Characterization of NGF, trkANGFR, and p75NTR in Retina of Mice Lacking Reelin Glycoprotein

Both Reelin and Nerve Growth Factor (NGF) exert crucial roles in retinal development. Retinogenesis is severely impaired in E-reeler mice, a model of Reelin deficiency showing specific Green Fluorescent Protein expression in Rod Bipolar Cells (RBCs). Since no data are available on Reelin and NGF cross-talk, NGF and trkANGFR/ p75NTR expression was investigated in retinas from E-reeler versus control mice, by confocal microscopy, Western blotting, and real time PCR analysis. A scattered increase of NGF protein was observed in the Ganglion Cell Layer and more pronounced in the Inner Nuclear Layer (INL). A selective increase of p75NTR was detected in most of RBCs and in other cell subtypes of INL. On the contrary, a slight trend towards a decrease was detected for trkANGFR, albeit not significant. Confocal data were validated by Western blot and real time PCR. Finally, the decreased trkANGFR/ p75NTR ratio, representative of p75NTR increase, significantly correlated with E-reeler versus E-control. These data indicate that NGF-trkANGFR/ p75NTR is affected in E-reeler retina and that p75NTR might represent the main NGF receptor involved in the process. This first NGF-trkANGFR/ p75NTR characterization suggests that E-reeler might be suitable for exploring Reelin-NGF cross-talk, representing an additional information source in those pathologies characterized by retinal degeneration.


Introduction
Reelin and Nerve Growth Factor (NGF) take part in retinogenesis, and their increased levels occur in inflamed/degenerating retina [1][2][3]. Reelin is a highly conserved extracellular glycoprotein, released by neurons/accessory cells and signals via specific surface receptors belonging to the Apolipoprotein E and the very low density lipoprotein families (ApoER2 and VLDLR) and via adaptor protein Dab1 [4,5]. Reelin expression peaks during retinogenesis, allowing migration/positioning and differentiation of retinal cells (physiological upregulation), returns to baseline levels in adulthood (physiological downregulation), and increases again following local injury/degeneration (pathological upregulation) [6,7]. Reelin deprivation causes a macroscopic modification of the retinal structure, with incorrect cells distribution and synaptic circuitry alteration, including a decrease in Rod Bipolar Cells (RBCs) density and an abnormal distribution of their processes in the Inner Nuclear Layer (INL), as well described in the reeler model [1,8].
In the visual system, NGF exerts pleiotropic effects during development and guarantees homeostasis during adulthood [2,3,[9][10][11]. NGF appears to exert multiple effects in both the neurons and accessory cells (proliferation, migration, differentiation, cytoskeletal reorganization, survival, apoptosis, etc.) [2,3,11]. NGF plays a crucial role during retinogenesis, influencing neuritic outgrowth, survival, and apoptosis, together with other neurotrophins and their related receptors, while in adulthood NGF is involved in several pathophysiological processes (homeostasis, ischemia, glaucoma, etc.) [2,3,10,11]. NGF is produced and used in an autocrine/paracrine fashion by Retinal Ganglion Cells (RGCs), Bipolar and other 2 International Journal of Cell Biology retinal cell types (horizontal and amacrine cells, Müller glia) [11]. The biological effects of NGF are directly dependent on the specific binding to two different cell surface receptors, the tyrosine kinase trkA NGFR and the glycoprotein p75 NTR [12,13]. NGF/trkA NGFR promotes the survival and recovery of RGCs, as observed in experimental models and after intraocular injection of NGF [2,11,14,15]. The crucial contribution of the trkA NGFR /p75 NTR ratio in the survival of RGCs has been recently envisaged [13,15]. Previous studies showed an impaired NGF expression in Reelin-deficient mice [16,17].
In view of all these findings a possible cross-talk between NGF and Reelin during retinal development might be hypothesized, suggesting an abnormal NGF pathway in Reelin-deficient retinas. To address this question, a homozygous reeler mice model was developed from founder couples and used to investigate both the biochemical and molecular expression of NGF-trkA NGFR /p75 NTR .

Ethics Statement.
All experiments were performed in compliance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Animal care procedures were conducted in conformity with the Intramural Committee and Institutional guidelines, in accordance with national and international laws and policies (EEC Council Directive 86/609, OJ L 358, 1, December 12, 1987; NIH Guide for the Care and Use of Laboratory Animals, NIH Publication 85- 23,1985). The experimental protocol was approved by the Ethical Committee of "Tor Vergata" University (Rome, Italy).

Animals and Genotyping.
Founder couples of reeler strain (B6C3Fe-a/a-rl; Jackson Laboratories, Bar Harbor, ME, USA), carrying the rl jx−/− mutation in a C57BL/6J background, were purchased from Charles River (Calco, Italy). The colony was housed at animal house facility ("Tor Vergata" University) under standard conditions (12 hrs light/dark cycle, temperature 21 ± 1 ∘ C, and relative humidity 60 ± 10%). Both water and food were freely available (Enriched Standard Diet, Mucedola, Settimo Milanese, Italy). To obtain the double-mutant reeler-L7-EGFP strain (referred to as Ereeler in this study), B6C3Fe-a/a-rl and B6-FVB-Tg (Pcp2-EGFP)2Yuza/J strains (Jackson Laboratories) were crossed and then backcrossed, according to a standard procedure [18]. The offspring from the 8th generation were used as donors that were hemizygotes for the Green Fluorescent Protein (GFP) locus. The B6C3Fe-L7-EGFP mice were used as control (wild-type rl jx+/+ genotype; E-control). Since GFP expression is under control of the L7 promoter, high levels of this live cell marker (lacking toxicity) characterize RBCs and allow their easy analysis at confocal microscopy (resistant to bleaching) [19].
For molecular and biochemical analysis, mice were sacrificed by cervical dislocation and the retinas were quickly processed for biochemical ( = 8 eyes each, for both E-reeler and E-control) or molecular ( = 14 eyes each, for both E-reeler and E-control) analysis. The dissection of retinas from fresh enucleated eyes was carried out under a dissector microscope (SMZ645; Nikon, Tokyo, Japan), equipped with cold-light optic fibers (PL2000 photonic; Axon, Vienna, Austria).

Immunofluorescence and Digital Analysis.
Postfixed and cryoprotected eyes were quickly frozen in dry ice, embedded in OCT medium (TissueTek; Leica, Heidelberg, Germany), and sectioned (CM3050 cryostat; Leica Microsystems, Rijswijk, The Netherlands). Serial sections (7 m) were placed ( = 3 sections/slide, = 6 slides/retina) onto gelatinized slides, preheated to increase tissue attachment, and stored at −20 ∘ C. Both antigen retrieval (0.05% trypsin-EDTA solution, 2 min) and blocking/permeabilizing (1% BSA and 0.5% Triton X100 in PBS, 15 min) steps were performed before the addition of specific antibodies: anti-NGF (sc-549), anti-trkA NGFR (sc-118), and anti-p75 NTR (sc-6188) antibodies, all from Santa Cruz Biotech (Santa Cruz, CA). A quenching passage was also carried out to minimize PFA background and slightly reduce the higher fluorescent expression of the E-construct. Cy3-conjugated donkey  species-specific antibodies (1 : 500-700; Jackson ImmunoResearch Europe Ltd., Suffolk, UK) were used to bind all primary antibodies. Nuclear counterstaining was performed with TOTO3-Iodide (Molecular Probes, Eugene, OR) while GFP expression specifically identified RBCs [19]. Isotypes (negative controls) were carried out in parallel with the omission of primary antibodies and used for appropriate channel series acquisition and related background subtractions. Slides were coverslipped using a glycerol gelatin mounting medium (Sigma-Aldrich, St. Louis, CA). Serial images were acquired by C1 software connected to an inverted microscope (Eclipse TE2000U, Nikon  [23]. Samples were amplified in duplicate and in parallel with negative controls (either without template or with mRNA as template). Real cycle thresholds (Cts) were recorded during linear amplification and normalized to those of referring genes run in parallel (nCts = Ct target − Ct referring ). Averages were calculated from these replicates and expressed as normalized Ct or as expression ratio of a normalized target gene (fold changes in log2-scale), according to REST © analysis [24]. Both primer sequence and amplification profile are accurately reported in  [22]. Kendall's rank coefficient (Tau) was also calculated to identify correlation between GFP versus p75 NTR and trkA NGFR versus p75 NTR . A probability of < 0.05 was presumed to reflect statistical significant difference between groups.

Results
A preliminary observation of E-reeler mice was carried out between P7 and P35 showing ataxia, eating complications (typical of reeler mice), and survival difficulties for the majority of animals, in the absence of appropriate handling [1]. Accordingly, the P21 time-point was selected for the following studies of characterization, and only E-reeler mice carrying both the reeler mutation and EGFP expression were included in the study. The genotyping is shown in Figure 1 According to Oberdick and coworkers procedure, both E-reeler and E-control retinas express GFP-specific fluorescence mainly localized in the dendrites, soma, and axon terminals of RBCs, allowing their easy recognition with confocal microscopy [20]. As shown in Figure 1(b) (merge), a decrease in the number (arrows) and dendrite length of RBCs populating the INL, as well as their synaptic buttons (arrowheads) in the GCL, was observed in E-reeler retinas. Digital analysis carried out on serial images showed a 27.2% decrease in GFP immunoreactivity in E-reeler retina (20899± 4663 versus 28727 ± 8134 IntDen, resp., E-reeler versus Econtrol; < 0.001). GFP-bearing RBCs were also counted in serial sections, showing a 27.8% decrease in the E-reeler retinas as compared to E-control ones (resp., 209 ± 4 versus 287 ± 81 cells/optic field; < 0.001; Figure 1(c)).

NGF Expression.
In order to understand whether NGF is affected in E-reeler retina, serial sections were probed with specific NGF antibody and NGF immunoreactivity was evaluated in the INL and GCL. As shown in Figure 2(a), NGF immunoreactivity was increased in RBCs, as well as in both GCL and other INL cells (arrowheads) of E-reeler retina. Therefore, the digital analysis carried out on these sections showed a 13.9% increase of NGF immunofluorescence in Ereeler retinas (32705±2197 versus 27756±1124 IntDen, resp., E-reeler versus E-control; < 0.05). In line with this result, Western blot and the related OD measurements showed a 37.4% NGF increase (both 12 and 15 kDa bands) in the Ereeler ( < 0.05; Figure 2(b)). Finally, the molecular analysis of E-reeler mRNA extracts did not show a significant effect on NGF mRNA expression (0.39 2 log ratio; 7.21 ± 1.05 versus 7.43 ± 1.89 nCts, resp., E-reeler versus E-control; > 0.05).
International Journal of Cell Biology

3.2.
Expression. Since trkA NGFR is widely considered the main NGF receptor involved in cell migration, proliferation, and differentiation, trkA NGFR protein expression was investigated by probing both E-reeler and E-control retinas. As shown in Figure 3(a), trkA NGFR immunoreactivity was detected in both the INL and GCL of control retinas (merge). A trend towards a 5.64% decrease of trkA NGFR immunoreactivity was quantified in the RGCs of E-reeler retinas (22601 ± 2391 versus 23953 ± 1278 IntDen, resp., E-reeler versus E-control; > 0.05). Accordingly, Western blot and OD analysis did not show any significant trkA NGFR differences between the E-reeler versus the E-control protein extracts (Figure 3(b)). These results were corroborated by molecular analysis (−0.14 2 log ratio; 17.04±1.81 versus 16.63± 0.54 nCts, resp., E-reeler versus E-control; > 0.05).

Discussion
This study was undertaken to verify the NGF-trkA NGFR / p75 NTR expression in retinas from E-reeler mice, a Reelindeprived model showing severe structural and functional changes in the retina, alongside the severe central nervous system alterations. To date, no data are available on the NGF pathway in Reelin deficient retina.
Retinogenesis is driven by a milieu of soluble factors, synergizing to obtain the final well-organized synaptic circuitry [1,8]. As observed in several experimental models, Reelin and NGF take part actively during retinogenesis and continue all over adulthood, contributing to retina homeostasis (synaptic plasticity) [1,3,7,8]. Structural and functional changes in the retina have been detected in Reelin deficient mice, showing an altered distribution of RBCs (both cell malpositioning and reduced dendrite density) and impaired synaptic circuitry [1]. RGC apoptosis has been quantified upon chemical NGF deprivation, while the recovery of damaged structures and particularly NGF/trkA NGFR -promoted RGC survival were observed upon the intraocular NGF injection, as observed in models of retinal degeneration (ischemia, glaucoma, and diabetes) [11,[26][27][28].
Since emerging data lead to a possible cross-talk between Reelin and NGF during retinogenesis and tissue remodeling, the well-characterized reeler-L7-EGFP model was developed (E-reeler) and Reelin-deprived retinas underwent NGF-trkA NGFR /p75 NTR confocal microscopy and biomolecular characterization [16,17]. The morphological analysis of Ereeler retina at P7-P35 showed structural changes characterized by altered arrangement and significant decrease of axon/dendrite density of residual RBCs (retina degeneration), in line with previous studies [1,6,8]. Ataxia and eating and survival difficulties were observed, and since the survival was highly reduced over P27 in the absence of appropriate management, the P21 was definitely selected for these studies. According to the digital analysis and the conventional cellcounting method, significant reductions in RBCs' soma, dendrites, and axons were quantified in E-reeler retina. This observation is in line with previous studies indicating that just RBCs represent the primary target of Reelin defect, even though both RBCs and RGCs express/react to Reelin [1,5]. Since RBCs carry the signal from the rod photoreceptors to the RGCs (visual function), it is reasonable to hypothesize that RBCs defect might interfere with the physiological activity of RGCs in E-reeler retina.
The principal finding of this study is the significant increase of NGF all over E-reeler retina, mainly localized in the layers populated by RBCs (INL) and RGCs (GCL). Indeed, NGF immunoreactivity was greatly expressed in accessory/glial cells, representing certainly the main NGF source in the damaged neighbourhood [29][30][31][32]. The observation of increased NGF protein, not corroborated by the molecular data, might be explained as possible transcriptional/posttranscriptional regulations (differential regulation, stability, and degradation of mRNA). In line with the well-known NGF pleiotropic effects, this NGF increase might be explained as an endogenous compensatory response to Reelin deficiency, either to limit abnormal/impaired cell distribution or counteract undesired apoptosis [12,26,33].
The statement that NGF exerts multiple effects depending on the surface (co)receptor appearance strength suggests the potential contribution of trkA NGFR and/or p75 NTR in this model. In the retina, p75 NTR privileges Müller glial cells while trkA NGFR is mainly expressed by RGCs [30,34]. While trkA NGFR tasks in neuronal survival, growth, and synaptic modulation are well established, p75 NTR ones are still an open debate due to p75 NTR -trkA NGFR coreceptor activity/complexity [15,33,35]. p75 NTR mediates a widespread range of cellular functions, depending on the cell-to-cell and/or cell-to-factor milieu as well as the repertoire of surface (co)receptors and can signal independently of trkA NGFR [12,36,37]. In the nervous system, p75 NTR mediates neuronal survival by facilitating trkA NGFR signal, increasing neuronal axon growth and reducing neuronal cell death [37][38][39][40][41][42]. As a second finding, a significant increase of p75 NTR and a trend towards  a decrease of trkA NGFR were observed in E-reeler retina, as detected by confocal microscopy, Western blot, and real time PCR analysis. Interestingly, p75 NTR was mainly localised in RBCs and cells populating the GCL, either RGCs or accessory ones. In particular for GCL, whether p75 NTR is expressed by immature/migrating or mature/positioned RGCs during retinogenesis (or in adult retina) remains to be verified, as accessory cells processed tightly surround RGCs [28,42,43]. This p75 NTR expression would imply that NGF from RGCs and nearby accessory cells might promote the survival of interneurons (and likewise RBCs) via NGF/p75 NTR -dependent mechanism, favouring a "rescue response" in an autocrine/paracrine fashion [11,27,28,39]. The contribution of p75 NTR in the regulation of RBCs survival has been reported in an experiment of exposure to NGF, BDNF, or Neurotrophin 3 [44]. Therefore, the possibility that p75 NTR might exert the herein reported actions through other growth factors cannot be excluded [12,27]. An attempt for cytoskeleton reorganization might be also prospected for NGF and p75 NTR overexpression, since p75 NTR is strictly required for appropriate axonal morphology/outgrowth and cell migration [38]. Actin reorganization and cell migration as well as neurite outgrowth are common properties of both NGF and Reelin [1, 4-7, 26, 27]. As reported, the loss of RGCs might be an outcome of the reduced crosstalk with the RBCs (synapses decline) [1]. NGF overexpression in RGCs or accessory/Müller cells might be also viewed as an attempt to stimulate RBC dendrites/arbors elongation, to (re)establish synaptic connections, or to provide a gradient for new dendrites [45][46][47][48]. As a support, p75 NTR binds both pro/mature forms of NGF and other neurotrophins (NTs, BDNF, NT3, and NT5), takes part in retrograde axonal transport of NTs (as survival or apoptotic factor), and works as a shuttle molecule for BDNF and NT4 (RBC survival NTs) [44,49]. As reported, NGF pathway contributes to the cytoskeleton reorganization, at least in structural cells as myofibroblasts [50].

International Journal of Cell Biology
By the way, the selective shift toward p75 NTR expression does not exclude the potential contribution of trkA NGFR signalling. As reported, p75 NTR acts as neuroprotective molecule, and trkA NGFR might work as a death receptor [43,51]. Upon NGF exposure, both homo-and heterodimerizations of membrane bound trkA NGFR and p75 NTR occur on receptive cells, and the ultimate signalling response is the result of some predominant cascade pathways [12,13]. The dogma "trkA NGFR mediates survival while p75 NTR triggers apoptosis" is strictly dependent on the cell type, microenvironment, and trkA NGFR /p75 NTR surface-expression ratio [12,15]. In line with our finding, a possible role of trkA NGFR /p75 NTR ratio in determining the fate of RBCs and RGCs might be prospected, as supported by studies on degeneration rescue [11,14,15]. Therefore the p75 NTR increase in RGCs and RBCs might also be interpreted as a proapoptotic effect. Studies aimed at verifying/quantifying apoptosis of RGC and/or RBC are actually under investigation.
As widely reported, a milieu of soluble factors synergize to allow for the correct position and functional activity of the RGCs, amacrine cells, RBCs, horizontal cells, Müller glial cells, and rod/cone photoreceptors during retinogenesis [8,52]. The possibility that upon Reelin deprivation other factors (cytokines and growth and angiogenic factors) might be upregulated to offset or facilitate the entire process cannot be excluded. Beside NGF, several other factors (such as BDNF, NT4, and GDNF) have been reported to increase RGC survival and regeneration [44,53]. These activities might be direct or indirect as observed for NGF-p75 NTR -induced TNFand TGF-activities in developing retina [54]. Any attempt to comprehend the mechanism underlying the pathogenesis of some retinal disorders represents a step forward in ophthalmology field characterized by severe, invalidating, and life-threatening diseases (glaucoma, ischemia, retinopathies, etc.). One of the major challenges of current eye-disease research is to develop models that mimic eyes pathologies, useful tools for studying cell-to-cell and cell-to-mediator mechanisms and providing the basis for novel therapeutic approaches to offset retinal degeneration. Despite recent advances, a clear comprehension of the mechanism underlying retinopathies is still needed and might require the understanding of some "missing" aspects during retinogenesis. For several decades, the reeler mutant has been used as a model for studying neurological disorders [1,16]. In line with recent and the herein presented data, we propose E-reeler mice as a good "retinal disease" model to explore the cross-talk between NGF and Reelin [16,17]. Growing data indicate that NGF provides a potential approach in the treatment of retinopathies, characterized by RGCs death and following optic nerve degeneration [2,3,10,11]. Studies are underway to discriminate p75 NTR -induced cell rescue or apoptosis and characterize the proNGF/NGF expression. These studies will contribute to better understanding of the relationship between NGF and Reelin in the retina under normal and pathological conditions.