Characterization and Localization of the Rabbit Ocular Calcitonin Gene-Related Peptide (CGRP)-Receptor Component Protein (RCP)

Mark I. Rosenblatt,1 Gerhard P. Dahl,2 and Ian M. Dickerson1,2

PURPOSE. To determine whether the calcitonin gene-related peptide (CGRP) receptor component protein (RCP), a novel signal transduction molecule, is required for CGRP signaling in the eye and to determine potential ocular sites of CGRP action.

METHODS. The cDNA for the rabbit ocular RCP homologue was cloned using a combination of reverse transcription-polymerase chain reaction (RT-PCR) and rapid ampliﬁcation of cDNA ends (RACE). Function of the rabbit ocular RCP was assessed using a sensitive oocyte-based assay, which utilizes the protein kinase A (PKA)-sensitive cystic ﬁbrosis transmembrane conductance regulator (CFTR) as a sensor of cAMP formation. RCP expression in the rabbit eye was localized using immunohistochemistry.

RESULTS. A 2063-bp cDNA for the rabbit ocular RCP was cloned and sequenced. Expression of the rabbit RCP cDNA confers CGRP responsiveness in a sensitive oocyte-based assay. Antisense oligonucleotides made to the ocular RCP abolishes CGRP responsiveness of ciliary body and iris mRNA in the oocyte–CFTR assay. Localization of RCP protein in the rabbit eye using immunohistochemistry demonstrated RCP immunoreactivity in the ciliary body and iris blood vessels, as well as in layers of the ciliary epithelium.

CONCLUSIONS. The rabbit ocular RCP appears to be required for signal transduction at ocular CGRP receptors and is localized to sites previously reported to bind CGRP, which affect intraocular pressure and neurogenic inﬂammation. (Invest Ophthalmol Vis Sci. 2000;41:1159–1167)

he neurogenic inflammatory response to ocular trauma neurogenic inflammation.7–9,13–15 In support of the role for is characterized by increased anterior uveal blood flow, CGRP action in the ciliary body and iris, CGRP binding sites are Televated intraocular pressure, a breakdown of the present in ciliary process membranes16 and 125I-CGRP binding blood–aqueous barrier, increased cAMP levels in the aqueous in ocular tissue sections has demonstrated the presence of 1–3 humor, and miosis. This ocular neurogenic inflammatory CGRP binding sites in the anterior uvea.17 Together, these response is initiated by traumatic insult to the eye with resul- results suggest the presence of CGRP and CGRP receptors in tant antidromic stimulation of sensory neurons, leading to the the anterior eye, which are responsible for mediating an in- 4,5 release of inflammatory mediators into the anterior uvea. flammatory response in the eye. Calcitonin gene-related peptide (CGRP), a 37-amino acid neu- We recently described the CGRP-receptor component 6 ropeptide that has potent vasodilator action, has been impli- protein (RCP) as a novel molecule required for CGRP signal cated as a mediator of this neurogenic inflammatory re- transduction.18,19 RCP is a novel hydrophilic 146-amino acid sponse.7–9 In the eye, CGRP has been colocalized with protein, which is colocalized with CGRP immunostaining neu- substance P (SP) in neurons that ensheathe the blood vessels of rons, and is required for CGRP receptor activity in cerebellum the anterior uvea.10,11 CGRP levels increase in the aqueous and cochlea.18 Additionally, expression of RCP correlates with humor after ocular trauma and initiation of a neurogenic inflammatory response.12 Furthermore, intracameral administra- CGRP function in the pregnant mouse uterus, where lowered tion of CGRP mimics the increased blood flow, elevated in- levels of RCP protein paralleled reduced inhibitory effects of 19 traocular pressure, and increased cAMP levels associated with CGRP on myometrial contraction. Because of its small size and hydrophilic nature, we do not think that RCP represents a CGRP receptor itself and instead hypothesize that RCP works in conjunction with a membrane-spanning protein to form a From the Departments of 1Biochemistry and Molecular Biology and 2Physiology and Biophysics, the University of Miami School of functional CGRP receptor complex. CGRP receptors thus rep- Medicine, Miami, Florida. resent one of the first examples of G protein–coupled recep- Supported in part by American Heart Association (AHA) Grant-In- tors that require an accessory protein. Here, we determined Aid 9810018FL (IMD), AHA Pre-Doctoral Fellowship 9604015 (MIR), that CGRP receptors in the eye share a similar requirement for and National Institutes of Health Grant DK52328 (IMD). Submitted for publication June 15, 1999; revised November 22, RCP by cloning a cDNA for RCP from the rabbit eye, demon- 1999; accepted November 30, 1999. strating that this cDNA confers CGRP responsiveness in an Commercial relationships policy: N. oocyte–cystic fibrosis transmembrane conductance regulator Corresponding author: Ian Dickerson, Department of Physiology and Biophysics, University of Miami School of Medicine, P.O. Box (CFTR) assay and localizing RCP expression to potential sites of 016430, Miami, FL 33101. [email protected] CGRP action in the eye.

Downloaded from iovs.arvojournals.org on 09/26/2021 1160 Rosenblatt et al. IOVS, April 2000, Vol. 41, No. 5

METHODS primer (RRES-4: 5Ј GAC AAT CTC AGG ACT CTG GTG CCG ACA TGG 3Ј) and adapter primer-1 (AP-1) were used for 5Ј Animal Care and Tissue Extraction RACE reactions and an upstream primer (RRES-3: 5Ј ACA AAG Animals were cared for in compliance with the ARVO State- TGG GAA GAA TAA ACA GAG TTC TGG 3Ј) and adap- ment for the Use Animals in Ophthalmic and Vision Research. ter primer AP-1 were used for 3Ј RACE reactions (Fig. 1). Male New Zealand albino rabbits were killed by intravenous RACE reactions were performed using the Expand PCR Kit injection of pentobarbital. Eyes were immediately enucleated (Boehringer-Mannheim, Indianapolis, IN). Initial 5Ј RACE prod- and transferred to 10 mM phosphate-buffered saline (PBS), pH ucts were reamplified using a downstream primer internal to 7.4. For RNA extraction, the anterior eye was removed, and the RRES-4 (RRES-6: 5Ј CCT ATG AAA CAT TAA AGT ACA TAT CAA ciliary body and iris were microdissected. For immunohisto- AAA 3Ј) and an adapter primer (AP-2) internal to AP-1. Initial 3Ј chemistry, eyes were cut tangentially and washed immediately RACE products were reamplified using a nested upstream with 10 mM PBS. primer (RRES-5: 5Ј GGC AAC AGA ACT TGA ATA CTA TCA CC 3Ј) and adapter primer AP-2. Primers RRES-5, RRES-6, and AP-2 RNA Extraction contained 12-nucleotide additions at their 5Ј ends, allowing final 5Ј and 3Ј RACE products to be cloned using the Clone- Total RNA was isolated from freshly dissected ciliary body and Amp System (GIBCO-BRL). RACE clones were sequenced using iris, which was homogenized in Tri-Reagent (Molecular Re- dideoxy double-stranded thermal cycle sequencing (Ampli- search, Cincinnati, OH) using a Polytron tissue homogenizer cycle Sequencing Kit; Perkin-Elmer). (Polytron; Brinkman, Westbury, NY). Total RNA was purified from the homogenate according to the manufacturer’s instruc- Expression of the Cloned Rabbit RCP tions. Poly(A)ϩ RNA was isolated from the total RNA using the PolyATtract mRNA System (Promega, Madison, WI). Primers RRES-10 (5Ј ATC TGC AGG CTG CTG GGG CCG AC 3Ј) and RRES-11 (5Ј GTG CAT CTG GCC CAG GCG TTT GAT GGA Isolation of a Rabbit RCP Amplimer by Reverse ATT CG 3Ј) were designed to flank the rabbit RCP open Transcription–Polymerase Chain Reaction reading frame (ORF) (Fig. 1A). Ciliary body and iris mRNA (150 ng) was reverse-transcribed using RRES-10. First-strand cDNA A portion of the rabbit RCP cDNA was isolated using degener- was subsequently amplified by PCR using RRES-9 and RRES-10. ate reverse transcription–polymerase chain reaction (RT- Primers were designed with restriction sites (RRES-10: PstI and PCR).19 RT of rabbit ciliary body and iris mRNA (150 ng) was RRES-11: EcoRI), which allowed ligation of the PCR amplimer primed using the primer DRES-1 (5Ј TCY TCN ACC ATN ARY into the low background vector, pZero (Invitrogen, Carlsbad, TGN ATY TCN AC 3Ј; IUPAC codes where Y ϭ CorT,R ϭ A CA), resulting in the plasmid pRCP.Zero. The ligation product or G, and N ϭ any), a degenerate oligonucleotide primer was transformed into Escherichia coli (One Shot Top10FЈ; designed to regions conserved between the guinea pig and Invitrogen), and the transformed E. coli was grown in liquid mouse RCP (Fig. 1), as previously described.20 The resulting culture in media containing 1 mM IPTG to lower non–insert- first-strand cDNA was used for PCR amplification with DRES-1 containing plasmids, as described by the manufacturer. Plas- and a second degenerate primer, DRES-3 (5Ј GTN TTY CAR mid DNA was isolated and tested using the oocyte–CFTR assay. YTN YTN ACN GAY YTN AA 3Ј) (Fig. 1). An aliquot of the initial PCR reaction was reamplified with a DRES-3 and a third Oocyte–CFTR Assay degenerate primer, DRES-2 (5Ј TGN ARY TTY TCN GCY TTN GTN ARY YYR TG 3Ј) (Fig. 1). Primers DRES-2 and DRES-3 Follicular cells were removed and Xenopus laevis oocytes were designed with 12-nucleotide additions (four copies of a were prepared for injection as described.21 Oocytes were coin- trinucleotide repeat containing a single deoxyuracil in each jected with 50 ng of CFTR cRNA and either 15 ng ciliary body repeat) at their 5Ј ends to allow rapid cloning using the and iris mRNA or 50 ng rabbit RCP cRNA. The CFTR cDNA CloneAmp System (GIBCO-BRL, Gaithersburg, MD), as de- (pACF23 obtained from J. Riordan, Mayo Clinic, Scottsdale, AZ) scribed.20 Positive clones were sequenced using dideoxy and rabbit RCP (pRCP.Zero) were transcribed in vitro using the double-stranded thermal cycle sequencing (Amplicycle Se- mMessage mMachine Kit (Ambion, Austin, TX) to synthesize quencing Kit; Perkin-Elmer, Foster City, CA). This sequence cRNA for injection. For antisense experiments, 50 ng of phos- was homologous to mouse and guinea pig RCP and was used to phorothioate oligonucleotide to the rabbit RCP (RCP-AS3: 5Ј design primers for subsequent rapid amplification of cDNA GCA GGC TCC TCT TCG TCC ATT GCC ACG T 3Ј) or rabbit ends (RACE) to acquire the complete RCP cDNA. calcitonin receptor (CTR-AS3: 5Ј TGA AGT AGA TTG CTC GAG TAA TAG CGT G 3Ј) were coinjected with RNA. After a 48- to Rapid Amplification of cDNA Ends 72-hour incubation to allow expression, oocytes were voltage RACE of the rabbit RCP was performed using the Marathon clamped at Ϫ50 mV and incubated with 100 nM CGRP RACE Kit (Clontech, Palo Alto, CA). Briefly, an adapter-ligated (Bachem Bioscience, King of Prussia, PA), as previously de- 18 cDNA library was created by reverse transcription of 1 ␮gof scribed. Maximum cAMP-induced CFTR response was deter- rabbit ciliary body and iris mRNA using an oligo-dT primer. mined by incubation with 20 ␮M forskolin (Sigma, St. Louis, Second-strand cDNA was synthesized using standard condi- MO). tions (Clontech). After blunting of the double-stranded cDNA with T4 DNA polymerase, adapters (Clontech) were ligated to Immunohistochemistry the cDNA ends using T4 DNA ligase. Freshly prepared eyes were fixed for 2 hours in 4% paraformal- The adapter-ligated cDNA was used as a template for 5Ј dehyde in 0.01 M PBS. Fixed eyes were passed through a and 3Ј RACE reactions using oligonucleotide primers designed graded series of ethanol washes (0%, 70%, 90%, 100%) and to the rabbit RCP fragment obtained by RT-PCR. A downstream paraffin embedded. Eight-micrometer sections were mounted

Downloaded from iovs.arvojournals.org on 09/26/2021 IOVS, April 2000, Vol. 41, No. 5 Rabbit Ocular RCP 1161

FIGURE 1. Rabbit RCP cDNA. (A) Rabbit RCP cDNA sequence. Top, nucleotide sequence for the rabbit ocular RCP cDNA; bottom, deduced amino acid sequence. Degenerate primers (DRES-1, DRES-2, and DRES-3) used for RT-PCR are indicated below the corresponding cDNA sequence (open arrows). Primers used for 5Ј RACE (RRES-4 and RRES-6) and 3Ј RACE (RRES-3 and RRES-5) as well as primers used to amplify the ORF (RRES-10 and RRES-11) are indicated below the corresponding cDNA sequence (ﬁlled arrows). (B) Protein alignment of rabbit RCP with guinea pig (GenBank accession no. U50188), mouse (GenBank accession no. AF028242), and human (GenBank accession no. AF073792) RCP homologues using the Clustal alignment matrix (DNASTAR). Residues match- ing the rabbit RCP are shaded black. Consensus phosphorylation sites for PKA (F), PKC (E), and casein kinase II () are indicated above the protein sequences.

on 3-triethoxysilylpropylamine (TESPA)-coated slides (Sigma, horse serum (Hyclone, Logan, UT) and 0.01% Triton X-100. The St. Louis, MO).22 Sections were incubated in 100% xylene for blocking solution was removed and sections were incubated in 10 minutes and rehydrated through a graded series of ethanol: primary (immune or pre-immune) antibody 1066 (1:1500 dilu- 100%, 90%, 70%, 0%. Sections were washed 10 minutes in 0.01 tion) in 0.01 M PBS/1% normal horse serum overnight at 4°C. M PBS and treated 15 minutes in 10% methanol/3% hydrogen Antibody 1066 was raised in chick (Aves Laboratories, Tigard, peroxide. After a 10-minute wash in 0.01 M PBS, sections were OR) against a synthetic peptide (EEQIEALLHTVT) conjugated blocked for 30 minutes in 0.01 M PBS containing 5% normal to keyhole limpet hemocyanin. Sections were washed in PBS,

Downloaded from iovs.arvojournals.org on 09/26/2021 1162 Rosenblatt et al. IOVS, April 2000, Vol. 41, No. 5

ﬁxed 20 minutes in 0.1% glutaraldehyde, washed again in PBS, endogenous CGRP receptor activity in CFTR-injected oocytes and incubated with a 1:400 dilution of biotin-conjugated don- without the addition of RCP. key anti-chicken secondary antibody (Jackson ImmunoRe- The oocyte–CFTR assay was used to test whether the search Laboratories, West Grove, PA) in 10 mM PBS. Immune ocular RCP could confer CGRP responsiveness. The rabbit RCP complexes were detected using avidin–biotin complex re- ORF was ampliﬁed from ciliary body and iris RNA using RT- agent coupled to horseradish peroxidase (ABC Reagent; Vector PCR, and the resulting amplimer ligated into the plasmid, Laboratories, Burlingame, CA) and developed with diamino- pZero (Invitrogen). Plasmid DNA was linearized and tran- benzidine as previously described.18 Sections were visualized scribed in vitro, the cRNA was coinjected with CFTR cRNA by light microscopy and images collected on a Pixera digital into oocytes, and RCP function (CGRP responsiveness) was camera (Pixera Corporation, Los Gatos, CA). tested using the oocyte–CFTR assay as described18,19 (Fig. 2A). Incubation of oocytes coinjected with rabbit RCP cRNA and CFTR cRNA with 100 nM CGRP resulted in CFTR currents of RESULTS 400 nA, indicating that the rabbit RCP could confer CGRP responsiveness to oocytes (Fig. 2B). Oocytes injected with Cloning of the Rabbit RCP from the CFTR cRNA alone did not responded CGRP (Fig. 2B) but did Ciliary Body/Iris respond to forskolin (data not shown).

We previously demonstrated that the CGRP receptor signaling RCP Antisense Experiments in the cochlea and uterus is dependent on RCP.18,19 To test our hypothesis that RCP is similarly required in the eye for CGRP- Our model predicts that RCP in the eye is required for signal mediated signal transduction, a cDNA encoding the rabbit transduction at CGRP receptors. To provide further evidence ocular RCP cDNA was isolated for testing in an oocyte-based for the requirement of RCP in CGRP signal transduction in the assay. A 206-bp fragment of the rabbit RCP ORF was amplified rabbit ciliary body and iris, we evaluated the ability of antisense by RT-PCR from ciliary body and iris mRNA using degenerate oligonucleotides made to the cloned rabbit ocular RCP to primers (Fig. 1A) designed to regions of protein homology inhibit the activity of rabbit ciliary body and iris mRNA in the between the guinea pig and mouse RCP. Sequencing indicated oocyte–CFTR assay. Messenger RNA isolated from ciliary body that the resulting amplimer shared a high degree of homology and iris conferred CGRP responsiveness in the oocyte–CFTR with the guinea pig and mouse RCP and coded for a fragment assay when coinjected with CFTR cRNA (Fig. 3A), similar to of the rabbit RCP homologue (Figs. 1A, 1B). the effect observed for RCP cRNA described above. These The sequence obtained from the RT-PCR amplimer was results indicate that the ciliary body and iris expresses mRNA used to design specific primers to the rabbit RCP to facilitate for a protein or proteins required for responsiveness to CGRP amplification of the 5Ј and 3Ј ends of the rabbit ocular RCP in oocytes. To determine whether RCP is responsible for the cDNA by RACE. Overlapping RACE amplimers were obtained, CGRP responsiveness conferred by ciliary body and iris mRNA and a 200-bp 5Ј RACE clone and an 1100-bp 3Ј RACE clone in the oocyte–CFTR assay, antisense oligonucleotides designed were isolated and sequenced. The rabbit RCP cDNA contains a against the RCP cDNA were coinjected with the CFTR cRNA 444-bp ORF coding for a 146-amino acid protein (Fig. 1A). The and ciliary body and iris mRNA to deplete RCP mRNA from the rabbit ocular RCP cDNA contains a consensus Kozak transla- pool of ciliary body and iris mRNA. Although oocytes injected tion initiation site and an in-frame stop codon 72 bp upstream with mRNA and CFTR cRNA responded to 100 nM CGRP (Fig. from the initiator methionine in the 5Ј untranslated region. The 3A), the addition of RCP antisense oligonucleotides abolished rabbit RCP does not contain any sites for N-linked glycosylation the CGRP-induced currents (Fig. 3B). Injection of unrelated but does contain several consensus phosphorylation sites that antisense oligonucleotides (to the calcitonin receptor) had no are conserved between guinea pig, mouse, and human (Fig. effect on CGRP-induced currents (Fig. 3C), indicating that the 1B). The rabbit RCP protein is homologous to previous cloned loss of signal associated with the RCP antisense oligonucleo- RCPs (Fig. 1B) sharing 81.5%, 87.1%, and 91.2% identity to the tides was specific for the CGRP receptor and not due to guinea pig, mouse, and human RCP, respectively. Similar to the nonspecific effects of phosphorothioate oligonucleotides. Data guinea pig, mouse, and human homologues, the rabbit RCP is for four such experiments (four different oocytes) are shown hydrophilic and contains no identifiable signal peptide se- in Figure 3D. These experiments demonstrated that the RCP quence. mRNA contained within the ciliary body and iris mRNA is a requisite factor for CGRP signal transduction in the oocyte. Functional Expression of the Rabbit Ocular RCP cDNA Immunohistochemical Localization or RCP in the Rabbit Eye The role of ocular RCP in CGRP-mediated signal transduction was determined using an oocyte-based assay originally de- Given the ability of CGRP to augment ocular blood flow, scribed by Uezono et al.23 In this assay the CFTR was used as increase intraocular pressure and cause breakdown of the a sensor for cAMP-mediated signal transduction. Elevated cAMP blood–aqueous barrier, CGRP most likely acts on the ciliary levels activate PKA, which in turn activates the CFTR, causing body and iris blood vessels that determine blood flow. CGRP this chloride channel to open and produces a measurable may influence aqueous humor dynamics as well by acting on inward current. We have previously demonstrated that RCP the ciliary epithelium, which is responsible for active produc- from cochlea and uterus confers responsiveness to CGRP using tion of aqueous humor and maintenance of the blood–aqueous this oocyte–CFTR assay.18,19 Xenopus are known to express barrier. If RCP were required for CGRP signal transduction in CGRP receptors,24,25 which were presumed to be inactive in the ciliary body and iris, then RCP would be expected to be in oocytes without the addition of RCP. We have not observed close association with CGRP binding sites. RCP protein expres-

Downloaded from iovs.arvojournals.org on 09/26/2021 IOVS, April 2000, Vol. 41, No. 5 Rabbit Ocular RCP 1163

FIGURE 2. The rabbit ocular RCP cDNA confers responsiveness to CGRP in the oocyte–CFTR assay. (A) Oocyte–CFTR assay. CGRP (CG) binds to receptor (R), and a G-Pro- tein (G) is activated, resulting in the stimulation of adenylate cyclase (AC). PKA is activated by cAMP formed by adenylate cyclase and phosphorylates the CFTR to produce an inward current when the oocyte is voltage-clamped at Ϫ50 mV. Po- tential sites of RCP interraction are indicated. (B) Expression of the rabbit ocular RCP. Oocytes were injected with either CFTR cRNA alone or coinjected with RCP cRNA and CFTR cRNA, voltage-clamped at Ϫ50 mV, and incubated with 100 nM CGRP. For all traces, the upward arrow indicates addition of CGRP and the downward arrow indicates washout. Inward current is indicated by upward deﬂection in the current trace.

sion in the eye was determined by immunohistochemistry, Immunohistochemistry with preimmune serum showed no using a polyclonal antibody 1066 raised in chicken against a staining, indicating that the immunoreactivity seen in the cili- synthetic peptide to RCP. Dense immunostaining for RCP was ary processes and iris is specific for RCP (Figs. 4C, 4F). observed in the ciliary processes along the ciliary epithelial Expression of RCP protein in the rabbit lens and retina also layer as well as blood vessels within the stroma of the ciliary was examined. RCP staining was observed in the lateral lens processes (Figs. 4A, 4B). Within the iris, the posterior epithe- fibers, but not in the most peripheral portions of the lens or in the lium is densely stained as well as blood vessels of varying lens epithelial layer (Figs. 5A, 5B). Immunoreactive regions of the caliber that are close to the epithelium (Figs. 4D, 4E). It ap- retina included the retinal ganglion cells and cells within the pears that the nonpigmented layer of the ciliary epithelium and amacrine layer as well as diffuse staining of the choroid, with the most superficial layer of the iris posterior epithelium stain more intense staining of the choroidal blood vessels (Figs. 5D, 5E). most densely for RCP expression (Fig. 4). The iris dilator Preimmune serum failed to stain either the lens or retina, con- muscle, anterior to the iris epithelium, also expresses RCP. firming the specificity of the RCP localization (Figs. 5C, 5F).

Downloaded from iovs.arvojournals.org on 09/26/2021 1164 Rosenblatt et al. IOVS, April 2000, Vol. 41, No. 5

DISCUSSION

CGRP has been well characterized as a mediator of the ocular neurogenic inflammatory response to injury,7–9 but the localization of CGRP receptors and the mechanisms by which CGRP causes this inflammation are not well understood. Studies on the role of CGRP in the physiology of the eye would be facilitated by the molecular identification of the CGRP receptor. However, the CGRP receptor has been difficult to identify, in part because the CGRP receptor appears to be one of the first examples of a G protein–coupled receptor that requires an accessory protein for function. This has complicated the isolation of the receptor cDNA and subsequent reconstitution of CGRP receptor activity by transfection in cell culture. We discovered an accessory protein for CGRP receptor activity as a result of expression cloning studies aimed at cloning the CGRP receptor using an oocyte-based assay. We cloned a cDNA that encoded a small hydrophilic protein named the CGRP-receptor component protein (RCP), which conferred CGRP responsiveness to X. laevis oocytes and is expressed in parallel with CGRP biological activity.18,19 RCP is not a receptor itself, but we hypothesize instead that RCP works in conjunction with a membrane-spanning receptor to constitute a functional CGRP receptor complex. To better understand the role of CGRP in ocular patho- physiology, we have cloned and characterized rabbit ocular RCP. Our hypothesis is that the ocular RCP also is necessary for signal transduction at CGRP receptors in the eye and is colocalized with CGRP receptor sites in the ciliary body and iris. The full-length rabbit RCP cDNA was cloned from the ciliary body and iris using a combination RT-PCR and RACE. Like the guinea pig and mouse RCP, the 148-amino acid rabbit RCP is hydrophilic, contains no identifiable signal peptide, and is not homologous to any other known signal transduction molecule. The cloned rabbit ocular RCP was sufficient to confer responsiveness to CGRP when expressed in the oocyte–CFTR assay, presumably working in conjunction with an endogenous CGRP receptor expressed in Xenopus,24,25 which is inactive without the addition of RCP. CGRP responsiveness also was conferred in the oocyte–CFTR assay by coinjection of eye mRNA with CFTR cRNA. This activity was abolished by coinjection of RCP antisense oligonucleotides, indicating that the CGRP receptor activity conferred by the eye mRNA required RCP for function. Antisense oligonucleotides to RCP did not affect the responsiveness of oocytes to forskolin, indicating that the antisense oligonucleotides did not have a general inhibitory effect on FIGURE 3. RCP is required for CGRP receptor function in ciliary body cAMP-mediated signal transduction. These data suggest that and iris. Messenger RNA from ciliary body and iris was coinjected with the rabbit RCP is a component of a CGRP signal transduction CFTR cRNA and either no antisense oligonucleotides, antisense oligo- pathway in the eye. nucleotides to RCP (anti-RCP) (A), or antisense oligonucleotides to a Previous studies have demonstrated CGRP binding sites in control mRNA (anti-control; calcitonin receptor) (B). Oocytes were voltage-clamped, and responses to 100 nM CGRP were determined. the eye using autoradiographic and membrane binding tech- 17 125 Results from multiple experiments with no antisense oligonucleotides niques. Heino et al. have reported I-CGRP binding in the and anti-RCP oligonucleotides are shown in (C). Data from separate ciliary processes, iris, ciliary muscle, chamber angle, limbal oocytes (n ϭ 4) were used to determine the current generated by conjuctiva, choroideae, and retina. Using immunohistochemis- CGRP as a percentage of the maximal current induced by 20 nM try, we found a correlation between RCP expression and these forskolin (% Max). Error bars, SEM. *Statistically significant (Student’s sites previously reported to bind CGRP. Our immunohisto- t-test) with P Ͻ 0.001. For all traces, the upward arrow indicates chemical results indicate that RCP is localized to numerous addition of CGRP and the downward arrow indicates washout. In- blood vessels in the iris as well as to the epithelial layers ward current is indicated by upward deflection in the current trace. coating the ciliary body, which are chiefly responsible for the active secretion of aqueous humor. The vascular localization of the RCP is consistent with the augmentation of blood flow by CGRP seen in the neurogenic inflammatory response8,26,27 and

Downloaded from iovs.arvojournals.org on 09/26/2021 IOVS, April 2000, Vol. 41, No. 5 Rabbit Ocular RCP 1165

FIGURE 4. RCP localization in the ciliary processes and iris. Immunohistochemistry was performed with antibody 1066 on 8-␮m rabbit eye sections. Cilliary process immunoreactivity was visualized at low (A) and high (B) power. Iris staining was visualized at low (D) and high (E) power. Control experiments for ciliary processes (C) and iris (F) were performed with preimmune 1066 serum at low power. Scale bars, (A, C, D, F) 100 ␮m; (B, E)30␮m.

correlates with the distribution of CGRP-immunoreactive may explain CGRP-induced elevation of intraocular pressure nerve ﬁbers around blood vessels in the iris and ciliary and breakdown of the blood–aqueous barrier.2,8,9,13,15,29–32 body.10,11,28 RCP staining of the ciliary epithelial layers also The nerve ﬁbers that release CGRP are in close proximity with

FIGURE 5. RCP localization in the lens and retina. Immunohistochemistry was performed with antibody 1066 on 8-␮m rabbit eye sections. Lens immunoreactivity was visualized at low (A) and high (B) power. Retina staining was visualized at low (D) and high (E) power. Control experiments for lens (C) and retina (F) were performed with pre-immune 1066 serum at low power. Scale bars, (A, C, D, F) 100 ␮m; (B, E)30␮m.

Downloaded from iovs.arvojournals.org on 09/26/2021 1166 Rosenblatt et al. IOVS, April 2000, Vol. 41, No. 5

the ciliary epithelial layers,10,11,28 and CGRP once released may irradiation of the iris. Q J Exp Physiol Cogn Med Sci. 1980;65:261– diffuse to these sites. It has been established that agents such 272. as CGRP that increase cAMP levels can increase the active 5. Jampol LM, Neufeld AH, Sears ML. Pathways for the response of the eye to injury. Invest Ophthalmol. 1975;14:184–189. secretion of aqueous humor by the ciliary epithelium, resulting 2,32 6. Brain SD, Williams TJ, Tippins JR, Morris HR, MacIntyre I. Calcito- in elevated intraocular pressure. The tight junctions be- nin gene-related peptide is a potent vasodilator. Nature. 1985;313: tween cells of the nonpigmented epithelium are the anatomic 54–56. sites of the blood–aqueous barrier, so that CGRP may be 7. Oksala O, Stjernschantz J. Effects of calcitonin gene-related pep- directly acting on these cells to cause disruption of the blood– tide in the eye. A study in rabbits and cats. Invest Ophthalmol Vis aqueous barrier. Sci. 1988;29:1006–1011. 8. Krootila K, Uusitalo H, Palkama A. Effect of neurogenic irritation RCP immunoreactivity also was seen in the lens and ret- and calcitonin gene-related peptide (CGRP) on ocular blood flow ina. Within the lens, RCP staining was evident in lateral lens in the rabbit. Curr Eye Res. 1988;7:695–703. fibers but not in lens fibers closest to the lens surface or in the 9. Unger WG, Terenghi G, Ghatei MA, et al. Calcitonin gene-related lens epithelium. CGRP is a growth factor in several sys- polypeptide as a mediator of the neurogenic ocular injury re- tems,33–35 and the patterned distribution of RCP in the lens sponse. J Ocul Pharmacol. 1985;1:189–199. suggests a developmentally regulated expression of RCP, al- 10. Terenghi G, Polak JM, Ghatei MA, et al. Distribution and origin of calcitonin gene-related peptide (CGRP) immunoreactivity in the though no role for CGRP in lens development has been stud- sensory innervation of the mammalian eye. J Comp Neurol. 1985; ied. This lens localization is also consistent with the isolation of 233:506–516. an expressed sequence tag for an RCP homologue (GenBank 11. Uusitalo H, Krootila K, Palkama A. Calcitonin gene-related peptide accession no. D26313) from a chick lens cDNA library. In the (CGRP) immunoreactive sensory nerves in the human and guinea retina, immunoreactive RCP was most apparent in the Mu¨ller pig uvea and cornea. Exp Eye Res. 1989;48:467–475. 12. Krootila K, Oksala O, von Dickhoff K, et al. Ocular irritative cells and in cells of the amacrine layer. CGRP has been re- response to YAG laser capsulotomy in rabbits: release of calcitonin ported to cause subtle changes in the rabbit electroretino- gene-related peptide and effects of methysergide. Curr Eye Res. gram,36 which is consistent with the localization of RCP in the 1992;11:307–314. retinal amacrine layer.37,38 RCP also is expressed by the vascu- 13. Krootila K, Uusitalo H, Palkama A. Effect of topical and intracam- lar choroidal layer of the retina, suggesting a possible role for eral methysergide on calcitonin gene-related peptide-induced irritative changes in the rabbit eye. J Ocul Pharmacol. 1992;8:121– CGRP in regulating retinal blood flow. 127. A candidate CGRP receptor named the calcitonin recep- 14. Osborne NN, Barnett NL. Calcitonin gene-related polypeptide stim- tor-like receptor (CRLR) was previously cloned by a PCR-based ulates c-AMP production in the iris/ciliary body complex (Letter). approach, but did not initially confer CGRP receptor activity Exp Eye Res. 1991;53:131–133. when transfected into cell culture.39,40 Aiyar et al.41 subse- 15. Krootila K. CGRP in relation to neurogenic inflammation and cAMP in the rabbit eye. Exp Eye Res. 1988;47:307–316. quently cloned a receptor they named CGRP1, which was 125 identical with CRLR and was active when stabley transfected 16. Malminiemi OI, Malminiemi KH. [ I]calcitonin gene-related peptide binding in membranes of the ciliary body-iris block. Curr Eye into HEK293 cells but not in COS cells, leading the authors and Res. 1992;11:1079–1085. others to suggest the requirement for a second protein for 17. Heino P, Oksala O, Luhtala J, Uusitalo H. Localization of calcitonin CRLR function.41,42 Although cotransfection of CRLR with RCP gene-related peptide binding sites in the eye of different species. into COS cells has not yielded functional CGRP receptors Curr Eye Res. 1995;14:783–790. (Rosenblatt MI, Dickerson IM, unpublished results, June 1997), 18. Luebke AE, Dahl GP, Roos BA, Dickerson IM. Identification of a protein that confers calcitonin gene-related peptide responsive- a second accessory protein recently has been cloned that does ness to oocytes by using a cystic fibrosis transmembrane conduc- work with CRLR. The receptor activity–modifying protein tance regulator assay. Proc Natl Acad Sci USA. 1996;93:3455– (RAMP1) was cloned using the oocyte–CFTR assay, acts as a 3460. chaperone for CRLR, and is required for CRLR activation.43 The 19. Naghashpour M, Rosenblatt MI, Dickerson IM, Dahl GP. Inhibitory inability of RCP to function in cotransfection experiments with effect of calcitonin gene-related peptide on myometrial contractil- ity is diminished at parturition. Endocrinology. 1997;138:4207– CRLR implies that RCP may work with a different receptor than 4214. CRLR. RCP is therefore a marker for this alternate CGRP recep- 20. Gomez-Saladin E, Wilson DL, Dickerson IM. Isolation and in situ tor complex present in the ciliary body and iris blood vessels localization of a cDNA encoding a Kex2-like prohormone conver- and ciliary epithelium and identifies a target for future studies tase in the nematode Caenorhabditis elegans. Cell Mol Neurobiol. on the pathophysiological action of CGRP in the eye. 1994;14:9–25. 21. Dahl G. Cell-Cell Interactions. A Practical Approach. In: Stevenson BR, Gallin WJ Paul DI, ed. New York: IRL; 1992:143–165. Acknowledgments 22. Luebke AE, Dickerson IM, Muller KJ. In situ hybridization reveals transient laminin B-chain expression by individual glial and muscle The authors thank Lake Paul for his technical assistance in the cloning cells in embryonic leech central nervous system. J Neurobiol. of the rabbit CGRP-RCP cDNA. 1995;27:1–14. 23. Uezono Y, Bradley J, Min C, et al. Receptors that couple to 2 References classes of G proteins increase cAMP and activate CFTR expressed in Xenopus oocytes. Receptors Channels. 1993;1:233–241. 1. Unger WG. Mediation of the ocular response to injury and 24. Adams JC, Mroz EA, Sewell WF. A possible neurotransmitter role irritation: peptides versus prostaglandins. Prog Clin Biol Res. for CGRP in a hair-cell sensory organ. Brain Res. 1987;419:347– 1989;312:293–328. 351. 2. Unger WG. Review: mediation of the ocular response to injury. J 25. Sewell WF, Starr PA. Effects of calcitonin gene-related peptide and Ocul Pharmacol. 1990;6:337–353. efferent nerve stimulation on afferent transmission in the lateral 3. Bill A. The Endre Balazs Lecture. Effects of some neuropeptides on line organ. J Neurophysiol. 1991;65:1158–1169. the uvea. Exp Eye Res. 1991;1990:53:3–11. 26. Almegard B, Andersson SE. Vascular effects of calcitonin gene- 4. Butler JM, Unger WG, Cole DF. Axon reflex in ocular injury: related peptide (CGRP) and cholecystokinin (CCK) in the monkey sensory mediation of the response of the rabbit eye to laser eye. J Ocul Pharmacol. 1993;9:77–84.

Downloaded from iovs.arvojournals.org on 09/26/2021 IOVS, April 2000, Vol. 41, No. 5 Rabbit Ocular RCP 1167

27. Oksala O. Effects of calcitonin gene-related peptide and substance calcitonin gene-related peptide. Am J Respir Cell Mol Biol. 1993; P on regional blood flow in the cat eye. Exp Eye Res. 1988;47: 8:592–596. 283–289. 36. Cao W, Drumheller A, Zaharia M, et al. Effects of calcitonin gene- 28. Terenghi G, Zhang SQ, Unger WG, Polak JM. Morphological related peptide on the rabbit electroretinogram. Neuropeptides. changes of sensory CGRP-immunoreactive and sympathetic nerves 1993;24:151–157. in peripheral tissues following chronic denervation. Histochemis- 37. Stindl G, Skofitsch G. Immunohistochemical localization of calci- try. 1986;86:89–95. tonin gene-related peptide in the rat eye. Response to capsaicin 29. Oksala O, Stjernschantz J, von Dickhoff K. Characterization of the pretreatment. Ann NY Acad Sci. 1992;657:464–465. mechanism of acute ocular irritation to YAG laser capsulotomy in 38. Klyama H, Katayama Y, Hillyard C. Occurrence of calcitonin gene- rabbits: effects of substance P antagonists, Met-enkephalin, tetra- related peptide in the chicken amacrine cells. Brain Res. 1985: caine and tetrodotoxin. Ophthalmic Res. 1989;21:360–368. 367–369. 30. Krootila K, Uusitalo H, Lehtosalo JI, Palkama A. Effect of topical 39. Chang CP, Pearse RVd, O’Connell S, Rosenfeld MG. Identification chemical irritation on the blood-aqueous barrier of the rat eye. of a seven transmembrane helix receptor for corticotropin-releas- Ophthalmic Res. 1986;18:248–252. ing factor and sauvagine in mammalian brain. Neuron. 1993;11: 31. Krootila K, Uusitalo H, Palkama A. Intraocular and cardiovascular 1187–1195. effects of calcitonin gene-related peptide (CGRP)-I and -II in the 40. Fluhmann B, Muff R, Hunziker W, Fischer JA, Born W. A human rabbit. Invest Ophthalmol Vis Sci. 1991;32:3084–3090. orphan calcitonin receptor-like structure. Biochem Biophys Res 32. Unger WG. Pharmacological and neural bases for eye irritation. Commun. 1995;206:341–347. Ann NY Acad Sci. 1992;641:176–186. 41. Aiyar N, Rand K, Elshourbagy NA, et al. A cDNA encoding the 33. Haegerstrand A, Dalsgaard CJ, Jonzon B, Larsson O, Nilsson J. calcitonin gene-related peptide type 1 receptor. J Biol Chem. Calcitonin gene-related peptide stimulates proliferation of human 1996;271:11325–11329. endothelial cells. Proc Natl Acad Sci USA. 1990;87:3299–3303. 42. Han ZQ, Coppock HA, Smith DM, et al. The interaction of CGRP 34. Cheng L, Khan M, Mudge AW. Calcitonin gene-related peptide and adrenomedullin with a receptor expressed in the rat pulmo- promotes Schwann cell proliferation. J Cell Biol. 1995;129:789– nary vascular endothelium. J Mol Endocrinol. 1997;18:267–272. 796. 43. McLatchie LM, Fraser NJ, Main MJ, et al. RAMPs regulate the 35. White SR, Hershenson MB, Sigrist KS, Zimmermann A, Solway J. transport and ligand specificity of the calcitonin- receptor-like Proliferation of guinea pig tracheal epithelial cells induced by receptor [In Process Citation]. Nature. 1998;393:333–339.

Downloaded from iovs.arvojournals.org on 09/26/2021