Lin Gan - US grants

DESCRIPTION (provided by applicant): The long-term objective of this proposal is to understand the molecular mechanisms of vertebrate retinal development and to identify the genes regulating the normal differentiation of retinal neurons. In mice, math5 and brn-3b are expressed early in the developing retinas starting at embryonic day 11 and 11.5 (E11 and E 11.5), respectively. Their temporal and spatial expression pattern in retina correlates with the onset of retinal ganglion cell (RGC) differentiation, suggesting their roles in the differentiation of RGCs. Previous studies have demonstrated that targeted deletion of brn-3b in mice causes the failure of axon formation and the programmed cell death of RGCs in early retinal development. The targeted mutation of math5 results in the absence of a large set of RGCs and abolishes the RGC-specific expression of brn-3b in developing retina. The expression and mutation studies have suggested that math 5 is the proneural gene for the development of RGCs and expression of math5 in the retinal progenitor cells defines the retinal ganglion cell fate. In addition, the absence of brn-3b expression in math5-null retinas suggests that brn-3b is the downstream effector gene of the math5 regulatory pathway in RGC development. Three specific aims are proposed to precisely define the role of math5 in retinal development and to establish relationship between math5 and brn-3b: (1) Mouse lines will be created to activate the stable expression of lacZ reporter gene in math5-expressing cells and the fates of these cells will be analyzed throughout development. Specifically, the effect of math5 mutation on RGC differentiation, migration, programmed cell death or converting to non-RGC fates will be assessed. (2) The ability of math5 to convert the retinal progenitors into RGCs will be tested by expressing math5 in retinal progenitor cells in vivo or in retina explant cultures. (3) The regulatory relationship between brn-3b and math5 will be analyzed using the math5 and brn-3b mutant mice and conventional promoter analysis approaches. Information obtained from these studies will significantly advance our understanding of the roles of transcriptional factors in normal retinal development and in retinal degenerative diseases. Results of these studies could provide the foundation of designing therapeutic approaches to maintain retinal cell viability and to regenerate retinal neurons.

DESCRIPTION (provided by applicant): Retinal degeneration diseases pose great health risks. Currently, the effective cures and prevention methods remain elusive. Identification of genes and genetic processes important for the formation and survival of retinal neurons will lead to the development of prevention drugs targeting these genes and processes and to develop approaches to replace the degenerated retinal neurons by transplantation or regeneration from neuronal stem cells. The long-term objective of this proposal is to understand the role of Brn-3 POU-domain transcription factors in retinal neurogenesis and to identify and characterize the roles of their downstream effect genes. The three brn-3 genes, brn-3a, brn-3b and brn-3c, share a highly conserved functional POU-domain and their expression during neurogenesis and in adult is largely overlapping. Targeted mutagenesis studies in mice have shown that deletion of each brn-3 genes lead to unique neuronal phenotypes with the loss of a selected group of neurons. Intriguingly, the uniqueness of knockout phenotypes in each brn-3 mutant closely correlated to its distinctive spatiotemporal expression pattern. In retina, expression of brn-3 genes is mostly overlapping in retinal ganglion cells (RGCs) and the onset of brn-3b expression precedes those of brn-3a and brn-3c. Deletion of brn-3b results in the terminal differentiation failure and apoptosis of approximate 70% of RGCs. Loss of brn-3c has similar effects on a small percentage of RGCs. To further understand the roles of brn-3 genes in retinal neurogenesis and to explore the common molecular mechanisms of brn-3 function, in this application, we will: 1) use the transgenic approach to test the functional equivalence of brn 3 genes. The coding regions of brn-3a and brn-3c will be used to replace brn-3b in the knock-in experiments. The ability of knock-in brn-3 genes to rescue the retinal phenotypes associated with brn-3b knockout will be examined; 2) determine the role of brn-3a in the development of RGCs. Defects in retina of brn-3a-null mice or mice null for brn-3b and brn-3a will be analyzed; 3) identify and characterize in vivo the role of brn-3b downstream genes. We have shown that loss of Brn-3b leads to the diminished RGC expression of transcription factors Gfi-1 and LMO2. Gfi-1 is a zinc-finger protein required for the survival of inner ear hair cells. Recently, we have identified LMO2 as the first transcription factor expressed in the GCL of developing retina in a low-nasal-to-high-temporal gradient. Transgenic approaches will be used to investigate their roles in retinal development, particularly the axon pathfinding and survival of RGCs.

DESCRIPTION (provided by applicant): A major cause of deafness disorders stem from the degeneration of hair cells within the inner ear although in many cases the mechanisms underlying these disorders is not understood. By defining the mechanisms controlling normal hair cell development, we will gain a better understanding of how these processes are disrupted in pathological situations and how the hair cells can be regenerated. The central hypothesis of this proposal is that the expression of the LIM-homeodomain transcription factor Isl1 provides the ventral cochlear epithelium with a competence to form the sensory organ and that the negative regulation of Isl1 function by the LIM-domain-only transcriptional regulators LMO3 and LMO4 restricts the competence to the presumptive OC (OC) region. In our preliminary studies, we have shown that during the sensory development in the cochlea at E12.5 to E16.5, Isl1 is expressed in a broad domain in the ventral cochlear epithelium, including the presumptive OC. Interestingly, the expression of LMO3 is detected in the lesser epithelial ridge (LER), whereas LMO4 expression is confined to the greater epithelial ridge (GER) and to the distal LER (dLER). The combined LMO3 and LMO4 expression domain overlaps with that of Isl1 except in the presumptive OC region where Isl1 is expressed alone. Consistent with our hypothesis, we have shown that loss of LMO4 results in the formation of supernumerary hair cells in the dLER, confirming a role for LMO4 as a negative regulator of sensory organ development. Thus, based on the established roles of LMO proteins in inhibiting LIM-HD proteins'function in transcriptional regulation, the combined action of Isl1, LMO3 and LMO4 could determine the formation of the OC region. In order to test this hypothesis and investigate the roles of LIM-domain factors in the inner ear development, we propose the following three specific aims: 1). To determine the requirement for LMO4 in the sensory and neuronal development in the cochlea and vestibule by targeted disruption of LMO4;2). To determine whether the ectopic expression of LMO4 in the presumptive prosensory region represses the sensory development by the conditional activation of LMO4 expression in the Isl1-expressing cells;and 3). To determine the role of Isl1 in the sensory and neuronal development of the inner ear by the conditional deletion of Isl1. PUBLIC HEALTH RELEVANCE Loss of the inner ear hair cells in the organ of Corti is the leading cause of hearing loss that affects 278 million people worldwide, including 28 million in the United States. However, since the loss of hair cells is an irreversible process and mammalian inner ear lacks the capability to regenerate hair cells, effective remedies to replace hair cells remain elusive. The studies proposed in this application will provide new insights into the molecular mechanisms underlying the sensory organ formation in the inner ear and could lead to novel approaches in the treatment and eventual cure of deafness by de novo hair cell regeneration. The central hypothesis of this proposal is that the expression of the LIM-homeodomain transcription factor provides the ventral cochlear epithelium with the competence to form the sensory organ and that the LIM-domain-only transcriptional regulators suppress the role of LIM-homeodomain factor in the ventral cochlea except the presumptive organ of Corti region. Thus, the combined function of LIM-homeodomain and LIM-domain-only factors regulates the competence in the ventral cochlear epithelium and determines the region of the presumptive organ of Corti.

DESCRIPTION (provided by applicant): Degenerative retinal diseases resulting from permanent loss of retinal neurons pose a great public health risk as no effective cell replacement therapies are currently available. De novo retinal regeneration from stem cells or retinal precursors could lead to a potential remedy for retinal degeneration diseases. Studies to identify genes and genetic processes essential for the formation of retinal neurons will provide the foundation for retinal regeneration. This proposal aims to reveal the molecular mechanisms governing the generation of retinal ganglion cells (RGCs), a group of neurons degenerated in glaucoma diseased eyes. The long-term objective of this proposal is to understand the role of an Atonal-class basic helix-loop-helix (bHLH) transcription factor, Math5, in the process of retinal neurogenesis and its interactions with other transcription factors in regulating the cell fate choices during retinal development. Knockout studies have shown that loss of Math5 leads to the absence of RGC formation and a concurrent increase in the generation of amacrine and cone photoreceptor cells, implying the dual roles of Math5 in promoting RGC differentiation pathway and inhibiting non-RGC differentiation pathways. Furthermore, we have shown that Math5 positively regulates the expression of RGC- specific transcription factors like Brn3b and Isl1, and negatively regulates non-RGC factors such as Bhlhb5, NeuroD, Math3, and Ngn2. The Aim 1 of this proposal is designed to characterize the role of Bhlhb5 in the development of selective subtypes of amacrine and cone bipolar cells. Lineage analysis using Bhlhb5-Cre (Cre recombinase) mice will be carried out to test if Bhlhb5 expression defines amacrine and cone bipolar subtypes. In Aim 2, the role of Isl1 in retinal genesis will be investigated in retina-specific Isl1 knockout mice. In addition, the synergistic effects of Math5 and Isl1 co-expression in promoting RGC formation from retinal precursors will be assessed by knocking-in Isl1 in Math5 locus. The Aim 3 is proposed to test the regulatory relationship of Math5 and its known downstream target genes. Chromatin immunoprecipitation (ChIP) experiments using anti- HA and genetically modified Math5-HA mice will be used to identify the direct target genes of Math5 as well as to discover novel Math5 target genes. Alternatively, a Tamoxifen-inducible Math5 (Math5ER, a fusion of Math5 and estrogen receptor ligand-binding domain) will be used to identify genes directly activated by Math5 in the presence of Tamoxifen and protein synthesis inhibitor, cycloheximide.

DESCRIPTION (provided by applicant): Our accurate vision depends on the flow of visual information through precisely wired connections between axons and dendrites of different retinal neurons. In the retina, neurons occupy spatial domains and arborize their dendrites, which require the proper distribution of their cell bodies and dendritic arbors. Cell bodies of the same type of neurons are spaced out in a process called mosaic patterning, and their dendrites establish a zone within which other cells of the same type are excluded, a process called tiling. In addition, the neurites from an individual cell display self-avoidance properties. In contrast to the excellent progress made in discovering genes and mechanisms of retinal cell fate determination and differentiation, relatively little is known about the molecular mechanisms underlying the mosaic patterning and tiling processes in the retina as well as in other nervous systems. Not until recently, studies show that in mice mutant for Down syndrome cell adhesion molecule (DSCAM), DSCAM-LIKE1 (DSCAML1), MEG10, and PCDH, certain types of retinal amacrine and ganglion cells exhibit defects in the spacing of cell bodies and in the dendritic arborization, which begins to implicate the roles of unique classes of cell adhesion molecules (CAMs) in regulating these processes in the vertebrate retina. Nevertheless, we have yet to uncover the other molecules involved in these processes in each of the nearly 80 retinal cell types and subtypes, and more importantly, to identify and characterize the entire genes and genetic pathways that govern the formation of functional neural circuitry. In the past, this question has been hard to address due to the lack of a suitable molecule. Here, we show that in mice lacking BARHL2, a BAR-homeodomain transcription factor, starburst amacrine cells in the ganglion cell layer have aggregated dendrites and clumped cell bodies, indicating Barhl2's role in self-avoidance. Being the first transcription factor implicated in neuronal mosaic patterning and tiling processes, BARHL2 offers a unique opportunity to ultimately identify genetic pathways of neuronal mosaic patterning and tiling formation. In this proposal, we will fully characterize th mosaic patterning and tiling phenotypes of starburst amacrine cells in the ganglion cell layer of the Barhl2-null retina. Second, to recover the genetic pathway of self-avoidance, we will perform RNA-Seq of Barhl2 wild type and null starburst amacrine cells and BARHL2 ChIP-Seq to screen for downstream target genes of Barhl2 and to identify the transcriptional network regulating the tiling and mosaic patterning processes of starburst amacrine cells. Together, these studies will define the role of Barhl2 in regulating the tiling and mosaic patterning processes of starburst amacrine cells and elucidate the transcriptional events that occur downstream of Barhl2.

? DESCRIPTION (provided by applicant): Our accurate vision depends on the flow of visual information through precisely wired synaptic connections among axons and dendrites of retinal neurons with unique morphological and functional properties. In the vertebrate retina, each of the six neuronal cell types: ganglion, amacrine, bipolar, horizontal, and rod and cone photoreceptor cells are further divided into subtypes based on location, morphology and function. Of all retinal neurons, amacrine cells are the most diverse group with >30 subtypes being identified so far. They represent ~40% of neurons both in the inner nuclear layer (INL) and the ganglion cell layer (GCL), make up a majority of synapses in the inner plexiform layer (IPL), and contribute to a majority of visual processing in the retina. One of the key questions i how the many retinal neuronal subtypes are produced and wired during development. In this proposal, we focus on the amacrine cells associated with the sublaminar layer 3 (S3) of the IPL. The S3 sublamina separates the ON and OFF laminas in the IPL but its cellular makeup and function is poorly understood. Here, we have demonstrated LHX9, a LIM-homeodomain transcription factor, is expressed early in retinogenesis and its expression is tightly confined toa few amacrine cells in the INL and the GCL. In our preliminary study, we have shown that these LHX9+ cells are a subgroup of GABAergic amacrine cells and express GAD67 but not GAD65. LHX9-expressing cells are also LHX2-expressing subgroup of amacrine cells. Targeted deletion of Lhx9 in mice results in a nearly complete loss of these LHX2-expressing amacrine cells and strikingly, in the absence of the S3 sublamina, suggesting that LHX9 could be expressed in and be required for the development of a unique, S3-stratifying amacrine subtype cells. Interestingly, our preliminary data show that bNOS expression is significantly down-regulated in the Lhx9-null retina, suggesting a loss of bNOS-subtype of amacrine cells that are known to project in the S3 sublamina. Being a transcription factor with a known function in neuronal subtype development in the central nervous system, LHX9 likely plays a critic role in amacrine subtype specification and offers us a unique opportunity to ultimately elucidate the genetic pathway governing the formation of the S3 sublamina and its associated neural circuitry. In this proposal, we will fully characterize the subtype identity of these LHX9-expressing amacrine subtypes and will identify its circuitry within the retina. Second, we will analyze the retinal defects of Lhx9-null mutation, particularly the effect on amacrine subtype specification, differentiation of these Lhx9-expressing S3 stratifying amacrine cells, and the change in the functional properties and circuitry of the Lhx9-lineage cells. To elucidate the LHX9 regulatory pathway in the S3-stratifying amacrine cells, we will perform RNA-Seq of control and Lhx9-null retinas and use LHX9 ChIP-Seq to screen for downstream target genes of LHX9 and to identify the transcriptional network. Together, these studies will define the role of LHX9 in regulating the formation and neural circuitry of S3 sublamina and elucidate the transcriptional events that occur downstream of LHX9.