Br J Ophthalmol 90:391-393 doi:10.1136/bjo.2005.085803
  • Letter

Sulphation patterns of keratan sulphate proteoglycan in sclerocornea resemble cornea rather than sclera

  1. R D Young1,
  2. A J Quantock1,
  3. C Sotozono2,
  4. N Koizumi2,3,
  5. S Kinoshita2
  1. 1Structural Biophysics Group, School of Optometry and Vision Sciences, Cardiff University, Cardiff, UK
  2. 2Department of Ophthalmology, Kyoto Prefectural University of Medicine, Kyoto, Japan
  3. 3Research Center for Regenerative Medicine, Doshisha University, Kyoto, Japan
  1. Correspondence to: Professor Shigeru Kinoshita Department of Ophthalmology, Kyoto Prefectural University of Medicine, Kawaramachi-hirokoji, kamigyo-ku, Kyoto 602-0841, Japan; shigeruk{at}
  • Accepted 24 November 2005

Sclerocornea is one of the most frequent causes of congenital blindness,1 in which the cornea is opaque and resembles sclera at birth.2 The stromal matrix in sclera and in cornea is composed of collagen fibrils, with proteoglycans (PGs) in the interfibrillar space fulfilling important roles in relation to tissue structure and function. Sclera and cornea contain distinct PG populations. Whereas the predominant corneal PG carries highly sulphated keratan sulphate (KS) glycosaminoglycans (GAGs), sulphated KSPGs are not major scleral components.3 Here, we investigate if sclerocornea is more like sclera or cornea in terms of its KSPG molecular profile.


An 8 month old girl, born at 39 weeks after an uncomplicated pregnancy and delivery, had bilateral corneal opacification and was referred to Kyoto Prefectural University of Medicine. There was no family history of eye disease. Slit lamp examination disclosed diffuse, full thickness opacification, which was more severe in her left eye (fig 1A), preventing examination of the anterior chamber. Ultrasound biomicroscopy detected synechiae between the iris and peripheral cornea which histology later showed to be attributable to absence of corneal endothelium and Descemet’s membrane. We diagnosed sclerocornea, and this was treated with a penetrating keratoplasty, lensectomy, and anterior vitrectomy. Postoperative medication was 1% methylprednisolone eye drops eight times daily, levofloxacin eye drops four times daily and oral prednisolone (5 mg/day for 3 months). Six months postoperatively the corneal graft was clear without any sign of rejection or infection (fig 1B).

Figure 1

 (A) Left eye of patient before operation showing severity of corneal opacification which precluded examination of the anterior chamber. (B) Corneal clarity is maintained in left eye at 6 months after penetrating keratoplasty. (C) Histological appearance of tissue removed at keratoplasty. Peripheral cornea in sclerocornea shows thinned epithelium (e) and disorganised stromal lamellae (arrows) with blood vessels (v). Descemet’s membrane and endothelium are absent (haematoxylin and eosin, bar  =  50 µm). (D) Electron micrograph of epithelial basement membrane in sclerocornea. Epithelial basal cells display desmosomes (d), hemidesmosomes (hd), and disrupted basal lamina associated with anchoring filaments (af). Bowman’s layer is absent (bar  =  500 nm). (E) A stromal cell in sclerocornea encloses a bundle of collagen fibrils (cf) (bar  =  1 μm). (F) Coarse collagen fibrils in sclerocorneal stroma exhibit fibril bundle rather than lamellar arrangement (bar  =  500 nm).

Tissue excised at surgery was taken for laboratory investigation in accordance with international and local ethical regulations. On histopathology vascular invasion was present in all layers of the corneal stroma (fig 1C), and lamellae were severely disorganised. In some regions the collagen appeared in bundles rather than in lamellar arrays, more resembling the tissue architecture in sclera than cornea.

Electron microscopy of the central cornea revealed that Bowman’s layer was absent. The basal lamina was discontinuous and sometimes irregularly thickened, and at these sites clusters of anchoring fibrils were evident (fig 1D). Stromal cells displayed a range of morphologies, sometimes with deep membraneous invaginations enveloping collagen bundles (fig 1E). Collagen fibrils appeared more coarse than those in normal cornea, and rarely exhibited lamellar arrangement (fig 1F).

In sclerocornea, as in normal human cornea (fig 2A), immunolabelling with antibody 5D4 identified highly sulphated KS epitopes4 in the tissue in close association with stromal collagen fibrils (fig 2C). No 5D4 labelling was seen in normal human sclera (fig 2E). Immunolabelling for the lower sulphated epitope of KS using antibody 1B4 was positive in normal human cornea and sclerocornea (fig 2B, D). Lower numbers of gold particles were present in sections exposed to 1B4 antibody. As with the 5D4 antibody, immunolabelling with 1B4 in normal human sclera was negative (fig 2F).

Figure 2

 Immunogold localisation of highly sulphated KS PGs with monoclonal antibody 5D4 (A, C, E) and lower sulphated KS PGs with monoclonal antibody 1B4 (B, D, F) in normal cornea (A, B), sclerocornea (C, D), and normal sclera (E, F). Gold particles labelling highly sulphated KSPGs are abundant in normal cornea and sclerocornea (A, C) with lesser numbers labelling lower sulphated KS PGs (B, D). Normal sclera is negative for both KSPG epitopes (E, F) (bar  =  500 nm).


Previous studies of sclerocornea report a marked disorganisation of stromal lamellae and collagen fibrils that are generally much larger than normal.5–9 Clinical and histological features support the view that sclerocornea is a congenital dysgenesis with failure of corneal-type tissue characteristics to develop, the tissue instead assuming scleral morphology. The nature of the PGs in sclerocornea has not been investigated previously. The data presented here indicate that sclerocornea expresses PGs with KS sulphation motifs that resemble those in cornea, rather than those in sclera. This finding was somewhat unexpected, particularly in view of the disorganisation of the collagenous matrix in sclerocornea, which resembles that of sclera.10 We conclude that sclerocornea extracellular matrix is like that of cornea in terms of expression of highly sulphated KSPGs, but like that of sclera in terms of its collagen architecture.


We thank the BBSRC for support.


  • Competing interests: none declared