Cyclic nucletides vary by area in the retina and pigmented epithelium f the human and mnkey David A. Neivsme,* R. Thedre Fletcher, and Gerald J. Chader Cyclic GMP and cyclic AMP are present in lwer cncentratins in the central (macular) regin f the neural retina f the human and mnkey than in ther areas. This pattern apprximates the distributin f rd phtreceptr cells. Surprisingly, an inverse gradient f cyclic GMP cncentratin is bserved in the pigmented epithelium. Levels in the central regin are ver furfld higher than in cells in the periphery, ffering the first evidence f bichemical differences in this embrylgically unifrm cell type. Key wrds: retina, pigmented epithelium, cyclic GMP, cyclic AMP, macula \e retina and pigmented epithelium (PE) frm a functining unit, translating a phtic stimulus int a neural respnse. Cyclic nucletides appear t have imprtant functins in this prcess. Cyclic guansine mnphsphate (cyclic GMP) in particular is cmpartmentalized in phtreceptr uter segments, with up t a 10-fld higher cncentratin in dark- than in light-adapted phtreceptrs. h 2 Cyclic AMP levels in phtreceptrs are lwer and are nt respnsive t light, 1-2 althugh dark adaptatin 2 and catechlamines 3 increase cyclic AMP in the inner layers f the retina. Nthing is knwn, hwever, cncerning cyclic nucletides in the retinal PE, the single layer f cells appsed t the retinal phtreceptr uter segments. In the present cmmunicatin, we reprt n differences in the cncentratin f cyclic GMP Frm the *Clinical Branch and the Labratry f Visin Research, Natinal Eye Institute, Natinal Institutes f Health, Bethesda, Md. Submitted fr publicatin Jan. 24, 1979. Reprint requests: Dr. Gerald J. Chader, Bldg. 6A, Rm B1A-07, Natinal Eye Institute, Natinal Institutes f Health, Bethesda, Md. 20205. and cyclic AMP in different areas f the light-adapted human retina and PE btained at autpsy and in the mnkey. Materials and methds Human eyes were btained at autpsy and dissected within 20 hr fllwing death. Dnr ages ranged frm 7 t 72 years; causes f death varied. Nne f the "nrmal" dnrs had a past histry f cular disease, and the eyes were anatmically nrmal n grss examinatin with the dissecting steremicrscpe. Eyes were als btained frm three patients, each with dcumented pathlgy f the retinchridal cmplex. Patient 1 was a 51-year-ld black man with a 25-year histry f prgressive visual lss due t retinitis pigmentsa. The eye used in this study had bare light perceptin with n prjectin. Patient 2 was a 63-yearld man with mderate prliferative diabetic retinpathy in ne eye which had been treated 2 years befre death with panretinal phtcagulatin. The ther eye had very early prliferative and macular changes and was untreated. Patient 3 was a 67-year-ld wman with an 8-year histry f mildly reduced visin (20/40 t 20/50) presumably due t macular changes, including drusen. There had been evidence f slw prgressin in the year prir t death. Retinas were dissected int three areas: (1) a 6 864
Vlume 19 Number 8 Reginal variatin f retinal cyclic nucletides 865 Table I. Variatin in cyclic nucletide by area in neural retina and pigmented epithelium A. Retina cyclic GMP: Midzne B. Retina cyclic AMP: Midzne C. PE cyclic GMP: Midzne D. PE cyclic AMP M idzne Cyclic iwcletide (Mean : t S.E.M.) 3.2 6.4 10.2 4.8 10.1 13.2 8.5 3.2 2.0 ± 0.7 ± 1.1 ± 1.8 ± 0.6 ± 1.2 ± 1.8 ± 2.1 ± 1.0 ± 0.6 2.6 2.2 ± 0.5 ± 0.4 3.3 ± 0.6 p value (vs. macula) 0.023 0.018 0.002 0.0005 0.024 0.031 Nt significant Nt significant Values given are frm 10 individual tissue samples. Duplicate radiimmundeterminatins at each f three different dilutins were averaged t get the final value fr each f the tissue samples. Each value given therefre is based n results frm up t 60 radiimmunassays. Statistical significance (p value) was determined with a "paired t statistical" cmputer prgram. Nrmal eyes were btained at autpsy (2 t 20 hr after death) frm the fllwing dnrs, listed by age (years), sex, and cause f death: (1) 65, F, cardiac; (2) 39, M, brain tumr; (3) 7, M, rhabdsarcma; (4) 48, M, melanma; (5) 52, M, at-cell carcinma; (6) 68, F, heart failure; (7) 47, F, melanma; (8) 29, M, melanma; (9) 72, F, heart failure; (10) 68, M, mycardial infarctin. mm disc centered n the fvea (macula), (2) a 10 mm annul us f equatrial retina exclusive f the ptic disc regin, and (3) apprximately 10 mm f peripheral retina. PE was gently scraped frm under each f these areas. Tissue samples were immediately placed int 0.5 ml f a 10% perchlric acid slutin. Rhesus mnkeys were maintained in the NIH animal clny and sacrificed by pentbarbital injectin. T btain a "zer" time value, the test animal was deeply anesthetized, the eye was pened in situ, and tissue samples were dissected and placed immediately int 0.5 ml f a 10% perchlric acid slutin (which stps all enzymatic reactins). The time lapse between pening the eye and placing the first sample (macula) int the acid slutin was 30 t 60 sec. Equatrial and peripheral retinal samples then were taken as well as samples frm the apprpriate underlying PE areas. This prcess required an additinal 2 t 3 min. Fr time perids ther than the zer time pint, mnkeys were sacrificed, eyes were remved, and samples were later dissected and "5 z cc \- z LU f \ LU Q ^11 I Dz (J _i O > 30 25 20 15 10 5 30 20 15 10 5 0 1.5 4 6 24 0 1.5 4 6 TIME (hr.) Fig. 1. Variatin in cyclic nucletide by area in retina and PE epithelium f rhesus mnkey as a functin f time after death. A, Retinal cyclic GMP. B, Retinal cyclic AMP. C, Pigmented epithelial cyclic GMP. D, Pigmented epithelial cyclic AMP. Areas f retina: 8 macula; ffl midzne; periphery. Time values refer t time after death (hurs); zer time values were btained with tissues remved frm deeply anesthesized animals. Cyclic nucletide values are averages f duplicate radiimmundeterminatins at each f three different dilutins. placed in perchlric acid slutin at stated intervals. Cyclic nucletides were subsequently separated, purified, 4 and succinylated. 5 Radiimmunassay was perfrmed by the prcedure f Steiner et al. 6 with minr mdificatin. 7 Prtein was determined by the methd f Lwry et al. 8 with bvine alubmin as standard. Results are expressed as picmles f cyclic nucletide per milligram f prtein. Results Cyclic nucletide levels varied by area in the human retina with lwest cncentratin bserved in the central (macular) regin f 24
866 Newsme, Fletcher, and Chader Invest. Ophthatml. Vis. Set. August 1980 Fig. 2. A, Light micrgraph {hematxlyn and esin) f a typical 18 hr pstmrtem specimen f neural retina demnstrating retentin f uter segments (asterisk). (x420.) B, Transmissin electrn micrgraph f PE frm the same area as in (A), demnstrating the clean separatin f uter segments frm the PE. (x35,000.)
Vlume 19 Number 8 Reginal variatin f retinal cyclic nucletides 867 the retina (Table I, A and B). This was fund t be true fr bth cyclic AMP and cyclic GMP, with a twfld t threefld increase in cncentratin as ne advanced thrugh the midzne t the periphery. The differences in cyclic nucletide cntent between macula and ther retinal areas were highly statistically significant. Parameters such as age, sex, cause f death, drug treatment, etc., did nt influence cyclic nucletide levels in the cases we examined. Data frm retinal samples btained later than 20 hr after death were nt included in the data analysis as discussed belw. Surprisingly, the cyclic GMP level in the PE fllwed an inverse pattern cmpared t its distributin in the retina (Table I, C), i.e., twfld t furfld higher in macula than in mre peripheral regins. N such distinct pattern was bserved with cyclic AMP in the different areas f the PE (Table I, D). A similar reginal distributin f cyclic nucletides was bserved in the mnkey retina and PE (Fig. 1). Mrever, the influence f the time lag between death and actual retinal dissectin n cyclic nucletide levels in retina and PE was als determined in this study. Fr this purpse, mnkeys were anesthesized, and eyes were dissected immediately (zer time) r enucleated and dissected at 1.5, 4, 6, and 24 hr after death. This study demnstrated that (1) the cncentratins f cyclic nucletides in the varius retinal areas (Fig. I, A t D) were, in fact, similar t thse in the human; (2) cyclic GMP levels in retina (Fig. 1, A) were relatively stable thrugh abut 24 hr but declined thereafter as autlysis f the retina prgressed; (3) cyclic AMP levels in retina (Fig. 1, B) were surprisingly stable in macula but suffered an initial drp (zer time t 1.5 hr) in ther retinal areas; (4) in PE, cncentratins f bth cyclic GMP (Fig. 1, C) and cyclic AMP (Fig. 1, D) drpped threefld t furfld within 4 hr, slwly declining thereafter; and (5) the same reginal pattern f distributin was evident as in the humanin retina, bth cyclic GMP and cyclic AMP lwest in macula; in PE, cyclic GMP highest in the macular regin; n pattern bserved with cyclic AMP. Time pints at 30 and 48 hr were als studied but are nt presented in Table I because they shw the pattern f prgressin already described. One pssible explanatin fr ur results wuld be that the retinal uter segments (rds and/r cnes) were brken ff during dissectin and were retained by the underlying PE cells. At the macula, this culd lead t artifactually lw cyclic nucletide levels and cncmitantly high levels in the underlying PE. Hwever, careful examinatin by bth light and electrn micrscpy (Fig. 2) revealed that the bulk f the uter segments in the macular as well as peripheral areas f the retina was retained by the retina and was nt artifactually assciated with the underlying PE. The sample shwn is a typical 18 hr pstmrtem human specimen and demnstrates that even thugh prgressive autlytic changes were evident, the bulk f the uter segments remained with the retina (Fig. 2, A). Apical pigmented epithelial prcesses were infrequently seen by this time (Fig. 2, B). In preliminary studies, we have fund that cyclic nucletide cncentratins actually d appear t be altered in the human retina in sme diseased states. Fr example, in the single case f advanced retinitis pigmentsa we have been able t examine (n = 1), cyclic GMP values were decreased fivefld t 10-fld (0.5, 0.6, and 2.0 pml/mg in macula, midzne, and periphery, respectively), whereas cyclic AMP values were virtually nrmal (6.7, 13.8, and 11.2 pml/mg in the three areas). In diabetes mellitus (ne retinal sample frm each eye f patient, n = 2), retinal cyclic AMP levels were greatly elevated (13.0, 39.3, and 33.2 pml/mg in the three areas), whereas cyclic GMP values were virtually nrmal (3.4, 6.7, and 8.8 pml/mg in the three areas). In an eye exhibiting many drusen, levels f bth cyclic nucletides were increased, and a reversed pattern f distributin bserved fr cyclic AMP, i.e., highest in macula (cyclic AMP 34.2, 23.9, and 9.7 pml/mg and cyclic GMP 12.6, 17.6, and 9.4 in the three areas).
868 Newsme, Fletcher, and Chader Invest. Ophthalml. Vis. Sci. August 1980 Discussin Only a few animals, including man, pssess a specialized retinal area, the macula, which subserves high-reslutin visual acuity. Althugh mrphlgical differences between macular and nnmacular areas f the retina are well knwn (e.g., rd/cne distributin 9 ), this reprt is the first t demnstrate bichemical differences between these areas. Such differences might be basic in understanding clinical cnditins which demnstrate specific reginal patterns. In cntrast t the cellular hetergeneity f the retina, the PE has been thught t be a relatively unifrm tissue because it is nly a single cell layer thick and all cells are thught t have similar functins. The finding f a high cyclic GMP cntent in PE cells in the macular regin was quite surprising, especially in light f the relatively lw cyclic GMP cntent f the verlying neural retina. Fr the first time, therefre, significant differences in cell metablism in different areas f the PE are indicated. We may ultimately be able t crrelate these bichemical changes with differences in PE cell mrphlgy. 10 ' 1J Of sme imprtance, besides the actual empirical values fr the tw nucletides, may be the cyclic GMP/cyclic AMP rati. 12 This is relatively unifrm in the varius areas f the retina (macula 0.7; midzne 0.6; periphery 0.8), but in the PE, the rati drps frm 3.3 in the macular zne t 1.5 in the midzne and nly 0.6 in the periphery. The unifrm rati in retina is als interesting because it indicates that there may be n preferential cncentratin f either nucletide in the cne-rich macula vs. the mre peripheral rd-rich areas. Frm ur present and previus 1 ' 7 results and mre detailed studies f Orr et al., 2 and Ferrendelli and Chen, 13 and Farber and Llley, 14 ' 15 it appears that tw distinct pls f cyclic nucletides can be discerned in retina. First, a relatively labile pl (visual cycle pl) appears t be present primarily in the phtreceptr regin and prbably is invlved in the rapid reactins f the visual cycle and ensuing neurtransmitter events. The present human and primate study was bviusly nt designed t investigate this pl, since studies n small rdent eyes 16 have shwn that there are very rapid changes in retinal cyclic nucletides after death. The secnd pl (basal pl) is ne that appears t be relatively stable and may be invlved in the mre general functins f the retina as in ther parts f the central nervus system. Perturbatins f this secnd pl in particular culd be assciated with retinal pathlgy. Our preliminary results n several human retinal cnditins indicate that study f cyclic nucletide perturbatins may yield valuable infrmatin abut the etilgy f sme f these diseases and pssibly frm a ratinal basis fr disease treatment. We thank Dr. Arnld Gldman fr aid in the statistical evaluatin and Dr. Merlyn Rdrigues fr perfrming the electrn micrscpy. REFERENCES 1. Fletcher RT and Chader GJ: Cyclic GMP: cntrl f cncentratin by light in retinal phtreceptrs. Bichem Biphys Res Cmmun 70:1297, 1976. 2. Orr H, Lwry O, Chen A, and Ferrendelli J: Distributin f3':5'-cyclic AMP and 3':5'-cyclic GMP in rabbit retina in viv: selective effects f dark and light adaptatin and ischemia. Prc Natl Acad Sci USA 73:4442, 1976. 3. Brwn J and Makman M: Stimulatin by dpamine f adenylate cyclase in retinal hmgenates and f adensine-3':5'-cyclic mnphsphate frmatin in intact retina. Prc Natl Acad Sci USA 69:539, 1972. 4. Krishnan N and Krishna G: A simple and sensitive assay fr guanylate cyclase. Anal Bichem 70:18, 1976. 5. Frandsen E and Krishna G: A simple ultrasensitive methd fr the assay f cyclic AMP and cyclic GMP in tissues. Life Sci 18:529, 1976. 6. Steiner A, Parker C, and Kipnis D: Radiimmunassay fr cyclic nucletides. I. Preparatin f antibdies and idinated cyclic nucletides. J Bil Chem 217:1106, 1972. 7. Krishna G, Krishnan N, Fletcher R, and Chader G: Effects f light n cyclic GMP metablism in retinal phtreceptrs. J Neurchem 27:717, 1976. 8. Lwry O, Rsebrugh N, Farr A, and Randall R: Prtein measurement with the Flin phenl reagent. J Bil Chem 193:265, 1951. 9. Osterberg G: Tpgraphy f the layer f rds and cnes in the human retina. Acta Ophthalml, Supp. 5-6, 1935. 10. Ts' M and Friedman E: The retinal pigment epithelium. I. Cmparative histlgy. Arch Ophthalml 78:641, 1967. 11. Sakuragawa M and Kuwabara T: The pigment epithelium f the mnkey: tpgraphic study by scan-
Vlume 19 Number 8 Reginal variatin f retinal cyclic nucletides 869 ning and transmissin electrn micrscpy. Arch Ophthalml 94:285, 1976. 12. Gldberg N and Haddx M: Cyclic nucletide metablism and invlvement in bilgical regulatin. Annu Rev Bichem 46:823, 1977. 13. Ferrendelli J and Chen A: The effects f light and dark adaptatin n the levels f cyclic nucletides in retinas f mice heterzygus fr a gene fr phtreceptr dystrphy. Bichem Biphys Res Cmmun 73:421, 1976. 14. Farber D and Llley R: Cyclic guansine mnphsphate: elevatin in degenerating phtreceptr cells f the C3H muse retina. Science 186:449, 1974. 15. Farber D and Llley R: Light-induced reductin in cyclic GMP f retinal phtreceptr cells in viv: abnrmalities in the degenerative diseases f RCS rats and rd mice. J Neurchem 28:1089, 1977. 16. Mitzel D, Hall I, DeVries G, Chen A, and Ferrendelli J: Cmparisn f cyclic nucletide and energy metablism f intact muse retina in situ and in vitr. Exp Eye Res 27:27, 1978. Infrmatin fr authrs Mst f the prvisins f the Cpyright Act f 1976 became effective n January 1, 1978. Therefre, all manuscripts must be accmpanied by the fllwing written statement, signed by ne authr: "The undersigned authr transfers all cpyright wnership f the manuscript (title f article) t The Assciatin fr Research in Visin and Ophthalmlgy, Inc., in the event the wrk is published. The undersigned authr warrants that the article is riginal, is nt under cnsideratin by anther jurnal, and has nt been previusly published. I sign fr and accept respnsibility fr releasing this material n behalf f any and all c-authrs." Authrs will be cnsulted, when pssible, regarding republicatin f their material.