Recent and Current Research Projects
Mature neurons display substantial plasticity (changes in structure and function) in support of adaptive changes in the nervous system, such as during learning, puberty, and seasonal reproduction, and following injury, ischemia, neurodegenerative disease, and chronic stress. Numerous cues contribute to neuronal plasticity. Among these, insulin/insulin-like growth factor signaling (IIS) and steroids are widely recognized as critical regulators of neuronal growth and survival. However, the processes by which these hormonal signals are integrated and translated into structural, dynamic changes in neuron morphology are poorly understood.
I am a developmental neurobiologist, and I have a long-standing interest in the cellular and molecular mechanisms underlying the hormonal regulation of neuronal plasticity, particularly in the context of neuropeptide signaling and the control of insect behavior. My current research is focused primarily on genetic and molecular pathways controlling neuronal remodeling, a process involving the extensive pruning of nerve cell connections followed by the formation of new ones. In addition, I am investigating the regulation of peptide hormone expression by steroids, and I have been part of a decade-long collaboration focused on the control of neuropeptide secretion.
The principal goal of my current research is to understand when and how mature, terminally differentiated neurons can respond to hormonal or other cues to reactivate and execute growth processes that are usually seen only in embryonic neurons. To address this question, my lab has pioneered a system for high-throughput screening for factors controlling remodeling of neuropeptide secreting (peptidergic) neurons during fruit fly (Drosophila melanogaster) metamorphosis. With this approach, we have identified key factors involved in regulating the outgrowth of nerve cell projections during remodeling. I describe these findings and our current research below.
In insects that undergo complete metamorphosis, most larval tissues are destroyed and are replaced by adult structures that are derived from stem cells or nests of undifferentiated cells set aside during earlier stages. The central nervous system is an exception, and many larval neurons are retained. In order to serve the markedly different functions of the adult nervous system, these larval neurons undergo extensive hormone-coordinated remodeling, involving the pruning back of larval axons and dendrites (neurites) followed by extensive outgrowth of adult neurites. We have been investigating this process in a set of D. melanogaster neurons that secrete the neurohormone bursicon (and as well as other neuropeptides) to control molting and post-molting behaviors such as wing expansion. We and others have shown that the bursicon neurons are essential for only two events in the life cycle , and this has provided us with a very powerful genetic foothold into the remodeling process. If the embryonic or larval development of these neurons is disrupted, animals may fail to complete head eversion at the molt from the larva to the pupa. If metamorphosis of the bursicon neurons is disrupted, adults may fail to expand their wings. Disruption of either of these behaviors results in permanent, morphological defects that are easily scored by eye, even days after these behaviors normally occur (Fig. 1) . With this insight, together with reagents that allow us to overexpress or knock down gene expression specifically in the bursicon neurons and to label these neurons to assess cellular morphology, we have performed several high-throughput genetic screens to identify factors that control key aspects of the remodeling process [1, 2, 5].
In 2008, we reported the results of the first of these genetic screens, based on overexpression of genes in specific peptidergic cells, to detect factors involved in neuronal remodeling . One of the strongest phenotypes involved FOXO, a negative regulator of the cellular signaling pathway that mediates cell growth stimulation by insulin. Upon binding to plasma membrane tyrosine kinase receptors, insulin and insulin-like growth factors activate a series of intracellular kinases that inhibit entry of the FOXO transcription factor into the nucleus and activate TOR, which stimulates cell growth. To confirm involvement of FOXO and the insulin/insulin-like growth factor signaling (IIS) pathway in metamorphic neuronal remodeling, we used cell-targeted RNA interference as well as gain- and loss-of-function gene constructs for several upstream and downstream components of this signaling pathway in the bursicon neurons. Bursicon-cell specific reduction of IIS resulted in reduced growth of the cell bodies and attenuated branching of peripheral axons, whereas IIS overstimulation had the opposite effect (Fig. 2). These results, which we published in Biology Open , demonstrated a critical role for IIS in metamorphic neuron growth.
The bursicon neurons also grow substantially as the nervous system more than doubles in size during larval development. In contrast to the metamorphic growth of the bursicon neurons, which we found was highly sensitive to the levels of IIS, this larval scaling was largely unaffected by decreased insulin signaling. Similarly, when we examined the regulation of cell body size in other neurons, every type examined during metamorphosis showed insulin-dependent growth, whereas the growth of most larval neurons was refractory to changes in IIS. These findings revealed a fundamental shift in control mechanisms as the nervous system changes from a maintenance class of growth, when neurons increase in size but maintain their gross morphologies, to an organizational class of growth, when neurons actively generate new neurite arbors . Our findings also provided indirect evidence to suggest that the increase in neuronal growth regulation by IIS involves an increase in secretion of insulin family peptides from local sources. To test this model, we are currently examining central nervous system neurons and glia to look for a local insulin source.
What factors interact with IIS to promote organizational neuron growth? We have begun to address this question in a genetic screen for chromosomal deficiencies that modify the wing expansion defects produced by foxo overexpression. We identified 14 suppressor deficiencies, as well as several enhancers, and we have mapped some of these deficiencies to single genes. These results implicate multiple signaling pathways and other molecular mechanisms in insulin-dependent organizational growth during metamorphosis.
In our 2008 gain-of-function screen, we identified a second factor, Split ends (SPEN), that caused a profound reduction in neurite outgrowth when overexpressed in the bursicon neurons during metamorphic remodeling  (Fig. 3). In subsequent genetic tests, we showed that SPEN and a closely related protein, Spenito (NITO), are required additively or synergistically to support axon outgrowth during metamorphic remodeling. We have submitted a paper describing these findings, and the manuscript is now in revision . While the spen gene was first identified as a factor controlling axon development in Drosophila embryos, our work is the first demonstration of a role for SPEN and NITO in the developmental plasticity of mature neurons. There are two subfamilies of SPEN family proteins that have been conserved from protists to plants to animals; the large SPEN proteins function as transcriptional coactivators and corepressors, whereas the small SPEN proteins are key regulators of posttranscriptional processing and nuclear export of mRNAs. Several recent studies, in a variety of systems, have shown either antagonistic or redundant functions of the large and small SPEN proteins, and our work is one of the best examples to date of redundant or synergistic functions for large and small SPEN proteins.
In parallel with the above study, we performed a genetic screen for modifiers of SPEN signaling in the bursicon neurons. We identified several genes that either suppress or enhance the mutant phenotypes produced by SPEN overexpression in the bursicon neurons. One of the strongest suppressors was Myosin binding subunit (Mbs), which dephosphorylates the myosin regulatory light chain to inactivate non-muscle Myosin II. Mbs is regulated by several Rho GTPases in different cellular contexts, and we have demonstrated genetic interactions among the genes encoding SPEN, the Rho GTPase Rac1, and the Rac1 effector PAK-kinase (PAK), as well as the genes encoding the large Myosin II subunit and MYPT-75D, another protein that inactivates Myosin II by dephosphorylating the myosin regulatory light chain. Based on these findings, we are developing a model for the interactions between SPEN and Rac1/PAK/Mbs/Myosin II in the regulation of axon outgrowth . SPEN, NITO, and their orthologs in plants and animals perform a wide variety of developmental functions, but this is the first example of a SPEN family protein controlling myosin-based movements.
Fig. 4.A 3D tracing in Imaris (projected to 2D) showing the organization of the subesophageal bursicon cell neurite arbors in brain (cyan) and subesophageal ganglia (magenta). SPEN promotes braching in the subesophagel region during metamorphosis. d, dorsal; p, posterior; r, right.
As a postdoc, I identified a key regulator of peptidergic neuron differentiation by virtue of its peptidergic cell-specific expression pattern . Recently, we employed a similar strategy to identify Alan shepard (SHEP), a protein that controls gene expression by inhibiting DNA insulators. Although we later determined that SHEP expression is widespread in (and largely restricted to) neurons, we observed wing expansion defects in shep mutants that led us to focus on the bursicon neurons. Like IIS, we found that SHEP promotes growth of the bursicon cell bodies, branching of the peripheral axon arbor, and branching/growth of central bursicon cell arbors during metamorphosis (Fig. 4), with little impact on the growth of these neurons during larval development. In addition, we found that SHEP controls metamorphic-specific growth of many other neurons. We published this work in Genetics , and we built upon these results with a deficiency-based shep modifier screen . To date, we have identified three shep suppressors, and we are now working to confirm the involvement of the implicated signaling pathways in organizational growth and to test for crosstalk between them and insulin signaling.
Nuclear receptors control transcriptional responses to steroid hormones and other small hydrophobic signaling molecules and are crucial regulators of development and homeostasis in diverse organisms. A key outstanding question about nuclear receptor signaling is how multiple receptor isoforms mediate tissue- and stage-specific responses to hormones. In many cases, nuclear receptor isoforms share common DNA- and ligand-binding regions but have distinct N-terminal sequences. These N-terminal sequences often contain hormone-independent transcriptional activation functions (AF1s) that direct distinct transcriptional responses.
The insect steroid hormones, ecdysteroids, coordinate molting and metamorphosis through binding to heterodimeric nuclear receptors containing the Ecdysone receptor (EcR) and Ultraspiracle. In D. melanogaster, there are three EcR isoforms (A, B1, and B2) that differ only in their N-terminal domains. In paper published in PLoS Genetics, we have recently shown that the Cryptocephal (CRC) basic leucine-zipper transcription factor is an isoform-specific coactivator for EcR-B2 . Our evidence for the EcR-B2/CRC interaction included yeast 2-hybrid and in vitro binding studies that were performed in the Cherbas lab at the University of Indiana, and genetic tests that we performed at OU. In addition, we showed that CRC and EcR/ecdysteroids regulate expression of a peptide hormone, Ecdysis triggering hormone (ETH). This followed our cloning of the crc gene  while I was a postdoc, and our subsequent demonstration at OU that CRC regulates ETH gene expression  (Fig. 5). Together, these experiments identified CRC as an isoform-specific transcriptional activator for EcR-B2 and help to explain how the different EcR isoforms mediate distinct tissue responses to ecdysteroids.
EcR contains two activation functions, AF1 and AF2. AF2 is common to the three EcR isoforms and is formed by ecdysteroid-induced folding of the ligand binding domain. Thus, in addition to the interaction between CRC and AF1, the AF2 domain of EcR-B2 may interact with other transcriptional coactivators to mediate ecdysteroid-dependent transcription. We are currently investigating one transcriptional coactivator that is expressed in the cells that produce ETH, and we have shown that loss-of-function alleles produce larval molting defects that resemble ETH mutants. We are currently examining how this coactivator interacts with CRC/EcR-B2 to increase ecdysteroid-dependent ETH transcription. These interactions may help to explain how multiple transcriptional coactivators for EcR can display ecdysteroid-dependent AF2 binding and yet control the expression of different subsets of ecdysteroid-responsive genes.
Acting as neuromodulators and hormones, neuropeptides are key regulators of diverse processes, including growth, reproduction, stress, energy balance, sleep, and circadian rhythms. They are also interesting because they are secreted by a process that is quite distinct from that used by classical neurotransmitters such as acetylcholine. Classical neurotransmitters are released rapidly in response to single action potentials, but neuropeptides are often only released after high frequency or repetitive neuronal firing. Because they are produced in the Golgi and endoplasmic reticulum and then transported to nerve terminals, neuropeptide stores can be depleted by nerve cell firing, whereas classical neurotransmitters are loaded locally into recycled vesicles within synapses, and their supplies are essentially limitless. Thus, in addition to investigating the mechanisms of peptidergic neuron remodeling, I also have a longstanding interest in understanding how neuropeptides are secreted and how that secretion is regulated [10-13].
I recently finished a 10-year, NIH-funded collaboration with Edwin Levitan (University of Pittsburgh School of Medicine) and David Deitcher (Cornell University) aimed at investigating the process of neuropeptide secretion. Levitan was the PI on these grants, and Deitcher and I were Co-PIs. The Deitcher lab created transgenic lines with new constructs that allow imaging of vesicle movement and secretory granule activity in living Drosophila cells in vivo. In my lab, we used these transgene insertions, and various stocks that we were employing to investigate peptidergic neuron development, to create stocks that drive transgene expression in various classes of peptidergic cells. The Levitan lab used these lines to image movements of secretory granules in living neurons. This collaboration resulted in several papers showing that: 1) secretory granules are immobile in resting boutons and then become motile in response to Ca++ influx to enable release , 2) Drosophila secretory granules (unlike their vertebrate counterparts) can support neuropeptide packaging and processing with a near neutral pH , 3) ryanodine-receptor mediated release of Ca++ from the endoplasmic reticulum and activation of Ca++/calmodulin-dependent protein kinase II support increased mobility of synaptic granules and post-tetanic potentiation of neuropeptide secretion , 4) synaptic neuropeptide release is triggered by octopamine through a synergistic mechanism involving a cAMP-dependent protein kinase and Ca++ secretion from the endoplasmic reticulum instead of Ca++ influx via the plasma membrane , and 5) synaptic granule capture in boutons varies from cell to cell and determines neuron-specific variation in the accumulation of neuropeptides in synaptic terminals .