Abstract
Vertebrates share common mechanisms in the control of development and in the maintenance of neural and retinal function. The secreted factor Noggin, a BMP inhibitor, plays a crucial role in neural induction during embryonic development. Moreover, we have shown its involvement in retinal differentiation of pluripotent cells. Here we show Noggin expression in the adult retina in three vertebrate species. Four Noggin genes are present in zebrafish (Danio rerio; ZbNog1, 2, 3, 5), three in frog (Xenopus laevis; XenNog1, 2 and 4), and one in mouse (Mus musculus; mNog). Quantitative RT-PCR experiments show the presence of ZbNog3 and ZbNog5 mRNAs, but not ZbNog1 and ZbNog2, in the adult zebrafish retina. All three genes are expressed in the frog retina, and mNog in the mouse. Immunohistochemistry data show that Noggin proteins are predominantly localized in the Golgi apparatus of photoreceptors and in the fibers of the outer plexiform layer. Lower expression levels are also found in inner plexiform layer fibers, in ganglion cells, in the ciliary marginal zone, and in retinal pigmented epithelium. Our results show that Noggin has a specific cellular and sub-cellular expression in the adult vertebrate retina, which is conserved during evolution. In addition to its established role during embryonic development, we postulate that Noggin also exerts a functional role in the adult retina.
Introduction
In all vertebrates, the eye is the organ devoted to the reception of luminous stimuli, the process that initiates vision. Within the eye, the retina represents the photosensitive structure. The anatomic structure of the retina is remarkably conserved among vertebrates. The mature retina has an ordered architecture, in which the seven retinal cell types are stratified in layers. The outer nuclear layer (ONL) contains cell bodies of cone and rod photoreceptors; the outer plexiform layer (OPL) contains the synapses of photoreceptors with horizontal and bipolar cells, whose cell bodies (together with those of amacrine and Müller glia cells) constitute the inner nuclear layer (INL); the inner plexiform layer (IPL) contains the synapses of bipolar and amacrine cells with ganglion cells, whereas the ganglion cell bodies are located in the ganglion cell layer (GCL).
In the past twenty years, several studies highlighted a remarkable evolutionary conservation, not only of eye and retina architecture, but also of the molecular mechanisms and gene regulatory networks involved in eye development and retina differentiation in vertebrates. The initial phases of eye specification are marked by the activation and cooperation of a network of transcription factors (Eye Field Transcription Factors, EFTFs), which allows the generation of a morphogenetic eye field in the anterior neural plate (Zuber et al. 2003). The morphogenetic events that follow, in concomitance with the closure of the neural tube, produce an evagination of the optic vesicles from the ventral forebrain and the subsequent formation of the neural retina and retinal pigmented epithelium (RPE). One of the initial steps in anterior neural tube specification is the inhibition of the Bone Morphogenetic Protein (BMP) cascade (Delaune et al. 2005). Many inhibitors of BMP were described to bind BMP proteins and block the ventralizing action of BMP during the early phases of embryonic development (Smith and Harland 1992; Lamb et al. 1993; Zimmerman et al. 1996; Thomsen 1997; Groppe et al. 2002; Krause et al. 2011). Among the many BMP inhibitors, Noggin is of particular interest. It was originally described as a factor secreted by the Spemann organizer of Xenopus laevis, participating with chordin and follistatin to induce neural structures that antagonize BMPs action (Smith and Harland 1992). Noggin was also described to control the induction of dorsal mesoderm during the later phases of Xenopus embryonic development (Smith et al. 1993). Importantly, these functions of Noggin were confirmed in other vertebrates, such as fish and mouse (Zimmerman et al. 1996; Furthauer et al. 1999). In recent years, Noggin has been implicated in eye and retina specification. When microinjected in Animal Cap Embryonic Stem (ACES) cells of Xenopus laevis, Noggin mRNA induces the expression of EFTFs (Zuber et al. 2003). More significantly, Noggin mRNA-microinjected ACES cells acquire a retinal fate in vitro, and are able to drive the formation of a normal eye after transplantation in the eye field region or in ectopic position in a host embryo at the neurula stage (Lan et al. 2009). Similar results were obtained by treating ACES cells with Noggin protein (Viczian et al. 2009). These studies clearly indicate that Noggin exerts a crucial role in the embryonic development of the Xenopus retina. However, no data are available so far regarding Noggin expression and function in the retina of other vertebrate species.
The genome of various vertebrate species contains different numbers of Noggin genes. Zebrafish has four Noggin genes (ZbNog1, 2, 3, 5), with similar biological activities, each functioning to antagonize ventralizing BMPs. On the other hand, their embryonic expression is different, with ZbNog2 the only gene expressed in the tele- and diencephalon (Furthauer et al. 1999). In Xenopus, three different Noggin genes have been found—XenNog1, 2 and 4 (Fletcher et al. 2004; Eroshkin et al. 2006). XenNog2 and 4 are expressed at a later stage than XenNog1 during embryogenesis and show expression in different neural territories (Eroshkin et al. 2006). It was recently shown that XenNog4 is unable to antagonize BMP (Molina et al. 2011). Moreover, activin/nodal- and Wnt-antagonizing functions have been reported at least for XenNog2 (Bayramov et al. 2011). Finally, only one Noggin gene is found in the mouse and human genomes (mNog, McMahon 1998; our unpublished observations; Valenzuela et al. 1995).
Here we describe for the first time that Noggin is expressed in the adult retina of zebrafish (Danio rerio), frog (Xenopus laevis) and mouse (Mus musculus), with a specific cellular and sub-cellular localization. Our results suggest an evolutionarily conserved function of Noggin in the vertebrate retina.
Materials & Methods
Ethics Statement
Experiments were carried out in conformity with the European Communities Directive 2010/63/EU for animal experiments. All experiments conformed to the Italian Ministry of Health guidelines on the ethical use of animals. All experiments were approved by the Ethics Committee of the University of Trento. Care was taken to minimize the number of animals used and their suffering.
Animals
Adult Danio rerio, Xenopus laevis and Mus musculus (C57BL/6 strain) animals were obtained from the Zebrafish International Research Center (ZIRC; University of Oregon, U.S.), NASCO (California, U.S.) and Charles Rivers Laboratories (Lecco, Italy), respectively. Animals of both sexes (n=3 per species) were used for the study. From each animal, both eyes were enucleated and retinae were dissected, one for RNA extraction, the other for immunohistochemistry. Retinae to be used for RNA extraction were rapidly frozen in liquid nitrogen, whereas those for immunohistochemistry were fixed in 4% paraformaldehyde, cryoprotected in 20% sucrose, embedded in OCT-TissueTek freezing medium (Sakura, Japan) and frozen on dry ice. All samples were stored at -80C until use.
Amino Acid Sequence Analysis
Amino acid sequence multialignment was performed using the NCBI public protein BLAST database (http://blast.ncbi.nlm.nih.gov). Phylogenetic relationships of Noggin proteins were determined using the free database TreeTop-Phylogenetic Tree Prediction (http://www.genebee.msu.su).
Quantitative RT-PCR (RT-qPCR)
Total RNA were extracted from adult retinae using NucleoSpin RNA XS columns (Macherey-Nagel; Düren, Germany). Five hundred ng of each total RNA were reverse transcribed with SuperScript VILO cDNA Synthesis Kit (Invitrogen; Carlsbad, CA). RT-qPCR experiments were performed using a Rotor-gene 6000™ thermal cycler with real-time detection of fluorescence (Corbett Research; Sydney, Australia). PCR reactions were assembled using the KAPA SYBR-Green Master mix (Resnova, Italy), according to manufacturer’s instructions. The expression levels of Noggin species-specific genes were quantified relative to the expression of housekeeping RNA [elongation factor EF1α for Danio rerio (Mich and Chen 2011) and Xenopus laevis (Eroshkin et al. 2006); and ribosomal protein L41 for Mus musculus (Tripathi et al. 2009)], using the Biogazelle QBASE PLUS v1.5 software (Integrated Sciences; Chatswood, Australia). The oligonucleotide primers used for the experiments are listed in Table 2.
Immunohistochemistry
Fifteen µm-thick cryostat sections were prepared from adult retinae and used for immunohistochemistry. Sections were washed in 1× PBS and blocked for 1 hr in 1× PBS containing 0.3% bovine serum albumin and 0.1% Triton X-100. Sections were incubated for 2 hr with primary antibodies (Table 3) diluted in blocking solution and then washed three times in 1× PBS containing 0.1% Tween-20. Sections were then incubated with secondary antibodies (Table 3) for 1 hr also diluted in blocking solution, and again washed three times in 1× PBS containing 0.1% Tween-20. Fluorescent signals were revealed directly (for fluorescent secondary antibodies) or after incubation with fluorescent streptavidin (for biotin-conjugated secondary antibodies). Noggin and synaptophysin1 double-labeling was performed as follows: Sections were incubated with anti-noggin and appropriate secondary antibodies, followed by streptavidin (as described), and then treated using a 5 min fixation step in 4% PFA. Sections were then washed in 1× PBS (three times) and anti-synaptophysin1 immunohistochemistry was performed. Nuclear staining was performed with DAPI (1:10,000 dilution in 1× PBS). Images were acquired at 20×, 40× and 63× primary magnification using a Zeiss Axio Observer z1 fluorescence microscope (Zeiss; Oberkochen, Germany) and processed using the Zeiss AxioVision software (v4.3.1).
Results & Discussion
The aim of this study was to investigate the evolutionarily conserved expression and cellular localization of Noggin in the adult retina of three vertebrate species: zebrafish (Danio rerio), frog (Xenopus laevis) and mouse (Mus musculus).
Phylogenetic Analysis of Noggin Proteins
Four functional Noggin orthologs are present in zebrafish (ZbNog1, 2,3 and 5; Furthauer et al. 1999), three in frog (XenNog1, 2 and 4; Fletcher et al. 2004; Eroshkin et al. 2006), one in mouse (mNog; McMahon, 1998; our unpublished observations) and one in humans (hNog, Valenzuela et al. 1995). Phylogenetic tree analysis of these proteins show that vertebrate Noggin proteins can be clustered in four major groups: one comprising hNog, mNog, XenNog1 and ZbNog1/3, the second composed of fish and frog Noggin2, whereas ZbNog5 and XenNog4 are evolutionarily quite distant from all others (Fig. 1). Multiple alignment of the amino acid sequences shows that ZbNog2 and XenNog2 share a lower homology with the other Noggin genes (Table 1). mNog displays a high degree of sequence homology with XenNog1 (86%), ZbNog1(76%) and ZbNog3 (74%) (Table 1), suggesting that these sequences have been subjected to pronounced evolutionary divergence from Noggin2 proteins, as already reported (Eroshkin et al. 2006). In humans, as in mouse, there is only one Noggin gene which shows a 99% homology with mNog. Although we have not performed any expression analysis in human tissues, we might speculate that, given the high conservation both in sequence and expression in the retina, human retinae could show a similar expression pattern to the one described here for the other three vertebrate species.
Phylogenetic tree of zebrafish (Zb), Xenopus (Xen) mouse (m) and human (h) Noggin (Nog) proteins. Four subgroups of Noggin proteins are clearly distinguishable, whose phylogenetic divergence corresponds to amino acid sequence distance.
Amino Acid Sequence Homology between Different Noggin Proteins.
Abbreviations: Zb, zebrafish; Xen, Xenopus; m, mouse; h, human; Nog, Noggin.
Noggin mRNAs in the Adult Vertebrate Retina
The mRNA expression of the different Noggin genes in the adult retina of fish, frog and mouse was analyzed by quantitative RT-PCR (RT-qPCR) experiments using specific primers (Table 2). In the zebrafish retina, ZbNog5 mRNA was the most abundantly expressed, whereas ZbNog3 mRNA was present at very low levels and ZbNog1 and ZbNog2 mRNA were not detected (Fig. 2). The mRNAs of all three Xenopus isoforms were present in the adult Xenopus retina, albeit at different levels. Finally, mNog mRNA was abundantly detected in the adult mouse retina (Fig. 2).
Oligonucleotides Used for RT-qPCR Experiments.
Accession numbers of sequences and expected amplicon sizes are indicated.
Expression of Noggin transcripts in the adult zebrafish, frog and mouse retina. RT-qPCR showing expression of Noggin isoforms in adult retina. ZbNog5, XenNog1 and mNog mRNAs are abundantly expressed as compared to ZbNog2, ZbNog3 and XenNog2 in adult retina. XenNog4 is expressed at levels comparable with those of XenNog2. No expression of ZbNog1 can be found. The levels of Noggin transcripts were normalized to the expression level of housekeeping genes (elongation factor EF1α for zebrafish and frog; ribosomal protein L41 for mouse; see Materials & Methods for details). Primers used are listed in Table 2. Error bars represent the standard deviation of three samples in three independent experiments. Abbreviations: Zb, zebrafish; Xen, Xenopus; m, mouse; Nog, Noggin.
The RT-qPCR experiments suggest that specific Noggin genes are expressed in retinal cells. The expression of ZbNog3 and ZbNog2 and the lack of expression of the ZbNog1 gene in head structures are in accordance with the results obtained by Furthauer and colleagues (Furthauer et al. 1999), who describe a later activation of ZbNog3 and ZbNog2 and a role for ZbNog1 only in the early phases of zebrafish development. Whereas no data was reported in relation to the expression and function of ZbNog5, Xenopus laevis data (Fletcher et al. 2004; Eroshkin et al. 2006) show an expression pattern for XenNog1 that overlaps with that of ZbNog1 and ZbNog3, and for XenNog2 with that of ZbNog2. XenNog4, which has no homologs in the other vertebrate species analyzed until now, is also expressed in the adult retina, in accordance with in situ hybridization experiments, which localize it to the developing retina during late embryogenesis (Eroshkin et al. 2006). On the contrary, only one gene is described in mouse, and its expression pattern overlaps with that of all of zebrafish and Xenopus homologs (McMahon et al. 1998). In this regard, the amino acid data analysis could suggest interesting speculations. There is strong divergence among the ZbNogs, which could indicate a divergence in function during zebrafish development. The uniqueness of the ZbNog5 sequence with respect to that of the other Noggin proteins (zebrafish, but also frog and mouse), could represent a redundancy in gene function and a negatively selected parameter during evolution. The very high homology between ZbNog1 and ZbNog3 could have rendered the presence of two almost identical Noggin genes redundant, thus resulting in a reduction in the number of the amphibian homologs to leave, in fact, only one, XenNog1. This speculation draws support from the overlap of XenNog1 expression and function with those of both ZbNog1 and ZbNog3. In contrast, we cannot speculate on ZbNog5, due to the lack of information on its distribution and function in the literature. Moreover, the highly conserved role of Noggin in vertebrates could have exerted a further negative selection on Noggin genes, producing an inactivation of Noggin2 in mammals, which results in the maintenance of only one active Noggin gene in higher vertebrates.
Noggin Proteins in the Adult Vertebrate Retina
The presence and localization of Noggin proteins were investigated by fluorescence immunohistochemistry on adult retinal sections, using two different Noggin polyclonal antibodies (Table 3). The antibodies recognize the first 100 amino acids of the human Noggin protein. This region is conserved among the different Noggin proteins, with homology ranging from 65–100% with respect to human, with the sole exception of XenNog4, which shows only 39% homology. Due to the relatively high homology among the different Noggin proteins, it is likely that the antibodies we used recognize all of them. The two antibodies revealed the same expression pattern of Noggin in the zebrafish, Xenopus and mouse retina, as shown in Supplemental Fig. S1. For all experiments, we used the antibody shown in Supplemental Fig. S1A–S1C, which had a lower background signal in the mouse retina (compare Supplemental Fig. S1C to Supplemental Fig. S1F). In all three species, Noggin protein was abundantly expressed in the ONL (containing the cell bodies of photoreceptors), OPL, IPL and ciliary marginal zone (CMZ). At lower levels, Noggin proteins were also found in ganglion cell bodies and in retinal pigmented epithelium (RPE) (Fig. 3A–3C). The RPE expression is very evident in Xenopus (Fig. 3B, arrows), but also visible in mouse (Fig. 3C, arrows) and zebrafish (Fig. 3A, arrows), albeit to a lesser extent. Supplementary Figure 2 shows control immunohistochemistry experiments performed using only secondary antibodies, in which no crossreactivity was visible in any of the retinal layers.
Primary and Secondary Antibodies Used for Immunohistochemistry Experiments.
Expression of Noggin protein in the adult zebrafish, frog and mouse central retina. (A–C) Immunodetection of Noggin (green) in the central retina of zebrafish (A), Xenopus (B) and mouse (C). Arrows point to Noggin expression in retinal pigmented epithelium. Nuclei are stained with DAPI (blue). (D–F) Co-localization of Noggin (green) and Rhodopsin (red) in zebrafish (D), Xenopus (E) and mouse (F). No overlap in the expression pattern of the two proteins can be found. (G–I) Co-localization of Noggin (green) and synaptophisin (red) in zebrafish (G), Xenopus (H) and mouse (I). The two proteins are expressed in synapses in both the outer plexiform layer (arrows) and inner plexiform layer (arrowheads), suggesting a possible secretion of Noggin at synaptic level. (J–L) Co-localization of Noggin (green) and TGN64 (red) in zebrafish (J), Xenopus (K) and mouse (L). The co-localization shown by the arrows suggests a specific sub-cellular localization of Noggin in the Golgi apparatus of photoreceptors. Abbreviations: ONL, outer nuclear layer; INL, inner nuclear layer; GCL, ganglion cell layer. Scale bars (A–I, J–L), 50 µm.
Double-labeling experiments were performed to better characterize the cellular and sub-cellular localization of Noggin in photoreceptors. Photoreceptor cells can be subdivided into three main compartments: the outer segment (OS, containing the photopigments, abutting the RPE), the inner segment (IS, in which the nucleus and organelles are located) and the synaptic segment (SS, containing the synapses between photoreceptors and bipolar and horizontal cells). We used anti-rhodopsin and anti-synaptophysin1 antibodies as specific markers of photoreceptors in the OS and SS, respectively (Spiwoks-Becker et al, 2001; Gao et al. 2002). Double immunostaining for Noggin and rhodopsin showed that Noggin is not present in the OS (Fig. 3D–3F, no overlap is shown). Noggin and synaptophysin1 immunoreactivity were instead co-localized in the OPL (Fig. 3G–3I; arrows), suggesting that Noggin is secreted at the level of photoreceptor processes. Noggin and synaptophysin1 were also co-localized outside photoreceptor synapses in the IPL (Fig. 3G–3I; arrowheads). The idea that Noggin is a secreted protein prompted us to investigate whether it has a specific sub-cellular localization in the IS of photoreceptors. In fact, double-immunostaining experiments showed that Noggin co-localizes with TGN46, a Golgi specific marker (Pfeffer 2011), suggesting that it could be accumulated in secretory vesicles inside photoreceptor cell bodies (Fig. 3J–3L; arrows). Further support for this co-localization with a trans-Golgi network-specific protein comes from published data showing the involvement of Rab3d—a Rab protein important for the secretory pathway from the trans-Golgi network to the plasma membrane (Kim and Han 2011)—in Noggin secretion.
The co-localization of Noggin with two specific markers, such as synaptophysin1 (synaptic terminals of fibers) and TGN46 (Golgi apparatus), could suggest a conserved mechanism of action for Noggin in the adult retina. Noggin is a secreted protein that acts in the extracellular space mainly by antagonizing the BMP signal transduction cascade (Thomsen 1997). The abundant expression in structures involved in secretion suggests that the principal activity of this protein in the retina may also occur in the extracellular space, with a function in cell-cell communication. The coexpression of BMPs and BMP antagonists with different diffusion rates in the same tissue may orchestrate the local fine-tuning of BMP signaling levels. Mathura and collaborators showed that BMPs are negative growth regulators of adult RPE (Mathura et al. 2000). The secretion of Noggin from photoreceptors towards RPE—hypothesized to occur based on the localization of Noggin in the extracellular space between photoreceptors and RPE (Fig. 3A–3C; arrows)—could antagonize BMP action in RPE cells. On the contrary, the release of BMPs from the RPE could modulate the gradient of expression of phototransduction pigments described by Satoh and colleagues (Satoh et al. 2009).
Noggin Expression in the Ciliary Marginal Zone
All vertebrates present a specific structure at the peripheral margin of the retina: the CMZ (ciliary body or ciliary margin in birds and mammals). In fish and frogs, tissue within this zone continually generates new retinal neurons throughout adult life (Perron and Harris 2000), whereas, in higher vertebrates, the neurogenic potential is more limited (Tropepe et al. 2000; Fischer and Reh 2003). Nonetheless, in all the three species we analyzed, Noggin is expressed in this region including its most peripheral part, which contains (at least in lower vertebrates) the bona fide stem cells (arrows; Fig. 4A, 4B). To confirm this idea, we compared the expression of Noggin to that of Pax6 (which is expressed throughout the ciliary marginal zone) by performing immunolabeling with the two antibodies on adjacent sections. As seen by comparing Figure 4A and 4B to Figure 4D and 4E, the Noggin expression domain in the ciliary margins of zebrafish and Xenopus is situated within that of Pax6. In the mouse, Pax6 is expressed in all layers of the ciliary margin (Fig. 4F), whereas we find Noggin localized only in the superficial layer (Fig. 4C)—the layer considered to contain stem cells (Tropepe et al. 2000).
Expression of Noggin protein in the adult zebrafish, frog and mouse ciliary marginal zone. (A–C) Immunodetection of Noggin (green) in the ciliary margin of zebrafish (A), Xenopus (B) and mouse (C). Nuclei are stained with DAPI (blue). Arrows in A and B show the Noggin expression domain in the ciliary margin of fish (A) and frog (B). (D–F) Immunodetection of Pax6 (red) in the ciliary margin of zebrafish (D), Xenopus (E) and mouse (F). Scale bar, 50 µm.
The dual inhibitory/regulatory of Noggin on BMP pathways can also be suggested in the ciliary body. In fact, BMP gradients contribute to the development of the mammalian ciliary body (Zhao et al. 2002). Furthermore, it is evident that, at least in fish and frogs, retinal stem cells reside in the CMZ. The existence of different subpopulations of cells within the fish and amphibian CMZ, based on transcription factor expression, argues for a spatial progression from uncommitted stem cells at the periphery to more committed, transient amplifying progenitors at the more central CMZ (Agathocleous and Harris, 2009). The co-localization of Noggin with Pax6 (which identifies retinal stem cells) in this area could support the hypothesis of a role for Noggin in retinal stem cell maintenance and identity acquisition (Lan et al. 2009; Viczian et al. 2009).
In this paper, we show that Noggin has a specific cellular and sub-cellular expression in the adult vertebrate retina, and this localization is conserved during vertebrate evolution. Although this is only a preliminary descriptive study, it raises new and intriguing questions about a possible role for Noggin in the maintenance of the adult functions in the vertebrate retina. As seen by its expression pattern in this tissue, Noggin could be implicated in the communication between photoreceptors and the RPE, which is a crucial step in retinal phototransduction activity. Moreover, its expression in the region in which stem cells reside is suggestive of a possible role in their generation or maintenance. Functional studies are needed to further investigate these issues.
Footnotes
Acknowledgements
We are grateful to Simone Bridi, Roberta Eccheli and Paola Sgadò for experimental support and helpful discussions. We thank Paolo Macchi and Laura Vidalino for the TGN64 antibody.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by University of Trento Startup Grant to Y.B. and S.C. and by Grant n. 2011.0251 of Cassa di Risparmio di Trento e Rovereto to S.C.