Endogenous arginine vasopressin–positive retinal cells in arginine vasopressin–eGFP transgenic rats identified by immunohistochemistry and reverse transcriptase–polymerase chain reaction

The arginine vasopressin–eGFP transgenic rat

Wild-type male Wistar rats (Charles River Laboratories, Yokohama, Japan) and heterozygous transgenic male Wistar rats that express an AVP-eGFP fusion gene [17] were bred and housed under standardized conditions (12 h:12 h light-dark cycle) with food and water. Five- to 12-week-old wild-type and AVP-eGFP transgenic rats were used for experiments. Light-adapted rats were deeply anesthetized with halothane (Takeda Pharmaceutical, Osaka, Japan) and sodium pentobarbital (Dainippon Pharmaceutical, Osaka, Japan), and then sacrificed by decapitation. Eyes were removed, transferred to carboxygenated Ames’ medium (Sigma-Aldrich, St. Louis, MO), and hemisected, and then the retinas were teased off the pigment epithelium. All procedures involving animals were approved in advance by the Ethics Review Committee for Animal Experimentation of the National Institute for Physiological Sciences, and were in accordance with the guidelines of the Physiological Society of Japan.

Immunostaining

The isolated retinas were fixed in phosphate buffered saline (PBS), pH 7.4, containing 4% paraformaldehyde for 1 h at room temperature. The retinal preparations were rinsed with PBS and incubated for 1 h at 4 °C in PBS with 4% donkey serum (Biowest, Nuaillé, France) and 0.3% Triton X-100 (Katayama Chemical, Osaka, Japan) for blocking before the addition of primary antibodies. Primary antibodies against AVP (1:800, rabbit polyclonal; Calbiochem, Darmstadt, Germany) and GFP (1:500, goat polyclonal; Rockland Immunochemicals, Gilbertsville, PA) were used. The preparations were incubated in primary antibody solution for at least 4 days at 4 °C (whole mounts). After several rinses in PBS, preparations were incubated with secondary antibodies (donkey antirabbit IgG coupled to Alexa Fluor 594 or donkey antigoat IgG coupled to Alexa Fluor 488, from Invitrogen) diluted 1:1,000 in blocking buffer overnight at 4 °C, rinsed, and coverslipped with Fluoromount-G (SouthernBiotech, Birmingham, AL). Images were captured using confocal laser scanning microscopy (A1R; Nikon). Nikon PlanApo VC20X and 40X lenses were used. The brightness and contrast of the images were adjusted by using ImageJ (National Institutes of Health, Bethesda, MD) or Photoshop CS (Adobe, San Jose, CA).

Density recovery profile

The spacing of cell bodies of AVP-eGFP-positive cells in the INL and the GCL (Figure 1) was analyzed using the density recovery profile (DRP) method, calculated in MatLab (The Mathworks, Lowell, MA) [18]. The concept of DRP analysis used the spatial autocorrelogram, as described previously [19].

Dissociation of GFP-positive cells from the arginine vasopressin–eGFP transgenic rat retina

Small pieces of retinas were cut off and dissociated in oxygenized 5 mM HEPES Hank’s balanced salt solution (HBSS; Invitrogen, Carlsbad, CA), pH 7.4, containing 15.6 U/ml papain (Worthington, Lakewood, NJ) and 12.5 μg/ml L-cysteine (Sigma-Aldrich) for 10 min at 37 °C. The retina pieces were washed once with 5 mM HEPES HBSS containing 200 U/ml DNase I (Sigma-Aldrich) and four times with 5 mM HEPES HBSS, and then stored in 5 mM HEPES HBSS bubbled with 100% O₂ at room temperature. The cells were dissociated by mechanical trituration through a Pasteur pipette. GFP expression was confirmed each time under a fluorescence microscope (TE300; Nikon, Tokyo, Japan).

Reverse transcriptase–polymerase chain reaction from GFP-positive cells

The techniques used were identical to those previously described [9,20]. Using a RNeasy Micro Kit (Qiagen, Hilden, Germany), we extracted total cellular RNA from cytosol suctioned into gigasealed patch pipettes from 20 eGFP-positive cells and pooled them. RNA samples were reverse-transcribed with random 6-mer primers using a PrimeScript 1st strand cDNA Synthesis kit (Takara Bio, Otsu, Japan) according to the manufacturer’s protocol. PCR was performed with 1 U of Blend Taq DNA polymerase (Toyobo, Osaka, Japan) in a 40 μl reaction mix contained 1× buffer, 0.2 mM deoxynucleotide triphosphates (dNTPs), and 3 μl cDNA template. Primer sequences are shown in Table 1. Note that we used two different pairs of primers to detect endogenous AVP in experiments for which the results are shown in Figure 2 and Figure 3. Amplification was performed in a thermal cycler (GeneAmp PCR System 9700; Applied Biosystems, Foster City, CA) under the following conditions: initial heating at 94 °C for 2 min followed by 45 cycles of denaturation at 94 °C for 30 s, annealing at 60 °C for 30 s, extension at 72 °C for 1 min, and then final extension at 72 °C for 5 min for endogenous AVP and transgene products of AVP-eGFP; and initial heating at 94 °C for 2 min followed by 35 cycles of denaturation at 94 °C for 30 s, annealing at 60 °C for 30 s, extension at 72 °C for 30 s, and then final extension at 72 °C for 5 min for glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as a positive control.

Whole-retina reverse transcriptase–polymerase chain reaction

Using TRIzol Reagent (Invitrogen), total RNA was extracted from the retinas of six-week-old male Wistar rats (Charles River Laboratories). RNA samples were treated with Recombinant DNase I (RNase-free; Takara Bio) and reverse-transcribed with oligo-dT primers using Superscript III RT (Invitrogen) according to the manufacturer’s instructions. AVP PCR primers designed by Glasgow et al. [21], V1a and GAPDH PCR primers designed by Sato et al. [9], and V2/a Rho GTPase-activating protein 4 (ARHGAP4) PCR primers designed by Vargas et al. [22] were employed. V1b PCR primers were designed with Primer3 software. The sequences of the primers are listed in Table 1. PCR was performed with Blend Taq DNA polymerase (Toyobo) under the following conditions: initial heating at 94 °C for 2 min followed by 25 cycles (GAPDH), 35 cycles (AVP, V1a, and V2/ARHGAP4), or 40 cycles (V1b) of denaturation at 94 °C for 30 s, annealing at 60 °C for 30 s, extension at 72 °C for 30 s, and then final extension at 72 °C for 5 min. The products of RT–PCR were electrophoresed on a 2% agarose gel, the bands were excised, and then the products were cloned into the pGEM-T Easy vector (Promega, Madison, WI) after purification with the Wizard SV Gel and PCR Clean-Up System (Promega). Plasmids were extracted with PI-200 (Kurabo, Osaka, Japan), treated with RNase (DNase free) Glycerol Solution (Nippon Gene, Tokyo, Japan), purified by phenol-chloroform extraction and ethanol precipitation, and used as templates for sequencing using a Big-Dye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems) and an ABI PRISM 310 Genetic Analyzer (Applied Biosystems).

Measurements of arginine vasopressin secretion

The amount of AVP secretion from whole retinas or dissociated retinal cells exposed to control solution or high potassium (50 mM) solution was measured at room temperature using an arg⁸-Vasopressin Enzyme Immunoassay Kit (Assay Designs, Ann Arbor, MI) according to the manufacturer’s protocol. The control solution contained (in mM): 125 NaCl, 2.5 KCl, 2 CaCl₂, 1 MgCl₂, 26 NaHCO₃, 1.25 KH₂PO₄, and 12 glucose (pH 7.4; 300 mosmol/kg-H₂O). The high potassium (50 mM) solution contained (in mM): 77.5 NaCl, 50 KCl, 2 CaCl₂, 1 MgCl₂, 26 NaHCO₃, 1.25 KH₂PO₄, and 12 glucose (pH 7.4; 300 mosmol/kg-H₂O).

For the measurements, whole retinas were isolated from four 5-week-old male wild-type Wistar rats (Charles River Laboratories). Then whole retinas were put in a carboxygenated custom-made chamber. After 1.5 h of preincubation in the control solution, retinas were rinsed once in the control solution and then each retina was incubated for 15 or 30 min in 5 ml control solution or high potassium (50 mM) solution. Each time, 0.5 ml was sampled for measurements.

Dissociation of retinal cells was performed using two retinas of a two-month-old wild-type Wistar rat (Charles River Laboratories), similar to the procedure for the dissociation of the AVP-eGFP transgenic rat retina in this study. After a 20 min preincubation of a batch of dissociated cells, the cells were incubated for 5, 15, or 30 min in the control solution or the high potassium (50 mM) solution. These solutions were then centrifuged at 500× g for 5 min at 4 °C and the supernatants were used for measurements.

Extraction of AVP was performed by application onto a C-18 SEP-COLUMN (Phoenix Pharmaceuticals, Burlingame, CA) according to the manufacturer’s protocol using Buffers A and B (Phoenix Pharmaceuticals). Extracted AVP was dried with a vacuum freeze dryer (VD-800F; Taitec, Koshigaya, Japan). The amount of AVP was estimated with a Microplate reader (Multiscan MS-UV; Labsystems, Vienna, VA) at a wavelength of 414 nm. The concentration of AVP was estimated each time under the condition that the concentration was set to 0 in the solution before incubation of whole retinas.

Endogenous arginine vasopressin–positive cells in the arginine vasopressin–eGFP transgenic rat retina

The AVP-eGFP transgenic rat was designed to express AVP-eGFP to recapitulate the expression of endogenous AVP in AVP-secreting cells under control of the AVP promoter [17]. In the transgenic rat, the AVP-eGFP fusion gene was expressed under the AVP promoter in cells that had endogenous AVP promoter activity [6,17]. In the retina of the AVP-eGFP transgenic rat, AVP-eGFP-positive cells were identified in the retinal GCL and the INL by immunoreactivity to the GFP antibody (Figure 1A,D). All of the GFP-positive cells in the GCL were also immunoreactive to the AVP antibody (Figure 1B,C), although only 5% of the GFP-positive cells in the INL were costained with the AVP antibody (Figure 1E,F). In the INL, there were two distinct types of GFP-positive cells: brighter cells and faint cells (Figure 1E, asterisks on faint cells). Almost all of the brighter GFP-positive cells in the INL were costained with an AVP antibody, although faint cells were not. The number of AVP-eGFP-positive cells in the GCL was 67.6 ±12.5 cells/mm² (mean±standard error of the mean [SEM], n=7 retina pieces). In the INL, the number of brighter AVP-eGFP-positive cells was 76.4±7.0 cells/mm² and the number of faint GFP-positive cells was 1,500±84 cells/mm². The DRP of GFP-positive cells in the GCL and brighter cells in the INL showed a mosaic distribution with a space size of ~50 μm (Figure 1G).

AVP-eGFP-positive cells in the GCL had no GFP-positive neurites/axons in the optical nerve layer identified by GFP immunoreactivity. In addition, the diameter of the soma of AVP-eGFP-positive cells in GCL was about 8 μm, similar to that of the brighter cells in the INL. Although some retinal ganglion cells can have somas of only ~11 μm diameter [23], small soma size is another good indication that these are a subtype of amacrine cells. Taken together, the results indicate that brighter AVP-eGFP-positive cells in both the GCL and INL are probably a subtype of amacrine cells with endogenous AVP.

Endogenous arginine vasopressin mRNA in dissociated GFP-positive cells

The next question is whether AVP-eGFP-positive cells have the ability to synthesize endogenous AVP. We assessed the endogenous AVP–synthesizing ability by RT–PCR from dissociated GFP-positive cells (Figure 2). We dissociated the transgenic rat retina enzymatically and collected the cytosol from brighter GFP-positive cells under a fluorescence microscope using Microglass pipettes (see Methods) [9]. RT–PCR was conducted to detect the transgene mRNA of AVP-eGFP and the endogenous AVP mRNA with primer pairs shown in Table 1 from a set of 20 GFP-positive cells [9,20]. After 45 cycles of RT–PCR, we found an endogenous AVP mRNA band, as well as a transgenic AVP-eGFP mRNA band (Figure 2B). To confirm that the wild-type retina has the ability to synthesize AVP, we checked the AVP mRNA expression in the wild-type whole retina. An endogenous AVP band was also detected in the wild-type retina (Figure 2C). Overall, AVP-eGFP-positive retinal cells have the ability to synthesize endogenous AVP.

Arginine vasopressin receptors in the whole retina

Next, we examined the possibility that AVP acts on the retina in a paracrine or autocrine manner. If the AVP peptide itself is synthesized and released from the retina to act on retinal cells, there should be AVP receptors in the retina. We examined the AVP receptor mRNA from the wild-type rat whole retina by RT–PCR (Figure 3). The whole-retina RT–PCR indicated that the retina has V1a and V1b AVP receptors. In regard to V2 receptors, our primer pairs could detect both the V2 receptor gene (301 base pairs) and opposite DNA strand of the ARHGAP4 gene, a Rho GTPase-activating protein 4 gene (462 bp) [24] that overlaps the V2 receptor gene in the rat (Figure 3, V2/ARHGAP4 lane). However, a V2 band could be distinguished by its size from an ARHGAP4 band. In addition, by cloning and sequencing the RT–PCR product, we concluded that the V2/ARHGAP4 band in Figure 3 was a band of ARHGAP4, not the V2 receptor. In contrast, no AVP receptor was detected by RT–PCR in dissociated AVP-eGFP-positive cells (Figure 3B, data from the same dissociated AVP-eGFP-positive cells shown in Figure 2B). Taken together, we concluded that the retina has an AVP-synthesizing and -receiving mechanism that can operate in a paracrine, not autocrine, manner through V1a and V1b AVP receptors.

Arginine vasopressin secretion was not detected by stimulation with high potassium solution

The remaining question is whether endogenous AVP is actually secreted in the retina under physiological conditions. To answer this question, we investigated endogenous AVP release from the retina by stimulating isolated wild-type retinas with high potassium solution, which has been shown to stimulate AVP release in the hypothalamus of the brain [25]. Isolated retinas were rinsed once in the control solution, and then each retina was incubated for 15 or 30 min in the control solution or the high potassium (50 mM) solution. The AVP concentration was assayed each time. Unfortunately, AVP levels were too low to detect AVP secretion from the retina by stimulation with the high potassium solution (Figure 4). We repeated the experiment using a batch of dissociated retinal cells instead of whole retinas, but we did not detect AVP secretion (data not shown).