Articles, Development/Plasticity/Repair

Onecut1 Is Essential for Horizontal Cell Genesis and Retinal Integrity

Fuguo Wu, Renzhong Li, Yumiko Umino, Tadeusz J. Kaczynski, Darshan Sapkota, Shengguo Li, Mengqing Xiang, Steven J. Fliesler, David M. Sherry, Maureen Gannon, Eduardo Solessio and Xiuqian Mu
Journal of Neuroscience 7 August 2013, 33 (32) 13053-13065; https://doi.org/10.1523/JNEUROSCI.0116-13.2013

Abstract

Horizontal cells are interneurons that synapse with photoreceptors in the outer retina. Their genesis during development is subject to regulation by transcription factors in a hierarchical manner. Previously, we showed that Onecut 1 (Oc1), an atypical homeodomain transcription factor, is expressed in developing horizontal cells (HCs) and retinal ganglion cells (RGCs) in the mouse retina. Herein, by knocking out Oc1 specifically in the developing retina, we show that the majority (∼80%) of HCs fail to form during early retinal development, implying that Oc1 is essential for HC genesis. However, no other retinal cell types, including RGCs, were affected in the Oc1 knock-out. Analysis of the genetic relationship between Oc1 and other transcription factor genes required for HC development revealed that Oc1 functions downstream of FoxN4, in parallel with Ptf1a, but upstream of Lim1 and Prox1. By in utero electroporation, we found that Oc1 and Ptf1a together are not only essential, but also sufficient for determination of HC fate. In addition, the synaptic connections in the outer plexiform layer are defective in Oc1-null mice, and photoreceptors undergo age-dependent degeneration, indicating that HCs are not only an integral part of the retinal circuitry, but also are essential for the survival of photoreceptors. In sum, these results demonstrate that Oc1 is a critical determinant of HC fate, and reveal that HCs are essential for photoreceptor viability, retinal integrity, and normal visual function.

Introduction

In the vertebrate retina, there are six types of neurons (rod and cone photoreceptors, horizontal cells (HCs), bipolar cells, amacrine cells, and retinal ganglion cells (RGCs)), and one glial cell type (Müller cells). Each of these cell types plays a distinct role in the process of vision (Wässle and Boycott, 1991; Masland, 2001). During development, all of the retinal cell types originate from a common pool of multipotent progenitor cells (Cepko et al., 1996). Both intrinsic and extrinsic pathways are involved in the process, and gene regulation is a major intrinsic mechanism controlling the formation of individual retinal cell types (Agathocleous and Harris, 2009; Xiang, 2013). However, a comprehensive understanding of the underlying gene regulatory pathways is still lacking. Many key regulators remain to be identified, and their functions to be characterized. Here we report the results of our investigation into the roles of a novel regulator, Onecut1 (Oc1; also known as Hnf6), in retinal development, and its function in the formation of HCs.

HCs are second-order neurons that process visual information laterally within the outer plexiform layer (OPL; Thoreson and Mangel, 2012). They receive input from rod and cone photoreceptors (Cervetto and MacNichol, 1972), and provide inhibitory input to photoreceptors and bipolar cells (Wu, 1992). HCs form during the first wave of retinal neurogenesis, which starts at embryonic day 11 (E11) and peaks at E14.5 in the mouse (Young, 1985; Cepko et al., 1996). Several transcription factors, including FoxN4, Prox1, Ptf1a, Lim1, Sall3, AP-2α, and AP-2β, have been reported to play roles in HC development (Dyer et al., 2003; Li et al., 2004; Fujitani et al., 2006; Nakhai et al., 2007; Pochè et al., 2007; de Melo et al., 2011; Bassett et al., 2012). Nevertheless, it remains unclear how and when fated HCs arise from the retinal progenitor pool and what factors determine the HC fate.

Oc1 belongs to the onecut transcription factor family, members of which have been shown to regulate development of the liver, pancreas, and the immune system (Lemaigre et al., 1996; Samadani and Costa, 1996; Landry et al., 1997; Jacquemin et al., 2003; Margagliotti et al., 2007; Furuno et al., 2008), as well as the CNS (Francius and Clotman, 2010; Espana and Clotman, 2012a, b; Roy et al., 2012; Wu et al., 2012). All three mouse Onecut factors (Oc1, Oc2, and Oc3) are also expressed in the developing retina, but only Oc1 and Oc2 are expressed at high levels (Wu et al., 2012). To investigate the role of Oc1 in retinal development, we specifically inactivated Oc1 in the developing mouse retina by Cre-mediated recombination. Our results indicate that Oc1 is essential for HC development and determines HC fate in collaboration with Ptf1a. Further, we show that when the number of HCs is severely reduced, the synaptic connections between photoreceptors and bipolar cells in the OPL are abnormal, and that photoreceptors degenerate as the mice age, suggesting a previously unrecognized role of HCs in maintaining synaptic structure, photoreceptor viability, and retinal integrity.

Materials and Methods

Animals.

The floxed Oc1 allele (Oc1flox) and the Six3-cre transgenic line have been described previously (Furuta et al., 2000; Zhang et al., 2009). The Ptf1a− (Ptf1acre) mice (Fujitani et al., 2006) were obtained from MMRRC-UNC (Stock number: 000435-UNC). All lines were maintained in C57BL/6×129 mixed background. All procedures using mice conform to the U.S. Public Health Service Policy on Humane Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committees of Roswell Park Cancer Institute and University at Buffalo.

Immunofluorescence labeling.

Immunofluorescence labeling of frozen retinal sections followed our previously published procedure (Wu et al., 2012). Whole-mount immunostaining of the retina followed a protocol used by Kay et al., (2011). In cases where the primary antibody was from mouse, the sections were first blocked by donkey anti-mouse IgG Fab fragment (50 μg/ml Jackson ImmunoResearch, 715-007-003) to reduce the background from endogenous mouse IgG. The sources and dilutions of antibodies against Pou4f2, HA tag, Onecut 1 (Hnf6), Onecut 2, calbindin, Isl1, Lim1 (mouse), Chx10, Pax6, syntaxin I, NF160, protein kinase C α (PKCα), FoxN4, and rhodopsin were as reported by us in previous publications (Li et al., 2004; Mu et al., 2005, 2008; Fu et al., 2006, 2009 ,Wu et al., 2012). Other antibodies used in these studies include: rabbit anti-Lim1 (1:100; Millipore, AB3200), mouse anti-Prox1 (1:200; Millipore, MAB5654), rabbit anti-Sox9 (1:500; Millipore, AB5535), rabbit anti-Pgp9.5 (1:500; Millipore, AB1761), rabbit anti-cone arrestin (CAR) (1:500; Millipore, AB15282), guinea pig anti-doublecortin (Dcx) (1:500; Millipore, AB2253), guinea pig anti-Ptf1a (1:400; Dr. Jane Johnson, Utah Southwestern Medical Center), rabbit anti-Ptf1a (1:800; Dr. Helena Edlund, Umeå University), mouse anti-Bassoon (1:100; Enzo Life Sciences, ADI-VAM-PS003-D), and rabbit anti-GFP (1:150; Abcam, AB290). Alexa Fluor 488-conjugated and Alexa Fluor 546-conjugated secondary antibodies against IgGs of various species (Invitrogen) were used at 1:800 dilution. When necessary, the nuclei were counterstained with propidium iodide. Images were collected using a Leica TCS SP2 confocal microscope. To collect whole-mount images, image stacks of equal thickness in the z-plane were collected for both wild-type and Oc1-null retinas.

Cell counting.

Cell counting was performed as described previously (Mu et al., 2005; Wu et al., 2012). Briefly, retinal sections from desired developmental stages were stained with cell type-specific antibodies as described above, counterstained with propidium iodide, and total cells and marker-positive cells were counted on arbitrary retinal section lengths in the central retinal regions. For each cell type, three to four sections from at least two individual animals were counted. To count cells in the whole-mount retinas, cells from corresponding regions of wild-type and Oc1-null retinas were counted. Statistical significance was assessed by unpaired Student's t test, assuming equal variance with a two-tailed distribution; the threshold for statistical significance was set at p < 0.05.

Histology.

Tissue processing and hematoxylin and eosin (H&E) staining of mouse retinal sections were performed as described previously (Mu et al., 2008). Briefly, eyes from wild-type and mutant mice were enucleated, fixed in buffered mixed aldehydes (3% paraformaldehyde and 2% glutaraldehyde, in PBS, pH 7.4), and embedded in paraffin. The tissues were then sectioned (7 μm in thickness), dewaxed, and stained with H&E sequentially. Images were collected using a Nikon Eclipse 80i microscope using a SPOT RT3 digital camera (Diagnostic Instruments).

In utero electroporation.

DNA constructs expressing Oc1 or Ptf1a under the CAG promoter were made by replacing the GFP coding sequence with Oc1 or Ptf1a full-length cDNA in the previously reported pCAG-GFP plasmid (Addgene; Plasmid 11150) (Matsuda and Cepko, 2004). In utero electroporation was performed following a previously published procedure (Petros et al., 2009). Briefly, pregnant female mice were anesthetized at E13.5 using vaporized isoflurane (2–5%) mixed with oxygen. A 2 cm vertical incision was made on the midline of the abdomen to expose the uterine horns, and 0.5 μl of DNA solution (2.5 μg/μl of each DNA construct, 0.5 μg/μl pCAG-GFP, and 0.025% Fast Green Dye) was injected through the uterine wall and amniotic sac into an embryonic retina with a fine-tipped micropipette. After injection, wet electroporation paddles were placed on the sides of the embryo head, and five 40 V, 50 ms square pulses were delivered by an electroporation device (ECM 830; Harvard Apparatus). The embryonic chain was then placed back into the abdominal cavity, the peritoneum and the skin were then closed, and the embryos were allowed to develop further to E17.5, and analyzed by immunofluorescence.

Transmission electron microscopy.

Mouse eyes were fixed with buffered mixed aldehydes, osmicated, serially dehydrated, and embedded in plastic resin in preparation for transmission electron microscopy (TEM) analysis, as described previously (Ding et al., 2004; Stricker et al., 2005). Thin sections (600–800 Å) were obtained with an ultramicrotome (Reichert–Jung Ultracut E Microtome; American Instruments) using a diamond knife, collected onto copper 75/300 mesh grids (Electron Microscopy Sciences), and stained with 2% (w/v) uranyl acetate and Reynolds' lead citrate. Sections were viewed using a JEOL 100CX electron microscope (JEOL) at an accelerating voltage of 60 keV, and digital images were collected and stored on a computer for subsequent viewing and analysis.

Electroretinogram recording.

Electroretinogram (ERG) recording was performed as previously reported (Umino et al., 2012). In brief, dark-adapted mice were placed in a light-proof cage and anesthetized with 60 mg/kg pentobarbital (Nembutal; Lundbeck). Pupils were dilated with a couple of drops of 1% tropicamide, corneas were kept moist with 0.3% glycerine/1.0% propylene glycol, and body temperature was maintained at 37°C with a heating pad. ERGs were recorded using the Espion E2 system and a ColorDome Ganzfeld stimulator (Diagnosys) in response to brief LED flashes (520 nm). The scotopic (dark-adapted) b-wave amplitude was measured from the a-wave trough to the peak of the corneal-positive b-wave. For light-adapted ERG, retinas were exposed to a steady adapting background (520 nm) of 10 cd/m2. The number of photoisomerizations/ rod/ s produced by the background was estimated as described previously (Umino et al., 2012).

Results

Retina-specific deletion of Oc1

To delete Oc1 specifically in the developing retina, we crossed the Oc1flox line with the Six3-Cre transgenic mice. The Oc1flox allele has two loxP sites flanking the first exon, which encodes the N-terminal part of the protein including the “cut” domain, which is the main DNA binding domain. Cre-mediated recombination leads to the deletion of this exon and results in a null allele (Zhang et al., 2009). The Six3-Cre transgene is specifically expressed in the neural retina, starting from E9, a time when retinal neurogenesis has yet to begin and Oc1 expression in the retina has not yet initiated (Cepko et al., 1996; Furuta et al., 2000; Wu et al., 2012). To examine the efficiency of deletion of Oc1 by Six3-cre, we examined Oc1 expression by immunofluorescence on Oc1flox/flox and Oc1flox/flox;Six3-cre retinal sections at E12.5 and E14.5 (Fig. 1A–D). Whereas Oc1 was readily detectable in control Oc1flox/flox retinas as previously reported for wild-type retinas (Wu et al., 2012), there was essentially no signal in Oc1flox/flox;Six3-cre retinas at either E12.5 or 14.5, indicating that Oc1 was efficiently deleted in the developing retina at an early stage. We refer to Oc1flox/flox;Six3-cre as Oc1-null (Oc1−/−) hereafter. Adult Oc1-null mice showed no gross health issues and were fertile. Since Oc1flox/flox mice exhibit the wild-type phenotype, both Oc1flox/flox and true wild-type littermates were used as controls and will be referred to as wild-type (Oc1+/+) hereafter. To first assess the consequences of Oc1 deletion, we examined postnatal day 16 (P16) retinas histologically (Fig. 1E,F). Retinas from Oc1-null mice exhibited normal cell layer lamination; the outer nuclear layer (ONL), inner nuclear layer (INL), inner plexiform layer (IPL), and ganglion cell layer (GCL) were of similar thickness to the corresponding layers of control retinas, and showed no overt morphological abnormalities. However, the OPL, which is composed of photoreceptor terminals and processes from HCs and bipolar cells, was markedly reduced and even absent in some locations (Fig. 1F). Since HCs are the only Oc1-expressing cell type that extends processes into the OPL (Wässle and Boycott, 1991; Masland, 2001; Wu et al., 2012), it was likely that changes in the OPL of the Oc1-null retina reflected effects on HCs.

Figure 1.
Figure 1.

Oc1 is efficiently deleted in the retina by Six3-Cre and its deletion leads to defects in the OPL. A–D, Immunofluorescence labeling using anti-Oc1 antibodies on Oc1flox/flox and Oc1flox/flox;Six3-Cre (Oc1-null) retinal sections from E12.5 and E14.5. Oc1 shows identical expression patterns in Oc1flox/flox retinas to those previously reported in wild-types at E12.5 (A) and E14.5 (C). In Oc1flox/flox;Six3-Cre retinal sections, almost no Oc1 could be detected at both E12.5 and E14.5. E, F, H&E staining of Oc1flox/flox and Oc1flox/flox;Six3-Cre retinal sections from P16 mice. Compared with the wild-type control retina, the OPL of the Oc1-null retina is greatly reduced. All other cell layers are comparable to those in the control retina. Scale bars: (in A), A–D, 150 μm; (in E), E, F, 75 μm.

Effect of Oc1 deletion on retinal cell differentiation

Next, we performed immunofluorescence labeling using cell type-specific markers to examine whether the different retinal cell types formed normally at P16. These markers included rhodopsin (rods; Fig. 2A,B), CAR (cones; Fig. 2C,D), Chx10, and PKCα (bipolar cells; Fig. 2E,F; data not shown), Pax6 (amacrine cells and RGCs; Fig. 2G,H), syntaxin I (amacrine cells; Fig. 2I,J), Sox9 and vimentin (Müller cells; Fig. 2K–N), Pou4f2 (RGCs; Fig. 2O,P), and calbindin (HCs and amacrine cells; see Fig. 4A,B). We observed no obvious changes in the locations and numbers of rods, cones, bipolar cells, Müller cells, amacrine cells, or RGCs in Oc1-null retinas compared with wild-type retinas (Fig. 2A–P). All these cell types had normal morphology, and cell counting confirmed that their individual proportions to the total cell number were not impacted by deletion of Oc1 (Fig. 2Q). However, consistent with the observed thinning of the OPL in Oc1-null retinas (Fig. 1F), the number of HCs detected by calbindin immunostaining was significantly reduced (Figs. 2, 4B; see details below).

Figure 2.
Figure 2.

Deletion of Oc1 does not affect formation of photoreceptors, bipolar cell, amacrine, ganglion, or Müller cells. Immunofluorescence labeling for cell type-specific markers was used to examine the formation of six retinal cell types in wild-type (Oc1+/+) (A, C, E, G, I, K, M, O) and Oc1-null (Oc1−/−) (B, D, F, H, J, L, N, P) retinal sections from P16 mice. A, B, Rhodopsin staining for rod photoreceptors. C, D, CAR staining for cone photoreceptors. E, F, Chx10 staining for bipolar cells. G, H, Pax6 staining for amacrine cells and retinal ganglion cells. I, J, Syntaxin I staining for amacrine cells. K, L, Sox9 staining for Müller cells. M, N, Vimentin staining for Müller cells. O, P, Pou4f2 staining for ganglion cells. A–P, Green is the cell-type marker and red is propidium iodide (nuclear) staining. Scale bar: (in A) A–P, 75 μm. Q, The proportion of the total retinal cell population accounted for by each cell type was calculated and statistical significance was determined using Student's t test (p < 0.05 criterion level for significance). There were no significant differences in the numbers of ganglion cells (GC), bipolar cells (BC), cone photoreceptors (CP), rod photoreceptors (RP), Müller cells (MC), or amacrine cells (AC) between wild-type and Oc1-null retinas. However, there were significantly fewer HCs (Fig. 4) in the Oc1-null retinas compared with wild-type retinas (p < 0.001).

Since Oc1 is also highly expressed in nascent RGCs and their precursors (Wu et al., 2012), we examined the expression of six RGC markers at E14.5, a time of peak RGC production, to ascertain whether or not RGCs were affected by deletion of Oc1 (Fig. 3). These markers included Pou4f2 (Fig. 3A,B), Isl1 (Fig. 3C,D), Oc2 (Fig. 3E,F), NF160 (Fig. 3G,H), Dcx (Fig. 3I,J), and Pgp9.5 (Fig. 3K,L). With the exception of Oc2, these markers showed no change in their expression patterns and levels in Oc1-null retinas compared with wild-type controls. In the case of Oc2, whereas its expression pattern did not change, there was a marked increase in its expression level in the GCL (Fig. 3E,F), suggesting that Oc2 may compensate and function redundantly with Oc1. These observations indicate that deletion of Oc1 did not affect the development and maintenance of RGCs. Therefore before P16, when retinal cell generation is essentially complete (Young, 1985; Cepko et al., 1996), only HCs were affected in the Oc1-null retinas.

Figure 3.
Figure 3.

RGC development is not affected in Oc1−/− Retina. Six RGC markers were examined in wild-type (Oc1+/+) (A, C, E, G, I, K) and Oc1-null (Oc1−/−) (B, D, F, H, J, L) retinal sections from E14.5 mouse embryos. The markers include Pou4f2 (A, B), Isl1 (C, D), Oc2 (E, F), NF160 (G, H), Dcx (I, J), and Pgp9.5 (K, L). Except for Oc2, all other markers show no changes in either the levels or patterns of expression in Oc1-null retinas compared with wild-type controls. The expression pattern of Oc2 does not change in Oc1-null retinas, but its level is increased (F). Scale bar, (in A) A–L, 150 μm.

Oc1 is required for HC development

Calbindin is expressed by HCs, which are located in the INL adjacent to the OPL, and by subsets of amacrine cells and RGCs found in the inner INL and GCL, respectively (Fig. 4A; Pochet et al., 1989). Deletion of Oc1 affected HCs, but not amacrine cells at P16 (Fig. 4B). The number of HCs in Oc1-null retinas was reduced by 75% (± 10.2%, n = 3) compared with wild-type retinas (Fig. 2Q). To further investigate HC deficiency in the Oc1-null retina, we examined the expression of several other HC-specific markers, including Oc2, Lim1, Pgp9.5, and NF160 (Fig. 4C–J). Consistent with the reduction in calbindin-positive HCs (Fig. 4B), the numbers of HCs positive for these other markers were similarly reduced in the Oc1-null retina (Fig. 4D,F,H,J). In addition to HCs, NF160 is expressed in RGCs (Fig. 4G), while Pgp9.5 is expressed in amacrine cells (Fig. 4I), but the levels of expression of these genes in non-HCs did not change in Oc1-null retinas (Fig. 4H,J). These data confirmed that Oc1 deletion selectively affected HCs and resulted in the loss of the majority of HCs. To further examine the residual HCs in Oc1-null retinas, we double immunolabeled for Oc1 and either calbindin or Oc2 (Fig. 4A–D). As we reported previously, Oc1 and Oc2 are expressed in all HCs of wild-type retinas, and they completely overlap with calbindin in these cells (Fig. 4A,C; Wu et al., 2012). In Oc1-null retinas, the vast majority (>95%) of residual HCs were negative for Oc1 (Fig. 4B,D), but all were positive for Oc2 (Fig. 4D; data not shown). These results suggest that Oc2, instead of undeleted Oc1 due to inefficient Cre activity, was responsible for the residual HCs. To further confirm these findings and to assess the distribution and morphology of the remaining HCs, we performed whole-mount staining of HCs with anti-Lim1 and anti-calbindin antibodies (Fig. 4K–P). The same degree of HC loss (80.5 ± 0.3%, n = 4) was observed in the Oc1-null retinas (Fig. 4L,N) compared with the wild-type retinas (Fig. 4K,M). In wild-type retinas, HCs are distributed in a nonrandom mosaic pattern (Fig. 4K,M; Cook and Chalupa, 2000). In the Oc1-null retinas, this mosaic distribution pattern was maintained for the remaining HCs, although the distances between neighboring HCs became larger due to reduction of HC numbers (Fig. 4L,N). In addition, immunostaining of calbindin revealed that the HC processes were longer and thicker in the Oc1-null retinas than those in the wild-type, but the overall density of HC processes was markedly reduced (Fig. 4M–P), indicating that the remaining HCs attempted to fill in the increased space between them with their processes, but unsuccessfully.

Figure 4.
Figure 4.

The great majority of HCs are missing in Oc1-null retina at P16. A, B, Coimmunolabeling for Oc1 (red) and calbindin (green). In wild-type retina (A), cells expressing both calbindin and Oc1 (arrows) are HCs, and those expressing only calbindin in the INL and GCL are amacrine cells and RGCs. In Oc1-null retinas, very few HCs remain and are Oc1-negative (B, arrows). Amacrine cells and RGCs are not affected (B). C, D, Colabeling for Oc1 (red) and Oc2 (green). Both are expressed in all HCs in the wild-type retina (C). In Oc1-null retina, only very few HCs remain, as indicated by Oc2 staining (D). Although occasional Oc1/Oc2 double-positive HCs are observed (yellow arrow in D), the vast majority of remaining HCs are negative for Oc1, but positive for Oc2 (white arrows in D). E, F, Immunofluorescence labeling for Lim1, which is expressed only in HCs in wild-type retina (E). Lim1+ cells (green) were greatly reduced in Oc1-null retinas (F, arrows). G, H, Immunolabeling (green) for NF160, which is expressed in HC processes in the OPL (G) and RGCs. The numbers of HC processes are reduced in the Oc1-null retina (H), but there is no change in RGCs. I, J, Staining for Pgp9.5, which is expressed in HCs (I, arrows) and a subset of amacrine cells in wild-type retinas (I). In Oc1-null retinas, Pgp9.5 expression in HCs (J, arrows), but not in amacrine cells, was greatly reduced. K–P, Whole-mount staining of HCs by anti-Lim1 (K, L) and anti-calbindin (M–P). Images were obtained from equivalent regions of wild-type and Oc1-null retinas. O, P, High-magnification images from anti-calbindin staining to show the processes of HCs. Scale bars: (in A) A–J, 25 μm; (in K) K–N, 12.5 μm; (in O) O, P, 2.125 μm.

The loss of HCs in the mature Oc1-null retina could be due to either initial failure of HC birth or defective differentiation/maintenance after their birth. To distinguish between these two possibilities, we examined HC formation at E14.5 by immunofluorescence staining for Lim1 and Prox1 (Dyer et al., 2003; Pochè et al., 2007). These markers are expressed in newly born HCs, which migrate basally to the boundary between the neuroblast layer (NBL) and the GCL in wild-type retinas by E14.5 (Fig. 5A,C,E,G). Prox1 is also expressed in retinal progenitor cells (Fig. 5E,G; Dyer et al., 2003) at E14.5, but its expression in newly formed HCs is much higher and colocalizes with Oc1 (Fig. 5E,G). In E14.5 Oc1-null retinas, the number of Lim1+ HC precursors was reduced by 85% (± 3.9%, n = 3) compared with wild-type controls (Fig. 5A–D). The numbers of HC precursors strongly expressing Prox1 showed a similar reduction (Fig. 5E–H). Thus, the loss of HCs in E14.5 Oc1-null retinas was comparable to that in mature retinas. The deficiency of HC formation continued at E17.5 (Fig. 5I–P). At this stage, HCs normally are scattered through the NBL while migrating apically to their final destination, and more HC marker genes begin to be expressed (Fig. 5I,K,M,O). Examination of some of these markers, including Lim1, Oc2, Prox1, and NF160 in Oc1-null retinas, showed that the numbers of HCs remained at ∼15% of those in wild-type retinas (Fig. 5I–P). The few remaining HCs migrated normally and all expressed Oc2 (Fig. 5L), but very few expressed Oc1 (Fig. 5N,P), further supporting the idea that Oc2 was responsible for the formation of these residual HCs. These data indicated that HC deficiency in Oc1-null retina occurred at very early stages of development, with a failure of HC differentiation. To confirm this, we performed terminal deoxynucleotidyl transferase-mediated biotinylated UTP nick end labeling (TUNEL) assay to detect apoptotic cells in wild-type and Oc1-null retinas at E14.5, E17.5, and P0, which span the time window for HC formation, and observed no change in apoptosis in Oc1-null retinas compared with wild-type controls (Fig. 5Q–V). Thus increased apoptosis was not the cause of HC loss in Oc1-null retinas. Together these data strongly suggest that in Oc1-null retinas, the majority of HCs failed to form during development.

Figure 5.
Figure 5.

Loss of HCs in Oc1-null retinas occurs at early developmental stages. A–H, Immunolabeling for Lim1 and Prox1 at E14.5. Lim1 and Prox1 are expressed in newly generated HCs located at the boundary between the GCL and the NBL (A, C, E, G). Prox1 also is expressed in the NBL, but its expression in the newly generated HCs is at a much higher level and overlaps with that of Oc1 (red in E and G). In the Oc1-null retina, the numbers of Lim1+ (B, D) and Prox1+ (F, H) HCs (arrow in D and H) are greatly reduced. I–P, Immunolabeling for four HC markers at E17.5. These markers show HCs migrating to their final destination scattered throughout the NBL in the wild-type retina (I–O). In the Oc1-null retina (J–P), the numbers of labeled HCs are greatly reduced, compared with wild-type controls, but these cells still migrate normally. Note that Oc2 and NF160 are also expressed in the GCL in wild-type retinas (K, O), and that their expression in RGCs does not change in the Oc1-null retina (L, P). In addition, all HCs at E17.5 express Oc1 in wild-type retina (M, O), but few are Oc1+ in Oc1-null retinas (N, P). Q–V, TUNEL assay for wild-type and Oc1-null retinas at E14.5, E17.5, and P0. No obvious differences were observed between wild-type and mutant retinas at any of the three stages. Scale bars: (in A) A, B, E, F, Q, R, 150 μm; (in C) C, D, G, H, 25 μm; (in I) I–P, S–V, 25 μm.

Oc1 is downstream of FoxN4, but acts in parallel with Ptf1a in HC genesis

Our finding that Oc1 was required for HC genesis, and that expression of Prox1 and Lim1 were dependent on Oc1, suggested that Oc1 functions upstream of Prox1 and Lim1 during HC development. To further decipher the underlying gene regulatory cascade, we examined the relationship of Oc1 to FoxN4 and Ptf1a, two transcription factors that are also essential for the HC lineage (Li et al., 2004; Fujitani et al., 2006; Nakhai et al., 2007).

We focused our analysis on E14.5, a stage when most HCs are generated (Young, 1985; Pochè et al., 2007). At E14.5, FoxN4 is expressed in a large subset, but not all, of retinal progenitor cells (Fig. 6A,C) and is critical for the formation of HCs and most amacrine cells (Li et al., 2004). Ptf1a is also expressed in the NBL at this stage, but there are far fewer Ptf1a+ cells than FoxN4+ cells (Fig. 6E,G,I,K; Fujitani et al., 2006; Nakhai et al., 2007). Oc1 is expressed in both the NBL and GCL at E14.5 (Fig. 6C,G,K,O). Cells expressing Oc1 in the GCL are RGCs since they also express Pou4f2 and Isl1, two RGC markers (Wu et al., 2012), but the identity of Oc1-expressing cells in the NBL at E14.5 has been uncertain, since they generally do not express markers for RGCs or their precursors (Wu et al., 2012). We observed that only a small subset of FoxN4+ cells expressed Oc1, while many Oc1+ cells in the NBL were FoxN4− (Fig. 6A,C). FoxN4 expression also showed limited overlap with Ptf1a localization; only a subset of FoxN4+ cells also expressed Ptf1a, while many Ptf1a+ cells expressed little or no FoxN4 (data not shown). Coimmunolabeling for Oc1 and Ptf1a revealed that these two factors were only partially overlapping in their expression, and that many cells in the NBL were positive for Ptf1a or Oc1 alone (Fig. 6E,G,I,K).

Figure 6.
Figure 6.

Relationship of Oc1 to other transcription factor genes involved in HC genesis. A–D, FoxN4 expression pattern overlaps partially with that of Oc1 in the NBL of E14.5 wild-type retinas, but its expression does not change in Oc1-null retinas. Arrows (C) indicate some Fox4/Oc1 double-positive cells. E–H, The expression pattern of Ptf1a only partially overlaps with that of Oc1 in the NBL and its expression is independent of Oc1. Some Ptf1a/Oc1 double-positive cells are indicated by arrows (G). I–L, Both Oc1 and Ptf1a require FoxN4 for expression in the NBL. All Ptf1a+ cells and most Oc1+ cells are lost in the NBL of FoxN4-null retinas (J). M–P, Newly formed HCs coexpress Lim1 and Ptf1a in wild-type retinas (arrows, O), but HCs fail to form in Ptf1a-null retinas, as indicated by the absence of Lim1+ cells (P). Expression of Oc1 in the NBL of Ptf1a-null retinas does not change in P. Q–T, Ectopic expression by in utero electroporation of Oc1 and Ptf1a did not induce the expression of each other. There was no increase in the number of GFP+Ptf1a+ cells when Oc1 and GFP were expressed (R) compared with when only GFP was expressed (Q). Similarly, when Ptf1a was expressed (T), there was no increase in GFP+Oc1+ cells compared with the control (S). Dotted lines demarcate the NBL (top) from the GCL (bottom) in C, D, G, H, K, L, O, P. Arrows in C, G, O, and Q–T indicate double-positive cells. Scale bars: (in A) A, B, E, F, I, J, M, N, 150 μm; (in C) C, D, G, H, K, L, O–T, 25 μm.

We then examined how deletion of the gene encoding one transcription factor affected the expression of the others. In Oc1-null retinas, there was no significant change in the expression of either FoxN4 or Ptf1a (Fig. 6B,D,F,H). In FoxN4-null retinas, as reported previously (Fujitani et al., 2006), Ptf1a expression was completely abolished (Fig. 6J,L). A marked decrease of Oc1+ cells in the NBL, but not in the GCL, also was observed in FoxN4-null retinas (Fig. 6I–L). The dependence of these Oc1+ cells on FoxN4 in the NBL indicates that they were likely HC precursors. These results suggest that Oc1, like Ptf1a, functions downstream of FoxN4 in the HC lineage. Consistent with FoxN4 being upstream of Ptf1a, FoxN4 expression does not change in Ptf1a-null retinas (Fujitani et al., 2006). Interestingly, we also did not observe any changes in Oc1 expression in the NBL of Ptf1a-null retinas (Fig. 6M–P). Similar to Oc1, Ptf1a functions before HC fate determination, since expression of Lim1, an early HC marker, is already diminished at E14.5 in the Ptf1a-null retina (Fig. 6M–P). These results indicate that, during HC genesis, FoxN4 functions upstream in a large set of retinal progenitor cells, perhaps as a competence factor. Oc1 and Ptf1a operate downstream of FoxN4. The observation that many Oc1+ or Ptf1a+ cells were FoxN4− also suggests that FoxN4 expression was extinguished once Oc1 or Ptf1a was expressed. Although both Oc1 and Ptf1a are essential for HC formation, they are expressed in two distinct, but overlapping, sets of cells in the NBL and largely do not regulate each other. Since Oc1 and Ptf1a are coexpressed in only a small subset of cells, it is possible that one transcription factor can regulate the other in those cells, but this might not be detected readily due to the small size of this cell population. To examine this possibility, we ectopically expressed either Oc1 or Ptf1a by in utero electroporation, using GFP to track cells that took up the DNA constructs, and then analyzed if one factor could induce the other (Fig. 6Q–T). When Oc1 was ectopically expressed, although GFP+Ptf1+ cells could be detected occasionally, no increase of these double-positive cells was observed compared with controls expressing GFP alone (Fig. 6Q,R). Similarly, when Ptf1a was ectopically expressed, there was no induction of Oc1 in the GFP+ cells compared with the controls (Fig. 6S,T). These results further support the conclusion that Oc1 and Ptf1a are independent of one other in expression.

Oc1 and Ptf1a codetermine the HC fate

Since Ptf1a is essential for both HC and amacrine cell formation (Fujitani et al., 2006; Nakhai et al., 2007), yet Oc1 is required only for HC fate, it is likely that only progenitor cells expressing both Ptf1a and Oc1 adopt the HC fate (i.e., Ptf1a and Oc1 may cooperate to specify HC fate). To test this hypothesis, we performed in utero electroporation of plasmids expressing Oc1 and Ptf1a into wild-type retinas. A plasmid-expressing GFP was coelectroporated in all experiments to track cells that had taken up the DNA. Since both Ptf1a+ and Oc1+ cells are sparse in the NBL and are postmitotic, most cells transfected by electroporation should have only expressed the ectopic genes, but not the endogenous Oc1 or Ptf1a. If our hypothesis is correct, cells expressing only ectopic Oc1 or Ptf1a should not adopt the HC fate efficiently, whereas cells transfected with both Oc1 and Ptf1a should become HCs. Retinal electroporation was performed on E13.5 embryos, a time when HC formation starts to peak, and the fate of electroporated retinal cells was examined at E17.5 by coimmunofluorescence labeling for GFP and Lim1 or Prox1 (Fig. 7). In retinas that were electroporated with the GFP plasmid alone, only a very small subset (4.7 ± 1.6%, n = 4) of all GFP+ cells were also Lim1+ (Fig. 7A,B). The Lim1+ cells (either GFP+ or GFP−) were HCs that formed as a process of normal development. In retinas electroporated with Oc1 and GFP or Ptf1a and GFP constructs, significantly more GFP+Lim1+ cells were observed in both cases (Fig. 7E,F,I,J). In retinas electroporated with Ptf1a and GFP, 13.3% (±0.6%; n = 4, p < 0.001) of electroporated cells (GFP+) were also Lim+ (Fig. 7E,F). In retinas electroporated with Oc1 and GFP, 29.1% (±4.4%; n = 4, p < 0.001) of GFP+ cells were Lim1+ (Fig. 7I,J). Critically, when Oc1 and Ptf1a were electroporated together, most GFP+ cells (70.8 ± 2.2%; n = 4, p < 0.001) were also Lim1+ (Fig. 7M,N), indicating these cells adopted the HC fate. Similar results were observed with Prox1 (Fig. 7C,D,G,H,K,L,O,P). In the control (GFP alone) retinas, 3.6 ± 2.9% (n = 4) of GFP+ cells were Prox1+ (Fig. 7C,D). In retinas expressing Ptf1a and GFP, 2.9 ± 0.4% (n = 4) of GFP+ cells were Prox1+(Fig. 7G,H). In retinas expressing Oc1 and GFP, 6.3 ± 1.5% (n = 4) of GFP+ cells were Prox1+ (Fig. 7K,L). In retinas expressing both Ptf1a and Oc1, 62.4 ± 6.8% (n = 4) of GFP+ cells were Prox1+ (Fig. 7O,P). We were unable to further track these cells and examine whether they further developed into mature HCs, since the GFP signals were lost at later stages. This could be due to either the loss of the expression constructs from the cells or loss of the cells themselves since they were ectopically produced. Nevertheless, we conclude that either Ptf1a or Oc1 is very inefficient when acting alone to promote HC fate. When both Ptf1a and Oc1 were present, however, most cells adopted the HC fate. Therefore, these two factors appear to be not only necessary but also sufficient to direct retinal progenitor cells to adopt the HC fate.

Figure 7.
Figure 7.

Oc1 and Ptf1a are sufficient to specify HC fate. Retinal sections from embryos electroporated with plasmid constructs expressing GFP, Oc1, and Ptf1a were double immunolabeled with anti-GFP (green) and anti-Lim1 or anti-Prox1 (red). A–D, In retinas electroporated with the construct for GFP alone, few GFP+ cells express Lim1 or Prox1. E–H, When constructs for GFP and Ptf1a were coelectroporated into the retina, some GFP+ cells are also Lim1+, but few were Prox+. I–L, Coelectroporation of constructs for Oc1 and GFP also resulted in only a few GFP+Lim1+ or GFP+Prox1+ cells. M–P, Electroporation of constructs for Oc1, Ptf1a, and GFP resulted in most GFP+ cells also being Lim1+ and Prox1+. Arrows point to double-positive cells. Scale bar, (in A) A–P, 25 μm.

Loss of HCs leads to defects in the synaptic connections in the OPL

The OPL is comprised of photoreceptor terminals, and the processes of HCs and bipolar cells. Photoreceptors form synapses with HCs and bipolar cells in a stereotypical fashion (Olney, 1968; Blanks et al., 1974), but the developmental mechanisms that regulate formation of the complex synaptic interactions remain unclear. Because the first synapses to develop in the OPL are those between photoreceptors and HCs (Olney, 1968; Blanks et al., 1974), an obvious question is what role HCs play in the formation of synapses between photoreceptors and bipolar cells. The deficiency of HCs in Oc1-null retinas allowed us to examine this question directly by TEM.

In the wild-type retina, rod and cone synaptic terminals stratify in the outer half of the OPL, with the cone terminals residing more proximally (vitread) than most rod terminals. The inner half of the OPL is occupied by the processes of HCs and bipolar cells (Fig. 8A; Olney, 1968; Blanks et al., 1974). We first performed TEM on P30 retinas. In the Oc1-null retinas, the rod and cone terminals did not exhibit normal stratification and were disorganized (Fig. 8B–D). Furthermore, photoreceptor terminals in the Oc1-null retina often penetrated deep into the inner portion of the OPL, sometimes extending all the way to the border with the INL (Fig. 8C). Rod cell bodies and cells from the INL also could be found penetrating into the OPL (Fig. 8B–D). In wild-type retinas, rods, bipolar cells, and HCs formed the classic “triad” ribbon synaptic complexes (Fig. 8E). However, in Oc1-null retinas, typical triad ribbon synapses were rarely seen, although they possibly could exist at very low frequency (Fig. 8F–I). However, synaptic ribbons were found in Oc1-null rod terminals, but much more rarely compared with the wild-type, and they typically were not anchored to the plasma membrane, i.e., they appeared to be “free floating” (Fig. 8G,H). Despite this, invaginations of postsynaptic processes sometimes were observed in rod terminals of the Oc1-null retina, although they did not show the classic triad organization (Fig. 8F). Wild-type cone terminals normally show multiple triadic ribbon synapses organized around short synaptic ribbons, in addition to flat contacts located on the base of the terminal that represent contacts with OFF-cone bipolar cell processes (Fig. 8J). In contrast, in Oc1-null cone terminals, no synaptic ribbons were seen, but invaginations of postsynaptic processes were present and flat contacts between cone terminals and OFF-cone bipolar cell dendrites also were observed (Fig. 8K,L). In addition, extensive tubulovesicular cisternae were observed in both rod and cone terminals (Fig. 8F,I,M). To rule out the possibility that the observed defects were secondary, we also performed TEM at P16, a stage immediately after completion of synapse formation in the OPL. At P16, HC plexus in the developing OPL was already severely reduced in the Oc1-null retina as show by anti-NF160 staining (Fig. 4H). Consistently, TEM revealed that the OPL of P16 Oc1-null retina showed defects comparable to those observed at P30, including less frequent and aberrantly positioned ribbons; absence of clear triadic organization to the postsynaptic complexes; and contact of at least some postsynaptic cells with photoreceptor terminals (data not shown).

Figure 8.
Figure 8.

The deficiency of HCs disrupts synaptic ultrastructure and organization of the OPL in Oc1-null retinas. A–D, Low-magnification (5000×) TEM images of the OPL from P30 wild-type (WT; A) and Oc1-null mutant mice (MT; B–D). The boundaries of the OPL are indicated by dotted lines. A, In the WT retina, the OPL is of uniform thickness and is well laminated, with the synaptic terminals of cones and rods localized to the outer (sclerad) portion and processes from secondary neurons localized to the inner (vitread) portion. B–D, In the MT retinas, this laminar structure of the OPL is disrupted, with aberrant localization of processes from photoreceptors and bipolar cells. The thickness of the OPL in the MT retina varies from region to region, resulting in irregular boundaries. In addition, intruding bipolar cell (B, C) and photoreceptor cell bodies (D) can be found in the OPL. E–I, High-magnification (30,000×) images of rod terminals. J–M, High-magnification images of cone terminals. E, WT rod terminals show normal synaptic ultrastructure with a large synaptic ribbon (arrowhead) in the presynaptic terminal and a postsynaptic “triad” comprised of two HC processes (H) flanking the ribbon laterally and a rod bipolar dendrite (B) occupying the central position. F, Rod terminals in the MT retina contain numerous synaptic vesicles but typically do not show synaptic ribbons. However, invaginations containing postsynaptic processes (arrowheads) are present. Arrows point to tubulovesicular cisternae. F–I, In rod terminals from MT retinas, “free-floating,” amorphous ribbon material (arrowhead in G), abnormal synaptic ribbons devoid of associated synaptic vesicles (arrowhead in H), and tubulovesicular cisternae (I, arrows) can be found. J, WT cone terminals show invaginations of postsynaptic processes forming multiple synaptic complexes organized around short synaptic ribbons (arrowheads). Flat contacts (FC) are present along the base of the cone terminal. K, L, Cone terminals in the MT retina. No synaptic ribbons are present, although invaginations of postsynaptic processes (arrowheads, in K) and FC (L) are present, but are sometimes ectopically positioned (L). M, Extensive tubulovesicular cisternae (arrows) in cone terminals in the MT retina. N–S, Colabeling using anti-Bassoon and PNA to visualize synaptic ribbons (red) and FCs (green), respectively, in the OPL of P16 WT and MT retinas. In the OPL of the MT retina, the number of synaptic ribbons is markedly reduced, compared with controls; there are also fewer FCs (arrowheads). T–Y, Colabeling of synaptic ribbons and rod bipolar cells using anti-Bassoon and anti-PKCα, respectively. In the WT retina, processes of rod bipolar cells contact rod terminals but do not extend beyond the OPL into ONL. In the MT retina, rod bipolar cell dendrites still make contacts with rod terminals, but in some areas project ectopically into the ONL. Scale bars: (in A) A–D, 2 μm; E–I, L, M, 500 nm; J, K, 2 μm; (in N) N–Y, 10 μm.

These results suggest that, in the reduction of HC processes, synaptic ribbons did not anchor properly to the plasma membrane, leading to the loss of the stereotypical triadic organization of postsynaptic processes associated with ribbon synapses of both rods and cones. In contrast, flat contacts between cone terminals and OFF-cone bipolar cells still formed. The appearance of the large cisternae in the rod and cone terminals of the Oc1-null retina may reflect a functional change in membrane trafficking in these terminals.

To extend our TEM observations, we immunolabeled synaptic ribbons with anti-Bassoon (Fig. 8N,Q,T,W) and flat contacts with peanut agglutinin (PNA; Fig. 8O,P,R,S). Consistent with the TEM data, many fewer synaptic ribbons were observed in the OPL of Oc1-null retinas (Fig. 8Q,W), and these ribbons tended to be shorter than those in wild-type retina. In addition, some regions in the OPL were devoid of synaptic ribbons (Fig. 8Q,W). Staining with PNA indicated that flat contacts between cones and OFF-cone bipolar cells were present in the OPL of Oc1-null retinas, but fewer in number, compared with wild-type controls (Fig. 8Q,R).

In addition, we performed immunolabeling with antibodies to PKCα to visualize rod bipolar cell dendrites in the OPL (Fig. 8U–Y). In the wild-type retina, rod bipolar cell dendrites contacted rod terminals and were confined within the OPL, as expected (Fig. 8U,V). In the Oc1-null retina, contacts between rod bipolar cell dendrites and photoreceptor terminals were present, but infrequently (Fig. 8X,Y). Furthermore, rod bipolar cell dendrites often showed abnormal sprouting and invasion of the ONL, consistent with disruption of the normal synaptic connections of the OPL (Fig. 8X,Y). These results indicated that although loss of HCs in the Oc1-null retinas severely disrupted photoreceptor terminals and their synaptic connections in the OPL, some contacts between photoreceptors and bipolar cells persisted.

HCs are required for retinal integrity and photoreceptor survival

To further evaluate how loss of HCs affected the function of Oc1-null retinas, we performed flash ERG recordings on 5-month-old mice under dark-adapted conditions (Fig. 9A). The corneal-negative a-wave of the ERG provides information about phototransduction (Pugh et al., 1998) and precedes the corneal-positive b-wave that originates from bipolar cell activity in the inner retina (Robson et al., 2004). The a- and b-wave amplitudes recorded from Oc1-null retinas were much smaller than those recorded from control retinas, but the kinetics of the responses, as inferred by the time to peak of the b-wave, was not affected. However, the Oc1-null retinas did not exhibit the robust oscillatory potentials, suggesting that the function of the inner retina also was impaired (Wachtmeister, 1998). The intensity-response functions (Fig. 9B,C) show that the a- and b-waves in Oc1-null retina were uniformly attenuated (by threefold) over much of the intensity range, indicating that the loss of HCs nonselectively impairs retinal function over the entire intensity range. The increase of the a-wave at –2 log sc cd s/m2 probably resulted from the relative reduction of the b-wave and unmasking of the a-wave.

Figure 9.
Figure 9.

Oc1-null mice exhibit abnormal ERGs and undergo retinal degeneration. A–C, Dark-adapted ERG. A, Average ERGs recorded from Oc1-null (MT) and littermate control (WT) retinas at 5 months of age in response to brief (4 ms) flashes. Flash intensities are indicated to the left of each trace. B, C, Intensity-response functions of b- and a-waves, respectively, recorded from Oc1-null mice (MT, red) and their wild-type (WT) control siblings (black) at 5 months of age. Gray dashed lines indicate the response of WT mice divided by a factor of 3. Error bars indicate SEM, n = 4 in each group. D, E, Light-adapted ERGs. D, Average ERGs recorded from Oc1-null (MT) and control retinas (WT) in response to brief (4 ms) flashes superimposed on a steady background of 10 cd/m2. Intensities of the flashes are indicated in the left column. E, Intensity-response functions of b-waves recorded from OC1-null mice (MT) and their wild-type (WT) siblings at 5 months of age. Gray dashed lines indicate the response of WT mice divided by a factor of 3. Bars indicate SEM, n = 4 in each group. F–K, Histological analysis of H&E stained wild-type (F, H, J) and Oc1-null (G, I, K) retinal sections from mice at 3 months (F, G), 5 months (H, I), and 8 months (J, K) of age shows progressive loss of photoreceptors in the Oc1-null retina. See text for detailed description. Scale bar, (in F) F–K, 25 μm.

HCs provide lateral inhibitory input to photoreceptors as well as bipolar cells and are thought to contribute to light adaptation (Werblin and Dowling, 1969; Wu, 1992; Thoreson and Mangel, 2012). To determine whether light adaptation was impaired in Oc1-null retinas we recorded flash ERG responses in the presence of steady background lights of 10 cd/m2. These light levels produce ∼8000 photoisomerizations/rod/s (Umino et al., 2012), which is sufficient to saturate rods (Nakatani et al., 1991) and induce light adaptation in cones (Dunn et al., 2007). We found that background illumination sped up the ERG response, reduced the amplitude of the b-wave, and limited the range of both wild-type and Oc1-null retinas to flash intensities >−1 log cd s/ m2 (Fig. 9D). However, much as in dark-adapted conditions, the b-waves of Oc1-null retinas were ∼3-fold smaller than those of control retinas over the entire intensity range (Fig. 9D,E). These results suggest that the cone-driven light responses in Oc1-null retinas are attenuated, although their light adaptive mechanisms appeared to remain intact.

Since we did not observe overt structural abnormalities in either the ONL or INL of Oc1-null retinas at P16, the impaired ERGs at 5 months of age prompted us to examine whether Oc1-null retinas underwent degeneration as the mice aged. At 3 months of age, the OPL was similarly defective as at P16. However, there were no obvious defects in the other layers of the Oc1-null retinas (Fig. 9F,G). In contrast, at 5 months, the OPL in Oc1-null retinas had virtually disappeared, while the thickness of the other retinal layers remained comparable to those in wild-type retinas. Some early signs of degeneration were noticeable in the ONL at this time: the ONL was slightly thinner and the nuclei in the ONL appeared less tightly packed than those in wild-type retinas (Fig. 9H,I). By 8 months of age, degeneration of photoreceptors was obvious in Oc1-null retinas, with the thickness of the ONL being only approximately one-third that of wild-type retinas (Fig. 9J,K). The degeneration of the ONL in Oc1-null retinas likely contributed to the reduction of the ERG a-waves under dark-adapted conditions. Nevertheless, the thickness of the INL did not change significantly, suggesting that the observed reduction in the ERG b-waves was due to diminished photoreceptor input, rather than bipolar cell loss. The observed age-dependent degeneration and loss of photoreceptors must be secondary to the loss of HCs in the Oc1-null retina, since photoreceptors do not express Oc1 themselves and formed normally. Thus, in addition to processing visual signals, HCs are also essential for the postnatal survival of photoreceptors.

Discussion

Roles of Oc1 in HC development and its potential redundancy with Oc2

We previously showed that Oc1 is expressed in developing HCs (Wu et al., 2012), but it was not clear from that study the stage at which Oc1 is involved in HC development. Nevertheless, at E14.5, Oc1+ cells, which are mostly postmitotic and typically do not coexpress factors that regulate the RGC lineage, exist in the NBL and their identities heretofore were not known (Wu et al., 2012). Our finding that deletion of Oc1 results in a loss of ∼80% of HCs and that this loss takes place early during HC development suggests that Oc1 is required for the initial genesis of HCs. The lack of increased apoptosis in the Oc1-null retinas supports this idea. Further, we found that the Oc1+ cells located in the NBL are dependent on FoxN4. This observation, in combination with the deficient HC development in both FoxN4-null and Oc1-null retinas, indicates that these Oc1+ cells in the E14.5 NBL are HC precursors and arise from FoxN4+ progenitor cells. Upon expressing Oc1, they become committed to the HC fate and migrate inward to the boundary of the NBL and GCL, where they undergo further differentiation and begin expressing mature HC markers, including Lim1 and Prox1. Oc2 exhibits very similar expression patterns to those of Oc1 in the developing retina (Wu et al., 2012). Since Oc1 and Oc2 are highly similar in their DNA-binding domains and bind to the same consensus sequence (Iyaguchi et al., 2007), it is possible that Oc1 and Oc2 function redundantly by regulating the same target genes in the developing retina. They both are expressed in all HCs during development and in adulthood. However, Oc2 is expressed in a much smaller cohort of HC precursors in the NBL during early development than is Oc1 (Wu et al., 2012). Consistent with this, only ∼20% of HCs still form in Oc1-null retinas and all of these remaining HCs express Oc2, suggesting that Oc2 is also involved in HC development.

Both Oc1 and Oc2 also are expressed in developing RGCs (Wu et al., 2012). However, we did not observe any RGC defects in the Oc1-null retina either during development or in adulthood. This also may be due to redundancy with Oc2, as suggested by the upregulation of Oc2 in Oc1-null retinas. Examining RGC development in the Oc2-null and Oc1/Oc2-double-null retinas in the future will help resolve this.

The gene regulatory cascade for HC development

The expression patterns and knock-out phenotypes of genes involved in HC development suggests a stepwise scenario of HC genesis (Fig. 10A). In this scenario, FoxN4 is most upstream and is expressed in a large population of, but not all, proliferating progenitor cells (Fig. 10A, 1). The function of FoxN4 is to render the progenitor cells competent for HC and amacrine cell fates. From these FoxN4-expressing cells emerge two distinct, but overlapping, cell populations: those expressing Ptf1a and those expressing Oc1 (Fig. 10A, 2). Although both Ptf1a and Oc1 are essential for the HC fate, neither is sufficient individually. As indicated by in utero electroporation, most, if not all, cells expressing both factors will eventually adopt the HC fate (Fig. 10A, 2,3), i.e., Oc1 and Ptf1a act together to determine the HC fate. Those precursor cells expressing Ptf1a alone become amacrine cells. In the Oc1-null retinas, precursor cells that normally adopt the HC fate may switch fate and become amacrine cells, since they now only express Ptf1a; however, this remains to be demonstrated unequivocally by further experiments. At present, it is not clear what fate(s) cells expressing only Oc1 adopt and what happens to them when Oc1 is inactivated. The HC fate-committed cells then migrate to the basal side of the retina at ∼E14.5 (Fig. 10A, 4) and begin to express transcription factors such as Lim1, Prox1, Sall3, and AP-2α and β, which regulate subsequent HC migration and differentiation (Dyer et al., 2003; Pochè et al., 2007; de Melo et al., 2011; Bassett et al., 2012). At later stages (e.g., E16.5), the immature HCs begin to migrate apically (Fig. 10A, 5), reach their final destination in the retina at ∼P0 (Fig. 10A, 6), and differentiate into mature HCs, ultimately becoming part of visual circuitry. Clearly, HC formation is subject to regulation by a hierarchical pathway comprised of factors that include FoxN4, Ptf1a, Oc1, Lim1, Prox1, SalI3, and AP-2α and β (Fig. 10B). Thus, our results identify Oc1 as an essential player in this pathway, placing it downstream of FoxN4, in parallel with Ptf1a, but upstream of Lim1, Prox1, Sall3, and AP-2α and β (Fig. 10B). It should be noted that Ptf1a inhibits RGC formation by repressing Math5 (Fujitani et al., 2006; Lelièvre et al., 2011), and is also required for amacrine cell formation(Fujitani et al., 2006; Nakhai et al., 2007; Jusuf et al., 2011), indicating that this regulatory cascade also interacts with gene regulatory pathways for other retinal cell types (Fig. 10B). Currently it is not clear how Oc1 and Ptf1a collaborate to specify the HC fate. Interestingly, both Ptf1a and Oc1 are expressed in the multipotent pancreatic progenitor cells and they coregulate the expression of Pdx1, a critical transcription factor for pancreas development (Jacquemin et al., 2006). Ptf1a and Oc1 may coregulate downstream genes through a shared mechanism in the two tissues.

Figure 10.
Figure 10.

Summary diagram, illustrating key transcriptional regulators that regulate HC genesis. A, Model for the genesis of HCs. 1, Expression of FoxN4 (blue) in a large subset, but not all, of retinal progenitor cells confers competence toward the HC fate. For simplicity, FoxN4− cells and the GCL are not depicted. 2, From HC-competent cells emerge two populations of cells: one expressing Ptf1a (red) and the other expressing Oc1 (green). Only a small subset of cells coexpress both of these factors (yellow cells). 3, Only those cells that express both Ptf1a and Oc1 (yellow) adopt the HC fate. 4, Fate-determined HC precursors migrate basally and begin to express Lim1, Prox1, and Sall3 (pink). 5, These cells then migrate apically and begin to express other HC markers. 6, Nascent migrating HCs reach their final destination and differentiate into mature HCs. B, Schematic illustration of the gene regulatory cascade underlying HC formation. FoxN4 is the most upstream factor, Ptf1a and Oc1 are downstream of FoxN4 and in parallel to each other, and Lim1 and Prox1 are downstream of Oc1 and Ptf1a. Ptf1a directs cells toward an amacrine cell fate in the absence of Oc1, and inhibits the ganglion cell fate by repressing Math5.

HCs are essential for retinal integrity and function, and photoreceptor viability

During normal development, HC processes are the first to contact photoreceptors; this leads to the recruitment of nascent synaptic ribbons to the site of contact, after which a second HC process is recruited and the synaptic complex begins to invaginate, followed by the subsequent addition of ON-type bipolar cell dendrites into the central position of the synaptic “triad” (Blanks et al., 1974). In Oc1-null retinas, this process is severely disrupted, although in some cases bipolar cell processes can still contact photoreceptor terminals and invaginations can still form. However, similar to the Bassoon knock-out retina (Dick et al., 2003), synaptic ribbons in the Oc1-null retina are rarely recruited to sites of contact with postsynaptic processes and formation of the stereotypical triadic synaptic complexes is impeded. The presence of these defects at p16, which coincides with the completion of photoreceptor synapse formation (Blanks et al., 1974) indicates that HCs are important for the establishment of normal synaptic connections in the OPL. The disruption of connectivity is entirely consistent with the abnormal ERG responses in the Oc1-null retina and indicates that the deficiency of HCs leads to severely compromised transmission of signals from photoreceptors to bipolar cells and the inner retina. Although it is clear that the synaptic organization of the OPL is disrupted in the Oc1-null retina, several important issues remain to be resolved including the extent of contacts, if any, that form between photoreceptors and HCs, and whether HC deficiency disrupts the specificity of contact formation between photoreceptors and the various classes of bipolar cells. All these issues may be addressed by further analyzing the Oc1-null retinas, as well as the Oc1/Oc2-double-null retinas we have just created, which are expected to have more severe HC defects, at stages of synaptogenesis in the OPL during development. These future studies may reveal the exact roles HCs play during synaptogenesis in the OPL.

In addition, HCs are also essential for the integrity and maintenance of the photoreceptors in the adult retina. Because Oc1-null retinas are largely normal when the mice are young, deficiency of HC formation does not affect the formation of the rest the retina on a large scale. However, severe retinal degeneration develops as the mice age. This is manifested mainly in the loss of photoreceptors, revealing a previously unrecognized but essential role for HCs in the survival of photoreceptors. Consequently, there were severe reductions in the a- and b-waves of the ERGs from these mice, although the change in the b-wave also likely arose from the severe compromise of synapse formation and the aberrant development of the presynaptic transmitter release machinery of the photoreceptor terminals due to reduction of HCs. Consistent with our findings, a recent study using a mouse model in which HCs were ablated by diphtheria toxin A revealed a similar degeneration of rod photoreceptors (Sonntag et al., 2012). The mechanism by which HCs promote and preserve photoreceptor viability is as yet unknown. It is possible HCs provide essential trophic factors for the survival of photoreceptors. Alternatively, electrical activity and/or feedback neurotransmission from HCs to photoreceptors may be required to maintain photoreceptor viability. Further investigations will be needed to identify the actual mechanisms involved. Once discovered, augmentation of these endogenous processes may provide a novel means of preventing photoreceptors from degenerating and dying in diseased retinas. Although retinal degeneration caused by HC deficiency has not been reported in humans, this model may have relevance to some types of hereditary retinal degenerations for which the underlying mechanisms are not yet understood and the genes involved have not been identified.

Footnotes

  • This project was supported by grants from the Whitehall Foundation (X.M.); the National Eye Institute (EY020545, X.M.; EY007361, S.J.F.; EY012190 to University of Oklahoma Health Sciences Center; EY020849 and EY012020, M.X.); the SUNY/RF Research Collaboration Fund (X.M., S.J.F., and E.S.); Unrestricted Grants from Research to Prevent Blindness to the Department of Ophthalmology of University at Buffalo (X.M. and S.J.F.) and to the Department of Ophthalmology at SUNY Upstate Medical University (E.S.); Oklahoma Center for the Advancement of Science and Technology (OCAST HR08-149S, D.M.S.); the Lions of Central New York (E.S.); and resources and facilities provided by the Veterans Administration Western NY Healthcare System (S.J.F.). The views expressed herein are not necessarily those of the Veterans Administration. We thank Drs. Jane Johnson and Helena Edlund for kindly providing the anti-Ptf1a antibodies, Barbara A. Nagel for expert technical assistance with TEM, and members of the Developmental Genomics Group at the New York State Center of Excellence in Bioinformatics and Life Sciences and the Department of Ophthalmology at the University of Buffalo for helpful discussions.

  • The authors declare no competing financial interests.

  • Correspondence should be addressed to Xiuqian Mu, 701 Ellicott Street, Buffalo, NY 14203. xmu{at}buffalo.edu

References

  1. ↵
    1. Agathocleous M,
    2. Harris WA
    (2009) From progenitors to differentiated cells in the vertebrate retina. Annu Rev Cell Dev Biol 25:45–69, doi:10.1146/annurev.cellbio.042308.113259, pmid:19575661.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Bassett EA,
    2. Korol A,
    3. Deschamps PA,
    4. Buettner R,
    5. Wallace VA,
    6. Williams T,
    7. West-Mays JA
    (2012) Overlapping expression patterns and redundant roles for AP-2 transcription factors in the developing mammalian retina. Dev Dyn 241:814–829, doi:10.1002/dvdy.23762, pmid:22411557.
    OpenUrlCrossRefPubMed
  3. ↵
    1. Blanks JC,
    2. Adinolfi AM,
    3. Lolley RN
    (1974) Synaptogenesis in the photoreceptor terminal of the mouse retina. J Comp Neurol 156:81–93, doi:10.1002/cne.901560107, pmid:4836656.
    OpenUrlCrossRefPubMed
  4. ↵
    1. Cepko CL,
    2. Austin CP,
    3. Yang X,
    4. Alexiades M,
    5. Ezzeddine D
    (1996) Cell fate determination in the vertebrate retina. Proc Natl Acad Sci U S A 93:589–595, doi:10.1073/pnas.93.2.589, pmid:8570600.
    OpenUrlAbstract/FREE Full Text
  5. ↵
    1. Cervetto L,
    2. MacNichol EF Jr.
    (1972) Inactivation of horizontal cells in turtle retina by glutamate and aspartate. Science 178:767–768, doi:10.1126/science.178.4062.767, pmid:5082843.
    OpenUrlAbstract/FREE Full Text
  6. ↵
    1. Cook JE,
    2. Chalupa LM
    (2000) Retinal mosaics: new insights into an old concept. Trends Neurosci 23:26–34, doi:10.1016/S0166-2236(99)01487-3, pmid:10631786.
    OpenUrlCrossRefPubMed
  7. ↵
    1. de Melo J,
    2. Peng GH,
    3. Chen S,
    4. Blackshaw S
    (2011) The Spalt family transcription factor Sall3 regulates the development of cone photoreceptors and retinal horizontal interneurons. Development 138:2325–2336, doi:10.1242/dev.061846, pmid:21558380.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    1. Dick O,
    2. tom Dieck S,
    3. Altrock WD,
    4. Ammermüller J,
    5. Weiler R,
    6. Garner CC,
    7. Gundelfinger ED,
    8. Brandstätter JH
    (2003) The presynaptic active zone protein bassoon is essential for photoreceptor ribbon synapse formation in the retina. Neuron 37:775–786, doi:10.1016/S0896-6273(03)00086-2, pmid:12628168.
    OpenUrlCrossRefPubMed
  9. ↵
    1. Ding XQ,
    2. Nour M,
    3. Ritter LM,
    4. Goldberg AF,
    5. Fliesler SJ,
    6. Naash MI
    (2004) The R172W mutation in peripherin/rds causes a cone-rod dystrophy in transgenic mice. Hum Mol Genet 13:2075–2087, doi:10.1093/hmg/ddh211, pmid:15254014.
    OpenUrlAbstract/FREE Full Text
  10. ↵
    1. Dunn FA,
    2. Lankheet MJ,
    3. Rieke F
    (2007) Light adaptation in cone vision involves switching between receptor and post-receptor sites. Nature 449:603–606, doi:10.1038/nature06150, pmid:17851533.
    OpenUrlCrossRefPubMed
  11. ↵
    1. Dyer MA,
    2. Livesey FJ,
    3. Cepko CL,
    4. Oliver G
    (2003) Prox1 function controls progenitor cell proliferation and horizontal cell genesis in the mammalian retina. Nat Genet 34:53–58, doi:10.1038/ng1144, pmid:12692551.
    OpenUrlCrossRefPubMed
  12. ↵
    1. Espana A,
    2. Clotman F
    (2012a) Onecut transcription factors are required for the second phase of development of the A13 dopaminergic nucleus in the mouse. J Comp Neurol 520:1424–1441, doi:10.1002/cne.22803, pmid:22102297.
    OpenUrlCrossRefPubMed
  13. ↵
    1. Espana A,
    2. Clotman F
    (2012b) Onecut factors control development of the Locus Coeruleus and of the mesencephalic trigeminal nucleus. Mol Cell Neurosci 50:93–102, doi:10.1016/j.mcn.2012.04.002, pmid:22534286.
    OpenUrlCrossRefPubMed
  14. ↵
    1. Francius C,
    2. Clotman F
    (2010) Dynamic expression of the Onecut transcription factors HNF-6, OC-2 and OC-3 during spinal motor neuron development. Neuroscience 165:116–129, doi:10.1016/j.neuroscience.2009.09.076, pmid:19800948.
    OpenUrlCrossRefPubMed
  15. ↵
    1. Fu X,
    2. Sun H,
    3. Klein WH,
    4. Mu X
    (2006) Beta-catenin is essential for lamination but not neurogenesis in mouse retinal development. Dev Biol 299:424–437, doi:10.1016/j.ydbio.2006.08.015, pmid:16959241.
    OpenUrlCrossRefPubMed
  16. ↵
    1. Fu X,
    2. Kiyama T,
    3. Li R,
    4. Russell M,
    5. Klein WH,
    6. Mu X
    (2009) Epitope-tagging Math5 and Pou4f2: new tools to study retinal ganglion cell development in the mouse. Dev Dyn 238:2309–2317, doi:10.1002/dvdy.21974, pmid:19459208.
    OpenUrlCrossRefPubMed
  17. ↵
    1. Fujitani Y,
    2. Fujitani S,
    3. Luo H,
    4. Qiu F,
    5. Burlison J,
    6. Long Q,
    7. Kawaguchi Y,
    8. Edlund H,
    9. MacDonald RJ,
    10. Furukawa T,
    11. Fujikado T,
    12. Magnuson MA,
    13. Xiang M,
    14. Wright CV
    (2006) Ptf1a determines horizontal and amacrine cell fates during mouse retinal development. Development 133:4439–4450, doi:10.1242/dev.02598, pmid:17075007.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Furuno K,
    2. Ikeda K,
    3. Hamano S,
    4. Fukuyama K,
    5. Sonoda M,
    6. Hara T,
    7. Sasazuki T,
    8. Yamamoto K
    (2008) Onecut transcription factor OC2 is a direct target of T-bet in type-1 T-helper cells. Genes Immun 9:302–308, doi:10.1038/gene.2008.18, pmid:18418398.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Furuta Y,
    2. Lagutin O,
    3. Hogan BLM,
    4. Oliver GC
    (2000) Retina- and ventral forebrain-specific Cre recombinase activity in transgenic mice. Genesis 26:130–132, doi:10.1002/(SICI)1526-968X(200002)26:2<130::AID-GENE9>3.3.CO%3B2-9, pmid:10686607.
    OpenUrlCrossRefPubMed
  20. ↵
    1. Iyaguchi D,
    2. Yao M,
    3. Watanabe N,
    4. Nishihira J,
    5. Tanaka I
    (2007) DNA recognition mechanism of the ONECUT homeodomain of transcription factor HNF-6. Structure 15:75–83, doi:10.1016/j.str.2006.11.004, pmid:17223534.
    OpenUrlCrossRefPubMed
  21. ↵
    1. Jacquemin P,
    2. Lemaigre FP,
    3. Rousseau GG
    (2003) The Onecut transcription factor HNF-6 (OC-1) is required for timely specification of the pancreas and acts upstream of Pdx-1 in the specification cascade. Dev Biol 258:105–116, doi:10.1016/S0012-1606(03)00115-5, pmid:12781686.
    OpenUrlCrossRefPubMed
  22. ↵
    1. Jacquemin P,
    2. Yoshitomi H,
    3. Kashima Y,
    4. Rousseau GG,
    5. Lemaigre FP,
    6. Zaret KS
    (2006) An endothelial-mesenchymal relay pathway regulates early phases of pancreas development. Dev Biol 290:189–199, doi:10.1016/j.ydbio.2005.11.023, pmid:16386727.
    OpenUrlCrossRefPubMed
  23. ↵
    1. Jusuf PR,
    2. Almeida AD,
    3. Randlett O,
    4. Joubin K,
    5. Poggi L,
    6. Harris WA
    (2011) Origin and determination of inhibitory cell lineages in the vertebrate retina. J Neurosci 31:2549–2562, doi:10.1523/JNEUROSCI.4713-10.2011, pmid:21325522.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. Kay JN,
    2. De la Huerta I,
    3. Kim IJ,
    4. Zhang Y,
    5. Yamagata M,
    6. Chu MW,
    7. Meister M,
    8. Sanes JR
    (2011) Retinal ganglion cells with distinct directional preferences differ in molecular identity, structure, and central projections. J Neurosci 31:7753–7762, doi:10.1523/JNEUROSCI.0907-11.2011, pmid:21613488.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    1. Landry C,
    2. Clotman F,
    3. Hioki T,
    4. Oda H,
    5. Picard JJ,
    6. Lemaigre FP,
    7. Rousseau GG
    (1997) HNF-6 is expressed in endoderm derivatives and nervous system of the mouse embryo and participates to the cross-regulatory network of liver-enriched transcription factors. Dev Biol 192:247–257, doi:10.1006/dbio.1997.8757, pmid:9441665.
    OpenUrlCrossRefPubMed
  26. ↵
    1. Lelièvre EC,
    2. Lek M,
    3. Boije H,
    4. Houille-Vernes L,
    5. Brajeul V,
    6. Slembrouck A,
    7. Roger JE,
    8. Sahel JA,
    9. Matter JM,
    10. Sennlaub F,
    11. Hallböök F,
    12. Goureau O,
    13. Guillonneau X
    (2011) Ptf1a/Rbpj complex inhibits ganglion cell fate and drives the specification of all horizontal cell subtypes in the chick retina. Dev Biol 358:296–308, doi:10.1016/j.ydbio.2011.07.033, pmid:21839069.
    OpenUrlCrossRefPubMed
  27. ↵
    1. Lemaigre FP,
    2. Durviaux SM,
    3. Truong O,
    4. Lannoy VJ,
    5. Hsuan JJ,
    6. Rousseau GG
    (1996) Hepatocyte nuclear factor 6, a transcription factor that contains a novel type of homeodomain and a single cut domain. Proc Natl Acad Sci U S A 93:9460–9464, doi:10.1073/pnas.93.18.9460, pmid:8790352.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    1. Li S,
    2. Mo Z,
    3. Yang X,
    4. Price SM,
    5. Shen MM,
    6. Xiang M
    (2004) Foxn4 controls the genesis of amacrine and horizontal cells by retinal progenitors. Neuron 43:795–807, doi:10.1016/j.neuron.2004.08.041, pmid:15363391.
    OpenUrlCrossRefPubMed
  29. ↵
    1. Margagliotti S,
    2. Clotman F,
    3. Pierreux CE,
    4. Beaudry JB,
    5. Jacquemin P,
    6. Rousseau GG,
    7. Lemaigre FP
    (2007) The Onecut transcription factors HNF-6/OC-1 and OC-2 regulate early liver expansion by controlling hepatoblast migration. Dev Biol 311:579–589, doi:10.1016/j.ydbio.2007.09.013, pmid:17936262.
    OpenUrlCrossRefPubMed
  30. ↵
    1. Masland RH
    (2001) The fundamental plan of the retina. Nat Neurosci 4:877–886, doi:10.1038/nn0901-877, pmid:11528418.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Matsuda T,
    2. Cepko CL
    (2004) Electroporation and RNA interference in the rodent retina in vivo and in vitro. Proc Natl Acad Sci U S A 101:16–22, doi:10.1073/pnas.2235688100, pmid:14603031.
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Mu X,
    2. Fu X,
    3. Sun H,
    4. Liang S,
    5. Maeda H,
    6. Frishman LJ,
    7. Klein WH
    (2005) Ganglion cells are required for normal progenitor- cell proliferation but not cell-fate determination or patterning in the developing mouse retina. Curr Biol 15:525–530, doi:10.1016/j.cub.2005.01.043, pmid:15797020.
    OpenUrlCrossRefPubMed
  33. ↵
    1. Mu X,
    2. Fu X,
    3. Beremand PD,
    4. Thomas TL,
    5. Klein WH
    (2008) Gene regulation logic in retinal ganglion cell development: Isl1 defines a critical branch distinct from but overlapping with Pou4f2. Proc Natl Acad Sci U S A 105:6942–6947, doi:10.1073/pnas.0802627105, pmid:18460603.
    OpenUrlAbstract/FREE Full Text
  34. ↵
    1. Nakatani K,
    2. Tamura T,
    3. Yau KW
    (1991) Light adaptation in retinal rods of the rabbit and two other nonprimate mammals. J Gen Physiol 97:413–435, doi:10.1085/jgp.97.3.413, pmid:2037836.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Nakhai H,
    2. Sel S,
    3. Favor J,
    4. Mendoza-Torres L,
    5. Paulsen F,
    6. Duncker GI,
    7. Schmid RM
    (2007) Ptf1a is essential for the differentiation of GABAergic and glycinergic amacrine cells and horizontal cells in the mouse retina. Development 134:1151–1160, doi:10.1242/dev.02781, pmid:17301087.
    OpenUrlAbstract/FREE Full Text
  36. ↵
    1. Olney JW
    (1968) An electron microscopic study of synapse formation, receptor outer segment development, and other aspects of developing mouse retina. Invest Ophthalmol 7:250–268, pmid:5655873.
    OpenUrlAbstract/FREE Full Text
  37. ↵
    1. Petros TJ,
    2. Rebsam A,
    3. Mason CA
    (2009) In utero and ex vivo electroporation for gene expression in mouse retinal ganglion cells. J Vis Exp pii:1333, doi:10.3791/1333, pmid:19779401.
    OpenUrlCrossRefPubMed
  38. ↵
    1. Poché RA,
    2. Kwan KM,
    3. Raven MA,
    4. Furuta Y,
    5. Reese BE,
    6. Behringer RR
    (2007) Lim1 is essential for the correct laminar positioning of retinal horizontal cells. J Neurosci 27:14099–14107, doi:10.1523/JNEUROSCI.4046-07.2007, pmid:18094249.
    OpenUrlAbstract/FREE Full Text
  39. ↵
    1. Pochet R,
    2. Blachier F,
    3. Malaisse W,
    4. Parmentier M,
    5. Pasteels B,
    6. Pohl V,
    7. Résibois A,
    8. Rogers J,
    9. Roman A
    (1989) Calbindin-D28 in mammalian brain, retina, and endocrine pancreas: immunohistochemical comparison with calretinin. Adv Exp Med Biol 255:435–443, doi:10.1007/978-1-4684-5679-0_46, pmid:2515762.
    OpenUrlCrossRefPubMed
  40. ↵
    1. Pugh EN Jr.,
    2. Falsini B,
    3. Lyubarsky AL
    (1998) in Photostasis and related phenomena, The origin of the major rod- and cone-driven components of the rodent electroretinogram and the effect of age and light-rearing history on the magnitude of these components, eds Williams TP, Thistle AB (Plenum, New York), pp 93–128.
  41. ↵
    1. Robson JG,
    2. Maeda H,
    3. Saszik SM,
    4. Frishman LJ
    (2004) In vivo studies of signaling in rod pathways of the mouse using the electroretinogram. Vision Res 44:3253–3268, doi:10.1016/j.visres.2004.09.002, pmid:15535993.
    OpenUrlCrossRefPubMed
  42. ↵
    1. Roy A,
    2. Francius C,
    3. Rousso DL,
    4. Seuntjens E,
    5. Debruyn J,
    6. Luxenhofer G,
    7. Huber AB,
    8. Huylebroeck D,
    9. Novitch BG,
    10. Clotman F
    (2012) Onecut transcription factors act upstream of Isl1 to regulate spinal motoneuron diversification. Development 139:3109–3119, doi:10.1242/dev.078501, pmid:22833130.
    OpenUrlAbstract/FREE Full Text
  43. ↵
    1. Samadani U,
    2. Costa RH
    (1996) The transcriptional activator hepatocyte nuclear factor 6 regulates liver gene expression. Mol Cell Biol 16:6273–6284, pmid:8887657.
    OpenUrlAbstract/FREE Full Text
  44. ↵
    1. Sonntag S,
    2. Dedek K,
    3. Dorgau B,
    4. Schultz K,
    5. Schmidt KF,
    6. Cimiotti K,
    7. Weiler R,
    8. Löwel S,
    9. Willecke K,
    10. Janssen-Bienhold U
    (2012) Ablation of retinal horizontal cells from adult mice leads to rod degeneration and remodeling in the outer retina. J Neurosci 32:10713–10724, doi:10.1523/JNEUROSCI.0442-12.2012, pmid:22855819.
    OpenUrlAbstract/FREE Full Text
  45. ↵
    1. Stricker HM,
    2. Ding XQ,
    3. Quiambao A,
    4. Fliesler SJ,
    5. Naash MI
    (2005) The Cys214–>Ser mutation in peripherin/rds causes a loss-of-function phenotype in transgenic mice. Biochem J 388:605–613, doi:10.1042/BJ20041960, pmid:15656787.
    OpenUrlCrossRefPubMed
  46. ↵
    1. Thoreson WB,
    2. Mangel SC
    (2012) Lateral interactions in the outer retina. Prog Retin Eye Res 31:407–441, doi:10.1016/j.preteyeres.2012.04.003, pmid:22580106.
    OpenUrlCrossRefPubMed
  47. ↵
    1. Umino Y,
    2. Herrmann R,
    3. Chen CK,
    4. Barlow RB,
    5. Arshavsky VY,
    6. Solessio E
    (2012) The relationship between slow photoresponse recovery rate and temporal resolution of vision. J Neurosci 32:14364–14373, doi:10.1523/JNEUROSCI.1296-12.2012, pmid:23055507.
    OpenUrlAbstract/FREE Full Text
  48. ↵
    1. Wachtmeister L
    (1998) Oscillatory potentials in the retina: what do they reveal. Prog Retin Eye Res 17:485–521, doi:10.1016/S1350-9462(98)00006-8, pmid:9777648.
    OpenUrlCrossRefPubMed
  49. ↵
    1. Wässle H,
    2. Boycott BB
    (1991) Functional architecture of the mammalian retina. Physiol Rev 71:447–480, pmid:2006220.
    OpenUrlFREE Full Text
  50. ↵
    1. Werblin FS,
    2. Dowling JE
    (1969) Organization of the retina of the mudpuppy, Necturus maculosus. II. Intracellular recording. J Neurophysiol 32:339–355, pmid:4306897.
    OpenUrlFREE Full Text
  51. ↵
    1. Wu F,
    2. Sapkota D,
    3. Li R,
    4. Mu X
    (2012) Onecut 1 and Onecut 2 are potential regulators of mouse retinal development. J Comp Neurol 520:952–969, doi:10.1002/cne.22741, pmid:21830221.
    OpenUrlCrossRefPubMed
  52. ↵
    1. Wu SM
    (1992) Feedback connections and operation of the outer plexiform layer of the retina. Curr Opin Neurobiol 2:462–468, doi:10.1016/0959-4388(92)90181-J, pmid:1525544.
    OpenUrlCrossRefPubMed
  53. ↵
    1. Xiang M
    (2013) Intrinsic control of mammalian retinogenesis. Cell Mol Life Sci 70:2519–2532, doi:10.1007/s00018-012-1183-2, pmid:23064704.
    OpenUrlCrossRefPubMed
  54. ↵
    1. Young RW
    (1985) Cell differentiation in the retina of the mouse. Anat Rec 212:199–205, doi:10.1002/ar.1092120215, pmid:3842042.
    OpenUrlCrossRefPubMed
  55. ↵
    1. Zhang H,
    2. Ables ET,
    3. Pope CF,
    4. Washington MK,
    5. Hipkens S,
    6. Means AL,
    7. Path G,
    8. Seufert J,
    9. Costa RH,
    10. Leiter AB,
    11. Magnuson MA,
    12. Gannon M
    (2009) Multiple, temporal-specific roles for HNF6 in pancreatic endocrine and ductal differentiation. Mech Dev 126:958–973, doi:10.1016/j.mod.2009.09.006, pmid:19766716.
    OpenUrlCrossRefPubMed
Back to top

In this issue

Email
Print
View Full Page PDF
Citation Tools
Respond to this article
Onecut1 Is Essential for Horizontal Cell Genesis and Retinal Integrity
Fuguo Wu, Renzhong Li, Yumiko Umino, Tadeusz J. Kaczynski, Darshan Sapkota, Shengguo Li, Mengqing Xiang, Steven J. Fliesler, David M. Sherry, Maureen Gannon, Eduardo Solessio, Xiuqian Mu
Journal of Neuroscience 7 August 2013, 33 (32) 13053-13065; DOI: 10.1523/JNEUROSCI.0116-13.2013
Twitter logo Facebook logo Mendeley logo

Jump to section

Responses to this article

Respond to this article

Jump to comment:

No eLetters have been published for this article.