Immunohistochemical Evaluation of Conjunctival Fibrillin-1 in Marfan Syndrome FREE

Marfan syndrome (Online Mendelian Inheritance in Man [OMIM] #154700) is one of the most common inherited disorders of connective tissue with an estimated prevalence of 2 to 3 per 10 000 individuals.¹ The diagnosis of Marfan syndrome is presently based on clinical evaluation and family history.² Although the majority of patients can be diagnosed on this basis, some patients pose a diagnostic dilemma because of variable involvement of different organ systems, considerable interfamilial and intrafamilial variability of manifestations, and many features of the condition occurring in other connective tissue disorders or in isolation. Some manifestations of Marfan syndrome are age-dependent, making it difficult to apply the clinical criteria in pediatric patients.³

The Marfan syndrome locus was mapped to chromosome 15q15-21⁴ and the gene fibrillin-1 (FBN1) subsequently cloned.⁵ A genetic linkage between isolated, autosomal-dominant ectopia lentis and the fibrillin-1 gene has also been suggested.^{6 - 7} Linkage studies may be used to diagnose Marfan syndrome, but linkage requires many available affected and unaffected family members who are willing to undergo DNA testing. Making a DNA-based diagnosis of Marfan syndrome by identifying a FBN1 mutation is also not a trivial matter because the FBN1 gene is extremely complex, spans about 200 kilobase pairs (kbp) of genomic DNA, and has 65 exons.⁸ Additionally, more than 150 mutations have been entered in the international Marfan database.⁹ These have been found in almost all exons of the gene and are often specific to a family or individual.¹⁰

Immunohistochemical staining with monoclonal antibodies confirms that fibrillin-1 is abnormal in Marfan syndrome.¹¹ Studies on dermal fibroblasts in culture^{11 - 13} and aortic smooth muscle cells from patients with Marfan syndrome¹⁴ have demonstrated abnormalities in fibrillin-1. However, 35% of all patients with Marfan syndrome may display normal fibrillin-1 immunostaining patterns in skin specimens,¹¹ perhaps because this organ system is not always involved. Fibrillin-1 is an integral component of ocular tissues, including the conjunctiva.¹⁵ Immunohistochemical analysis revealed differences in fibrillin-1 staining patterns in the lens capsule and zonules of patients with Marfan syndrome.^{16 - 17}

To our knowledge, the status of conjunctival fibrillin in Marfan syndrome has not been investigated. Our study was designed to evaluate fibrillin-1 staining in conjunctival tissue of patients with Marfan syndrome who are known to have involvement of the ocular structures by virtue of the presence of ectopia lentis.

The study, conducted in 2001 through 2002, was approved by the research ethics board of The Hospital for Sick Children, Toronto, Ontario, and the Toronto Western Hospital, Toronto. Informed consent was obtained from all participants.

Six patients with Marfan syndrome with ectopia lentis (group 1) and 15 patients who were systemically healthy and undergoing strabismus surgery (group 2) at The Hospital for Sick Children or Toronto Western Hospital were recruited for the study. The 2 study groups were age-matched (Table).

Patients in group 1 were identified using the clinical records and databases of the Division of Clinical Genetics, the Ocular Genetics Program, and the health records department at The Hospital for Sick Children. Patients were selected based on availability, certainty of clinical diagnosis, presence of ectopia lentis, age old enough to comply with conjunctival biopsy when awake, and consent. The diagnosis of Marfan syndrome was ascertained through application of the Revised Berlin Criteria² by a consultant staff geneticist specializing in skeletal dysplasias from the Division of Clinical Genetics at The Hospital for Sick Children (S.U.) if the diagnosis had not been made by a geneticist previously. Additional diagnostic validation was obtained by performing molecular genetic (mutation) analysis on all patients with a clinical diagnosis of Marfan syndrome (M.G.). Patients in group 2 were recruited prospectively and consecutively from the practices of 2 strabismus surgeons (R.B. and S.K.). The pathologist (C.S.) reviewing the specimens was masked to the patient diagnosis.

For all groups, patients were excluded from the study if they refused to give or could not give informed consent or if they had ocular conditions suspected to have potential fibrillin-1 abnormalities (chronic conjunctivitis, anterior segment dysgenesis, prior ocular surgery). Patients were excluded from group 2 if they had personal or family history or physical findings suggestive of any heritable connective tissue disorders.

Among the inclusion criteria were an age older than 8 years and the ability of the patient or guardian to read and understand English. All patients had levels of visual acuity that allowed them to read the consent form. All patients were able to assent to the research even if they were not old enough to give full informed consent. Inclusion in the study was not dependent on completion of a specified follow-up.

Genomic DNA was extracted from peripheral blood mononuclear cells using standard techniques. The coding region of the FBN1 gene was amplified using all 65 exons, and further analysis was done using a combination of denaturing high-performance liquid chromatography and sequencing.

For group 1, conjunctival biopsy was performed in patients with Marfan syndrome under topical anesthesia. After obtaining informed consent, the surgeon excised an approximately 1 × 1–mm specimen from the inferior bulbar conjunctiva using toothed forceps and scissors with the patient receiving proparacaine hydrochloride 0.5%.

For group 2, the control patients, conjunctival biopsy was an extension of strabismus surgery and performed at the wound margin prior to closing the conjunctival incision. All patients received a short course of topical antibiotics postoperatively and were followed up to ensure lack of complications from the procedure.

Conjunctival specimens from Marfan and strabismus patients were processed using either formalin or tissue-freezing medium. (1) Conjunctival samples from 6 patients with Marfan syndrome and 6 control subjects were fixed and embedded in neutral buffered formalin. Tissue sections with a thickness of 5 μm were placed on gelatin-coated slides and air-dried. The tissue sections were then dewaxed and rehydrated in water.

(2) Conjunctival specimens from 6 patients with Marfan syndrome (same patients as formalin patients) and 9 control subjects (different from formalin patients) were embedded in a tissue-freezing medium and stored at −80°C until sectioning. Prior to immunostaining, the specimens were sectioned to a thickness of 5 μm and placed on sialynated slides.

The formalin-fixed paraffin sections were dewaxed in xylene and hydrated to water through a series of alcohols. The sections were then blocked with 3% hydrogen peroxide in absolute methanol for 30 minutes. The frozen section slides were fixed for 15 minutes in acetone at room temperature and washed with phosphate-buffered saline for 10 minutes.

All slides were incubated at room temperature with normal horse serum (Vector Laboratory, Burlingame, Calif) to block nonspecific staining. The slides were then incubated at room temperature with mouse antihuman fibrillin-1 antibody-2 clone 12A5.18 at a 1:100 dilution (Lab Vision, Fremont, Calif) for 60 minutes in a humidity chamber. After they were washed with phosphate-buffered saline, the sections were incubated with biotinylated horse antimouse IgG (Vector Laboratory) for 30 minutes. An avidin-biotin peroxidase reagent (Vector Laboratory) was applied for 45 minutes at room temperature. After 2 phosphate-buffered saline washes, the reaction was developed in 3-3-aminobenzidine tetrahydrochloride (Sigma, St Louis, Mo) containing 0.01% hydrogen peroxide for 10 minutes. The slides were then rinsed twice in distilled water, counterstained with Harris hematoxylin, and rinsed 2 more times in distilled water. Coverslips were applied. Qualitative analysis of the fibrillin-1 staining pattern in the conjunctival samples from the 3 groups was performed by studying the samples under a light microscope.

A mutation was detected in the FBN1 gene of all patients with a clinical diagnosis of Marfan syndrome, thereby confirming the diagnosis of Marfan syndrome (Table).

None of the conjunctival specimens that had been fixed in formalin stained positively for fibrillin-1. This was confirmed by repeat staining with different concentrations of the antibody and by using a new batch of the antibody. All frozen samples stained intensely for fibrillin-1.

All frozen conjunctival samples from the control subjects (group 2) revealed a characteristic and reproducible pattern of fibrillin-1 staining. A woven network of thin fibrils with a consistent thickness was observed outlining many of the collagen bundles in the conjunctival stroma (Figure 1).

The approximate amount of fibrillin-1 in the frozen conjunctival samples from patients with Marfan syndrome with ectopia lentis (group 1) matched that of the controls. However, consistent qualitative differences in the pattern of fibrillin-1 staining in the conjunctival stroma of all specimens were observed when compared with samples from control subjects. The fibrillin-1 fibrils were longer and slender. They showed fewer tendencies to form a meshwork (Figure 2A) and also exhibited a variation in caliber, resulting in a pattern of segmentation (Figure 2B).

Morphometric analysis of fibrillin-1 could not be performed because of variability in staining of the conjunctival sections. The variation in staining was attributed to inherent variations in the thickness of sections obtained by frozen microtomy. However, on additional sections, the staining patterns confirmed the qualitative differences.

Marfan syndrome is an autosomal disorder of connective tissue that predominantly affects the cardiovascular system, musculoskeletal system, and eye. Several ocular abnormalities are found in patients with Marfan syndrome. Ectopia lentis, which occurs in up to 80% of patients with Marfan syndrome, is probably congenital in most cases and almost always bilateral.¹⁸ Affected individuals may also present with pulmonary, dural, and skin manifestations. The diagnosis of Marfan syndrome is made when a patient satisfies the Revised Berlin Criteria.¹ Equivocal cases and pediatric patients often pose a diagnostic dilemma. Molecular diagnostic methods may be used to resolve these cases as well as define affectation in a family. However, to attain sufficient power, linkage requires families with many available affected and unaffected patients willing to undergo DNA testing. In small families and sporadic cases, which constitute up to 25% of cases,¹⁹ linkage analysis is often not possible. Additionally, linkage studies are costly and time-consuming and may not be readily available on a clinical basis. Making a DNA-based diagnosis of Marfan syndrome by identifying a specific FBN1 mutation is also not a trivial matter because of the large and complex nature of the gene.⁶ With comprehensive screening of all the exons for mutations, current techniques have a high sensitivity (66%-100%) but a high rate of false positives (52%).^{20 - 21}

Molecular genetic analysis revealed a FBN1 mutation in all our patients with a clinical diagnosis of Marfan syndrome. These mutations are likely to be pathogenic because all of them affect cysteine residues or produce stop codons, resulting in significantly altered or truncated fibrillin-1 protein. Segregation analysis, which could confirm the pathogenicity, has not yet been performed in our patients. To our knowledge, all of the mutations we found are novel and have not been reported previously. Mutations of FBN1 in Marfan syndrome are known usually to be private. This has been attributed to the high mutation rate of the FBN1 gene and the impaired reproductive fitness of patients with Marfan syndrome due to early morbidity and mortality.²²

Fibrillin-1 is a glycoprotein that is a critical component of extracellular microfibrils and serves as a scaffold for deposition of elastin and decorin.²³ There is widespread distribution of fibrillin-1–containing microfibrils in the connective tissue matrices of the skin, lung, kidney, vasculature, cartilage, tendon, and muscle.²⁴ Immunohistochemical staining with monoclonal antibodies showed fibrillin-1 to be the protein that is responsible for Marfan syndrome when abnormal.⁹ Microfibrillar abnormalities have been demonstrated by immunofluorescent techniques in dermal, lenticular, and cardiovascular tissues from patients with Marfan syndrome.²⁵ However, because of a high false-negative rate with skin biopsy and the inaccessibility of intraocular and cardiac tissue, these tissues are not desirable for diagnostic testing.^{14 ,23}

Conjunctival fibrillin-1 has not been studied in Marfan syndrome. Wheatley et al¹⁵ reported the pattern of immunohistochemical localization of fibrillin-1 in normal ocular tissues. They observed intense staining in the conjunctival stroma for fibrillin-1.¹⁵ Our study was conducted to evaluate the status of conjunctival fibrillin-1 in patients with Marfan syndrome with ectopia lentis. We are unaware of any conjunctival manifestations due to altered fibrillin-1. Further, our patients with Marfan syndrome did not have any clinically evident conjunctival disease. We hypothesized they might demonstrate altered fibrillin-1 in their conjunctiva because the ocular system is known to be involved; in contrast, the high rate of negative skin biopsies presumably relates to a lack of skin involvement in some patients. Conjunctiva is easily accessible, and conjunctival biopsy is an outpatient office procedure that can be done with most patients awake, even children, and with only a single drop of topical anesthetic.

Qualitative abnormalities in fibrillin-1 immunostaining were detected in the conjunctival stroma in all of our patients with Marfan syndrome and ectopia lentis. However, fibrillin-1 staining could only be demonstrated in frozen sections of conjunctiva. Formalin fixation appears to be unsuitable for studying fibrillin-1 staining characteristics in conjunctival specimens. This may be related to the antibody used in the study (mouse antihuman fibrillin-1 antibody-2 clone 12A5.18; Lab Vision). Other fibrillin-1 antibodies might recognize fibrillin-1 in fixed tissues.

Isolated, autosomal-dominant ectopia lentis may also be associated with underlying fibrillin-1 abnormalities. Fibrillin-1 protein abnormalities¹⁰ have been demonstrated in 2 patients with isolated ectopia lentis. It is likely that some of the cases of isolated ectopia lentis represent subdiagnostic clinical variants of Marfan syndrome. Identifying these cases has potentially important implications for families because the diagnosis prompts the physician to look for other manifestations of Marfan syndrome and provide appropriate genetic counseling.

Our study has indicated qualitative differences in fibrillin-1 staining pattern in the conjunctiva of patients with Marfan syndrome with ectopia lentis. Both fibrillin-1 molecules and fibrillin-rich microfibrils are susceptible to proteolytic degradation in ocular aging and disease.^{26 - 28} Structural consequences of microfibrillar alterations due to aging in zonules of human lens capsules are well known, namely anterior shifting of zonular insertion onto the lens capsule and disruption of zonular bundles.^{26 - 27} We do not believe that age was a factor in our study because our control subjects and patients with Marfan syndrome were age-matched and all younger than 40 years old. Our patients did not have other ocular surface disorders known to affect conjunctival fibrillin-1 (eg, chronic inflammation, prior surgery). Our antibody is not known to cross-react with fibrillin-2 or with other proteins.²⁹ The appearance of Marfan syndrome fibrillin-1 was readily identifiable by our pathologist, who was masked to patient clinical status.

One limitation of our study is that we were unable to develop a quantitative technique to measure more objectively the differences between normal and Marfan fibrillin-1 staining patterns. Because this seems to be a limitation of frozen section, and because formalin fixation was unsatisfactory for fibrillin-1 analysis, we must expand our understanding of qualitative variation. To verify the results of our study and test the hypothesis that conjunctival biopsy may be a new diagnostic modality for Marfan syndrome, we would require a larger sample of patients that includes patients with isolated ectopia lentis, ectopia lentis due to causes other than Marfan syndrome, and Marfan syndrome without ectopia lentis. This would perhaps allow us to define sensitivity and specificity parameters, identify age-related variations unique to Marfan syndrome, and recognize other ocular disorders that might produce false-positive results.

Should our results be validated, conjunctival biopsy could be a significant advancement in the diagnosis of Marfan syndrome at a time when molecular genetic tests are unsatisfactory and clinical diagnosis at times difficult.

Correspondence: Alex Levin, MD, Department of Ophthalmology, M158, The Hospital for Sick Children, 555 University Ave, Toronto, Ontario, Canada M5G 1X89 (alex.levin@sickkids.ca).

Submitted for Publication: July 29, 2004; final revision received March 22, 2005; accepted March 26, 2005.

Financial Disclosure: None.

Funding/Support: This study was supported in part by a grant from the Canadian Marfan Association, Toronto, Ontario.

Previous Presentations: This study was presented in part at the annual meeting of the American Association of Pediatric Ophthalmology and Strabismus; March 22, 2002; Seattle, Wash; the 44th Annual Research Day, Department of Ophthalmology; May 17, 2002; University of Toronto, Toronto, Ontario; the annual meeting of the Canadian Ophthalmological Society; June 14, 2002; Ottawa, Ontario; and the 7th meeting of the International Society of Genetic Eye Diseases; May 19, 2003; Paris, France.

Acknowledgment: We thank Professor Dr Ing. P. J. Coucke (Collagen Lab, Centrum Medische Genetica, Ghent, Belgium) for his help with the molecular genetic analyses. We thank Dr Venita Jay for her assistance in the initiation of this work. We are also grateful for the participation of our research assistants, LaToya Austin and Enza Perruzza; the Canadian Marfan Association, who provided funding; and our patients and their families.