SEARCH

SEARCH BY CITATION

Keywords:

  • avascular;
  • development;
  • EdU;
  • glia;
  • Müller glia;
  • regeneration;
  • Sox2;
  • zebrafish

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Müller cells in the chick retina are generally thought to be a homogeneous population. We show that the transcription factor Pax2 is expressed by Müller cells in the central chick retina and its expression was first observed at stage 32 (embryonic day [E] 7.5). Birth-dating indicated that the majority of Pax2-positive Müller cells are generated between stage 29 and 33 (E5.5–E8). At stage 42 (E16), several Müller cell markers, such as Sox2 and 2M6, had reached the peripheral retina, while the Pax2 labeling extended approximately half-way. A similar pattern was maintained in the 6-month-old chicken. Neither the Pax2-positive nor the Pax2-negative Müller cells could be specifically associated to proliferative responses in the retina induced by growth factors or N-methyl-D-aspartate. Pax2 was not detected in Müller cells in mouse, rat, guinea-pig, rabbit, or pig retinas; but the zebrafish retina displayed a similar pattern of central Pax2-expressing Müller cells. Developmental Dynamics 239:1858–1866, 2010. © 2010 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Glial cells are essential for the function of the vertebrate retina. Depending on the species, the types of glia in the retina differ, but Müller cells are the dominant glia in all vertebrate retinas. Astrocytes are normally also present in vertebrate species with a vascularized retina, which is not the case in avascular retinas of chick and guinea pig (Dreher et al.,1992; Won et al.,2000). Müller cells have a pleiotropic role in maintaining structure and homeostasis of the retina (reviewed by Newman and Reichenbach,1996; Bringmann et al.,2009). They extend radially from the outer to the inner limiting membrane, and their nuclei are located in the inner nuclear layer (INL). Based on the number and thickness of processes, Prada et al. (1989a,b) distinguished two types of Müller cells in the chick retina, denoted type I cells, with numerous thin processes, and type II cells, with fewer and thicker processes. These morphological differences were further characterized by Anezary et al. (2001) and the type I cells were in majority throughout the retina, whereas type II were mainly found in the peripheral parts of the retina (Prada et al.,1989a,b; Anezary et al.,2001). The morphological differences have in part been suggested to be related to the adaptation of the Müller cell growth to the thinner peripheral retina (Prada et al.,1989a,b; Anezary et al.,2001).

Pax2, a member of the paired homeobox family, is critical for several developmental processes in the eye. Initially, Pax2 is involved in the morphogenic process of the early optic region, defining the ventral region of the optic cup (Nornes et al.,1990; Schwarz et al.,2000). Pax2 is then confined to a more proximal part destined to give rise to the optic nerve, followed by its involvement in the closure of the optic fissure (Torres et al.,1996; Mansouri et al.,1996). However, Pax2 also plays an important role in the differentiation process, both in the case of forming the optic nerve from the optic stalk and during the closure of the fissure by generating astrocytes (Sehgal et al.,2008). Pax2 is expressed in astrocyte precursor cells and mature astrocytes of the optic stalk/nerve and glial cells of a vascular structure in avian and reptile species, called the pecten (Sehgal et al.,2008; Chan-Ling et al.,2009). It has been suggested that Pax2 is able to regulate a switch between neuronal and glial fates (Soukkarieh et al.,2007). Pax2 seems to act as both an inhibitor of neurogenesis and an inducer of gliogenesis in the optic nerve.

Although some morphological differences between Müller cells in the chick retina have been described, no overt molecular markers identifying Müller cells as a topically heterogeneous population have hitherto been reported. In this work, we show that Pax2 expression was restricted to centrally located Müller cells in late developing and adult retina. We also investigated the induced proliferative response after injection of growth factors or after an excitotoxic lesion in relation to Pax2-positive Müller cells. These results describing Müller cell subpopulations defined by the presence or absence of Pax2 expression reveal further complexity for this important retinal cell type.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Pax2 Labeling in the Chick Retina

The classic nonoverlapping Pax6/Pax2 dorsal–ventral pattern was observed in the early developing eye (stage, st18, Hamburger and Hamilton,1951; Supp. Fig. S1A, is available online) (Nornes et al.,1990; Schwarz et al.,2000). Later, at st29 (embryonic day [E] 5.5–6), Pax2 labeling was observed in nuclei in the optic nerve (Supp. Fig. S1B) previously described as astrocytes (Parrilla el al.,2009). Pax2 labeling in neural retina was first observed at st32 in a band of nuclei within the INL in a central region close to the optic nerve (Fig. 1A). At st35–38, the band of Pax2-positive nuclei became narrower. At st42, the Pax2 labeling was in a single row of elongated nuclei located approximately in the middle of the INL; this pattern remained at st46 and in the 6 months/adult chick retina. The 6-month retina repeatedly displayed labeling outside of the nucleus in the plexiform layers that could not be blocked out and we do not exclude that this pattern is specific even though the immunoreactivity of the transcription factor Pax2 is expected to be nuclear (Fig. 1A; Supp. Fig. S2). Co-labeling with known Müller cell markers such as the 2M6 antigen (Schlosshauer et al.,1991; Linser et al.,1997), vimentin (Lemmon and Rieser,1983), and Sox2 (Taranova et al.,2006) at st42 revealed that the Pax2-expressing cells in the central retina were Müller cells (Fig. 1B–D). All Pax2 cells were Sox2-positive, and co-labeling of Pax2 and neuronal markers could not be detected (data not shown). Neuronal markers used were Lim3 (bipolar cells), Ap2α (amacrine cells), and Isl1 (subsets of retinal ganglion cells, horizontal, amacrine, and bipolar cells; Edqvist et al.,2006). Sox2 also labels a subpopulation of amacrine cells and a recently described nonastrocytic inner retinal glial cell (Taranova et al.,2006; Fischer et al.,2010; Fig. 1D).

thumbnail image

Figure 1. Pax2 labeling in Müller cells during chick retinal development. Epifluorescence micrographs of immunohistochemical labeling of Pax2 during retinal development. A: A series of st32 (embryonic day [E] 7.5) to adult chick retina. B–D: Co-labeling of Pax2 and Müller cell markers: 2M6 (B), vimentin (C), and Sox2 (D). The boxed regions in B–D are shown as single-color channel images at higher magnification. ONL, outer nuclear layer; INL, inner nuclear layer; GCL, ganglion cell layer. Scale bar = 50 μm in A, 25 μm in B–D.

Download figure to PowerPoint

Birth-dating of the Pax2-Expressing Müller Cells

We undertook a birth-dating analysis, using incorporation of nucleotide analogues to identify the time when the Pax2-expressing Müller cells in the centrotemporal retina undergo their last replication (S-phase). We injected embryos with the nucleotide analogue 5-ethynyl-2′-deoxyuridine (EdU). EdU is an alternative to 5-bromo-2′-deoxyuridine (BrdU) that can more easily be combined with immunohistochemistry because it does not require DNA denaturation for its detection (Buck et al.,2008). Injection of EdU and BrdU at st35 and analysis at st36 produced a complete overlap between the EdU and the BrdU labeling (Supp. Fig. S3). Analysis of EdU incorporation showed that the Pax2-labeled Müller cells underwent their final S-phase after st27 and before st35 (Fig. 2A). When EdU injection was performed at st27, less than 2% (6 of 872 counted Pax2 cells) of the Pax2-positive cells were EdU positive when analyzed at st42; and when injection was performed at st29, approximately 35% (471 of 1,360 cells) were labeled. The majority of Pax2-positive cells were generated between st31 and st33, as shown by that 57% (542 of 941 cells) of the Pax2 cells were labeled with EdU when embryos were injected at st31 and 36% (322 of 875 cells) when embryos were injected at st33. After st35, less than 1% (2 of 988 cells) of the Pax2-positive cells incorporated EdU. The cumulative percentage of the number of EdU, Pax2 double-positive cells at st42 is displayed to visualize the generation of Müller cells (Fig. 2B). An example of the results obtained when a st33 embryo was injected with EdU is presented in Figure 2C. Many more EdU-positive cells than the Pax2-positive cells were seen in the INL of both the central and the peripheral retina at st36 (E10), indicating late birth of cells other than the Pax2-positive Müller cells (Supp. Fig. S3). EdU was incorporated in Sox2-positive, Pax2-negative Müller cells in the peripheral retina when eyes were injected at st33 and st35 (data not shown). By analyzing and comparing phosphohistone-3 (PH3), a marker for late G2-phase and mitosis, with Pax2 expression in the Müller cells, it was possible to asses if mitosis occurred after or before onset of Pax2 expression. Co-labeling of PH3 and Pax2 could not be observed, suggesting that the onset of Pax2 expression was after the terminal mitosis (data not shown).

thumbnail image

Figure 2. Birth-dating of Pax2-positive Müller cells. The nucleotide analogue 5-ethynyl-2′-deoxyuridine (EdU) was injected at st27 (embryonic day [E] 5), 29 (E5.5–E6), 31 (E7), 33 (E8), and 35 (E9), and the retinas were analyzed at st42 (E16). A: Graph visualizing the percentage of EdU, Pax2 double-positive cells at st42 when embryos were injected at the various time-points. B: The cumulative percentage value for Pax2, EdU double-positive Müller cells at the actual and preceding stages. C: Example of a retina injected at st33 and analyzed at st42. Pax2, EdU double-positive and Pax2-positive, EdU-negative Müller cells at are marked by white arrowheads and white arrows, respectively. ONL, outer nuclear layer; INL, inner nuclear layer; GCL, ganglion cell layer. Scale bar = 25 μm.

Download figure to PowerPoint

Spatial Restriction of Pax2 in the Chick Retina

When comparing the Pax2 labeling pattern at st32, 35, and 38, a central to peripheral progression of the pattern was observed. However, at st42 this progression had stopped, with 7.0 mm of the dorsal periphery and 3.2 mm of the ventral periphery displaying no Pax2 labeling in the section shown in Figure 3A. The central Pax2 expression spread 7.7 mm in the st42 retina. In contrast, the Sox2 labeling continued to the most peripheral part of the st42 retina, indicating that Müller cells are present throughout the entire retina at this stage (Fig. 3A). Figure 3B depicts the extent of Pax2 and Sox2 labeling in a schematic flat-mount of a st42 retina. To corroborate the Pax2 expression, small pieces of st42 retina were dissected from central and peripheral regions and used for quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR) analysis. The results showed a much higher expression of Pax2 mRNA in the central retina compared with the barely detectable levels in the peripheral retina, confirming the results from immunohistochemistry (Fig. 3C). The mRNA levels of Sox2 were similar in the central and peripheral samples. In the central retina, all Pax2-positive cells co-labeled with Sox2 at st42, whereas no Sox2-positive cells were Pax2 positive in the peripheral retina. The border between Pax2-positive and Pax2-negative Müller cells contained intermixed 2M6-positive cells that were either labeled for Pax2 or not; thus, no graded expression of Pax2 in this region could be seen (Fig. 3D). A similar spatial pattern was visible in the adult chick retina with 11.5 mm of the dorsal periphery and 7.0 mm of the ventral periphery displaying no Pax2 labeling (Fig. 3E; Supp. Fig. S2). The central Pax2 expression was 14 mm in the adult retina.

thumbnail image

Figure 3. Spatial labeling of Pax2 in the st42 (embryonic day [E] 16) retina. A,B: Schematic diagrams of the localization of Pax2 and Sox2 labeling in a cross-section of a st42 (E16) retina through the optic nerve (A), and a flat-mount (B). For the st42 retina depicted in A, the distance from the upper arrow to the periphery was 7.0 mm, the distance from the lower arrow to the periphery was 3.2 mm, and the length of the Pax2-expressing region in between was 7.6 mm. Inserts (a–e) depicts epifluorescence micrographs of immunohistochemical labeling of Sox2 (green) and Pax2 (red) of the boxed areas in A. C: Quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR) analysis comparing Pax2 and Sox2 mRNA expression in central and peripheral parts of the st42 retina. D: Intermediate zone between the central Pax2-positive and peripheral Pax2-negative Müller cells, showing both 2M6, Pax2 double-positive and 2M6-positive, Pax2-negative Müller cells. White arrows indicate 2M6-positive cells, and the white arrowhead indicates a Pax2, 2M6 double-positive cell. E: Central and peripheral expression of Pax2 and Sox2 in the adult chick retina. ONL, outer nuclear layer; INL, inner nuclear layer; GCL, ganglion cell layer; t, temporal; n, nasal; d, dorsal; v, ventral; on, optic nerve; ns, nonsignificant; *P < 0.05 t-test. Scale bar = 50 μm in A (applies to E), 25 μm in D.

Download figure to PowerPoint

Pax2 Expression in Retinas From a Few Other Species

We analyzed the Pax2 expression in retinas from a few other species to investigate whether the subdivision of Pax2-positive and Pax2-negative Müller cells was universal among species. Retinas from pig, mouse, rat, guinea pig, and rabbit as well as zebrafish were analyzed with respect to Pax2 and Müller cells markers. Pax2 labeling of cells that were co-labeled for the Müller cells markers could only be found in zebrafish among the analyzed species (Fig. 4). Of interest, species with a vascularized retina such as mouse, rat, and pig displayed Pax2 labeling in astrocytes present in the retinal ganglion cell- and optic fiber layers, but without any labeling in the inner nuclear layer as was seen in chick and zebrafish (Supp. Fig. S4). The guinea-pig and rabbit retinas are avascular and lacked astrocytes in most parts of the retina. Neither guinea pig nor rabbit had any Pax2-labeled cells in any layer of the retina except for cells in the optic nerve. In the case of rabbit, a few Pax2-positive cells were also seen within a narrow band of vessels that extends along the myelinated streak in the region of the former choroid fissure (Chan-Ling,1997). These few Pax2-positive cells were in the vicinity of the blood vessels and were most likely astrocytes (data not shown). The expression of Pax2 in the zebrafish retina was similar to that observed in the chick retina, with Pax2-positive Müller cells in the central retina and Pax2-negative Müller cells in the peripheral retina (Fig. 4A). The Pax2-positive cells were verified as being Müller cells by co-labeling with glutamine synthetase (Prada et al.,1998; Fig. 4B).

thumbnail image

Figure 4. Pax2 labeling in the zebra fish retina. A: Epifluorescence micrograph from a whole section of adult retina showing Pax2 labeling. Arrows indicate the limits of the central Pax2 labeling. a,b: Magnifications of the boxed regions in A. a′,b′: The green channel of the boxed regions in (a) and (b). B: Pax2 co-labeling with the Müller cell marker glutamine synthetase (GS). Arrows indicate Pax2, GS double-positive cells. ONL, outer nuclear layer; INL, inner nuclear layer; GCL, ganglion cell layer. Scale bar = 25 μm in B (applies to (a) and (b)).

Download figure to PowerPoint

Pax2 Expression in Relation to Induced Cell Proliferation

Müller cells in retinas of posthatch chickens have been shown to proliferate in response to excitotoxic lesions or administration of exogenous growth factors (Fischer and Reh,2001,2003; Fischer,2005). We analyzed 10-day posthatch (P10) retinas from eyes injected with fibroblast growth factor-2 (FGF2) and insulin to investigate the proliferative response of Pax2-expressing Müller cells. In eyes that had received 3 daily injections of FGF2 and insulin, followed by an injection of EdU at P9 and analysis at P10, we saw abundant EdU-labeled cells extending 0.5–1 mm from the retinal margin including the ciliary marginal zone (Fig. 5A). In agreement with Fischer et al. (2002) the majority of these EdU-labeled cells were also labeled for Sox2 (Fig. 5B). The control-injected eyes displayed only sparse EdU labeling in the ciliary marginal zone and the ciliary body (Fig. 5C). The region with proliferating cells in the FGF2- and insulin-treated retinas was distant from the Pax2-positive Müller cells, which appeared more than 10 mm from the retinal margin.

thumbnail image

Figure 5. Induced proliferation in chicken retina. Epifluorescence and confocal micrographs of chicken retina showing 5-ethynyl-2′-deoxyuridine (EdU) incorporation in combination with immunohistochemistry after treatment with growth factors or excitotoxin. A–C: Peripheral region of posthatch day (P) 10 retina including the ciliary marginal zone. A: Eyes injected daily with fibroblast growth factor-2 (FGF2) and insulin at P6, 7, and 8 followed by an EdU injection at P9 and analysis at P10. B: Adjacent section to that shown in A also labeled for Sox2. The boxed region is magnified and shown as single-color channel images. Arrows mark some of the double-labeled cells. C: Vehicle-injected control eye analyzed at P10. D: Central retina from eye injected with N-methyl-D-aspartate (NMDA) and EdU at st40 (embryonic day [E] 14) and analyzed 2 days later at st42 (E16) showing the distribution of EdU. E: Confocal micrograph of a central region of a st42 retina exposed to NMDA and EdU, and labeled for Sox2 and Pax2. The boxed region is magnified and shown as single-color channel images. CB, ciliary body; CMZ, ciliary marginal zone; CTRL, vehicle-injected control eye; GCL, ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear layer; P, posthatch day; pe, pigment epithelium. Scale bars = 200 μm in A, 50 μm in C (applies to B,D), 50 μm in E.

Download figure to PowerPoint

Fischer and Reh (2003) found that the zone of proliferating Müller cells in response to excitotoxic injury was not confined to the peripheral margin at younger ages. We, therefore, injected N-methyl-D-aspartate (NMDA) in the eye of st40 (E14) chicken to trigger proliferation in Müller cells located in more central regions of the retina. Analysis at st42 (E16) showed that EdU-labeled cells were found throughout the retina with more labeled cells toward the periphery. The retinal morphology was affected in the NMDA-injected eyes with degeneration of the inner plexiform and ganglion cell layer. Control-injected eyes had EdU-labeled cells in the retina only a few millimeters in from the retinal margins and an intact inner plexiform layer (data not shown). EdU-labeled cells in the central regions of the NMDA-treated retinas were located both in the INL and in what remained of the GCL (Fig. 5D). Among the EdU labeled cells within the neural retina, were cells with elongated EdU-labeled nuclei located in the middle of the INL (<10% of EdU labeled INL cells). These elongated nuclei were Sox2-positive, but none were labeled for Pax2 (Fig. 5E). The position, nuclear shape, and Sox2 labeling suggested that these cells were Müller cells. Neighboring cells with elongated nuclei were Sox2, Pax2 double-positive (Fig. 5E). The majority of the EdU cells located in the INL were either close to the outer or inner plexiform layers, and these cells had round nuclei and did not overlap with Sox2 or Pax2 labeling. However, some EdU cells in the GCL were also found to be Sox2 positive (Fig. 5E) and could be Sox2-positive nonastrocytic inner retinal glial cell (Fischer et al.,2010).

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

We have identified Pax2 as a novel marker for a population of centrally located Müller cells in the chick retina. This report is to our knowledge the first report of a marker identifying Müller cells as a topically heterogeneous population based on gene expression. Müller cells in chick and zebrafish retina expressed Pax2 but a similar expression could not be detected in mouse, rat, guinea-pig, rabbit, or pig retina.

The retina in chicken and zebrafish is avascular, and the question was raised if avascular inner retinas were associated with Pax2 expression in central Müller cells. Guinea-pig and rabbit retinas are also avascular and did not have Pax2 expression in their central Müller cells; thus, a direct relation between Pax2-expressing Müller cells and avascular retinas was not supported (Fig. 4; Supp. Fig. 4).

Pax2 is a pivotal gene during early development of the vertebrate eye defining the ventral region of the optic cup and guiding the formation of the optic nerve (Nornes et al.,1990; Torres et al.,1996). Our results replicated previously described patterns of Pax2 expression in the optic cup, early eye and optic nerve (Supp. Fig. S1). Pax2 labeling was first observed in Müller cells in the st32 chick retina (Fig. 1). Our data suggest that the onset of Pax2 expression in Müller cells occurred after their withdrawal from the cell cycle as shown by that the majority of Pax2 cells had their last S-phase before st33 and more importantly that there was no overlap between the mitosis marker PH3 and the Pax2-positive Müller cells. This suggests an involvement of Pax2 in the Müller cell differentiation, maturation, or maintenance processes rather than in the fate determination process.

Only few EdU, Pax2 double-positive Müller cells were found when embryos were injected at st27 (E5) while injection at st29 (E6) labeled one third of the cells (Fig. 2). Therefore, the first Pax2-expressing Müller cells progressed through their last S-phase between st27 and st29. In mice, a bolus injection of BrdU (50 μg/gr body weight) is available for as little as 2–3 hr (Packard et al.,1973) while an injection of 25 μCi [6-3H] thymidine (23 Ci/mmol) can sufficiently label all subsequent S-phases (Kahn,1974; Prada et al.,1991). It is not known how long a time after a bolus injection cells become labeled in S-phase by EdU. The large difference between retinas injected at st27 and st29, suggest that the duration of the EdU pulse was less than 1 day long and likely shorter. The graph with the cumulative percentage of Pax2, EdU double-positive cells shows that the majority of Pax2-positive Müller cells are generated between st29 to st35 (E6–E9). These results are overall in agreement with previous results (Kahn,1974; Prada and Ramirez,1983; Prada et al.,1991). Previous data showed that the period of Müller cell generation for the whole chick retina was more extended in time (E4–E11). In this study, we focused on the birth of centrally located Pax2, Sox2 double-positive Müller cells, but our data suggest that the peripherally located Pax2-negative, Sox2-positive Müller cells are born after the Pax2-positive ones, which would explain some of the discrepancy in the birth-dating. Furthermore, we do not exclude the possibility that additional Pax2-negative Müller cell types exist.

Pax2 expression persisted in the adult retina, suggesting a role in mature Müller cells (Fig. 3E). The restriction of Pax2 to the centrally located Müller cells in the adult retina suggests that these cells possess hitherto unknown properties, not required or present in the peripheral Müller cells. Pax2 is expressed in astrocyte precursors and it is continuously expressed throughout their cellular life-span (Chen-Ling et al.,2009). Therefore, the expression of Pax2 in subpopulations of chicken and zebrafish Müller cells may be related to astrocyte functions carried out by these cells. Pax2 deficiency is associated with coloboma, fewer astrocytes in the optic nerve and retina, and abnormal axonal pathfinding of the ganglion cell axons at the optic chiasm (Sanyanusin et al.,1995; Favor et al.,1996). Ectopic expression of Pax2 in the optic cup mimics Pax2 deficiency and gives choroid fissure closure defects. The mechanism is associated with a switch of cell fate from ventral retina and retinal pigment epithelium to an astroglial cell fate (Sehgal et al.,2008). Pax2-expressing astroglial cells in the optic nerve down-regulate Pax2 and up-regulate neuronal markers when explanted in vitro, and overexpression of Pax2 by electroporation in the optic nerve or ectopic expression of Pax2 in the neural tube is sufficient to block neuronal differentiation and allow glial development (Soukkarieh et al.,2007). Together this shows that Pax2 is able to inhibit the neuronal fate and promote astroglial development, which also supports the idea that Pax2-expressing Müller cells are astrocyte-like in the chicken and zebrafish retinas.

The Müller cells have similarities to retinal progenitor cells with neurogenic potentials and can, under certain conditions such as acute injury, reenter the cell cycle and be a source of new cells including neurons (Dyer and Cepko,2000; Fischer and Reh,2001; Fischer,2005; Bernardos et al.,2007). The degree of proliferative potential seems to vary with the location of the Müller cells (Fischer,2005; Jadhav et al.,2009). Injection of growth factors in the eyes of young chicken initiates a wave of proliferation in Müller cells that begins in the retinal margin and spreads inward in the peripheral retina (Fischer et al.,2002). Similarly to the experiments by Fischer et al. (2002), we induced proliferation by injection of growth factors (Fig. 5A–C). The results showed that the zone with proliferating cells did not extend as far into the retina as to the border between Pax2-positive and Pax2-negative Müller cells. We, therefore, concluded that Pax2 expression was not directly involved in the proliferative response by Müller cells. Fischer and Reh (2003) also showed that the zone of proliferation extended further into the retina in younger chickens. We injected NMDA in st40 (E14) eyes, a time after the onset of Pax2 expression in Müller cells, and analyzed the retinas 2 days later (Fig. 5D,E). Robust proliferation was seen throughout the neural retina. Pax2 expression could still be seen in central Müller cells, and we concluded that the excitotoxic lesion did not cause a general down-regulation of Pax2 expression in Müller cells at the analyzed time point. However, cells identified by their elongated Sox2-positive nucleus that had incorporated EdU were Pax2 negative. This suggests that Pax2 expression is down-regulated in cells that have undergone S-phase. Both central (Pax2-positive) and peripheral Müller cells express Sox2. Sox2 contributes to and maintains neural progenitor competence (Graham et al.,2003; Lin et al.,2009), and the expression of Pax2 in the central chick Müller cells may thus modulate such competence during normal development or after injury.

We show in this work that the transcription factor Pax2 is expressed in Müller cells in a similar central to peripheral pattern in which morphologically distinct populations in the chick retina have been described. This novel discovery of molecularly distinct Müller cell subpopulations that are topically organized adds a new level of complexity that needs to be considered in studies of genesis and maturation as well as function of Müller cells in the vertebrate retina.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Tissue Preparation and Immunohistochemistry

Fertilized White Leghorn eggs were obtained from OVA Produktion AB (Västerås, Sweden) and incubated at 38°C in a humidified incubator. The animal experiments were done according to the guidelines from Association for Research in Vision and Ophthalmology and the experiments were approved by the local animal ethics committee in Uppsala. Embryos were staged according to Hamburger and Hamilton (1951). The eyes were dissected and fixed in 4% paraformaldehyde (PFA) for 15–20 min, washed 10 min in phosphate buffered saline (PBS) and cryoprotected in 30% sucrose for 3–4 hr before being frozen in OCT (Sakura). Retinas were cryosectioned in an orientation parallel to the center of the lens and through the optic nerve exit containing dorsal and ventral retina and collected on Superfrost Plus glasses (Menzel-Gläser). Sections were washed in PBS for 5 min and incubated in blocking solution (PBS containing 1% fetal calf serum and 0.1% Triton X-100) for 60 min. Primary and secondary antibodies were diluted in blocking solution. Primary antibodies were allowed to react with the sections overnight at 4°C, and secondary antibodies for 2 hr at room temperature. Primary antibodies used in this study were against the 2M6 antigen (Linser et al.,1997), Pax2 (Biosite, PRB-276P), Pax6 (Hybridoma bank, Pax6), Sox2 (Santa Cruz, sc-17320), Vimentin (Hybridoma bank, H5), GS (Chemicon, MAB302), BrdU (AbD Serotec, OBT0030F), and PH3 (Millipore, 06-570). Secondary antibodies were obtained from Vector Laboratories, Jackson Immunoresearch Laboratories or Molecular Probes. The Pax2 signal was amplified using a horseradish peroxidase (HRP) -conjugated secondary antibody in combination with a Tyramide Signal Amplification (TSA) kit (Invitrogen, T20922, Alexa Fluor 488). Tyramide working solution was prepared by diluting the tyramide stock solution 1:200 in amplification buffer/0.0015% H2O2 just before labeling. The tyramide solution was allowed to react for 10 min at room temperature. The immunohistochemistry was analyzed using a Zeiss Axioplan2 microscope equipped with Axiovision software or Zeiss LSM 510 confocal microscope. Images were formatted, resized, enhanced and arranged for publication using Axiovision and Adobe Photoshop.

Birth-dating Analysis

Click iT EdU imaging kit (Buck et al.,2008, C10084, Invitrogen) was used to visualize cells passing through S-phase (Salic and Mitchison,2008). A total of 50 μg of EdU (in PBS, Boije et al.,2009) was injected into the yolk of embryos st27, 29, 31, 33, and 35. We determined the smallest amount of Edu that was needed to clearly label cells after a single bolus injection in ovo to 50 μg of EdU (data not shown). Eyes were fixed and frozen at st42 as described in the immunohistochemistry section. Pax2-positive cells were counted in dorsoventral retinal cross-sections through the optic nerve from three animals. All Pax2-positive and EdU, Pax2 double-positive cells on four sections were counted, and the percentage of EdU-positive cells were calculated for each stage (Fig. 2A), and the cumulative percentage value of labeled cells for the actual and preceding stages was plotted. The total number of labeled cells from all analyzed stages was set to 100% (Fig. 2B).

The Pax2 labeling was visualized as described followed by detection of the EdU signal using the EdU imaging kit according to manufacturer's protocol. An additional experiment was performed comparing the EdU labeling to more classic BrdU by co-injection at st35 and analysis at st36. The BrdU was visualized by a fluorescein isothiocyanate (FITC) -conjugated BrdU antibody after 10 min denaturation with 4 M HCl, 0.1% Triton X-100.

Intraocular Injections

Chickens were anesthetized by inhalation of 2% isoflurane. Intraocular injections were made in the dorsal quadrant of the vitreous chamber using a 100-μl Hamilton syringe and a 26-gauge needle with a stop. Lidokainchloride (0.5%) for local anesthesia and Chloromycetin (5 mg/ml) were dropped into the eye. Ten microliter vehicle with human FGF-2 (Peprotech; 300 ng/dose) and bovine insulin (Sigma-Aldrich; 1 μg/dose) was injected into experimental (left) eye and vehicle (saline with bovine serum albumin, 50 μg/ml) in control (right) eye. Chickens (n = 4) were injected once a day starting on posthatch day 6 (P6). On P9, both eyes received an injection of 2 μg of EdU and retinas were analyzed 18 hr later.

Fertilized eggs (E14) were opened at the blunt end and a small hole was made in the egg shell and chorioallantoic membranes. The head was pulled toward the membranes with a bent glass rod and intra-ocular injection was done through the membranes. A total of 200 ng of N-methyl-D-aspartate (NMDA) and 2 μg of EdU in 5 μl of vehicle was injected in one eye (n = 4). Control injections with vehicle and 2 μg of EdU were done in separate eggs (n = 4). The injected eggs were sealed and incubated in a humidified incubator and analyzed after 48 hr.

Quantitative RT-PCR

Small pieces (2 × 2 mm) of central and peripheral retina from stage 42 embryos were cut using a pair of scissors and total RNA was extracted using the Trizol reagent (Invitrogen). cDNA was prepared for three biological replicates from 1 μg of RNA using GeneAmp (Applied Biosystems). Samples were run in duplicate using IQ SyBr Green Supermix (Bio-Rad) and normalized to β-actin and TATA-box binding protein. Control reactions containing primers but no RNA were analyzed in parallel. Primer sequences were as follows: β-actin (NM205518; 5′-aggtcatcaccattggcaatg-3′ and 5′-cccaagaaagatggctggaa-3′), TATA-box binding protein (NM2005103; 5′-tagcccgatgatgccgtat-3′ and 5′-gttccctgtgtcgcttgc-3′), Sox2 (NM205188; 5′-gtcacctcctcgtctcattcga-3′ and 5′-ggcagctggttctggtacttca-3′), and Pax2 (NM204793; 5′-acctgtgaccatcgtcctgttt-3′ and 5′-ggtccttctcttgtttgccaga-3′).

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
DVDY_22309_sm_suppFig-S1.tif5294KSupp. Fig. S1. Early Pax2 expression in the chick eye. A: Pax2, Pax6 co-labeling in a st18 eye. B: Pax2 labeling in a st29 eye. l, lens; nr, neural retina; on, optic nerve; os, optic stalk; pe, pigment epithelium. Scale bar = 100& mgr;m and is valid for A and B.
DVDY_22309_sm_suppFig-S2.tif5161KSupp. Fig. S2. Pax2 and Sox2 expression in an adult chick retina. A,B: High resolution image of an entire cross-section of an adult chick retina labeled with Pax2 (A) and Sox2 (B). Magnified inserts depict the boxed regions. Distances in A are given from the arrow to the end of the peripheral retina. The spread of the Pax2-expressing region in between the arrows was 14.4 mm. INL, inner nuclear layer; IPL, inner plexiform layer; GCL, ganglion cell layer. Scale bar = 400 μm.
DVDY_22309_sm_suppFig-S3.tif4244KSupp. Fig. S3. The 5-ethynyl-2′-deoxyuridine (EdU) and 5-bromo-2′-deoxyuridine (BrdU) co-labeling. Chick eye was injected with EdU and BrdU at st35, and the retina was analyzed at st36. ONL, outer nuclear layer; INL, inner nuclear layer; GCL, ganglion cell layer. Scale bar = 50 μm.
DVDY_22309_sm_suppFig-S4.tif4294KSupp. Fig. S4. Pax2 expression in guinea-pig, pig, mouse, rat, and rabbit retina. A–E: Pax2 labeling in guinea-pig (A), pig (B), mouse (C), rat (D), and rabbit (E) retina. ONL, outer nuclear layer; INL, inner nuclear layer; GCL, ganglion cell layer. Scale bar = 50 μm.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.