently, the general hypothesis for my doctoral work is that if Arp2/3 is a major regulator of dendritic spine morphology, then Arp2/3 will be enriched in dendritic spines heads and inhibition of Arp2/3 activity will alter the number and/or morphology of dendritic spines. Our results show Arp2/3 localization within the dendritic spine heads of cultured hippocampal neurons. However, we observed Arp2/3 redistribution within dendritic shafts in response to induced synaptic activity. Temporal inhibition of Arp2/3 function during dendritic spine development showed severe morphological consequence in mature cultures. Finally, in collaboration with Dr. Robert Meller of Dow Neurobiology Laboratory at Legacy Research, Legacy Health in Portland Oregon, we show that Arp2/3 has a neuroprotective role of in ischemia tolerance.

Dendritic spines are the post-synaptic sites of most excitatory synapses in the brain, and changes in their morphology are implicated in synaptic plasticity and long-term memory. F-actin dynamics are thought to be a basis for both the formation of dendritic spines during development and their structural plasticity (Ethell and Pasquale, 2005; Pontrello and Ethell, 2009). We have shown that the F-actin-severing protein cofilin, which is regulated by phosphorylation, can induce remodeling of mature dendritic spines in hippocampal neurons (Shi et al., 2009). We also demonstrate that β-Arrestins play an important role in spatial control over cofilin activity in dendritic spines, which underlies NMDA-mediated dendritic spine remodeling. Cofilin activity in dendritic spines is regulated through CaMKII-mediated suppression of cofilin activity by phosphorylation, and calcineurin-dependent cofilin activation through its dephosphorylation. In addition, while the EphB receptor promotes spine stabilization through cofilin inactivation, the NMDA receptor prevents EphB-mediated cofilin inactivation through calcineurin. NMDAR activation also promotes the translocation of cofilin to dendritic spines, an event that requires cofilin dephosphorylation and is also dependent on β-Arrestins, which have recently been shown to scaffold cofilin with its regulators, LIM kinase and slingshot phosphatase (Zoudilova et al., 2010). Our studies demonstrate that cofilin clustering in the spines is affected in both β-Arrestin1- and β-Arrestin2- deficient neurons under normal synaptic activity, and a constitutively-active cofilinS3A mutant fails to translocate to spines in response to NMDA in β-Arrestin2 KO neurons. Moreover, while wt neurons display dendritic spine remodeling in response to NMDA or with over-expression of cofilinS3A, β-Arrestin2-deficient neurons are resistant to both NMDA-induced and cofilinS3A-induced spine remodeling. In contrast, dominant-negative cofilinS3D prevents NMDA-induced dendritic spine remodeling in wt neurons, and also rescues a mature spine phenotype that is lost in β-Arrestin1 KO neurons. In addition, over-expression of β-Arrestin in the KO neurons rescues spine abnormalities. β-Arrestin1-deficient neurons also develop immature spines in vivo, whereas hippocampal neurons lacking β-Arrestin2 develop normal mature spines, but fail to remodel in response to NMDA. Our studies demonstrate novel functions of β-Arrestin1 in the development of mature dendritic spines, and β-Arrestin2 in NMDAR-mediated dendritic spine plasticity through spatial control over cofilin activity.