Novel Mutation in the TIMP3 Gene Causes Sorsby Fundus Dystrophy

OPHTHALMIC MOLECULAR GENETICS Novel Mutation in the TIMP3 Gene Causes Sorsby Fundus Dystrophy

Samuel G. Jacobson, MD, PhD; Artur V. Cideciyan, PhD; Jean Bennett, MD, PhD; Ronald M. Kingsley, MD; Val C. Sheffield, MD, PhD; Edwin M. Stone, MD, PhD

Objective: To determine the molecular basis of a retinopa- Conclusions: The Y172C mutation in the TIMP3 gene thy previously described as dominant macular subretinal is another cause of Sorsby fundus dystrophy. The ex- neovascularization with peripheral retinal degeneration. pression of this form of the disease, as in other C- terminal TIMP3 mutations, is speculated to be second- Methods: The TIMP3 gene was analyzed in family mem- ary to mutant TIMP-3, causing a decreased turnover of bers, and 4 mutation-positive patients were studied us- the extracellular matrix. ing psychophysics and electroretinography. Clinical Relevance: The molecular clarification of in- Results: Cosegregating with disease in the family was a herited retinal degeneration involving abnormal extra- single base pair change in the TIMP3 gene, altering a con- cellular matrix turnover in and around Bruch’s mem- served tyrosine to cysteine at amino acid position 172 brane should provide clues to the pathogenesis of not only (Y172C). There was psychophysical and electroretino- these particular diseases but also forms of age-related graphic evidence of rod dysfunction greater than cone macular degeneration. dysfunction. Dark adaptometry showed abnormalities with regional retinal variation in degree. Arch Ophthalmol. 2002;120:376-379

MONG MANY monogenic over of mutant TIMP-3 are aided by the retinal degenerative dis- knowledge of disease-causing TIMP3 mu- eases, only Sorsby fundus tations. Apart from the founder mutation in dystrophy (SFD) com- codon 181 from the British Isles, only a few monly manifests as a hem- other disease-causing TIMP3 gene muta- orrhagic maculopathy secondary to cho- tions have been identified.6,8 A 1 roidal neovascularization (CNV). This More than a decade ago, a family was feature of SFD is interesting because of its described as having “dominant macular clinical similarity to some forms of age- subretinal neovascularization with periph- related macular degeneration (AMD). The eral retinal degeneration.”9 Among the di- hypothesis that there may be both molecu- agnostic possibilities considered was SFD, lar and clinical similarity to AMD became but the age at disease onset was thought testable when it was determined that SFD to be too early compared with other re- was caused by mutations in the gene en- ports in the literature at that time.10,11 We coding tissue inhibitor of metalloprotein- reevaluated the family and report that this From the Department of ases-3 (TIMP-3),2 an extracellular matrix retinopathy is a severe form of SFD caused Ophthalmology, Scheie Eye by a novel mutation in the TIMP3 gene. Institute, University of (ECM) protein. However, candidate gene Pennsylvania, Philadelphia screening with the TIMP3 gene in AMD (Drs Jacobson, Cideciyan, populations did not reveal a simple ge- RESULTS and Bennett); Department of netic relationship between the rare inher- Ophthalmology, University ited disease and the common age-related The PCR product from exon 5 of the of Oklahoma Health Sciences cause of CNV.3,4 TIMP3 gene amplified from the proband Center, Dean A. McGee Eye The complex pathways leading from (patient III-5) migrated in an aberrant elec- Institute, Oklahoma City the TIMP3 mutation to ECM disturbance trophoretic pattern compared with nor- (Dr Kingsley); and and CNV are still being studied.5,6 Results mal control DNA samples run in parallel Departments of Ophthalmology (Dr Stone) and Pediatrics and are beginning to reveal a pathogenetic se- during SSCP analysis. Direct sequencing the Howard Hughes Medical quence that may explain abnormal sub- of the PCR product revealed a single base Institute (Dr Sheffield), RPE (retinal pigment epithelium) deposi- pair change, altering a conserved tyro- University of Iowa College tion in a number of retinopathies.5-7 Studies sine to cysteine at amino acid position 172. of Medicine, Iowa City. to understand the effects on ECM turn- This Y172C TIMP3 gene mutation coseg-

(REPRINTED) ARCH OPHTHALMOL / VOL 120, MAR 2002 WWW.ARCHOPHTHALMOL.COM 376

II Informed consent was obtained from the study par- 1 ticipants for all procedures. Mutation screening of the III TIMP3 gene was performed. Venous blood samples were 123 4 5 6 7 collected from 13 family members (generations III and IV in Figure 1A), and DNA was extracted. Tech- IV niques for single-strand polymorphism (SSCP) analy- 1-7 8 9 10 11 12 13 14 15 16 17 18 sis and direct sequencing of polymerase chain reaction (PCR) products have been described elsewhere.12 B A subset of 4 patients (III-4, 5, and 6 and IV-14), III-6 IV-15 IV-16 III-5 IV-12 IV-13 III-3 IV-14 III-2 III-4 IV-11 IV-10 all positive for a TIMP3 gene mutation, were assessed with routine ophthalmic examinations and specialized tests of visual function. Static threshold perimetry in dark-adapted (500- and 650-nm stimuli) and light- adapted (600-nm stimulus on a 10 candela/m2 white background) states was performed using a modified au- tomatedperimeterandanalyzedforrodandconethresh- old elevation. Dark adaptometry was tested with 500- and 650-nm stimuli after a retinal exposure of 7.8 log +–+++–++–+–+ scotopic-troland-seconds,estimatedtobleachabout99% of the rhodopsin present, and recovery was measured C untilprebleachbaselinedark-adapted(Ͼ3-hour)thresh- Patient III-5 oldswereattained.Detailsoftheseprocedureshavebeen published previously.13-15 For the patients with central scotomas, bleaching and testing were performed using infrared visualization of the fundus with a modified fundus photoperimeter. Electroretinogram (ERG) photoresponses were evoked in the dark-adapted state with high-energy blue (2.3-4.6 log scotopic-troland- seconds) and red (1.4-3.6 log photopic-troland- seconds) stimuli and in the light-adapted (3.2 log troland white background) state with red (2.2-4.1 log photopic-troland-seconds) stimuli. A model of photo- Figure 1. A, Pedigree of the family in this study. Filled symbols indicate transduction activation consisting of the sum of rod and affected individuals; shaded symbols, individuals molecularly affected but cone components was used to quantify the leading edges clinically unaffected to date; open symbols, unaffected individuals; and of dark-adapted waveforms; a model of cone photo- slashed symbols, deceased individuals. Patient III-5 is the proband. transduction was used to quantify the leading edges B, Single-strand polymorphism analysis of TIMP3 exon 5 polymerase chain reaction products in available family members. Each lane of this silver of light-adapted waveforms. Details of recording and nitrate–stained 6% polyacrylamide gel contains the amplified DNA from the 16,17 analysis methods have been published previously. individual whose pedigree number is above it. Individuals with the mutation (+) have 2 extra bands (arrows), indicating a heterozygous exon 5 mutation. Four individuals (−) lack the extra bands, representing a homozygous normal sequence in exon 5. C, Fundus photograph of patient III-5 (at age 45 years) showing central retinal scarring and degeneration. regates with the disease in generation III, which has the only living clinically affected members at this time; there were both mutation-positive and mutation-negative mem- Rod- and cone-mediated thresholds throughout the field bers in generation IV (Figure 1B). of vision are shown for a 39-year-old patient (III-6) with a The fundus photograph of patient III-5 at age 45 years large central scotoma. Throughout the peripheral visual field (Figure 1C) is representative of the central retinal scarring were significant rod threshold elevations (mean, 2.0 log from hemorrhagic macular degeneration found in affected units); beyond the central scotoma, most cone thresholds members in the fourth and fifth decades of life. Fundus ap- were within normal limits. Patients III-3 and III-5 also pearance and fluorescein angiography in patients III-3, 4, showed more rod than cone threshold elevations in the pe- 6, and 7 documenting RPE abnormalities in the macula and ripheral visual field beyond their central scotomas. An midperipheral retina were previously reported.9 Five of the asymptomatic 20-year-old heterozygote (patient IV-14) had 6 affected patients had an acute loss of central vision in one normal results on eye examination and visual function tests. eye between ages 26 and 29 years, with the second eye los- Rod and cone ERG photoresponses suggest that the ing vision within 2 years of the first event. The affected abnormalities on psychophysical testing have a photore- half brother (patient III-1) lost central vision between ages ceptor basis (Figure 2B). Patient IV-14 had normal rod and 35 and 36 years. In generation IV, 4 younger asymptom- cone responses, whereas 3 older patients (patients III-3, 5, atic family members (ages 7-21 years) carry the mutation and 6) had rod maximum amplitudes reduced to approxi- and are at risk for developing the disease (Figure 1A). mately half of the mean normal value and normal or slightly Visual function studies in a subset of heterozygotes reduced cone maximum amplitudes. Rod photoresponse for the Y172C TIMP3 mutation are shown in Figure 2. sensitivities were within normal limits in 2 patients (III-6

(REPRINTED) ARCH OPHTHALMOL / VOL 120, MAR 2002 WWW.ARCHOPHTHALMOL.COM 377

24 Eccentricity, degrees Eccentricity,

I 3 log units T 72 48 24 0 24 48 72 48 24 0 24 48 N Eccentricity, degrees B

ROD CONE III-3 0 0 III-5 III-6 IV-14

–100 –20 V µ –200 –40 Amplitude,

–300 –60

–400 –80

2468 2468 D C Time, ms Time, ms 0 0 III-5 III-3 III-6 IV-14 –1 –1

–2 –2

–3 –3

Threshold, log units 50N –4 –4 33N ~30T 33T

–5 –5

01020 30 40 50 01020 30 40 50 Time, min Time, min

Figure 2. Functional phenotype of heterozygotes with the Y172C TIMP3 mutation. A, Static threshold dark- and light-adapted perimetry in a patient representing a late stage of the disease. Elevation of rod and cone thresholds across the visual field are shown as gray-scale maps, with 16 levels representing 0 to 3 log units of threshold elevation. Black squares indicate no detection of stimuli. T, N, I, and S indicate temporal, nasal, inferior, and superior visual fields, respectively. B, Rod- and cone-isolated photoresponses (symbols connected by thin black lines) fitted with a model of phototransduction activation (thicker gray lines). Photoresponses evoked with blue 4.6 log scotopic-troland-second (ROD) and red 4.1 log photopic-troland-second (CONE) flashes are shown for each patient. Arrows on the y-axis denote the lower limit (mean−2 SD) of normal maximum amplitude. C and D, Dark adaptation with 500-nm stimulus after a full bleach exposure in patients (symbols and thin lines) compared with the normal range (gray, thicker lines). Prebleach thresholds are shown preceding time zero; numbers represent test location in degrees in the T and N visual fields.

(REPRINTED) ARCH OPHTHALMOL / VOL 120, MAR 2002 WWW.ARCHOPHTHALMOL.COM 378

©2002 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021 and IV-14), whereas 2 other heterozygotes (patients III-3 Submitted for publication July 17, 2001; final revision re- and 5) had sensitivity losses of about 0.4 log units. Cone ceived October 25, 2001; accepted November 16, 2001. photoresponse sensitivities were within normal limits for This study was supported in part by grants EY-05627, all 4 patients. EY-13203, and EY-12156 from the National Institutes of Dark adaptation was tested at several different reti- Health, Bethesda, Md; the Macular Disease Foundation, nal loci in the 4 heterozygotes; representative functions Virginia Beach, Va; the Macula Vision Research Founda- obtained at Ն30° eccentric to the anatomical fovea are tion, West Conshohocken, Pa; Foundation Fighting Blindness shown (Figure 2C and D). In patients IV-14 and III-6, Inc, Owings Mills, Md; and the F. M. Kirby Foundation, Mor- dark-adaptation functions at 3 different loci were nor- ristown, NJ. mal; in patient III-5, dark adaptation was borderline nor- Dr Jacobson is a Research to Prevent Blindness (New mal at 4 loci; and in patient III-3, it ranged from normal York, NY) Senior Scientific Awardee, Drs Bennett and Ci- to abnormal at 4 loci tested. The abnormality, when pres- deciyan are Research to Prevent Blindness Special Schol- ent, consisted mainly of threshold elevation (Figure 2D). ars, and Dr Sheffield is an associate investigator of the Howard Hughes Medical Institute, Iowa City, Iowa. COMMENT We are grateful to M. Benegas, D. Hanna, J. Emmons, L. Gardner, Y. Huang, D. Marks, J. Huang, and C. Taylor Following the first association between a TIMP3 mutation for help with the conduct of this study. and SFD,2 several reports of other TIMP3 mutations in SFD Corresponding author and reprints: Samuel G. Jacob- have appeared.6,8 Almost all mutations documented in SFD son, MD, PhD, Scheie Eye Institute, 51 N 39th St, Phila- to date, including Y172C, would be expected to alter resi- delphia, PA 19104 (e-mail: [email protected]). dues in the C-terminal domain of the molecule. There is still no biological explanation of how mutation in an in- REFERENCES hibitor of matrix metalloproteinase may be leading to de- 1. Zhang K, Yeon H, Han M, Donoso LA. Molecular genetics of macular dystro- creased ECM turnover and extreme thickening of Bruch’s phies. Br J Ophthalmol. 1996;80:1018-1022. membrane, the histopathological hallmark of SFD.18,19 Re- 2. Weber BHF, Vogt G, Pruett RC, Sto¨hr H, Felbor U. Mutations in the tissue inhibitor of metalloproteinases-3 (TIMP3) in patients with Sorsby’s fundus dystro- cent in vitro studies of 4 SFD mutations suggest that the phy. Nat Genet. 1994;8:352-356. pathogenesis is probably not caused by haploinsuffi- 3. Felbor U, Doepner D, Schneider U, Zrenner E, Weber BHF. Evaluation of the gene ciency but by persistent increased function. Evidence has encoding the tissue inhibitor of metalloproteinases-3 in various maculopathies. Invest Ophthalmol Vis Sci. 1997;38:1054-1059. appeared that mutant TIMP-3 may be accumulating as a 4. De La Paz MA, Pericak-Vance MA, Lennon F, Haines JL, Seddon JM. Exclusion dimer and functionally contributing to a slowed turnover of TIMP3 as a candidate locus in age-related macular degeneration. Invest Oph- of the ECM.6 Immunocytochemical studies in a donor retina thalmol Vis Sci. 1997;38:1060-1065. 5. Langton KP, Barker MD, McKie N. Localization of the functional domains of hu- with SFD previously demonstrated an accumulation of man tissue inhibitor of metalloproteinases-3 and the effects of a Sorsby’s fun- TIMP-3 in Bruch’s membrane.19 Recently, the need for a dus dystrophy mutation. J Biol Chem. 1998;273:16778-16781. 6. Langton KP, McKie N, Curtis A, et al. A novel tissue inhibitor of metalloprotein- greater understanding of ECM biology to interpret dis- ases-3 mutation reveals a common molecular phenotype in Sorsby’s fundus dys- ease was further emphasized when another counterintui- trophy. J Biol Chem. 2000;35:27027-27031. tive observation was made in patients with multicentric os- 7. Kamei M, Hollyfield JG. TIMP-3 in Bruch’s membrane: changes during aging and in age-related macular degeneration. Invest Ophthalmol Vis Sci. 1999;40:2367-2375. teolysis and arthritis syndrome, disorders characterized by 8. Felbor U, Benkwitz C, Klein ML, Greenberg J, Gregory CY, Weber BH. Sorsby an enhanced ECM breakdown. Causative mutations were fundus dystrophy: reevaluation of variable expressivity in patients carrying a TIMP3 found in the gene encoding matrix metalloproteinase-2,20 founder mutation. Arch Ophthalmol. 1997;115:1569-1571. 9. Balyeat RM, Kingsley RM. Dominant macular subretinal neovascularization with suggesting that like SFD this was another ECM disorder peripheral retinal degeneration. Ophthalmology. 1987;94:1140-1147. with a complex mechanism.21 10. Sorsby A, Mason MEJ, Gardener N. A fundus dystrophy with unusual features. Br J Ophthalmol. 1949;33:67-97. Considering the onset ages reported for hemor- 11. Hoskins A, Sehmi K, Bird AC. Sorsby’s pseudo-inflammatory macular dystro- rhagic macular degeneration in SFD families with other phy. Br J Ophthalmol. 1981;65:859-865. causative mutations,8 the family with the Y172C TIMP3 12. Nishimura DY, Swiderski RE, Alward WLM, et al. The forkhead transcription fac- tor gene FKHL7 is responsible for glaucoma phenotypes which map to 6p25. mutation is among those with an earlier onset of macu- Nat Genet. 1998;19:140-147. lar disease. The clinical manifestations in SFD are all likely 13. Jacobson SG, Voigt WJ, Parel JM, et al. Automated light- and dark-adapted perim- to be secondary to the slowly progressive thickening of etry for evaluating retinitis pigmentosa. Ophthalmology. 1996;93:1604-1611. 14. Jacobson SG, Cideciyan AV, Regunath G, et al. Night blindness in Sorsby’s fun- Bruch’s membrane from the ECM imbalance discussed dus dystrophy reversed by vitamin A. Nat Genet. 1995;11:27-32. previously. Until the disease threshold is reached, the reti- 15. Cideciyan AV, Pugh EN Jr, Lamb TD, Huang Y, Jacobson SG. Rod plateaux during dark adaptation in Sorsby’s fundus dystrophy and vitamin A deficiency. nas of these patients are clinically and functionally nor- Invest Ophthalmol Vis Sci. 1997;38:1786-1794. mal. Choroidal neovascularization could be secondary 16. Cideciyan AV, Jacobson SG. An alternative phototransduction model for human to a hypoxic stimulus from the retina, which becomes rod and cone ERG a-waves. Vision Res. 1996;36:2609-2621. 17. Cideciyan AV. In vivo assessment of photoreceptor function in human diseases caused separated by a critical distance from its choriocapillaris by photoreceptor-specific gene mutations. Methods Enzymol. 2000;316:611-626. blood supply by ECM buildup. The photoreceptor ab- 18. Capon MRC, Marshall J, Krafft JI, Alexander RA, Hiscott PS, Bird AC. Sorsby’s normalities leading to extramacular vision loss may be fundus dystrophy. Ophthalmology. 1989;96:1769-1777. 14 19. Fariss RN, Apte SS, Luthert PJ, Bird AC, Milam AH. Accumulation of tissue in- secondary to RPE dysfunction. The finding of inhibitor of metalloproteinases-3 in human eyes with Sorsby’s fundus dystrophy creased TIMP-3 content of Bruch’s membrane with nor- or retinitis pigmentosa. Br J Ophthalmol. 1998;82:1329-1334. 7,22 20. Martignetti JA, Aqeel AA, Al Sewairi WA, et al. Mutation in the matrix metallo- mal aging and in AMD raises the suspicion that this proteinase 2 gene (MMP2) causes multicentric osteolysis and arthritis syn- molecule may be a marker for pathological processes that drome. Nat Genet. 2001;28:261-265. lead to a disturbed ECM turnover. Abnormal binding and 21. Vu TH. Don’t mess with the matrix. Nat Genet. 2001;28:202-203. 22. Fariss RN, Apte SS, Olsen BR, Iwata K, Milam AH. Tissue inhibitor of metallo- activity of TIMP-3 may be one of the aberrant pathways proteinases-3 is a component of Bruch’s membrane of the eye. Am J Pathol. 1997; leading to many diseases of Bruch’s membrane. 150:323-328.

(REPRINTED) ARCH OPHTHALMOL / VOL 120, MAR 2002 WWW.ARCHOPHTHALMOL.COM 379