Localization of cGMP-Dependent Protein Kinase Isoforms in Mouse Eye

Abstract

purpose. To examine the expression of the major isoforms of cyclic guanosine monophosphate (cGMP)-dependent protein kinase (cGK) in mouse eye.

methods. Immunohistochemical localization of cGMP in mouse eye cryosections was performed using an anti-cGMP antibody, followed by visualization with indirect fluorescence microscopy. The presence of types Iα, Iβ, and II cGK mRNAs in mouse eye extracts was determined initially by RNase protection analysis. Further localization of cGK I and II mRNAs on cryosections was accomplished by in situ hybridization using digoxigenin-labeled cRNA probes and an alkaline phosphatase-conjugated anti-digoxigenin antibody. Finally, cGK I protein was localized to subcellular areas within the retina using an anti-cGK I–specific primary antibody.

results. In initial immunohistochemical experiments cGMP was present in numerous regions and layers within the eye and retina. Subsequent RNase protection studies demonstrated that cGK Iα, Iβ, and II mRNAs were present in mouse eye and that type Iβ mRNA were 6.6 and 30 times more abundant than type Iα and type II, respectively. By in situ hybridization, cGK I mRNA was localized to photoreceptor inner segments and the ganglion cell and inner nuclear layers of the retina, and lesser amounts were found in the ciliary epithelium, lens, and cornea. The cGK II mRNA expression pattern was similar but not identical with that of cGK I. Finally, within the retina, cGK I protein was most abundant in the inner plexiform layer, with significant amounts in ganglion cells and photoreceptor inner segments as well.

conclusions. The presence of these cGK isoforms in discrete areas throughout the eye suggests multiple roles for the cGMP-dependent signal transduction system in the regulation of physiologic and pathologic ocular processes.

Since its discovery in 1963, ¹ cyclic guanosine monophosphate (cGMP) has been found to be an important signaling molecule in a number of cellular systems and physiologic events. ^{2 3} One example is found in the outer segments of rod photoreceptors, where phototransduction requires the phosphodiesterase-mediated hydrolysis of cGMP and subsequent inactivation of cGMP-gated cation channels. ⁴ This unique and critical role of cGMP in the outer retina has led to the search for additional cGMP-dependent signaling mechanisms throughout the eye. ^{5 6} Nitric oxide and/or cGMP has been implicated in the regulation of numerous ocular processes, including glaucoma, ^{7 8} diabetic retinopathy, ⁹ uveitis, ¹⁰ corneal wound healing, ⁶ and retinal neurotransmission. ^{5 11 12 13} Few of these studies, however, attempted to identify the cGMP receptor involved in the signaling cascade. The identification of the specific effectors of cGMP in ocular tissues could prove valuable in efforts to develop new therapies for ocular diseases. 

One reason for the paucity of information regarding cGMP effector mechanisms is the complicated array of potential cGMP receptors within cells. Unlike cAMP, which acts primarily on the ubiquitous cAMP-dependent protein kinase, cGMP can bind to three separate classes of receptors: ion channels, phosphodiesterases, and cyclic nucleotide-dependent protein kinases. ³ Within the eye, little is known about the distribution and role(s) of the cGMP-dependent protein kinase (cGK) isoforms, despite their importance in the regulation of cellular events similar to those that occur there. ^{14 15} In contrast, the ocular expression patterns of many other potential cGMP-dependent signal transduction components have already been described. ^{16 17}  

Among the known cGK isoforms, many important differences exist that may further complicate efforts to study the consequences of cGMP production. The type Iα and Iβ cGK isoforms are cytosolic and differ structurally only at their amino termini. ³ Phosphorylation by either cGK Iα or Iβ can stimulate such cellular events as smooth muscle relaxation, platelet aggregation, apoptosis, and neurotransmission. However, the activation, expression patterns, and regulatory properties of the type I cGKs are distinct. ³ The type II cGK is membrane-associated and demonstrates both cyclic nucleotide affinities and substrate specificities distinct from the type I isoforms. ^{18 19} cGK II is highly expressed in intestinal microvilli where it has been shown to regulate chloride ion secretion. ²⁰ Mice carrying a null mutation of the gene encoding cGK II were refractory to enterotoxin-stimulated intestinal fluid secretion. ²¹ Together, the physical, biochemical, and physiological differences between cGK Iα, Iβ, and II suggest that these isoforms, if present, may serve separate and varied functions in the eye. Therefore, as an initial step toward further investigations of this possibility, we determined the presence and localization of cGK Iα, Iβ, and II in the mouse eye, by using RNase protection analysis, in situ hybridization, and immunohistochemistry. 

Methods

Preparation of Mouse Eye Sections

Mouse eye sections were prepared essentially as described. ²² Mice were reared in an environment with a 12-hour light period followed by 12 hours of darkness. At the end of a dark period, 2-month-old 129/Sv mice were exposed for 5 minutes to room light, in an effort to reduce the disproportionately high cGMP signal in photoreceptor outer segments anticipated in the cGMP immunohistochemistry experiments. Eyes were then cryoprotected with sucrose solutions and embedded in sucrose phosphate and optimal cutting temperature medium (OCT; Miles, Kankakee, IL). Sections were cut at 5-μm thickness and placed on RNase-free slides previously coated with gelatin and poly-l-lysine. The slides were stored at− 90°C for later processing. All procedures conformed to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Immunohistochemistry

Immunohistochemical experiments were performed as described by Raymond and Barthel. ²³ Cryosections were blocked with 20% normal goat serum in 0.01 M phosphate-buffered saline before the application of 1:200 diluted rabbit anti-cGMP ^{24 25} (provided by W.M. Steinbusch, Maastricht University, The Netherlands) or 1:200 rabbit anti-mouse cGK I antibody (StressGen, Victoria, British Columbia, Canada). Visualization of primary antibody binding was achieved using a goat anti-rabbit fluorescein isothiocyanate (FITC)–labeled secondary antibody (Vector, Burlingame, CA) diluted 1:50. To control for nonspecific binding of secondary antibody, some sections were treated with secondary antibody in the absence of preincubation with primary antibody. Further controls for the anti-cGMP immunohistochemistry included preadsorption of the primary antibody for 2 hours at room temperature with 20 μg thyroglobulin-cGMP ²⁵ before incubation on sections. Thyroglobulin-cGMP was generated as described by de Vente et al. ²⁴ Similar controls for the anti-cGK I immunohistochemistry included preadsorption of the anti-cGK I antibody with the cGK I peptide (peptide–antibody molar ratio, 100:1) used in the production of this antibody. Of note, equal photographic exposure times were used for matched nonpreadsorbed and preadsorbed slides. Comparison of the indirect fluorescence signal produced by the preadsorbed and nonpreadsorbed primary antibodies was then used to determine the specific cGMP and cGK I signals within the eye. No distinction was made between cGMP levels or cGK expression in rods versus cones in any of these experiments. The relative levels of fluorescent immunoreactivity reported in Table 1 , columns A and D, are based on subjective rating by several investigators. 

Plasmid Constructs

The template for the cGK I cRNA probes used in the RNase protection assays and in situ hybridizations was constructed by inserting a PstI-NcoI fragment of mouse cGK Iβ ²⁶ (encoding amino acids 43–175) into a pGEM-5Z vector (Promega, Madison, WI), creating p5Z-CGKI. This 403-bp fragment spans the cGK Iβ mRNA splice site and therefore contains 213 bp of DNA sequence shared by cGK Iα and Iβ in addition to 190 bp of DNA sequence specific for cGK Iβ. The template for the cGK II cRNA probes was constructed by digesting pCGKI.2 ²⁷ with SstI and NsiI and inserting the resultant 353-bp fragment into a pSP73 vector (Promega) previously digested with SstI and PstI, creating pSP73-CGKII. The integrity of the p5Z-CGKI and pSP73-CGKII templates was verified by DNA sequencing and restriction enzyme digestion. p5Z-CGKI was linearized with NdeI or NcoI and used to synthesize antisense or sense cGK I cRNA probes, respectively. In the RNase protection assays, the presence of cGK Iα and/or cGK Iβ could be distinguished by virtue of the difference in the expected size of the protected fragments (described later). In the in situ hybridization experiments, however, no such distinction could be made between these two isotypes. pSP73-CGKII was linearized with EcoRI and XhoI and used to synthesize antisense or sense cGK II cRNA probes, respectively. 

RNase Protection Analysis

Synthesis of sense RNA strands and antisense cRNA probes for RNase protection experiments was performed essentially as described ²⁸ using ³²P-UTP as the labeling isotope. Briefly, total RNA was purified from whole mouse eye, lung, intestine, and brain using an acid guanidinium isothiocyanate-phenol-chloroform protocol. ²⁹ The RNA was quantitated by spectrophotometry and verified by formaldehyde-agarose gel electrophoresis followed by ethidium bromide staining. In separate control reactions, yeast tRNA was mixed with varying concentrations of cGK I or cGK II sense RNA, which also provided a standard curve for later quantification of signal intensity. Individual samples containing either 20 μg of total RNA from a mouse tissue or 0, 0.33, 1, 3.3, 10, or 33 pg of sense RNA were hybridized with the appropriate antisense cRNA probe. After treatment with RNase A and T1, samples were incubated with proteinase K and sarkosyl, precipitated, resuspended, and denatured before they were electrophoresed. The predicted sizes for the protected fragments from samples containing mouse tissues were 213, 399, and 353 nucleotides for cGK Iα, Iβ, and II, respectively. Protected fragments from the samples containing cGK I or cGK II sense RNA strands used in the standard curves had predicted lengths of 405 and 379 nucleotides, respectively. Full length antisense probes produced with the p5Z-CGKI or pSP73-CGKII templates had predicted lengths of 441 or 394 nucleotides, respectively. Quantification of signal intensity was performed using a PhosphorImager (with ImageQuant software; Molecular Dynamics, Phoenix, AZ), as previously described. ²⁸ Absolute amounts of cGK Iα, Iβ, and II mRNA present in the tissues were then determined by interpolation from sense RNA standard curves. Values listed in Figure 2C represent the average of two experiments and are expressed as picograms mRNA per milligram total RNA using mRNA sizes of 8.5 and 6.0 kb for cGK I and II, respectively. ²⁷  

In Situ Hybridizations

Digoxigenin-labeled cGK I and II sense and antisense RNA probes were prepared by in vitro transcription (Genius system; Boehringer–Mannheim, Indianapolis, IN) and hybridized to mouse eye cryosections as described previously, ³⁰ with the exception that 4 μg of probe was used in each hybridization. After treatment with RNase, any remaining hybridized probe was visualized with alkaline phosphatase (AP)–conjugated anti-digoxigenin antibodies followed by an enzymatic color reaction using the substrate 5-bromo-4-chloro-3-indolyl phosphate (BCIP) and nitroblue tetrazolium (NBT) salt. All experiments using antisense RNA probes were compared with matched controls using the appropriate sense RNA probe to determine signal specificity. Table 1 , columns B and C, compares relative signal intensities among eye regions, detected by a single probe. Values presented in this table are not meant to reflect relative differences between cGK I and II signal intensities in a given eye region. As noted, the cGK I probe spans the mRNA splice junction responsible for the production of the Iα and Iβ isotypes; therefore, the specific localization of these cGK I isotypes cannot be distinguished in these experiments. 

Western Blot Analysis

To obtain whole mouse eye extract, 10 mouse eyes were quick-frozen in liquid nitrogen, pulverized, and added to 500 μl of homogenization buffer (250 mM sucrose, 1 mM EDTA, 1 mM dithiothreitol, 10 mM Na₂HPO₄ [pH 7]) containing 1 mM phenylmethylsulfonyl fluoride, 5 μg/ml leupeptin, and 1 μg/ml pepstatin A (Boehringer–Mannheim). Protein concentrations were determined (reagent from Bio-Rad, Hercules, CA) on an automated workstation (Biomek-1000; Beckman, Berkeley, CA). Duplicate samples containing 100 μg total protein from the extract were separated by 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE; MiniProtean II apparatus; Bio-Rad) and transferred to nitrocellulose (BA85; Schleicher & Schuell, Keene, NH) overnight in 20 mM Tris (pH 8.2), 150 mM glycine, and 20% methanol using the MiniProtean II apparatus. The following day, the nitrocellulose membranes were blocked for 2 hours in TBST buffer (50 mM Tris-HCl [pH 7.5], 150 mM NaCl, Tween 0.05%) containing 5% nonfat dried milk and 1% bovine serum albumin. Nitrocellulose membranes were then divided into identical sections and incubated at room temperature for 4 hours in anti-mouse cGK I primary antibody diluted to 1:200 in TBST blocking buffer or an identical dilution preadsorbed with a cGK I peptide as described. After three washings in TBST (10 minutes each), the membrane samples were incubated for 1 hour at room temperature in a 1:10,000 dilution of AP-conjugated rabbit anti-mouse antibody (in TBST blocking buffer). The membranes were washed three more times in TBST (10 minutes each) before visualization of color product using the AP substrate BCIP (0.4 mM) and NBT salt (0.4 mM) diluted in a buffer containing 100 mM Tris-HCl (pH 9.5), 100 mM NaCl, and 5 mM MgCl₂. The expected size of the cGK I monomers is approximately 76 kDa. ³  

Results

cGMP Immunohistochemistry

Before pursuing histochemical analyses of cGK expression, we sought to qualitatively determine tissue levels of cGMP using a comparable method and identical physiological conditions. This in turn might offer insight into potential interactions between cGMP and cGK interactions in ocular tissues. As determined by indirect fluorescence microscopy, the outer segments of photoreceptors contained the highest amounts of specific cGMP immunoreactivity (Fig. 1A 1B 1C ; Table 1 , column A). Other retinal layers that exhibited cGMP-specific fluorescence included (in order of greatest to least intensity) the inner plexiform layer (IPL), the ganglion cell layer (GCL), the outer plexiform layer (OPL), and the pigmented epithelium. No specific signal was detected in the inner nuclear layer (INL), the outer nuclear layer (ONL), or the inner segments. Outside the retina, cGMP was found in greatest abundance within the corneal endothelium (Fig. 1I) , followed by the lens epithelium (Fig. 1G) , ciliary epithelium (Fig. 1E) , and choroid (Figs. 1B 1C) . No specific signal was detected within the corneal epithelium (Fig. 1F) . Signal specificity was determined by comparing matched cryosections treated with either the nonpreadsorbed or preadsorbed anti-cGMP antibodies. 

RNase Protection Analysis of cGK Isoform Expression

The presence and relative abundance of cGK isoforms (Iα, Iβ, and II) in various mouse tissues was examined by RNase protection analysis (Fig. 2A 1B 1C ). Whole mouse eye had the largest amount of cGK Iβ mRNA (5.9 pg mRNA per microgram total RNA), possessing a 20-, 10-, and 3.7-fold higher level than lung, intestine, or brain, respectively (Figs. 2A 2C) . The presence of multiple bands below the predicted cGK Iβ band in all tissues examined represents binding by incompletely synthesized probe, because matching bands were also seen in sense control lanes hybridized with the same probe (data not shown). However, because of this internal experimental control, quantitation of absolute cGK mRNA levels was possible. ²⁸ Expression of cGK Iα mRNA was also detected in mouse eye (0.9 pg mRNA per milligram total RNA), but the level was 6.6 times lower than that of cGK Iβ (Fig. 2A 2C) . Levels of cGK Iα mRNA expression were similar in all tissues examined. The greatest amount of cGK II mRNA was in mouse intestine, the tissue from which this isoform was originally isolated ³¹ (Figs. 2B 2C) . In contrast, the level of cGK II mRNA (0.2 pg of mRNA per milligram total RNA) in whole mouse eye was 2.5, 53, and 10 times lower than in mouse lung, intestine, and brain, respectively (Figs. 2B 2C) . Negative control samples containing only yeast tRNA and antisense cRNA probes were fully degraded after RNase treatment (data not shown). In summary, all three major cGK isoforms were expressed in mouse eye. 

In Situ Hybridization of cGK I and II

The mouse cGK I probe (which recognizes both cGK Iα and Iβ mRNAs) hybridized strongly and selectively to cells in multiple layers of the retina (Figs. 3A 3B ; and Table 1 , column B). The signal was most intense in the GCL, INL, and photoreceptor inner segments, with trace amounts of signal in the ONL. The apparent restriction of cGK I and II mRNA to nuclear layers and photoreceptor inner segments reflects the subcellular localization of the transcriptional and translational machinery within the retina. Outside the retina, expression of cGK I was detected in the lens epithelium (Fig. 3G) , ciliary epithelium (Fig. 3E) , choroid (Figs. 3A 3B ), and corneal epithelium and endothelium (Fig. 3I) . The mouse cGK II probe also hybridized selectively within various retinal layers and ocular tissues (Figs. 3C 3D 3F 3H 3J ; Table 1 , column C). The highest level of cGK II signal was in the INL (Figs. 3C 3D) , with lower levels in the GCL, photoreceptor inner segments, ciliary epithelium (Fig. 3F) , and corneal epithelium (Fig. 3J) . Lower levels of cGK II expression were observed in the ONL, choroid and corneal endothelium, whereas the lens epithelium (Fig. 3H) and photoreceptor outer segments had no detectable cGK II mRNA. Although cGK II mRNA appeared to be considerably less abundant than cGK I mRNA in these studies, quantification of message levels is not possible using the Genius system (Boehringer–Mannheim), because of differences in the sizes and nucleotide compositions of the cGK I and II probes. However, data from the RNase protection analyses suggest a generally lower level of expression in mouse eye of cGK II mRNA than of cGK I mRNA (Fig. 2) . The specificity of signal in these in situ hybridization studies was determined by comparing results derived from the antisense and sense probes for cGK I and II, as described in the Methods section. 

Detection of cGK I in Whole Mouse Eye by Western Blot Analysis

To verify the presence of cGK I protein in whole mouse eye extracts, we used Western blot analysis with an antibody directed against a carboxyl terminus cGK I peptide. Because cGK Iα and Iβ differ only at their amino termini, this antibody does not discriminate between these two isotypes. As shown in Figure 4 , lane 1, the anti-cGK I antibody recognizes a single 76-kDa protein corresponding to the previously reported size of cGK Iα and Iβ. ³ No cross-reactivity with the type II cGK monomer (86 kDa) ¹⁹ was observed on these blots. Preadsorbing the anti-cGK I antibody with the carboxyl terminal cGK I peptide eliminated the 76-kDa band (Fig. 4 , lane 2), demonstrating the specificity of this antibody. 

cGK I Immunohistochemistry

To delineate the subcellular sites of cGK I activity within the retina, immunohistochemical localization was performed using the anti-mouse cGK I antibody described. The IPL had the highest level of specific cGMP immunoreactivity, determined by indirect fluorescence microscopy (Figs. 5A 5B 5C ; Table 1 , column D). Other retinal layers with significant cGK I-specific fluorescence include the photoreceptor inner segments and the GCL. Minimal fluorescence was detected in the INL, OPL, pigmented epithelium, and choroid, and there was no specific signal in the photoreceptor outer segments or the ONL. Signal specificity was determined by comparing matched cryosections treated with either nonpreadsorbed or preadsorbed anti-cGK I antibodies. The absence of nonspecific immunoreactivity observed in the Western blot (Fig. 4 , lane 2) compared with the presence of nonspecific binding seen in immunohistochemistry (Fig. 5C) was probably due to differences in the sensitivities, local antibody binding conditions, and antigen presentations inherent in these two techniques. 

Discussion

It is well known that ocular tissues vary in baseline cGMP level and cGMP response to certain common stimuli. ^{5 32 33} Our immunohistochemical results also revealed differing amounts of cGMP within various retinal layers and ocular tissues in eyes briefly exposed to light before harvest. Although these findings may portend the existence of multiple cGMP systems throughout the eye, the downstream signaling components are yet to be completely understood. Evidence for the existence of a cGK in whole retinal extracts was provided by Thompson and Khorana, ³⁴ using cGMP photoaffinity labeling. More recent studies have demonstrated cGK activity in retinal amacrine ¹¹ and horizontal cells, ³⁵ pigmented epithelium, ³⁶ and corneal epithelium. ⁶ Our data further define the presence of the individual cGK isoforms in mouse eye and provide a more comprehensive localization of cGK I and II mRNA and cGK I protein in specific ocular tissues and retinal layers. It should be noted, however, that the subcellular distribution of soluble retinal proteins such as cGKI can vary with different light conditions. ³⁷  

The differences observed in the relative amounts of cGMP and cGK isoform expression are probably due in part to the more diverse physiologic role of cGMP in ocular tissues. Therefore, it is not surprising that cGMP levels do not parallel cGK expression under the limited conditions examined in this study. Of particular note is the absence of specific cGK I protein immunoreactivity in the photoreceptor outer segments, which suggests that cGK I is compartmentalized and may not play a significant role in phototransduction. However, that there was no detectable cGMP or cGK in certain eye regions does not preclude their existence at very low levels or under some different manner of stimulation (e.g., light versus dark adaptation). Furthermore, tissue abundance of mRNA and/or protein should not necessarily be construed as a measure of physiologic importance. 

Despite these cautions, the presence of individual cGK isoforms in certain eye regions may provide insights into their potential roles, particularly when physiologic data obtained from other tissues are considered. For example, cGK II is known to modulate chloride ion efflux and subsequent fluid secretion from intestinal microvilli through phosphorylation of a specific ion channel. ²⁰ Similarly, chloride channels in the basolateral membranes of ciliary epithelial cells are partly responsible for producing aqueous humor. ³⁸ The finding of cGK II mRNA in ciliary epithelial cells suggests that this isoform may catalyze reactions that influence intraocular fluid homeostasis. Of related interest, the injection of nitric oxide donors in rabbit eye has been shown to produce a dramatic decrease in intraocular pressure. ⁸ In contrast, the cGK I isoforms, which regulate neurotransmission elsewhere in the nervous system, ³⁹ may play a greater role in retinal neurotransmission. A recent report provided evidence that upregulation of cGK activity by nitric oxide analogs depressed γ-aminobutyric acid receptor function in cultured retinal amacrine cells. ¹¹ The functional importance of this observation and the specific cGK isoform(s) involved are not yet known, but our results showed that cGK I was highly expressed in the inner nuclear layer and neighboring plexiform (synaptic) layers of the retina. Finally, the postulated role of cGK I in apoptosis, ⁴⁰ combined with its presently described localization in retinal ganglion cells, suggests a potential contribution to the pathogenesis of glaucoma. 

A complete understanding of the consequences of differential expression patterns of cGK isoforms requires knowledge of their physiologic substrates, but few substrates have been identified that are preferentially phosphorylated by cGK. ¹⁴ Recently, however, the mRNA transcript of G-substrate, a protein phosphatase inhibitor and specific cGK substrate, was discovered in mouse eye. ⁴¹ Ongoing efforts to characterize additional substrates will aid in determining the functional impact of cGK phosphorylation within the eye. Furthermore, the use of cGK I– or II–selective cyclic nucleotide analogues ^{3 19} may help elucidate the roles of these individual isoforms in cell types or tissues where they are colocalized. This information then may lead to more effective treatments for the growing list of ocular diseases believed to be influenced by cGMP-dependent signaling systems. 

²  Present address: Department of Ophthalmology, University of Wisconsin, 2870 University Avenue, Madison, WI 53705-3611.

 Supported by National Institutes of Health Grants GM50791 (MDU) and EY04318 (PAR).

 Submitted for publication May 27, 1999; revised November 29, 1999 and February 28, 2000; accepted March 17, 2000.

 Commercial relationships policy: N.

 

Table 1.

 

cGMP Abundance and cGK I and II Expression Patterns in Mouse Eye

	A cGMP Abundance	B cGK I mRNA Expression	C cGK II mRNA Expression	D cGK I Protein Expression
Ganglion cell layer	++	++++	+++	++
Inner plexiform layer	+++	ND	ND	++++
Inner nuclear layer	ND	++++	++++	+
Outer plexiform layer	+	ND	ND	+
Outer nuclear layer	ND	+	+	ND
Inner segments	ND	++++	+++	++
Outer segments	++++	ND	ND	ND
Pigmented epithelium	+	+	+	+
Choroid	+	+	+	+
Ciliary epithelium	++	++	+++	NS
Corneal epithelium	ND	+	++	NS
Corneal endothelium	+++	+	+	NS
Lens epithelium	++	+++	ND	NS

 Results shown in Figures 1 3 and 5 were evaluated subjectively by several investigators. Levels of (A) cGMP, (B) cGK I mRNA, (C) cGK II mRNA, or (D) cGK I protein are rated + to ++++, from lowest to highest. Region(s) possessing no detectable specific signal were rated ND (not detected), whereas those noted as NS were not studied. These ratings reflect differences within individual columns and are not meant to indicate relative differences between columns.

Figure 1.

 

Immunocytochemical localization of cGMP in mouse eye cryosections. Nomarski image of mouse retina (A) and corresponding fluorescence images revealing nonpreadsorbed (B) and preadsorbed (C) anti-cGMP antibody binding within the retina. Nomarski images (left) and corresponding fluorescence images (right) showing (D, E) nonpreadsorbed anti-cGMP antibody binding within ciliary processes, (F, G) nonpreadsorbed anti-cGMP antibody binding within lens, and (H, I) nonpreadsorbed anti-cGMP antibody binding within cornea. Fluorescence images of ciliary processes, lens, and cornea incubated with preadsorbed anti-cGMP antibody (not shown) were used to determine signal specificity as shown in the retina images in (B) and (C). Indirect fluorescence for these experiments was generated through use of an FITC-labeled secondary antibody. Comparative results are listed in Table 1 , column A. IS, inner segments; OS, outer segments; RP, retinal pigmented epithelium; CH, choroid; CE, ciliary epithelium; EP, epithelium; EN, endothelium.

Figure 2.

 

RNase protection analysis of (A) cGK Iα and Iβ and (B) cGK II mRNA transcripts in various mouse tissues. Antisense cRNA probes (cGK I: 441 nucleotides; cGK II: 394 nucleotides) were hybridized to the appropriate sense standards (0–30 pg; not shown) or to 20 μg of total RNA isolated from whole mouse eye, lung, intestine, or brain. The 399-nucleotide fragment near the top of (A) corresponds to protected cGK Iβ transcripts, and the 213-nucleotide fragment near the bottom corresponds to protected cGK Iα transcripts. Similarly, the 353-nucleotide fragment identified in (B) corresponds to protected cGK II transcripts. The sizes indicated were determined by comparison with DNA size standards. Autoradiograms were exposed for 5 days at −80°C with intensifying screens. (C) Quantitation of the RNase protection results after phosphorimaging and analysis by computer. Standard curves (not shown) were generated to allow interpolation and determination of picograms of protected fragment per milligram of total RNA. Results were then expressed as picograms mRNA per milligram total RNA to correct for differences in the lengths of the protected fragments. Values depicted represent the average of two experiments.

Figure 3.

 

In situ hybridization of mouse eye cryosections using digoxigenin-labeled cRNA probes. (A) cGK I antisense probe, retina; (B) cGK I sense probe, retina; (C) cGK II antisense probe, retina; (D) cGK II sense probe, retina; (E) cGK I antisense probe, ciliary processes; (F) cGK II antisense probe, ciliary processes; (G) cGK I antisense probe, lens; (H) cGK II antisense probe, lens; (I) cGK I antisense probe, cornea; and (J) cGK II antisense probe, cornea. As with (B) and (D), matched cryosections hybridized with either the cGK I or II sense probes produced essentially no reaction product in all eye regions examined (data not shown). However, varying amounts of nonspecific deposition of color product were observed in the corneal stroma (see I and J). This artifact was also inconsistently seen in cryosections treated with only the AP-conjugated, anti-digoxigenin secondary antibody. Comparisons of relative levels of cGK I and II in ocular tissues are presented in Table 1 , columns B and C. Abbreviations, see Figure 1 .

Figure 4.

 

Western blot detection of cGK I protein in mouse eye extracts. One hundred micrograms of total protein was electrophoresed in duplicate lanes on an SDS-PAGE gel and transferred to nitrocellulose. Nitrocellulose strips containing a single lane of transferred protein were then incubated with either nonpreadsorbed anti-cGK I antibody (lane 1) or preadsorbed anti-cGK I antibody (lane 2). The molecular masses (in kilodaltons) are indicated to the left.

Figure 5.

 

Immunocytochemical localization of cGK I in mouse retina. (A) Nomarski image of mouse retina and corresponding fluorescence images revealing (B) nonpreadsorbed and (C) preadsorbed anti-cGK I antibody binding within the retina. Indirect fluorescence for these experiments was accomplished in the same manner as described in Figure 1 . Comparative results are listed in Table 1 , column D. Abbreviations, see Figure 1 .

The authors thank Yunde Zhao for aid in the construction of the plasmid constructs and Jim Beals for assistance in the figure preparation.