Immunohistochemical Localization of Somatostatin Receptor Subtypes sst₁ and sst₂ in the Rat Retina

Abstract

purpose. To investigate the distribution of somatostatin receptor subtypes sst₁ and sst₂ in the rat retina by immunohistochemistry and to characterize further the neurotransmitters of the sst₁- and sst₂-immunoreactive cells.

methods. Polyclonal antibodies raised against sst₁ and sst₂ receptors were applied to 12-μm cryostat sections of rat retinas fixed in paraformaldehyde. Further, immunofluorescence double labeling was performed for the sst₁ and sst₂ receptors with somatostatin, tyrosine hydroxylase (TH) and glutamate decarboxylase (GAD).

results. Immunoreactivity for sst₁ was present in somatostatinergic amacrine cells located in the inner nuclear layer (INL) and in displaced amacrine cells in the ganglion cell layer of the retina. Also, a small number of ganglion cells were sst₁ immunoreactive. Immunoreactivity for sst₂ was observed in many medium-sized amacrine cells in the middle part of the INL, with a central process projecting to the sublaminae of the inner plexiform layer. Furthermore, sst₂ immunoreactivity was found in large amacrine cells of the INL. These cells also contained TH. Inner segments of cone receptors were stained with the sst₂ antiserum. Immunostaining for sst₂, and to a minor extent for sst₁, was found in Müller cell fibers. None of the somatostatin receptors colocalized with GAD.

conclusions. These findings suggest that the sst₁ receptor may function as an autoreceptor on retinal somatostatinergic cells. The presence of sst₂ receptors on the TH-immunoreactive amacrine cells indicates an influence of somatostatin on the secretion of dopamine in rat retina.

The vertebrate retina is rich in neurotransmitters and neuromodulators. Among these, the neuropeptide somatostatin has been detected by immunohistochemistry in the retina of a number of mammalian species, including rat, ^{1 2 3} rabbit, ^{1 4 5} and human. ^{6 7 8} In these studies, somatostatin-like immunoreactivity was confined mainly to cells in the ganglion cell layer (GCL), to amacrine and interplexiform cells in the inner nuclear layer (INL), and to cellular processes located in one or more sublayers of the inner plexiform layer (IPL). 

Somatostatin has been isolated from the hypothalamus as an inhibitor of growth hormone release. ⁹ Moreover, this 14- or 28-amino acid cyclic peptide has been found to act as a neurotransmitter or neuromodulator in the brain. ¹⁰ Evidence suggests that somatostatin also has a neuromodulatory function in the retina. In rabbit retina, somatostatin was found to have effects on the signal-to-noise pattern and center–surround balance of ganglion cells. ¹¹ These actions are most likely mediated through somatostatin receptors, of which five subtypes (sst₁ through sst₅) have been identified by molecular cloning in human and rat. ^{12 13 14 15 16 17 18 19 20}  

Using ligand binding, somatostatin receptors have been demonstrated in retinas of mouse, ²¹ rat, ²² sheep, ²³ and rabbit. ²⁴ Binding sites are confined mainly to the IPL and outer plexiform layer (OPL). ^{21 22} Furthermore, by reverse transcription–polymerase chain reaction, mRNAs encoding all five receptor subtypes have been shown to be present in rat retina, with sst₂ being the major subtype expressed. ²⁵ This is matched by a recent immunohistochemical study by Johnson et al. ²⁶ who found sst_2A immunoreactivity in bipolar cells and in wide-field amacrine cells of the rabbit retina. 

We have raised antibodies against the five somatostatin receptor subtypes. ²⁷ In the present study, we have used two of these antibodies to delineate the distribution of sst₁ and sst₂ receptor immunoreactivity in rat retina. We have further characterized the immunolabeled cells, by performing double localization with somatostatin itself, with tyrosine hydroxylase (TH), and with glutamate decarboxylase (GAD). We found sst₁ immunoreactive somatostatinergic cell bodies in the INL and GCL. Immunoreactivity for sst₂ was observed in large TH-positive amacrine cells and in medium-sized perikarya in the INL, ramifying in the IPL. Furthermore, sst₂ immunoreactivity was observed in the inner segments of cone receptors. 

Methods

Tissue Preparation

Male Wistar rats weighing 250 g were anesthetized by intraperitoneal injection of tribromoethanol (400 mg/kg) and fixed by vascular perfusion with 4% cold paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) for 15 minutes. The animals were kept under a 12–12-hour light–dark cycle and killed during the light phase. The eyes were postfixed overnight in the same fixative and changed to phosphate buffered saline (PBS). The eyes were cryoprotected in 30% sucrose in PBS and were sectioned at 12 μm in a cryostat. Sections were thaw mounted on gelatin-coated slides. All experiments with animals were performed in accordance with the Principles of Laboratory Animal Care (National Institutes of Health Publ. No. 86-23, revised 1985), with Danish national laws, and with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Immunohistochemistry

For studying somatostatin receptor immunoreactivity, we used rabbit polyclonal antisera raised against the C-terminal parts of the human sst₁ and sst₂ receptors, expressed as fusion proteins with glutathione S-transferase. ²⁷ In Western blot analysis, the anti-sst₁ and anti-sst₂ antisera showed no cross-reactivity with the other somatostatin receptor subtypes. 

Endogenous peroxidase activity was quenched by incubating the sections in 0.5% H₂O₂ in PBS for 10 minutes, followed by a 20-minute preincubation with 5% normal swine serum in a solution of 1% bovine serum albumin and 0.3% Triton X-100 in PBS (PBS-BT). Sections were incubated overnight at 4°C with the anti-sst₁ antiserum diluted 1:20,000 or anti-sst₂ antiserum diluted 1:10,000 in PBS-BT. The sections were washed three times in PBS containing 0.1% Triton X-100 and then incubated for 1 hour with biotinylated swine anti-rabbit immunoglobulins at 1:500 (E353; Dako, Glostrup, Denmark). After washing, the sections were incubated for 20 minutes with a blocking buffer (supplied with the biotinylated tyramide kit; NEL700; DuPont–NEN, Boston, MA), followed by 45 minutes’ incubation with horseradish peroxidase–conjugated streptavidin–biotin complex (HRP-streptABC; PK-6100; Vector, Burlingame, CA). After sections were washed, biotinyl-tyramide (DuPont NEN) was applied to the sections at 1:100 for 6 minutes, and after washing, the sections were finally incubated for a further 45 minutes with HRP-streptABC. Immunoreactivity was visualized with 0.05% diaminobenzidine and 0.01% H₂O₂. 

For controls, 1 ml diluted antiserum was preabsorbed before incubation of the sections, with 50 μg sst₁ or sst₂ fusion proteins against which the antibodies were raised. 

Double Immunofluorescence Labeling

For the double localization of sst₁ and sst₂ receptors with TH, somatostatin, GAD, protein kinase C (PKC), calbindin D, or vimentin sections were incubated overnight at 4°C with anti-sst₁ antiserum at 1:10,000 or with anti-sst₂ antiserum at 1:5000, together with mouse monoclonal TH antibody in a 1:250 dilution (d108460; Incstar, Stillwater, MN), or sheep somatostatin antibody diluted 1:1000 (13-2366; American Research Products, Belmont, MA), or mouse monoclonal GAD antibody at 0.5 μg/ml (4670-6559; Biogenesis, Poole, UK), or mouse monoclonal PKC antibody in a 1:10 dilution (RPN536; Amersham Research Products, Little Chalfont, UK), or mouse monoclonal calbindin D antibody in a 1:200 dilution (C8666; Sigma, St. Louis, MO), or mouse monoclonal vimentin antibody in a 1:200 dilution (MS-129; NeoMarkers, Union City, CA). The sections were then washed three times in PBS containing 0.1% Triton X-100, followed by incubation with biotinylated swine anti-rabbit antibodies 1:500 for 1 hour, blocking buffer for 20 minutes, HRP-streptABC for 45 minutes, and biotinylated tyramide 1:50 for 8 minutes. The sections were washed three times after each incubation period. The sections were then incubated with streptavidin fluorescein at 1:50 (RPN1232; Amersham Research Products) or Cy2-labeled streptavidin at 1:400 (PA42001; Amersham Research Products) together with Texas red–conjugated donkey anti-mouse IgG (715-075-150; Jackson ImmunoResearch, West Grove, PA) at 1:100 for visualization of TH, GAD, PKC, calbindin, and vimentin or with Texas red–conjugated donkey anti-sheep IgG (713-076-147; Jackson ImmunoResearch) at 1:100 for visualization of somatostatin. For double visualization of sst₂ immunoreactivity and cone photoreceptors, sections processed for sst₂ were finally subjected to streptavidin Texas red at 1:50 (RPN1233; Amersham Research Products) together with fluorescein-labeled peanut agglutinin (PNA) at 100 μg/ml (343249; Calbiochem, La Jolla, CA). 

The sections were mounted with fluorescent mounting medium (Dako) and examined by microscope (Axiophot; Carl Zeiss, Oberkochen, Germany) equipped with an epifluorescence system and interference filters. Identical fields of sections were photographed for the two fluorescent markers. 

To eliminate the possibility of cross-reaction between primary and secondary antibodies in double-labeling studies, control sections were made by omitting either of the primary antibodies. 

Results

Localization of sst₁ Receptors

Intense sst₁ immunoreactivity was observed in sparsely distributed perikarya located in the inner part of the INL (Fig. 1 B). From these cells, a few immunoreactive processes were seen extending into the IPL. Similarly, labeled perikarya were observed occasionally in the GCL, extending an immunoreactive process into the IPL. Immunostained fibers were present in sublamina 1 of the IPL, and a less dense fiber layer was located in sublamina 3 (Fig. 1B) . Immunoreactivity was also seen in scattered fibers traversing the IPL. 

A small number of ganglion cells were also revealed by sst₁ immunostaining (Fig. 2 A). Moreover, thick processes in the GCL, which were oriented toward the inner surface of the retina, were strongly sst₁ immunoreactive (Fig. 1B) . Double staining for vimentin showed that these sst₁-immunoreactive processes represented the end feet of Müller cells (Fig. 2A) . 

The OPL and the outer nuclear layer (ONL) were unstained, as was the photoreceptor layer. Immunostaining was absent in control sections (Fig. 1A) . 

Localization of sst₂ Receptors

In the INL, sst₂ immunoreactivity was confined to many medium-sized perikarya located mainly in the middle part of the layer (Fig. 1D) . Immunostained processes from these cells (one per cell) were clearly observed to project into the IPL. There, they were seen to dichotomize in sublaminae 1 through 4 of the IPL. Stained processes were not seen to extend to sublamina 5, although this layer also contained immunolabeled fibers. The strongest labeling was seen in sublaminae 1 and 4, where immunoreactive varicose fibers formed dense layers; less dense fiber staining was observed in sublaminae 2, 3, and 5. Double labeling with antibodies directed against PKC revealed no colocalization of sst₂ and PKC (Fig. 3 D), indicating that these cells were not rod bipolar cells. 

In addition to the medium-sized perikarya, large immunoreactive cell bodies were present in the amacrine cell layer of the INL, with perikarya located immediately adjacent to the IPL (Fig. 1E) . Thick immunoreactive processes emerged from these perikarya and terminated mainly in sublamina 1 of the IPL. 

According to double labeling, calbindin-immunoreactive horizontal cells were not stained for the sst₂ receptor (Fig. 2C) . However, sst₂-immunoreactive processes were observed to encircle some of the horizontal cells. Calbindin-positive amacrine cells did not exhibit sst₂ immunoreactivity. 

Intensely stained radial fibers were detected in the GCL. To a minor extent, fibers traversing the INL and ONL were also stained. These fibers were observed to belong to vimentin-immunoreactive Müller cells (Fig. 2B) . Vimentin staining was also present in the OPL. Similarly, the OPL exhibited a diffuse immunostaining for the sst₂ receptor (Figs. 1D 2B) . 

Strong immunostaining was observed in the outer part of inner segments of some of the photoreceptor cells (Fig. 1F) . To characterize these structures further, we performed double labeling with fluorescein-conjugated PNA, which binds specifically to cone photoreceptors. We found an almost complete colocalization of PNA with sst₂ immunoreactivity (Fig. 2D) . The ONL was devoid of staining. In control sections, sst₂ immunoreactivity was not observed (Fig. 1C) . 

Double Staining for Somatostatin with sst₁ and sst₂

A few strongly somatostatin-immunoreactive amacrine perikarya were observed in the inner part of INL and in the GCL, with processes extending into the IPL. Punctate immunoreactive fibers were present in sublamina 1 of the IPL. Occasionally, single varicose fibers were encountered in the other laminae of the IPL. 

All somatostatin-positive perikarya examined were also immunoreactive for sst₁ (Fig. 3A) . This colocalization of immunoreactivity was also observed in varicose fibers in sublamina 1 and to a minor extent in other laminae of the IPL. Although most of these fibers were oriented horizontally, some fibers were also seen to traverse the IPL. Some sst₁-positive fibers in sublamina 1 did not stain for somatostatin. 

No colocalization was found between somatostatin and sst₂ in either perikarya or fibers in any part of the retina (data not shown). No immunoreaction was obtained when omitting either of the primary antibodies in the control sections. 

Double Staining for Tyrosine Hydroxylase with sst₁ and sst₂

Large amacrine perikarya in the INL showed strong immunoreactive labeling for TH. These perikarya were located at the border of the IPL. Immunoreactive processes were observed mainly in sublamina 1 of the IPL, where they formed a thick, strongly labeled layer. 

Double labeling revealed that sst₂ immunoreactivity colocalized with TH in the large amacrine perikarya as well as in fibers of sublamina 1 (Fig. 3B) . All TH-immunoreactive structures were observed to exhibit sst₂ staining, whereas no TH immunoreactivity was observed in the sst₂-positive perikarya in the middle of the INL or in fibers in sublaminae 2, 4, and 5. 

None of the large TH-positive amacrine cells or their processes was seen to stain for sst₁ (data not shown). No cross-reactivity was observed in control sections, when either of the primary antibodies was omitted. 

Double Staining for Glutamate Decarboxylase with sst₁ and sst₂

Many GAD-immunostained medium-sized perikarya were present in the INL and to a lesser extent in the GCL. These cells possessed the characteristics of amacrine cells, with immunoreactive processes projecting to the IPL. Varicose fibers showing strong immunoreactivity for GAD were observed in all five sublaminae of the IPL. With double labeling, GAD was not observed to colocalize with either sst₁ (data not shown) or sst₂ immunoreactivity (Fig. 3C) in any of the cells examined. 

Figure 4 summarizes our results on sst₁ and sst₂ immunoreactivity. 

Discussion

It is interesting to note that although the number of somatostatin-containing cells in the rat retina is low, the distribution of somatostatin receptors is quite widespread. For the first time, we have demonstrated the presence of sst₁ and sst₂ immunoreactivity in the rat retina. Whereas sst₁ immunoreactivity was found in only a few amacrine and ganglion cells, the sst₂ receptor showed a more widespread distribution. Scant information is available on the physiological significance of somatostatin in the mammalian retina. However, assuming that the widespread distribution of somatostatin receptor immunoreactivity in the rat retina represents functional receptors, our data suggest that somatostatin plays an important role in modulating visual signals. Electrophysiological studies on rabbit retina suggest that somatostatin may modulate retinal function by enhancing the signal-to-noise ratio and producing a shift in the center–surround balance of the ganglion cells. ¹¹ In that study, somatostatin affected both ganglion cells and interneurons and was acting as a slow modulator. However, to our knowledge, no functional studies have been reported on somatostatin in rat retina. 

Amacrine cells, and cone bipolar cells, have been shown to terminate in all sublaminae of the IPL in the rat. ^{28 29} We observed processes originating in perikarya of the INL that ramified clearly in sublaminae 1 through 4, whereas we did not observe processes terminating directly in sublamina 5, although many stained fibers were present in this layer. We could not establish the exact nature of these sst₂-immunoreactive cells in the INL. Most of these cells exhibited a single immunostained process projecting into the IPL with no peripheral processes visible. Furthermore, we could not colocalize sst₂ immunoreactivity with that of PKC, which has been shown to label rod bipolar cells selectively and also some amacrine cells in the rat. ^{30 31 32} Because we observed only one process on these medium-sized sst₂-immunolabeled cells in the INL, they may represent unistratified amacrine cells. 

Somatostatin binding sites have been demonstrated by autoradiography in the OPL and IPL of rat ²² and mouse ²¹ retina. This finding is in accordance with our demonstration of mainly sst₂ labeling and some sst₁ labeling in these layers. In the present study, sst₂ immunoreactivity was also found in the inner segments of the photoreceptors. This observation is matched by autoradiographic studies that have revealed the presence of receptors in the same location. ²² We analyzed this finding further by performing double staining with PNA, which is known to bind specifically to cone photoreceptors. ³³ We found an almost complete colocalization of PNA and sst₂ immunoreactivity, suggesting that the light perception through cone photoreceptors could be influenced by somatostatin. 

Calbindin stains horizontal cells and some amacrine cells in the rat retina. ³⁴ Calbindin-labeled horizontal cells were not immunoreactive for either somatostatin receptor. However, sst₂-positive fibers were sometimes observed to encircle horizontal cell bodies, possibly representing Müller cell fibers. 

The immunohistochemical localization of sst₂ receptors has been described recently in the rabbit retina. ²⁶ Our localization of sst₂ immunoreactivity in the rat was clearly different from the pattern observed in the rabbit retina. In contrast to our observation in the rat, immunoreactive cells in the INL of the rabbit were shown to be of the rod bipolar type. In both rat and rabbit, sst₂ immunostaining was present in large amacrine cells in the INL adjacent to the IPL. However, whereas the rabbit immunoreactive amacrine cells gave rise to processes located in sublaminae 2 and 4, those in the rat were seen to send thick processes into sublamina 1. In the rat, sst₂-labeled Müller cell fibers were detected in the GCL and nuclear layers, but no such staining was reported in the rabbit. The apparent species-dependent distribution of sst₂ receptors in retinal cells may indicate a difference in somatostatin function in the retina between the two species or perhaps merely a difference in the somatostatin receptor subtypes involved. 

The cellular localization of somatostatin in amacrine cells in the INL and GCL agrees with previous reports on the rat retina. ^{1 2 3} It is an interesting and new finding in our study that the sst₁ receptor was present on somatostatinergic cells. This suggests that somatostatin in the amacrine cells may be able to regulate its own release through the sst₁ receptor subtype. There are no previous reports on autoregulation of somatostatin in the retina, but autoreceptors for glutamate and γ-aminobutyric acid (GABA) have been localized by immunohistochemistry in rat retina. ^{35 36} However, we have shown recently by immunohistochemistry that the sst₁ receptor is present in somatostatinergic neurons of the rat hypothalamus, especially in the nerve terminals located in the external lamina of the median eminence. ³⁷ As was the case in rat hypothalamus, sst₂ immunoreactivity was not detected in somatostatinergic perikarya or fibers in the retina; it is thus unlikely that sst₂ receptors take part in any autoregulation of somatostatin release in these tissues. 

Immunoreactivity for TH has been detected in amacrine cells and interplexiform cells in the rat retina. ^{38 39} Our observation of TH-immunostaining in large amacrine cells in the proximal INL is in agreement with these studies. In our study, TH-immunoreactive perikarya and fibers were found to express the sst₂ receptor as well, whereas there was no colocalization of TH with sst₁. This suggests that the sst₂ receptor may mediate the regulation of dopamine release by somatostatin in the rat retina. It remains to be established whether receptors sst₃, sst₄, and sst₅ are present in dopaminergic retinal cells. In contrast, Johnson et al. ²⁶ did not observe such colocalization of TH and sst₂ in rabbit retina. In chicken retina, amacrine cells coexpressing enkephalin, neurotensin, and somatostatin (ENSLI) are active in the dark and appear to exert an inhibitory effect on dopaminergic amacrine cells. ⁴⁰ The equivalents of ENSLI cells have not been found in the mammalian retina, and whether somatostatin exerts an inhibition on dopaminergic cells is not known. However, the results presented here suggest that somatostatin may be able to influence dopaminergic amacrine cells in the rat, at least through the sst₂ receptor. Interestingly, somatostatin has been shown to increase the release of dopamine in rat striatum. ⁴¹  

Glutamate decarboxylase is present in amacrine cells in the INL and GCL and is distributed in the five sublayers of IPL, ^{42 43 44} findings matched by the observations of this study. We found no colocalization of GAD immunoreactivity with either sst₁ or sst₂ receptors. In the rat retina, GABAergic amacrine cells make synaptic contacts in dyads with rod bipolar cells and the same amacrine cell makes reciprocal synapses onto the bipolar cells. ⁴⁵ Therefore, GABA is considered a major modulator for signal transmission through the rod bipolar cells. Our study of somatostatin receptor localization indicates that this system is not influenced by somatostatin, at least not through sst₁ or sst₂ receptors. 

In conclusion, the distribution of sst₁ and sst₂ receptors in the rat retina suggests that somatostatin is able to modulate visual signals at different levels. Indeed, sst₁ may act as an autoreceptor in somatostatinergic amacrine cells in the INL and GCL. Somatostatin may influence the secretion of dopamine through the sst₂ receptor and, furthermore, may exert an as yet unknown effect on amacrine cells in the INL. Somatostatin also may affect cone photoreceptors through the sst₂ receptor. The detection of sst₁ receptors in ganglion cells may indicate a role for somatostatin in the transmission of retinal signals to the brain. 

 Supported by a grant from the Biotechnology Centre for Cellular Communication and a grant from the Lundbeck Foundation.

 Submitted for publication August 10, 1998; revised February 5, 1999; accepted May 20, 1999.

 Proprietary interest category: N.

 Corresponding author: Lone Helboe, Institute of Medical Anatomy, Section B, Panum Institute, University of Copenhagen, Blegdamsvej 3, DK-2200N Copenhagen, Denmark. E-mail: [email protected]

 

Figure 1.

 

(A) Control section of rat retina reacted with the preabsorbed sst₁ antiserum showing no immunostaining. (B) The sst₁-immunoreactive amacrine perikarya were present in the INL sending processes into the IPL. Strongly stained radial fibers were present in the GCL. Immunoreactive punctate fibers were present in sublamina 1 and to a lesser extent in sublamina 3 of the IPL. (C) In control sections reacted with preabsorbed sst₂ antiserum, no immunostaining was observed. (D) In the INL, medium-sized perikarya were sst₂-immunoreactive and are shown with single processes ramifying in sublaminae 1 through 4 of the IPL. Immunolabeled fibers were also present in sublamina 5. Radial fibers in the GCL were strongly stained, and the OPL shows moderate staining. (E) In the INL, a large amacrine perikaryon sending thick processes to sublamina 1 exhibits intense sst₂ immunoreactivity. (F) Strong sst₂ immunoreactivity was detected in inner segments in a subset of photoreceptors. Ph, photoreceptor layer; 1 through 5, sublaminae of the inner plexiform layer. Scale bars, 25μ m.

Figure 2.

 

Double-fluorescence staining of rat retina for sst₁ (A) and vimentin (A*). The sst₁ immunoreactivity colocalized with vimentin in thick Müller cell fibers in the GCL (arrows). An sst₁-stained ganglion cell is seen (A; open arrow). (B) Immunoreactivity for sst₂ was present in vimentin-stained Müller cell fibers (B*) in the GCL (arrows) and also in the nuclear layers (arrowheads). Double labeling for sst₂ (C) and calbindin (C*). Some calbindin-immunoreactive horizontal cells (arrows) were encircled by sst₂-immunoreactive fibers. However, there was no direct colocalization of sst₂ and calbindin. Other horizontal cells were devoid of a ring of sst₂-stained fibers (open arrow). In weakly calbindin-stained amacrine cells (arrowheads), no sst₂ immunoreactivity was present. In the photoreceptor layer, sst₂ immunoreactivity was present in the outer parts of inner segments (D). By double labeling for PNA (D*), a complete colocalization was observed (arrows). Ph, photoreceptor layer. Scale bar, 25 μm.

Figure 3.

 

Immunofluorescence double labeling of rat retina. The sst₁ and sst₂ immunoreactive cells were visualized by fluorescein (left). Somatostatin-, TH-, GAD- and PKC-immunoreactive cells were visualized by Texas red (right). (A) An sst₁-immunoreactive amacrine cell in the INL. An immunostained process is observed in the IPL. (A*) The same cell stained positively for somatostatin. (B) Immunoreactivity for sst₂ in a large amacrine perikaryon in the INL. (B*) Immunoreactivity for TH colocalized with sst₂ in the amacrine perikaryon and in fibers in sublamina 1 of the IPL. (C) An sst₂-immunoreactive medium-sized perikaryon was visible in the INL (arrow). (C*) GAD immunoreactivity was present in amacrine perikarya in the INL (open arrows) but did not colocalize with sst₂-positive perikarya (arrow), which, however, showed a weak fluorescence through the interference filter used for Texas red. (D) Amacrine perikarya immunoreactive for sst₂ in the INL (arrows). (D*) No colocalization was observed with PKC in sst₂-stained perikarya (arrows). Rod bipolar cells and a few amacrine cells were PKC-immunoreactive (open arrows). Scale bar, 25 μm.

Figure 4.

 

Schematic diagram showing the distribution of sst₁ and sst₂ immunoreactivity in rat retina. Ph, photoreceptor layer; 1 through 5, sublaminae of the inner plexiform layer.