Clinicopathologic Reports, Case Reports, and Small Case Series
January 2004

Isolated Ectopia Lentis: Potential Role of Matrix Metalloproteinasesin Fibrillin Degradation

Author Affiliations



Copyright 2004 American Medical Association. All Rights Reserved.Applicable FARS/DFARS Restrictions Apply to Government Use.2004

Arch Ophthalmol. 2004;122(1):111-114. doi:10.1001/archopht.122.1.111

Isolated ectopia lentis is a rare disorder in which no underlying causefor lens subluxation can be found, and it remains primarily a clinical diagnosis.Patients with this disorder may have a variety of ocular complaints, mostcommonly decreased vision due to lens subluxation.

Matrix metalloproteinases (MMPs) are proteolytic enzymes important inphysiologic and pathologic connective tissue remodeling. Proteolytic activityis stringently controlled by a family of natural antagonists, the tissue inhibitorsof metalloproteinases (TIMPs). The MMPs and TIMPs are present in the aqueoushumor in normal eyes1 and may interact withthe lens zonules.

We describe a patient with lens subluxation associated with positiveMMP expression and no demonstrable TIMP immunoreactivity within the lens.To our knowledge, this is the first report of MMP activity, which we postulateleads to isolated ectopia lentis. Understanding the role of these proteasesmay lead to novel therapies to reduce the progressive nature of lens subluxationaffecting the other eye.

Report of a Case

A 60-year-old man was examined across 12 years of follow-up in the OphthalmologyClinic at Prince of Wales Hospital (Sydney, Australia). The patient initiallywas seen at our department in June 1990 with a 2-week history of reduced peripheralvision in his right eye because of a rhegmatogenous retinal detachment thatwas primarily superotemporal. He had sought treatment elsewhere 9 months previouslybecause of sudden reduction in vision; the right crystalline lens was totallydislocated. After the manifestation of right retinal detachment, the best-correctedvisual acuity was counting fingers OD and 20/20 OS. In the left eye, a superonasallens coloboma was noted in the dilated pupil (Figure 1, A). The day after he was first seen in our clinic, heunderwent retinal detachment repair by means of external drainage of subretinalfluid and a scleral buckling procedure. The dislocated right lens remainedin the vitreous cavity.

Image not available

A, Clinically observed dislocation of the lens in the left eye viewedwith a slitlamp. B, Macroscopic view of extracted lens (original magnification×5). C, Macroscopic view of a normal age-matched and sex-matched lens(original magnification ×5). Immunolocalization of ectopic lens zonulematrix metalloproteinase (MMP)-1 (D, arrow), MMP-3 (E, arrows), and MMP-9(F, arrow) (original magnification ×250). Control lens (D, inset, arrow)negative for MMP-1 (original magnification ×125). Lack of ectopic lenszonule staining for tissue inhibitor of metalloproteinase (TIMP)-1 (G, arrow),TIMP-2 (H, arrows), and TIMP-3 (I, arrow) (original magnification ×250).Positive fibrillin immunoreactivity in the patient's lens zonule (G, inset,arrow) and skin biopsy results (H, inset) (original magnification ×125).A normal age-matched and sex-matched lens revealed no immunoreactivity forMMP-1 (J, arrow), MMP-3 (K, arrow), or MMP-9 (L, arrow) but positive stainingfor TIMP-1 (M, arrow), TIMP-2 (N, arrow), and TIMP-3 (O, arrow) (originalmagnification ×250). Control lenses were negative for MMP-1 (J, inset,arrow) and TIMP-1 (M, inset, arrow) (original magnification ×125).

By September 1998, the dislocated right lens was cataractous, and thepatient had abandoned regular use of a contact lens. In anticipation of thepossibility of spontaneous total subluxation of the left lens, removal ofthe right lens before intraocular lens placement was recommended. Right vitreolensectomyand placement of a scleral fixated posterior chamber intraocular 21 OD lens(CZ70BD; Alcon Laboratories Inc, Ft Worth, Tex) were successfully performed.Postoperatively, the best-corrected visual acuity eventually was 20/30 OD.

In May 2001, the patient sought treatment because of a 5-week historyof reduced vision in the left eye. There was no history of ocular trauma.Ophthalmological examination results revealed total dislocation of the leftcrystalline lens. In August 2002, a pars plana vitrectomy was performed, andthe dislocated lens was removed from the eye in toto via a superior limbalincision. An intraocular lens (CZ70BD; Alcon Laboratories Inc) was locatedin the ciliary sulcus and fixed with polypropylene sutures (Alcon LaboratoriesInc) to the sclera. The removed lens (Figure1, B) was immediately washed twice with phosphate-buffered salineand prepared at room temperature in buffered saline for analysis.

The patient's general medical history included mild aortic regurgitation,gastroesophageal reflux, hypothyroidism after a left thyroid lobectomy, andleft-sided deafness that began spontaneously in December 1996. There was nofamily history of systemic or ocular disease. Findings from a thorough systemicexamination and investigation by a cardiologist and general physician andbiochemical urinalysis revealed no features of homocystinuria, and the diagnosticcriteria for Marfan syndrome were not met2;ectopia lentis was the only major criterion present when 2 are needed. Thepatient's aortic regurgitation was not associated with aortic root dilatation;therefore, it did not fulfill a minor criterion for Marfan syndrome.

In addition to analysis of the extracted left lens, a skin punch biopsywas performed for fibrillin immunoreactivity. The lens was compared with lensesfrom age-matched and sex-matched normal postmortem eyes (n = 10) obtainedfrom the Lions Eye Bank (Sydney, Australia). Enucleated lenses and the patient'slens were processed as previously described.3 Eyeswith posterior synechiae or posterior subcapsular, nuclear, or other formsof cataract were excluded. In addition, patients with a recent history oftrauma, steroid treatment, alcohol abuse, or premature cataract formationwere excluded. The 10 subjects from whom the control eyes were obtained hadno previous or family history of ophthalmic disease and had not been receivingany medication known to influence cataract formation.

Macroscopically, the lenses (n = 11) were oriented, retroilluminated,and photographed under a microscope. Slitlamp examination (Figure 1, A) was used to orient the patient's extracted lens (Figure 1, B). Three experienced and maskedobservers assessed all photographs for opacification. The photographs werescanned (OpticPro 4830; Plustek, Taipei, Taiwan, Republic of China) and analyzedwith a Nikon 2.0 (Nikon Corp, Tokyo, Japan) in Windows 98 (Microsoft Corp,Redmond, Wash).

After orientation, the lenses were weighed and placed on a grid fordissection into 4 quadrants under a microscope. Each quadrant of each lenswas weighed individually before and after a small region from each quadrantof the anterior capsule periphery was carefully dis sected under a microscopeand placed in 4% paraformaldehyde to immunolocalize zonular proteins MMP-1,MMP-3, and MMP-9 and TIMP-1, TIMP-2, and TIMP-3, as previously described.3 Lens quadrants were individually homogenized, andthe supernatant was analyzed with enzyme-linked immunosorbent assay in triplicateto quantify MMP-1, MMP-3, MMP-9, TIMP-1, and TIMP-2.


The mean ± SD time from death to fixation of the control lenseswas 5.5 ± 1.4 hours (range, 4-9 hours). None of the control lenseswas subluxated. There was no significant difference in the weight of eachquadrant in all 11 lenses examined. All 10 control lenses were normal macroscopically(Figure 1, C), and the capsuleswere normal histologically at hematoxylin-eosin staining. The patient's leftlens macroscopically appeared compacted in the superonasal quadrant and apartfrom nuclear sclerosis appeared otherwise normal (Figure 1, B).

None of the normal lens zonules (n = 10) had MMP-1 (Figure 1, J), MMP-3 (Figure 1,K), or MMP-9 (Figure 1, L) immunoreactivity,and TIMP-1 (Figure 1, M), TIMP-2(Figure 1, N), and TIMP-3 (Figure 1, O) staining was observed in all10 normal lens zonules. Positive staining of the zonules immunolocalized onlyto the capsular insertion point and lens epithelium in all eyes. Enzyme-linkedimmunosorbent assay analysis of the normal age-matched and sex-matched lensesrevealed low MMP-1, MMP-3, and MMP-9 activity levels (Table 1), and the results were not different among quadrants. Levelsof TIMP-1 and TIMP-2 activity for the whole lenses were more than 11 timeshigher in normal lenses, as compared with that in the subluxated lens, andwere not different among quadrants.

Image not available
Proteolytic Activity in 11 Lenses*

In contrast, MMP-1 (Figure 1,D), MMP-3 (Figure 1, E), and MMP-9(Figure 1, F) were localized tothe degraded lens zonule and lens epithelium in the superonasal area in thesubluxated lens. No TIMP-1 (Figure 1,G), TIMP-2 (Figure 1, H), or TIMP-3(Figure 1, I) staining was foundin the dislocated lens. A fibrillin monoclonal antibody was included to identifythe lens zonule in the patient's lens (Figure1, G, inset). Sections incubated with isotype control antibodiesdemonstrated no reactivity in normal lenses or the patient's lens (Figure 1, D, inset, Figure 1, J, inset, and Figure 1, M, inset). Enzyme-linked immunosorbent assay analysis (Table 1) revealed a MMP-1 level 5.6 timeshigher in the superonasal quadrant of the dislocated lens than in normal lensesand 4.1 times higher than the mean of the other 3 quadrants. In the dislocatedlens, there was an unequal distribution among quadrants (P<.001). Similarly, there were 8.0 times greater MMP-3 and 4.2 timesgreater MMP-9 levels in the superonasal quadrant between the dislocated lensand normal lenses and unequal distribution among quadrants (P<.001) in the dislocated lens. Little TIMP-1 or TIMP-2 was foundin the dislocated lens, and levels were not significantly different amongquadrants. Skin biopsy results revealed normal fibrillin morphology (Figure 1, H, inset).


To our knowledge, this is the first article in which lens-associatedimmunoreactivity to MMPs and TIMPs in isolated ectopia lentis is quantified.Authors of previous articles have shown that MMPs degrade fibrillin4 and may play a role in lens subluxation in Marfansyndrome.3 The MMPs appear to be stablefor as long as 24 hours post mortem in removed ocular tissue.5 Toour knowledge, no authors have examined whether MMPs can diffuse into thelens from the aqueous humor or vitreous body or quantified the levels withinhuman lenses. The fibrillin in some cases of ectopia lentis is abnormal.6 We hypothesize that (1) the fibrillin in patientswith isolated ectopia lentis is more prone to degradation by MMPs than isnormal fibrillin; (2) there is an abnormal zonular protein in ectopia lentisthat predisposes fibrillin to MMP degradation, as in Marfan syndrome; or (3)dysregulation of MMPs and TIMPs results in the progressive destruction oflens zonules and subsequent lens subluxation. These data support existingstudy results that show that excessive proteolysis of the lens zonule resultsin fibrillin degradation.7 The presenceof excessive MMP-1, MMP-3, and MMP-9 and low TIMP-1 and TIMP-2 levels in thesuperonasal quadrant in this patient correlate with the inferotemporal directionof lens displacement observed clinically before total subluxation.

The lens zonule consists of a series of fibers composed of microfibrils8 to 12 nm in diameter. The fibrils consist largely of a cysteine-rich microfibrillarcomponent of the elastin system, fibrillin. In other tissues, fibrillin providesa template for elastin deposition.8 Understandingof the functions of the fibrillin-containing microfibrils is still incomplete;correspondingly, no comprehensive theory of the pathogenesis of lens dislocationhas emerged.

Both fibrillin molecules and fibrillin-rich microfibrils are susceptibleto degradation by serine proteases, and the amino acid substitutions foundin Marfan syndrome change the fragmentation patterns.4 Fibrillindegradation products generated by MMP activity provide conclusive evidencethat these enzymes cause specific changes to assembled microfibrils.4 In Marfan syndrome, most of the mutations in fibrillin-1are found within epidermal growth factor–like motifs and are predictedto disrupt calcium binding. These mutations may render fibrillin-1 more susceptibleto proteolytic cleavage.9 Patients withisolated ectopia lentis may also have an increased susceptibility to zonulardegradation. Structural modifications in fibrillin-rich microfibrils occurduring aging of the human ciliary zonule.10 Theseage-related changes may account for the increased incidence of ocular diseasein older patients with ectopia lentis. The hypothesis regarding the role ofMMPs in lens subluxation implies that an imbalance of lens proteases and theirantagonists may be involved in the development of ectopia lentis.

We thank ophthalmologists Malcolm Capon, FRANZCO, and John Downie, FRANZCO;physician Warren Kidson, FRACP; cardiologist Anthony Freeman, MD; and maskedobservers Timothy Nolan, MBBS, Jeanie Chui, MBBS, and Jenny Lan, PhD.

The authors have no relevant financial interest in this article.

Corresponding author: Michael P. Hennessy, MBiomedE, FRANZCO, Deptof Ophthalmology, Prince of Wales Hospital, University of New South Wales,Randwick, Sydney, Australia (e-mail:

Di Girolamo  NVerma  MJMcCluskey  PJ  et al.  Increased matrix metalloproteinases in the aqueous humor of patientsand experimental animals with uveitis. Curr Eye Res. 1996;151060- 1068
Pyeritz  RE The Marfan syndrome. Annu Rev Med. 2000;51481- 510
Sachdev  NHDi Girolamo  NMcCluskey  PJ  et al.  Lens dislocation in Marfan syndrome: potential role of matrix metalloproteinasesin fibrillin degradation. Arch Ophthalmol. 2002;120833- 835
Ashworth  JLMurphy  GRock  MJ  et al.  Fibrillin degradation by matrix metalloproteinases: implications forconnective tissue remodelling. Biochem J. 1999;340171- 181
Kamei  MHollyfield  JG TIMP-3 in Bruch's membrane: changes during aging and in age-relatedmacular degeneration. Invest Ophthalmol Vis Sci. 1999;402367- 2375
Tsipouras  PDel Mastro  RSarfarazi  M  et al.  Genetic linkage of the Marfan syndrome, ectopia lentis, and congenitalcontractural arachnodactyly to the fibrillin genes on chromosomes 15 and 5:the International Marfan Syndrome Collaborative Study. N Engl J Med. 1992;326905- 909
Booms  PTiecke  FRosenberg  T  et al.  Differential effect of FBN1 mutations on in vitro proteolysis of recombinantfibrillin-1 fragments. Hum Genet. 2000;107216- 224
Robinson  PNGodfrey  M The molecular genetics of Marfan syndrome and related microfibrillopathies. J Med Genet. 2000;379- 25
Reinhardt  DPOno  RNSakai  LY Calcium stabilizes fibrillin-1 against proteolytic degradation. J Biol Chem. 1997;2721231- 1236
Hanssen  EFranc  SGarrone  R Fibrillin-rich microfibrils: structural modifications during agingin normal human zonule. J Submicrosc Cytol Pathol. 1998;30365- 369