Ultrastructure of the Posterior Corneal Stroma Ursula Schlötzer-Schrehardt, PhD,1 Bjoern O. Bachmann, MD,1 Theofilos Tourtas, MD,1 Andre A.M. Torricelli, MD,2 Arun Singh, MD,2 Sheyla González, PhD,3 Hua Mei, PhD,3 Sophie X. Deng, MD, PhD,3 Steven E. Wilson, MD,2 Friedrich E. Kruse, MD1 Purpose: To reinvestigate the ultrastructure of the posterior stroma of the human cornea and to correlate the findings with the stromal behavior after big-bubble creation. Design: Observational consecutive 3-center case series. Specimens: Fresh corneoscleral buttons from human donors (n ¼ 19) and organ-cultured corneoscleral buttons (n ¼ 10) obtained after Descemet’s membrane endothelial keratoplasty. Methods: Corneal specimens were divided into central (3 mm), mid peripheral (8 mm), and peripheral parts by trephination and processed for transmission electron microscopic and immunohistochemical analyses. A big bubble was created by air injection into the stroma of organ-cultured corneas before fixation. Main Outcome Measures: The distance of keratocytes to Descemet’s membrane, number of collagen lamellae between keratocytes and Descemet’s membrane, diameter and arrangement of collagen fibrils, thickness of stromal lamella created by air injection, and immunopositivity for collagen types III, IV, and VI. Results: Stromal keratocytes were observed at variable distances from Descemet’s membrane, increasing from 1.5 to 12 mm (mean, 4.972.19 mm) in the central, 3.5 to 14 mm (mean, 8.032.47 mm) in the midperipheral, and 4.5 to 18 mm (mean, 9.772.90 mm) in the peripheral regions. The differences in mean distances were significant (P < 0.0001). The number of collagen lamellae between Descemet’s membrane and most posterior keratocytes varied from 2 to 10 and the diameter of collagen fibrils averaged 23.51.8 nm and corresponded with that of the remaining stroma. A thin layer (0.5e1.0 mm thick) of randomly arranged, unaligned collagen fibers, which was positive for collagen types III and VI, was observed at the Descemetestroma interface. The residual stromal sheet separated by air injection in 8 of 10 donor corneas varied in thickness from 4.5 to 27.5 mm, even within individual corneas (3-fold), and was composed of 5 to 11 collagen lamellae that revealed keratocytes on their anterior surface and in between. Conclusions: Barring an anchoring zone of interwoven collagen fibers at the Descemetestroma interface, the findings did not provide any evidence for the existence of a distinctive acellular pre-Descemet’s stromal layer in the human cornea. The intrastromal cleavage plane after pneumodissection seems to be nonreproducibly determined by the intraindividually and interindividually variable distances of keratocytes to Descemet’s membrane. Ophthalmology 2014;-:1e7 ª 2014 by the American Academy of Ophthalmology.
With a more widespread application of new lamellar corneal transplantation techniques, such as deep anterior lamellar keratoplasty (DALK), Descemet’s stripping automated endothelial keratoplasty, and Descemet membrane endothelial keratoplasty (DMEK), the structural and biomechanical characteristics of the posterior corneal stroma and Descemet’s membrane (DM), as well as the DMestroma interface become increasingly important. Detailed ultrastructural features of the DMestroma interface have been previously described, including a thin intermediary meshwork of randomly arranged collagen fibrils projecting into the anterior DM zone and extracellular matrix complexes formed by keratoepithelin (transforming growth factorbeinduced) and collagen type VI, suggesting a firm connection between DM and the posterior stroma.1e4 We have continued to learn about structural and biochemical characteristics of DM, which contains increased amounts of 2014 by the American Academy of Ophthalmology Published by Elsevier Inc.
adhesive glycoproteins in its most anterior zone, the interfacial matrix, mediating stromal adherence, and about its physiologic cleavage plane after stripping, as well as its interindividual morphologic and biochemical variations preventing stripping in about 2% of donor corneas.5e7 Recently, the existence of a novel, well-defined, acellular layer of the pre-DM corneal stroma, which can be separated by air injection into the stroma (big-bubble technique),8 has been reported.9 This distinct layer was reported to measure 6 to 16 mm (mean, 10.153.6 mm) in width and was characterized to lack any keratocytes and to consist of 5 to 8 tightly packed collagen lamellae containing collagen fibers with a slightly thinner diameter (21.72.4 nm) than that of collagen fibers in the remaining stroma (24.22.6 nm). Immunohistochemistry showed a more pronounced staining of this layer for collagen types III, IV, and VI.10 Although detailed analyses of the architecture of the corneal stroma
http://dx.doi.org/10.1016/j.ophtha.2014.09.037 ISSN 0161-6420/14
1
Ophthalmology Volume -, Number -, Month 2014 have been published previously,11e15 a distinctive pre-DM stromal layer had not been reported. Therefore, the description of this hypothesized new anatomic layer was critically commented on in the literature.16e18 However, because structural features of the posterior cornea are likely relevant for lamellar surgical procedures, we performed a detailed ultrastructural reinvestigation of the posterior stroma of human corneas by combining corneoscleral tissue specimens and morphologic expertise from 3 different ophthalmologic laboratories and tried to correlate the ultrastructural findings collated in fresh intact corneas with observations made in organ-cultured corneas after bigbubble creation.
Images were captured and analyzed using CCD cameras (Olympus, Tokyo, Japan; Gatan Inc, Pleasanton, CA). Measurement of structural parameters (distance of keratocytes from DM, number of collagen lamellae between keratocytes and DM, diameter of collagen fibrils) were performed with automated imageprocessing systems and integrated software packages (Analysis, Soft Imaging Systems, Münster, Germany; Soft Imaging Solutions, Olympus; Digital Micrograph, Gatan Inc). Ten measurements were performed per corneal region (central, mid peripheral, peripheral) in 10 fresh and 4 short-term cultured corneal specimens, yielding a total of 420 measurements for each parameter. The thickness of the stromal layer created by air injection was measured at 20 measuring points in each of the organ-cultured corneal specimens (n ¼ 10), yielding a total of 200 measurements.
Immunohistochemistry
Materials and Methods Tissue Specimens For electron microscopic analysis, fresh corneoscleral tissue specimens (n ¼ 5) were obtained from normal human donor eyes (mean age, 66.312.8 years) without any known ocular diseases, that were not suitable for transplantation, and were fixed within 10 hours after death. Another 5 fresh corneoscleral tissue specimens from human eyes (mean age, 59.814.9 years) enucleated for intraocular melanoma were obtained and fixed in transmission electron microscopy fixative19 immediately after enucleation. In addition, 4 donor buttons (mean age, 52.010.5 years) within 10 hours of death stored in Optisol-GS at 4 C were obtained from eye banks (SightLife, Seattle, WA). Two buttons were stored for 6 days and another 2 for 15 days before fixation. Each participating ophthalmologic laboratory contributed 4 corneas each for ultrastructural analysis. All corneal specimens were divided into central (3 mm diameter), midperipheral (8 mm diameter), and peripheral parts by trephination after fixation to avoid any mechanical tissue damage and were further processed for transmission electron microscopy. Another 10 corneal donor tissues (mean age, 68.58.9 years), which had been organ-cultured in Dulbecco’s Modified Eagle’s Medium containing streptomycin, penicillin, and fetal calf serum (Biochrom, Berlin, Germany) for about 4 weeks, were obtained after DMEK. In these corneal buttons, a big bubble was created by air injection into the stroma8 using a 30-gauge needle before fixation and processing for light and transmission electron microscopy. For immunohistochemistry, additional corneal tissues were obtained from 5 normal donor eyes (mean age, 67.610.3 years) without any known ocular disease within 10 hours after death. Informed consent for tissue donation and use in research was obtained from the relatives and the study adhered to the tenets of the Declaration of Helsinki for experiments involving human tissue. Institutional review board/ethics committees at the institution of each laboratory provided approval for this study.
Transmission Electron Microscopy Corneal specimens were fixed in 2.5% glutaraldehyde in 0.1 mol/l phosphate buffer, postfixed in 2% buffered osmium tetroxide, dehydrated in graded alcohol concentrations, and embedded in epoxy resin according to standard protocols. We stained 1-mm semithin sections for orientation with toluidine blue. Ultrathin sections (80 nm) were stained with uranyl acetate and lead citrate and examined with transmission electron microscopes (EM 906E or EM910, Carl Zeiss, Oberkochen, Germany; Tecnai G2 Spirit BT, FEI Company, Hillsboro, OR) in the 3 participating centers.
2
For indirect immunofluorescence, corneal specimens were embedded in optimal cutting temperature compound and frozen in isopentane-cooled liquid nitrogen. Cryostat-cut sections (6 mm) were fixed in cold acetone for 10 minutes, blocked with 10% normal goat serum, and incubated in primary monoclonal antibodies against collagen type III (clone 1E7-D7; Millipore, Schwalbach, Germany), collagen type IV (clone 2F11; SouthernBiotech, Birmingham, AL), and collagen type VI (clone 3C4; Millipore) diluted in phosphate-buffered saline overnight at 4 C. Antibody binding was detected by Alexa 488-conjugated secondary antibodies (Molecular Probes, Eugene, OR) and nuclear counterstaining was performed with propidium iodide (SigmaAldrich, St Louis, MO). In negative control experiments, the primary antibodies were replaced by phosphate-buffered saline or equimolar concentrations of an irrelevant isotypic primary antibody.
Statistical Evaluation Statistical evaluation was performed using SPSS version 21 (IBM Corp, Armonk, NY). The normal distribution of tested values was assessed by the KolmogoroveSmirnov method. Differences between groups were evaluated by the t-test for independent samples. The significance level was set at P ¼ 0.05.
Results The structure of the pre-DM stroma did not differ from that of the central corneal stroma by light microscopic examination and did not reveal any demarcation of a distinct layer. Stromal keratocytes were observed at variable distances from DM, with focally close approximation. Transmission electron microscopy confirmed variable distances of stromal keratocytes from DM, which ranged from 1.5 to 12 mm (mean, 4.972.19 mm) in central areas, and 3.5 to 14 mm (mean, 8.032.47 mm) in midperipheral areas to 4.5 to 18 mm (mean, 9.772.90 mm) in peripheral areas (Fig 1). Thus, keratocytes approached DM 1.5 mm in the center of the cornea, while keeping greater distances with larger standard deviations in the midperipheral and peripheral cornea. The number of collagen lamellae between DM and the last row of keratocytes varied from 2 to 10 (mean, 5.652.02) in central areas, and 4 to 9 (mean, 6.311.38) in midperipheral areas, to 5 to 9 (mean, 6.621.17) in peripheral areas of the cornea. The differences between mean distances of keratocytes, which increased from central to midperipheral and peripheral corneal regions, were statistically significant (P < 0.0001; Fig 2A), whereas there were no differences between mean numbers of collagen lamellae (Fig 2B). The individual collagen fibrils in the posterior stroma, which were interspersed with long-spacing collagen fibers, had a uniform diameter
Schlötzer-Schrehardt et al
Posterior Corneal Stroma
Figure 1. Transmission electron micrographs showing varying distances of keratocytes from Descemet’s membrane (DM) in central, mid-peripheral, and peripheral regions of the cornea obtained from a 54-year-old human donor.
(23.51.8 nm) throughout all corneal regions, which corresponded with that of the remaining stroma (data not shown). There were no differences in the mean values and standard deviations of parameters measured between the 3 participating centers. At the DMestroma interface, an approximately 0.5 to 1 mm thin intermediary layer consisting of randomly arranged collagen fibrils of slightly smaller diameter (21.52.1 nm) was observed (Fig 3A, C, D). The collagen fibrils were not aligned in bundles, but closely packed in an irregular interwoven lattice, which closely resembled the ultrastructure of Bowman’s layer with its characteristic crisscross arrangement of slightly thinner collagen fibrils at the epitheliumestroma interface (Fig 3B). The collagen fibrils of this posterior “Bowman-like” layer, which could be best visualized in oblique sections (Fig 3E), were partly merging with the anterior zone of DM. By immunofluorescence, this layer was found to be strongly positive for collagen type III (Fig 4A) and focally also for collagen type VI (Fig 4D) in all specimens examined, whereas collagen type IV did not show any specific staining reactions in the posterior stroma (Fig 4C). Partial stripping of DM confirmed that collagen type III immunolocalized to this most posterior stromal layer (Fig 4B). Air injection into the corneal stroma of organ-cultured corneas obtained after DMEK resulted in the formation of a big bubble, which measured 6.0 to 8.5 mm in diameter, in 8 of the 10 specimens (Fig 5A). The wall of the bubble was composed of a thin stromal sheet varying in thickness from 4.5 to 27.5 mm (mean, 14.96.5 mm). The thickness of the wall was usually thinnest in the central portion (mean, 8.32.9 mm) and thickest in the peripheral parts (mean, 22.63.4 mm) of the bubble (Fig 5B, C). Therefore, in individual corneas, dissection did not occur at a single reproducible stromal plane but resulted in an approximately 3-fold variation in wall thickness. Accordingly, the stromal sheet separated by air injection was composed of 5 to 11 collagen lamellae (mean, 7.41.8) including the most posterior collagen fiber lattice as described, and revealed remnants of keratocytes on its anterior surface and in between the collagen lamellae (Fig 5DeF). The variable width of the residual stromal sheet seemed to reflect the increasing spacing between keratocytes and DM from center to periphery but mostly exceeded the minimum distance of keratocytes from DM in thickness.
Discussion The description of a new, previously unrecognized layer in the posterior corneal stroma, which could be separated by air injection (big-bubble technique)8 into organ-cultured
donor corneas,9 has raised considerable controversy in the scientific community.16e18 This pre-DM stromal layer, which was noted to only be present in the central and midperipheral (9 mm) regions of the stroma and to vary in thickness from 6 to 16 mm (mean 10.153.6 mm), was characterized by its acellularity, its composition of 5 to 8 tightly packed collagen lamellae consisting of collagen fibers with a slightly thinner diameter of 21.72.43 nm with a greater interfibril spacing, and its immunopositivity for collagen types III, IV, and VI. It has been suggested that the cleavage plane was localized along the last row of keratocytes separating the acellular layer of pre-DM stromal collagen from the remaining stroma. These observations, particularly the designation of this residual stroma as a new anatomic layer of the cornea, initiated the present 3-center study aimed at reinvestigating the ultrastructure of the posterior stroma of the human cornea, which fit well with our previous aims to ultrastructurally analyze the DMestroma interface and the cleavage plane during DM stripping for DMEK.5,6 Using fresh, intact corneas fixed immediately or within 10 hours after death and 3 independent examiners, we observed (1) a variable distance of keratocytes from DM approaching DM of 1.5 mm in the central cornea and increasing toward the periphery, (2) a rather constant number of collagen lamellae of different thickness between the last row of keratocytes and DM throughout all corneal regions, (3) a uniform collagen fiber diameter and interfibril spacing throughout the corneal stroma, and (4) a 0.5 to 1 mm thin, delicate intermediary layer of randomly arranged, unaligned collagen fibers at the DMestroma interface. These findings clearly argue against the existence of a distinctive acellular posterior stromal layer, approximately 10 mm in width, adjacent to DM and are consistent with observations by Jester et al,16 who have also observed keratocytes within 5 mm of DM. Moreover, in a study using in vivo confocal imaging, keratocyte density in the posterior 10% of the stroma (pre-DM region) was not different from the remainder of the posterior third of the stroma.20 The occurrence of a minimum distance (1.5 mm) of stromal keratocytes from DM is in agreement with general biological principles, requiring direct connections of cellular basement membranes, such as DM, with collagen fibers of
3
Ophthalmology Volume -, Number -, Month 2014
Figure 2. Distances of keratocytes from Descemet’s membrane (A) and number of collagen lamellae between the last row of keratocyte and Descemet’s membrane (B) in central, midperipheral, and peripheral regions of intact human corneas (n ¼ 14) analyzed by 3 different centers. Figures are based on 140 measurements per corneal region and statistical differences between mean values were assessed by t test.
connective tissues, such as the corneal stroma, as a crucial determinant for maintaining tissue-tissue interaction and anchorage.5,21 Any direct apposition of stromal keratocytes to DM would disrupt this interaction and compromise mechanical adhesion of DM to the posterior stroma. If any structure of the posterior stroma deserves the assignation of a “layer,” it would rather be the 0.5- to 1-mm thin intermediary meshwork of interwoven collagen fibrils at the DMestroma interface, which can be clearly differentiated from the regularly aligned stromal collagen lamellae by its crisscross pattern of collagen fibrils, which are slightly smaller in diameter than those of the regular stromal collagen lamellae and partly extend into DM, serving a connecting function. Because this intermediary layer closely resembled Bowman’s layer, which is also composed of randomly dispersed, slightly thinner collagen fibrils and also represents
4
an intermediary, although essentially thicker, layer at the epithelial basement membraneestroma interface, we have designated the corresponding layer at the DMestroma interface “Bowman’s-like” layer. This delicate layer, which can be only visualized by electron microscopy, has been identified previously and described in other studies, and has been suggested to mediate mechanical anchorage between DM and posterior stroma.1,2,4 Binder et al1 used high-voltage electron microscopy of the human cornea and observed small collagen fibrils 22.3 nm in diameter running perpendicularly to DM and penetrating 0.16 to 0.21 mm into the anterior zone of the DM. Lectin histochemical studies also indicated the presence of a posterior filamentous collagenous network.22 The most exact description has been provided by Komai and Ushiki2 using scanning and transmission electron microscopy, showing a thin layer (about 0.5 mm) of irregularly arranged collagen fibrils partly extending into DM and forming a dense meshwork between DM and posterior stroma. Scanning electron microscopy confirmed the presence of this fine collagenous meshwork, which was only observed in human corneas.4 The findings of the present study provide additional information about its biochemical composition showing a pronounced immunopositivity for collagen type III and to a lesser extent for collagen type VI, which are typically observed in interstitial flexible tissues. Whereas collagen type VI has been previously shown to colocalize with transforming growth factor-beinduced along the DMestroma interface,3,5 immunolocalization of collagen type III in the human cornea has not yielded consistent results.23 Although some authors reported an absence of collagen III in adult corneal tissue,24 other investigators found it to be present in Bowman’s layer and on the edges of stromal lamellae,25 or throughout the corneal stroma.26 The immunohistochemical findings of the present study suggest that collagen type III is most abundantly expressed at the DMestroma interface as well as in the corneal epithelial basement membrane, but only weakly in the corneal stroma proper. Hence, immunolabeling for collagen types III, IV, and VI also failed to demarcate a distinct pre-DM stromal layer, approximately 10 mm in width, in this study. Moreover, we could not observe any evidence for the separation of a well-defined, keratocyte-free, pre-DM stromal layer after air injection into the stroma of organ-cultured donor corneas. Instead, the wall of the big bubble, which could be created in the majority (8/10) of donor corneas and corresponded to the type I bubble described by Dua et al,9 consisted of a stromal sheet varying considerably in thickness (4.5e27.5 mm) and revealing remnants of keratocytes on top of its anterior surface, but also inside in between the collagen lamellae. Even in individual corneas, dissection did not occur at a single stromal plane but at multiple planes resulting in an approximate 3-fold variation in wall thickness. The frequent observation of keratocytes on the sheet’s anterior surface suggests that pneumodissection of the posterior corneal stroma mostly occurs along rows of keratocytes, which may offer the least resistance to nonphysiologic mechanical forces. The thickness of the detached stromal sheet was usually thinnest in the central and thickest in the peripheral portions of the bubble, and thus seemed to reflect the increasing distances of keratocytes to DM from central to
Schlötzer-Schrehardt et al
Posterior Corneal Stroma
Figure 3. Transmission electron micrographs showing the presence of a thin layer of randomly arranged collagen fibrils (between dotted lines) at (A) the Descemet’s membrane (DM)estroma interface, (B) ultrastructurally resembling Bowman’s layer (BL). This meshwork of interwoven collagen fibrils was consistently observed both in fresh corneas obtained from older (age 81 years; C) and younger donors (age 54 years; D) and in organ-cultured corneas obtained after Descemet’s membrane stripping for Descemet’s membrane endothelial keratoplasty (E). Magnification bars ¼ 0.5 mm in AeD, and 1 mm in E.
peripheral corneal regions, but still exceeded the minimum distance of keratocytes (1.5e5 mm) in thickness. Thus, intrastromal cleavage after pneumodissection appears to separate a variable sheet of keratocyte-containing pre-DM stroma in many but not all cases, which has been well documented as “residual stroma” in previous studies providing evidence that the big-bubble technique in DALK is not consistently a DM-baring technique.27e30 In accordance with the present findings, this residual stroma adhering to DM in many cases has been reported to show considerable variations in thickness ranging from 2.6 to 25.8 mm. In conclusion, the ultrastructural and immunohistochemical findings of the present study suggest that there is no distinctive acellular pre-DM stromal zone justifying the term “layer” apart from the thin intermediary “Bowman’slike” zone of randomly arranged collagen fibers at the DM-stroma interface, which has been already described in previous publications and may serve an anchoring function between DM and posterior stroma. By reconciling the
observations made in intact corneas and in corneas after bigbubble creation, we therefore suggest that intrastromal cleavage after pneumodissection cannot be correlated to any distinctive anatomic pre-DM stromal layer,9 but seems to be consistent with normal anatomic and biomechanical conditions of the posterior corneal stroma, as suggested previously.16e18 As a consequence of potential surgical relevance, the physiologic cleavage plane using the bigbubble technique in DALK seems, in contrast with the uniform cleavage plane in DMEK,5,6 to occur at multiple stromal levels, which are nonreproducibly determined by the intraindividually and interindividually variable distances of keratocytes from DM, resulting in considerable variations in width of the residual stromal lamella. Thus, detailed knowledge of keratocyte distances to DM showing regional differences in central and peripheral corneal regions may be valuable in understanding the behavior of corneal tissue during DALK and other lamellar transplantation techniques.
5
Ophthalmology Volume -, Number -, Month 2014
Figure 4. Immunofluorescence labelling of collagen type III (A, B), collagen type IV (C), and collagen type VI (D) of Descemet’s membrane (DM) and posterior stroma of 2 normal human donor corneas (AeD, 54 years; A0 eD0 , 76 years). The arrow in (B) marks stripping of DM revealing immunolocalization of collagen type III to the innermost stromal collagen lamella, and the arrow in (D) marks focal immunolocalization of collagen type VI to the same stromal layer. Nuclear counterstain, propidium iodide; magnification bar ¼ 15 mm.
Figure 5. Light (AeE) and transmission electron micrographs (F) of an organ-cultured cornea obtained after Descemet’s membrane stripping for Descemet’s membrane endothelial keratoplasty showing big-bubble formation (A) after air injection into the corneal stroma. The boxed areas in (A) are shown in higher magnification in (BeF) and display the stromal sheet forming the bubble wall of variable thickness and containing remnants of keratocytes (arrows) within and on top of it. Magnification bars ¼ 2 mm in A, 40 mm in BeE, and 10 mm in F.
6
Schlötzer-Schrehardt et al
Posterior Corneal Stroma
References 16. 1. Binder PS, Rock ME, Schmidt KC, Anderson JA. Highvoltage electron microscopy of normal human cornea. Invest Ophthalmol Vis Sci 1991;32:2234–43. 2. Komai Y, Ushiki T. The three-dimensional organization of collagen fibrils in the human cornea and sclera. Invest Ophthalmol Vis Sci 1991;32:2244–58. 3. Streeten BW, Qi Y, Klintworth GK, et al. Immunolocalization of beta ig-h3 protein in 5q31-linked corneal dystrophies and normal corneas. Arch Ophthalmol 1999;117:67–75. 4. Hayashi S, Osawa T, Tohyama K. Comparative observations on corneas, with special reference to Bowman’s layer and Descemet’s membrane in mammals and amphibians. J Morphol 2002;254:247–58. 5. Schlötzer-Schrehardt U, Bachmann BO, Laaser K, et al. Characterization of the cleavage plane in Descemet’s membrane endothelial keratoplasty. Ophthalmology 2011;118: 1950–7. 6. Schlötzer-Schrehardt U, Bachmann BO, Tourtas T, et al. Reproducibility of graft preparations in Descemet’s membrane endothelial keratoplasty. Ophthalmology 2013;120:1769–77. 7. Gordon SR. Fibronectin antibody labels corneal stromal collagen fibrils in situ along their length and circumference and demonstrates distinct staining along the cell and stromal interfaces of Descemet’s membrane. Curr Eye Res 2014;39:312–6. 8. Anwar M, Teichmann KD. Big-bubble technique to bare Descemet’s membrane in anterior lamellar keratoplasty. J Cataract Refract Surg 2002;28:398–403. 9. Dua HS, Faraj LA, Said DG, et al. Human corneal anatomy redefined: a novel pre-Descemet’s layer (Dua’s layer). Ophthalmology 2013;120:1778–85. 10. Dua HS, Faraj LA, Branch MJ, et al. The collagen matrix of the human trabecular meshwork is an extension of the novel preDescemet’s layer (Dua’s layer). Br J Ophthalmol 2014;98:691–7. 11. Daxer A, Misof K, Grabner B, et al. Collagen fibrils in the human corneal stroma: structure and aging. Invest Ophthalmol Vis Sci 1998;39:644–8. 12. Radner W, Zehetmayer M, Aufreiter R, Mallinger R. Interlacing and cross-angle distribution of collagen lamellae in the human cornea. Cornea 1998;17:537–43. 13. Müller LJ, Pels E, Vrensen GF. The specific architecture of the anterior stroma accounts for maintenance of corneal curvature. Br J Ophthalmol 2001;85:437–43. 14. Meek KM, Boote C. The organization of collagen in the corneal stroma. Exp Eye Res 2004;78:503–12. 15. Petsche SJ, Pinsky PM. The role of 3-D collagen organization in stromal elasticity: a model based on X-ray diffraction data
17. 18.
19.
20.
21. 22. 23. 24.
25. 26. 27. 28.
29. 30.
and second harmonic-generated images. Biomech Model Mechanobiol 2013;12:1101–13. Jester JV, Murphy CJ, Winkler M, et al. Lessons in corneal structure and mechanics to guide the corneal surgeon. Ophthalmology 2013;120:1715–7. Schwab IR. Who’s on first? Ophthalmology 2013;120:1718–9. McKee HD, Irion LC, Carley FM, et al. Re: Dua et al.: Human corneal anatomy redefined: a novel pre-Descemet layer (Dua’s layer) (Ophthalmology 2013;120:1778e85) [letter online]. Ophthalmology 2014;121:e24–5. Torricelli AA, Singh V, Agrawal V, et al. Transmission electron microscopy analysis of epithelial basement membrane repair in rabbit corneas with haze. Invest Ophthalmol Vis Sci 2013;54:4026–33. Patel SV, McLaren JW, Hodge DO, Bourne WM. Normal human keratocyte density and corneal thickness measurement by using confocal microscopy in vivo. Invest Ophthalmol Vis Sci 2001;42:333–9. Adachi E, Hopkinson I, Hayashi T. Basement-membrane stromal relationships: interactions between collagen fibrils and the lamina densa. Int Rev Cytol 1997;173:73–156. Kratz-Owens K, Schanzlin DJ, Hageman GS. Identification and characterization of a posterior corneal filamentous network. Invest Ophthalmol Vis Sci 1990;31 [Suppl]:31. Marshall GE, Konstas AG, Lee WR. Collagens in ocular tissues. Br J Ophthalmol 1993;77:515–24. Ben-Zvi A, Rodrigues MM, Krachmer JH, Fujikawa LS. Immunohistochemical characterization of extracellular matrix in the developing human cornea. Curr Eye Res 1986;5: 105–17. Newsome DA, Foidart JM, Hassell JR, et al. Detection of specific collagen types in normal and keratoconus corneas. Invest Ophthalmol Vis Sci 1981;20:738–50. Nakayasu K, Tanaka M, Konomi H, Hayashi T. Distribution of types I, II, III, IV and V collagen in normal and keratoconus corneas. Ophthalmic Res 1986;18:1–10. Jafarinasab MR, Rahmati-Kamel M, Kanavi MR, Feizi S. Dissection plane in deep anterior lamellar keratoplasty using the big-bubble technique. Cornea 2010;29:388–91. McKee HD, Irion LC, Carley FM, et al. Residual corneal stroma in big-bubble deep anterior lamellar keratoplasty: a histological study in eye-bank corneas. Br J Ophthalmol 2011;95:1463–5. McKee HD, Irion LC, Carley FM, et al. Donor preparation using pneumatic dissection in endothelial keratoplasty: DMEK or DSEK? Cornea 2012;31:798–800. Kim SY, Muftuoglu O, Hogan RN, et al. Histopathology and spectral domain OCT findings of pneumatic-assisted dissection in DALK. Cornea 2012;31:1288–93.
Footnotes and Financial Disclosures Originally received: April 29, 2014. Final revision: September 30, 2014. Accepted: September 30, 2014. Available online: ---.
The authors have no proprietary or commercial interest in any materials discussed in this article. Manuscript no. 2014-653.
1
Department of Ophthalmology, University of Erlangen-Nürnberg, Erlangen, Germany.
2
The Cole Eye Institute, Cleveland Clinic, Cleveland, Ohio. 3 Jules Stein Eye Institute, Department of Ophthalmology, David Geffen School of Medicine, University of California, Los Angeles, California.
Abbreviations and Acronyms: DALK ¼ deep anterior lamellar keratoplasty; DM ¼ Descemet’s membrane; DMEK ¼ Descemet membrane endothelial keratoplasty. Correspondence: Ursula Schlötzer-Schrehardt, PhD, Department of Ophthalmology, University of Erlangen-Nürnberg, Schwabachanlage 6, D-91054 Erlangen, Germany. E-mail: [email protected].
Financial Disclosure(s): S.E.W.: Supported in part by the National Eye Institute, Bethesda, Maryland (grant no. EY10056).
7
With a more widespread application of new lamellar corneal transplantation techniques, such as deep anterior lamellar keratoplasty (DALK), Descemet’s stripping automated endothelial keratoplasty, and Descemet membrane endothelial keratoplasty (DMEK), the structural and biomechanical characteristics of the posterior corneal stroma and Descemet’s membrane (DM), as well as the DMestroma interface become increasingly important. Detailed ultrastructural features of the DMestroma interface have been previously described, including a thin intermediary meshwork of randomly arranged collagen fibrils projecting into the anterior DM zone and extracellular matrix complexes formed by keratoepithelin (transforming growth factorbeinduced) and collagen type VI, suggesting a firm connection between DM and the posterior stroma.1e4 We have continued to learn about structural and biochemical characteristics of DM, which contains increased amounts of 2014 by the American Academy of Ophthalmology Published by Elsevier Inc.
adhesive glycoproteins in its most anterior zone, the interfacial matrix, mediating stromal adherence, and about its physiologic cleavage plane after stripping, as well as its interindividual morphologic and biochemical variations preventing stripping in about 2% of donor corneas.5e7 Recently, the existence of a novel, well-defined, acellular layer of the pre-DM corneal stroma, which can be separated by air injection into the stroma (big-bubble technique),8 has been reported.9 This distinct layer was reported to measure 6 to 16 mm (mean, 10.153.6 mm) in width and was characterized to lack any keratocytes and to consist of 5 to 8 tightly packed collagen lamellae containing collagen fibers with a slightly thinner diameter (21.72.4 nm) than that of collagen fibers in the remaining stroma (24.22.6 nm). Immunohistochemistry showed a more pronounced staining of this layer for collagen types III, IV, and VI.10 Although detailed analyses of the architecture of the corneal stroma
http://dx.doi.org/10.1016/j.ophtha.2014.09.037 ISSN 0161-6420/14
1
Ophthalmology Volume -, Number -, Month 2014 have been published previously,11e15 a distinctive pre-DM stromal layer had not been reported. Therefore, the description of this hypothesized new anatomic layer was critically commented on in the literature.16e18 However, because structural features of the posterior cornea are likely relevant for lamellar surgical procedures, we performed a detailed ultrastructural reinvestigation of the posterior stroma of human corneas by combining corneoscleral tissue specimens and morphologic expertise from 3 different ophthalmologic laboratories and tried to correlate the ultrastructural findings collated in fresh intact corneas with observations made in organ-cultured corneas after bigbubble creation.
Images were captured and analyzed using CCD cameras (Olympus, Tokyo, Japan; Gatan Inc, Pleasanton, CA). Measurement of structural parameters (distance of keratocytes from DM, number of collagen lamellae between keratocytes and DM, diameter of collagen fibrils) were performed with automated imageprocessing systems and integrated software packages (Analysis, Soft Imaging Systems, Münster, Germany; Soft Imaging Solutions, Olympus; Digital Micrograph, Gatan Inc). Ten measurements were performed per corneal region (central, mid peripheral, peripheral) in 10 fresh and 4 short-term cultured corneal specimens, yielding a total of 420 measurements for each parameter. The thickness of the stromal layer created by air injection was measured at 20 measuring points in each of the organ-cultured corneal specimens (n ¼ 10), yielding a total of 200 measurements.
Immunohistochemistry
Materials and Methods Tissue Specimens For electron microscopic analysis, fresh corneoscleral tissue specimens (n ¼ 5) were obtained from normal human donor eyes (mean age, 66.312.8 years) without any known ocular diseases, that were not suitable for transplantation, and were fixed within 10 hours after death. Another 5 fresh corneoscleral tissue specimens from human eyes (mean age, 59.814.9 years) enucleated for intraocular melanoma were obtained and fixed in transmission electron microscopy fixative19 immediately after enucleation. In addition, 4 donor buttons (mean age, 52.010.5 years) within 10 hours of death stored in Optisol-GS at 4 C were obtained from eye banks (SightLife, Seattle, WA). Two buttons were stored for 6 days and another 2 for 15 days before fixation. Each participating ophthalmologic laboratory contributed 4 corneas each for ultrastructural analysis. All corneal specimens were divided into central (3 mm diameter), midperipheral (8 mm diameter), and peripheral parts by trephination after fixation to avoid any mechanical tissue damage and were further processed for transmission electron microscopy. Another 10 corneal donor tissues (mean age, 68.58.9 years), which had been organ-cultured in Dulbecco’s Modified Eagle’s Medium containing streptomycin, penicillin, and fetal calf serum (Biochrom, Berlin, Germany) for about 4 weeks, were obtained after DMEK. In these corneal buttons, a big bubble was created by air injection into the stroma8 using a 30-gauge needle before fixation and processing for light and transmission electron microscopy. For immunohistochemistry, additional corneal tissues were obtained from 5 normal donor eyes (mean age, 67.610.3 years) without any known ocular disease within 10 hours after death. Informed consent for tissue donation and use in research was obtained from the relatives and the study adhered to the tenets of the Declaration of Helsinki for experiments involving human tissue. Institutional review board/ethics committees at the institution of each laboratory provided approval for this study.
Transmission Electron Microscopy Corneal specimens were fixed in 2.5% glutaraldehyde in 0.1 mol/l phosphate buffer, postfixed in 2% buffered osmium tetroxide, dehydrated in graded alcohol concentrations, and embedded in epoxy resin according to standard protocols. We stained 1-mm semithin sections for orientation with toluidine blue. Ultrathin sections (80 nm) were stained with uranyl acetate and lead citrate and examined with transmission electron microscopes (EM 906E or EM910, Carl Zeiss, Oberkochen, Germany; Tecnai G2 Spirit BT, FEI Company, Hillsboro, OR) in the 3 participating centers.
2
For indirect immunofluorescence, corneal specimens were embedded in optimal cutting temperature compound and frozen in isopentane-cooled liquid nitrogen. Cryostat-cut sections (6 mm) were fixed in cold acetone for 10 minutes, blocked with 10% normal goat serum, and incubated in primary monoclonal antibodies against collagen type III (clone 1E7-D7; Millipore, Schwalbach, Germany), collagen type IV (clone 2F11; SouthernBiotech, Birmingham, AL), and collagen type VI (clone 3C4; Millipore) diluted in phosphate-buffered saline overnight at 4 C. Antibody binding was detected by Alexa 488-conjugated secondary antibodies (Molecular Probes, Eugene, OR) and nuclear counterstaining was performed with propidium iodide (SigmaAldrich, St Louis, MO). In negative control experiments, the primary antibodies were replaced by phosphate-buffered saline or equimolar concentrations of an irrelevant isotypic primary antibody.
Statistical Evaluation Statistical evaluation was performed using SPSS version 21 (IBM Corp, Armonk, NY). The normal distribution of tested values was assessed by the KolmogoroveSmirnov method. Differences between groups were evaluated by the t-test for independent samples. The significance level was set at P ¼ 0.05.
Results The structure of the pre-DM stroma did not differ from that of the central corneal stroma by light microscopic examination and did not reveal any demarcation of a distinct layer. Stromal keratocytes were observed at variable distances from DM, with focally close approximation. Transmission electron microscopy confirmed variable distances of stromal keratocytes from DM, which ranged from 1.5 to 12 mm (mean, 4.972.19 mm) in central areas, and 3.5 to 14 mm (mean, 8.032.47 mm) in midperipheral areas to 4.5 to 18 mm (mean, 9.772.90 mm) in peripheral areas (Fig 1). Thus, keratocytes approached DM 1.5 mm in the center of the cornea, while keeping greater distances with larger standard deviations in the midperipheral and peripheral cornea. The number of collagen lamellae between DM and the last row of keratocytes varied from 2 to 10 (mean, 5.652.02) in central areas, and 4 to 9 (mean, 6.311.38) in midperipheral areas, to 5 to 9 (mean, 6.621.17) in peripheral areas of the cornea. The differences between mean distances of keratocytes, which increased from central to midperipheral and peripheral corneal regions, were statistically significant (P < 0.0001; Fig 2A), whereas there were no differences between mean numbers of collagen lamellae (Fig 2B). The individual collagen fibrils in the posterior stroma, which were interspersed with long-spacing collagen fibers, had a uniform diameter
Schlötzer-Schrehardt et al
Posterior Corneal Stroma
Figure 1. Transmission electron micrographs showing varying distances of keratocytes from Descemet’s membrane (DM) in central, mid-peripheral, and peripheral regions of the cornea obtained from a 54-year-old human donor.
(23.51.8 nm) throughout all corneal regions, which corresponded with that of the remaining stroma (data not shown). There were no differences in the mean values and standard deviations of parameters measured between the 3 participating centers. At the DMestroma interface, an approximately 0.5 to 1 mm thin intermediary layer consisting of randomly arranged collagen fibrils of slightly smaller diameter (21.52.1 nm) was observed (Fig 3A, C, D). The collagen fibrils were not aligned in bundles, but closely packed in an irregular interwoven lattice, which closely resembled the ultrastructure of Bowman’s layer with its characteristic crisscross arrangement of slightly thinner collagen fibrils at the epitheliumestroma interface (Fig 3B). The collagen fibrils of this posterior “Bowman-like” layer, which could be best visualized in oblique sections (Fig 3E), were partly merging with the anterior zone of DM. By immunofluorescence, this layer was found to be strongly positive for collagen type III (Fig 4A) and focally also for collagen type VI (Fig 4D) in all specimens examined, whereas collagen type IV did not show any specific staining reactions in the posterior stroma (Fig 4C). Partial stripping of DM confirmed that collagen type III immunolocalized to this most posterior stromal layer (Fig 4B). Air injection into the corneal stroma of organ-cultured corneas obtained after DMEK resulted in the formation of a big bubble, which measured 6.0 to 8.5 mm in diameter, in 8 of the 10 specimens (Fig 5A). The wall of the bubble was composed of a thin stromal sheet varying in thickness from 4.5 to 27.5 mm (mean, 14.96.5 mm). The thickness of the wall was usually thinnest in the central portion (mean, 8.32.9 mm) and thickest in the peripheral parts (mean, 22.63.4 mm) of the bubble (Fig 5B, C). Therefore, in individual corneas, dissection did not occur at a single reproducible stromal plane but resulted in an approximately 3-fold variation in wall thickness. Accordingly, the stromal sheet separated by air injection was composed of 5 to 11 collagen lamellae (mean, 7.41.8) including the most posterior collagen fiber lattice as described, and revealed remnants of keratocytes on its anterior surface and in between the collagen lamellae (Fig 5DeF). The variable width of the residual stromal sheet seemed to reflect the increasing spacing between keratocytes and DM from center to periphery but mostly exceeded the minimum distance of keratocytes from DM in thickness.
Discussion The description of a new, previously unrecognized layer in the posterior corneal stroma, which could be separated by air injection (big-bubble technique)8 into organ-cultured
donor corneas,9 has raised considerable controversy in the scientific community.16e18 This pre-DM stromal layer, which was noted to only be present in the central and midperipheral (9 mm) regions of the stroma and to vary in thickness from 6 to 16 mm (mean 10.153.6 mm), was characterized by its acellularity, its composition of 5 to 8 tightly packed collagen lamellae consisting of collagen fibers with a slightly thinner diameter of 21.72.43 nm with a greater interfibril spacing, and its immunopositivity for collagen types III, IV, and VI. It has been suggested that the cleavage plane was localized along the last row of keratocytes separating the acellular layer of pre-DM stromal collagen from the remaining stroma. These observations, particularly the designation of this residual stroma as a new anatomic layer of the cornea, initiated the present 3-center study aimed at reinvestigating the ultrastructure of the posterior stroma of the human cornea, which fit well with our previous aims to ultrastructurally analyze the DMestroma interface and the cleavage plane during DM stripping for DMEK.5,6 Using fresh, intact corneas fixed immediately or within 10 hours after death and 3 independent examiners, we observed (1) a variable distance of keratocytes from DM approaching DM of 1.5 mm in the central cornea and increasing toward the periphery, (2) a rather constant number of collagen lamellae of different thickness between the last row of keratocytes and DM throughout all corneal regions, (3) a uniform collagen fiber diameter and interfibril spacing throughout the corneal stroma, and (4) a 0.5 to 1 mm thin, delicate intermediary layer of randomly arranged, unaligned collagen fibers at the DMestroma interface. These findings clearly argue against the existence of a distinctive acellular posterior stromal layer, approximately 10 mm in width, adjacent to DM and are consistent with observations by Jester et al,16 who have also observed keratocytes within 5 mm of DM. Moreover, in a study using in vivo confocal imaging, keratocyte density in the posterior 10% of the stroma (pre-DM region) was not different from the remainder of the posterior third of the stroma.20 The occurrence of a minimum distance (1.5 mm) of stromal keratocytes from DM is in agreement with general biological principles, requiring direct connections of cellular basement membranes, such as DM, with collagen fibers of
3
Ophthalmology Volume -, Number -, Month 2014
Figure 2. Distances of keratocytes from Descemet’s membrane (A) and number of collagen lamellae between the last row of keratocyte and Descemet’s membrane (B) in central, midperipheral, and peripheral regions of intact human corneas (n ¼ 14) analyzed by 3 different centers. Figures are based on 140 measurements per corneal region and statistical differences between mean values were assessed by t test.
connective tissues, such as the corneal stroma, as a crucial determinant for maintaining tissue-tissue interaction and anchorage.5,21 Any direct apposition of stromal keratocytes to DM would disrupt this interaction and compromise mechanical adhesion of DM to the posterior stroma. If any structure of the posterior stroma deserves the assignation of a “layer,” it would rather be the 0.5- to 1-mm thin intermediary meshwork of interwoven collagen fibrils at the DMestroma interface, which can be clearly differentiated from the regularly aligned stromal collagen lamellae by its crisscross pattern of collagen fibrils, which are slightly smaller in diameter than those of the regular stromal collagen lamellae and partly extend into DM, serving a connecting function. Because this intermediary layer closely resembled Bowman’s layer, which is also composed of randomly dispersed, slightly thinner collagen fibrils and also represents
4
an intermediary, although essentially thicker, layer at the epithelial basement membraneestroma interface, we have designated the corresponding layer at the DMestroma interface “Bowman’s-like” layer. This delicate layer, which can be only visualized by electron microscopy, has been identified previously and described in other studies, and has been suggested to mediate mechanical anchorage between DM and posterior stroma.1,2,4 Binder et al1 used high-voltage electron microscopy of the human cornea and observed small collagen fibrils 22.3 nm in diameter running perpendicularly to DM and penetrating 0.16 to 0.21 mm into the anterior zone of the DM. Lectin histochemical studies also indicated the presence of a posterior filamentous collagenous network.22 The most exact description has been provided by Komai and Ushiki2 using scanning and transmission electron microscopy, showing a thin layer (about 0.5 mm) of irregularly arranged collagen fibrils partly extending into DM and forming a dense meshwork between DM and posterior stroma. Scanning electron microscopy confirmed the presence of this fine collagenous meshwork, which was only observed in human corneas.4 The findings of the present study provide additional information about its biochemical composition showing a pronounced immunopositivity for collagen type III and to a lesser extent for collagen type VI, which are typically observed in interstitial flexible tissues. Whereas collagen type VI has been previously shown to colocalize with transforming growth factor-beinduced along the DMestroma interface,3,5 immunolocalization of collagen type III in the human cornea has not yielded consistent results.23 Although some authors reported an absence of collagen III in adult corneal tissue,24 other investigators found it to be present in Bowman’s layer and on the edges of stromal lamellae,25 or throughout the corneal stroma.26 The immunohistochemical findings of the present study suggest that collagen type III is most abundantly expressed at the DMestroma interface as well as in the corneal epithelial basement membrane, but only weakly in the corneal stroma proper. Hence, immunolabeling for collagen types III, IV, and VI also failed to demarcate a distinct pre-DM stromal layer, approximately 10 mm in width, in this study. Moreover, we could not observe any evidence for the separation of a well-defined, keratocyte-free, pre-DM stromal layer after air injection into the stroma of organ-cultured donor corneas. Instead, the wall of the big bubble, which could be created in the majority (8/10) of donor corneas and corresponded to the type I bubble described by Dua et al,9 consisted of a stromal sheet varying considerably in thickness (4.5e27.5 mm) and revealing remnants of keratocytes on top of its anterior surface, but also inside in between the collagen lamellae. Even in individual corneas, dissection did not occur at a single stromal plane but at multiple planes resulting in an approximate 3-fold variation in wall thickness. The frequent observation of keratocytes on the sheet’s anterior surface suggests that pneumodissection of the posterior corneal stroma mostly occurs along rows of keratocytes, which may offer the least resistance to nonphysiologic mechanical forces. The thickness of the detached stromal sheet was usually thinnest in the central and thickest in the peripheral portions of the bubble, and thus seemed to reflect the increasing distances of keratocytes to DM from central to
Schlötzer-Schrehardt et al
Posterior Corneal Stroma
Figure 3. Transmission electron micrographs showing the presence of a thin layer of randomly arranged collagen fibrils (between dotted lines) at (A) the Descemet’s membrane (DM)estroma interface, (B) ultrastructurally resembling Bowman’s layer (BL). This meshwork of interwoven collagen fibrils was consistently observed both in fresh corneas obtained from older (age 81 years; C) and younger donors (age 54 years; D) and in organ-cultured corneas obtained after Descemet’s membrane stripping for Descemet’s membrane endothelial keratoplasty (E). Magnification bars ¼ 0.5 mm in AeD, and 1 mm in E.
peripheral corneal regions, but still exceeded the minimum distance of keratocytes (1.5e5 mm) in thickness. Thus, intrastromal cleavage after pneumodissection appears to separate a variable sheet of keratocyte-containing pre-DM stroma in many but not all cases, which has been well documented as “residual stroma” in previous studies providing evidence that the big-bubble technique in DALK is not consistently a DM-baring technique.27e30 In accordance with the present findings, this residual stroma adhering to DM in many cases has been reported to show considerable variations in thickness ranging from 2.6 to 25.8 mm. In conclusion, the ultrastructural and immunohistochemical findings of the present study suggest that there is no distinctive acellular pre-DM stromal zone justifying the term “layer” apart from the thin intermediary “Bowman’slike” zone of randomly arranged collagen fibers at the DM-stroma interface, which has been already described in previous publications and may serve an anchoring function between DM and posterior stroma. By reconciling the
observations made in intact corneas and in corneas after bigbubble creation, we therefore suggest that intrastromal cleavage after pneumodissection cannot be correlated to any distinctive anatomic pre-DM stromal layer,9 but seems to be consistent with normal anatomic and biomechanical conditions of the posterior corneal stroma, as suggested previously.16e18 As a consequence of potential surgical relevance, the physiologic cleavage plane using the bigbubble technique in DALK seems, in contrast with the uniform cleavage plane in DMEK,5,6 to occur at multiple stromal levels, which are nonreproducibly determined by the intraindividually and interindividually variable distances of keratocytes from DM, resulting in considerable variations in width of the residual stromal lamella. Thus, detailed knowledge of keratocyte distances to DM showing regional differences in central and peripheral corneal regions may be valuable in understanding the behavior of corneal tissue during DALK and other lamellar transplantation techniques.
5
Ophthalmology Volume -, Number -, Month 2014
Figure 4. Immunofluorescence labelling of collagen type III (A, B), collagen type IV (C), and collagen type VI (D) of Descemet’s membrane (DM) and posterior stroma of 2 normal human donor corneas (AeD, 54 years; A0 eD0 , 76 years). The arrow in (B) marks stripping of DM revealing immunolocalization of collagen type III to the innermost stromal collagen lamella, and the arrow in (D) marks focal immunolocalization of collagen type VI to the same stromal layer. Nuclear counterstain, propidium iodide; magnification bar ¼ 15 mm.
Figure 5. Light (AeE) and transmission electron micrographs (F) of an organ-cultured cornea obtained after Descemet’s membrane stripping for Descemet’s membrane endothelial keratoplasty showing big-bubble formation (A) after air injection into the corneal stroma. The boxed areas in (A) are shown in higher magnification in (BeF) and display the stromal sheet forming the bubble wall of variable thickness and containing remnants of keratocytes (arrows) within and on top of it. Magnification bars ¼ 2 mm in A, 40 mm in BeE, and 10 mm in F.
6
Schlötzer-Schrehardt et al
Posterior Corneal Stroma
References 16. 1. Binder PS, Rock ME, Schmidt KC, Anderson JA. Highvoltage electron microscopy of normal human cornea. Invest Ophthalmol Vis Sci 1991;32:2234–43. 2. Komai Y, Ushiki T. The three-dimensional organization of collagen fibrils in the human cornea and sclera. Invest Ophthalmol Vis Sci 1991;32:2244–58. 3. Streeten BW, Qi Y, Klintworth GK, et al. Immunolocalization of beta ig-h3 protein in 5q31-linked corneal dystrophies and normal corneas. Arch Ophthalmol 1999;117:67–75. 4. Hayashi S, Osawa T, Tohyama K. Comparative observations on corneas, with special reference to Bowman’s layer and Descemet’s membrane in mammals and amphibians. J Morphol 2002;254:247–58. 5. Schlötzer-Schrehardt U, Bachmann BO, Laaser K, et al. Characterization of the cleavage plane in Descemet’s membrane endothelial keratoplasty. Ophthalmology 2011;118: 1950–7. 6. Schlötzer-Schrehardt U, Bachmann BO, Tourtas T, et al. Reproducibility of graft preparations in Descemet’s membrane endothelial keratoplasty. Ophthalmology 2013;120:1769–77. 7. Gordon SR. Fibronectin antibody labels corneal stromal collagen fibrils in situ along their length and circumference and demonstrates distinct staining along the cell and stromal interfaces of Descemet’s membrane. Curr Eye Res 2014;39:312–6. 8. Anwar M, Teichmann KD. Big-bubble technique to bare Descemet’s membrane in anterior lamellar keratoplasty. J Cataract Refract Surg 2002;28:398–403. 9. Dua HS, Faraj LA, Said DG, et al. Human corneal anatomy redefined: a novel pre-Descemet’s layer (Dua’s layer). Ophthalmology 2013;120:1778–85. 10. Dua HS, Faraj LA, Branch MJ, et al. The collagen matrix of the human trabecular meshwork is an extension of the novel preDescemet’s layer (Dua’s layer). Br J Ophthalmol 2014;98:691–7. 11. Daxer A, Misof K, Grabner B, et al. Collagen fibrils in the human corneal stroma: structure and aging. Invest Ophthalmol Vis Sci 1998;39:644–8. 12. Radner W, Zehetmayer M, Aufreiter R, Mallinger R. Interlacing and cross-angle distribution of collagen lamellae in the human cornea. Cornea 1998;17:537–43. 13. Müller LJ, Pels E, Vrensen GF. The specific architecture of the anterior stroma accounts for maintenance of corneal curvature. Br J Ophthalmol 2001;85:437–43. 14. Meek KM, Boote C. The organization of collagen in the corneal stroma. Exp Eye Res 2004;78:503–12. 15. Petsche SJ, Pinsky PM. The role of 3-D collagen organization in stromal elasticity: a model based on X-ray diffraction data
17. 18.
19.
20.
21. 22. 23. 24.
25. 26. 27. 28.
29. 30.
and second harmonic-generated images. Biomech Model Mechanobiol 2013;12:1101–13. Jester JV, Murphy CJ, Winkler M, et al. Lessons in corneal structure and mechanics to guide the corneal surgeon. Ophthalmology 2013;120:1715–7. Schwab IR. Who’s on first? Ophthalmology 2013;120:1718–9. McKee HD, Irion LC, Carley FM, et al. Re: Dua et al.: Human corneal anatomy redefined: a novel pre-Descemet layer (Dua’s layer) (Ophthalmology 2013;120:1778e85) [letter online]. Ophthalmology 2014;121:e24–5. Torricelli AA, Singh V, Agrawal V, et al. Transmission electron microscopy analysis of epithelial basement membrane repair in rabbit corneas with haze. Invest Ophthalmol Vis Sci 2013;54:4026–33. Patel SV, McLaren JW, Hodge DO, Bourne WM. Normal human keratocyte density and corneal thickness measurement by using confocal microscopy in vivo. Invest Ophthalmol Vis Sci 2001;42:333–9. Adachi E, Hopkinson I, Hayashi T. Basement-membrane stromal relationships: interactions between collagen fibrils and the lamina densa. Int Rev Cytol 1997;173:73–156. Kratz-Owens K, Schanzlin DJ, Hageman GS. Identification and characterization of a posterior corneal filamentous network. Invest Ophthalmol Vis Sci 1990;31 [Suppl]:31. Marshall GE, Konstas AG, Lee WR. Collagens in ocular tissues. Br J Ophthalmol 1993;77:515–24. Ben-Zvi A, Rodrigues MM, Krachmer JH, Fujikawa LS. Immunohistochemical characterization of extracellular matrix in the developing human cornea. Curr Eye Res 1986;5: 105–17. Newsome DA, Foidart JM, Hassell JR, et al. Detection of specific collagen types in normal and keratoconus corneas. Invest Ophthalmol Vis Sci 1981;20:738–50. Nakayasu K, Tanaka M, Konomi H, Hayashi T. Distribution of types I, II, III, IV and V collagen in normal and keratoconus corneas. Ophthalmic Res 1986;18:1–10. Jafarinasab MR, Rahmati-Kamel M, Kanavi MR, Feizi S. Dissection plane in deep anterior lamellar keratoplasty using the big-bubble technique. Cornea 2010;29:388–91. McKee HD, Irion LC, Carley FM, et al. Residual corneal stroma in big-bubble deep anterior lamellar keratoplasty: a histological study in eye-bank corneas. Br J Ophthalmol 2011;95:1463–5. McKee HD, Irion LC, Carley FM, et al. Donor preparation using pneumatic dissection in endothelial keratoplasty: DMEK or DSEK? Cornea 2012;31:798–800. Kim SY, Muftuoglu O, Hogan RN, et al. Histopathology and spectral domain OCT findings of pneumatic-assisted dissection in DALK. Cornea 2012;31:1288–93.
Footnotes and Financial Disclosures Originally received: April 29, 2014. Final revision: September 30, 2014. Accepted: September 30, 2014. Available online: ---.
The authors have no proprietary or commercial interest in any materials discussed in this article. Manuscript no. 2014-653.
1
Department of Ophthalmology, University of Erlangen-Nürnberg, Erlangen, Germany.
2
The Cole Eye Institute, Cleveland Clinic, Cleveland, Ohio. 3 Jules Stein Eye Institute, Department of Ophthalmology, David Geffen School of Medicine, University of California, Los Angeles, California.
Abbreviations and Acronyms: DALK ¼ deep anterior lamellar keratoplasty; DM ¼ Descemet’s membrane; DMEK ¼ Descemet membrane endothelial keratoplasty. Correspondence: Ursula Schlötzer-Schrehardt, PhD, Department of Ophthalmology, University of Erlangen-Nürnberg, Schwabachanlage 6, D-91054 Erlangen, Germany. E-mail: [email protected].
Financial Disclosure(s): S.E.W.: Supported in part by the National Eye Institute, Bethesda, Maryland (grant no. EY10056).
7
