Expression of Type XVIII Collagen during Healing of Corneal Incisions and Keratectomy Wounds

Abstract

purpose. To determine the distribution of type XVIII collagen in mouse ocular tissues and to investigate the expression of type XVIII collagen during healing of corneal incisions and keratectomy wounds.

methods. Immunohistochemical analysis of type XVIII collagen was performed in mouse ocular tissue, with polyclonal antibodies to the hinge domain. For wound-healing experiments, excimer laser keratectomy and single linear incisions were performed on mouse corneas. The corneas were harvested at various time points after wounding and processed for immunohistochemistry, in situ hybridization, competitive reverse transcription–polymerase chain reaction (RT-PCR), and Western blot analysis.

results. In the unwounded mouse cornea, type XVIII collagen was expressed by the corneal epithelial cells. Type XVIII collagen was immunolocalized to the mouse corneal epithelium, epithelial basement membrane, Descemet’s membrane, ciliary epithelium, lens capsule, retinal inner limiting membrane, and Bruch’s membrane. In the early stages of wound healing after excimer laser keratectomy (days 3 and 7), type XVIII collagen staining of the epithelial basement membrane was absent, whereas its localization to Descemet’s membrane was unchanged. After linear corneal incisions, however, type XVIII collagen was clearly seen in the stroma and in the epithelial basement membrane. Type XVIII collagen immunolocalization to the subepithelial stromal wound region peaked at 1 week after wounding, and its mRNA showed a corresponding temporal increase in expression within the same region after linear corneal incisions.

conclusions. The results suggest that type XVIII collagen is broadly expressed in ocular tissues and that it may have a role in wound healing, especially after incisional corneal wounds.

Type XVIII collagen is a relatively new member of the collagen superfamily and is associated with heparan sulfate side chains. ¹ The collagen superfamily comprises more than 20 distinct types of collagen, encoded by at least 38 genes. ^{2 3} Type XVIII collagen was identified by screening cDNA libraries with probes for collagen-like proteins. ^{4 5 6} Elucidation of the complete primary structure of the mouse α1(XVIII) collagen chain has revealed a polypeptide consisting of 11 noncollagenous domains (NC1–NC11) and 10 collagenous domains (COL1–COL10). Type XVIII collagen is broadly expressed in a wide variety of basement membranes ⁷ and is homologous to type XV collagen. Thus, it has been suggested that type XV and XVIII collagens form a subfamily of multiplexins (multiple triple helix domains and interruptions) within the collagen superfamily. ⁴  

Type XVIII collagen has recently become of clinical interest, because the COOH-terminal 20-kDa fragment of the collagen, termed endostatin, has been shown to be a specific inhibitor of endothelial cell proliferation and angiogenesis. ⁸ The role of type XVIII collagen in the angiogenic process has been studied extensively. ^{9 10} Previous studies in our laboratory have indicated that type XVIII collagen is present in the human cornea and that MMP-7 cleaves type XVIII collagen to generate a 28-kDa, endostatin-like fragment in the cornea. ¹¹  

Information regarding the distribution of type XVIII collagen in the eye is relatively limited, and the role of this molecule during wound healing is not known. The present studies were undertaken to evaluate the distribution of type XVIII collagen in mouse ocular tissues and to determine its expression during corneal wound healing, by using immunohistochemistry, competitive polymerase chain reaction (PCR), and in situ hybridization. 

Methods

Animals

Female C57BL6 mice, 6 to 10 weeks of age, were used. All animal studies were conducted in accordance with the Animal Care and Use Committee guidelines of the Massachusetts Eye and Ear Infirmary and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Antibodies and Reagents

Rabbit anti-mouse polyclonal antibodies to type XVIII collagen (anti-NC1-hinge and anti-endostatin antibodies) were generated as previously described (Table 1) . ¹¹ Fluorescein isothiocyanate (FITC)–labeled donkey anti-rabbit IgG antibody was purchased from Jackson ImmunoResearch Laboratories (West Grove, PA). Propidium iodide (Vector Laboratories, Burlingame, CA) was used for nuclear staining. 

Surgical Procedures

Mice were anesthetized with an intramuscular injection of a mixture of ketamine (200 mg/kg) and xylazine (10 mg/kg). An excimer laser keratectomy wound was created (2.0 mm, 180 mJ/cm², 120 pulses per eye; Summit Technology, Waltham, MA) using a number 15 Beaver blade. The mice were killed with an overdose of intraperitoneal pentobarbital sodium. The corneas were harvested at 3 days and 1, 2, and 4 weeks after surgery. 

Confocal Laser Scanning Microscopy

Corneal sections from four specimens per time point were processed for immunoconfocal microscopy. Confocal microscopic analysis was performed as described previously. ^{11 12} Briefly, cryosections (8 μm) were mounted on albumin-coated slides, air dried at room temperature for 1 hour, and fixed in acetone. Sections were incubated with 1% bovine serum albumin to block nonspecific binding of antibody, followed by incubation with anti-type XVIII collagen antibody (anti-hinge antibody) ¹¹ for 60 minutes. After a wash with PBS, the sections were incubated for 30 minutes with FITC-conjugated donkey anti-rabbit IgG antibody in PBS and viewed with a confocal laser scanning microscope (TCS 4D; Leica, Heidelberg, Germany). Negative control experiments were performed, with preimmune IgG or PBS used in place of the primary antibody. 

RNA Extraction and cDNA Synthesis

Normal corneas were dissected from mouse eyes. The epithelial cell layer, stroma, and endothelial cell layer were dissected with a surgical blade under a surgical microscope. Each sample was homogenized in 500 μL of RNA extraction reagent (TRIzol; Life Technologies, Gaithersburg, MD). The homogenates were incubated for 5 minutes at room temperature, 0.2 mL of chloroform was added, and the mixture was centrifuged at 12,000g for 15 minutes at 4°C. The upper aqueous phase was isolated, 0.5 mL of isopropanol was added, and the mixture was centrifuged at 12,000g for 10 minutes at 4°C. The resultant RNA pellet was precipitated with ethanol, air dried, resuspended in RNase-free water, and quantified by measurement of absorbance at 260 nm. RNA samples (2 μg from five mouse corneas) were then incubated at room temperature for 20 minutes with (1 U/10 μL reaction solution) DNase (Life Technologies) to rule out interference with genomic DNA contamination, after which first-strand cDNA was generated with a preamplification system (SuperScript; Life Technologies). Total RNA (2 μg) was dissolved in diethylpyrocarbonate-treated water to a volume of 11 μL, 1 μL of random hexamers was added, and the samples were incubated at 70°C for 10 minutes in a thermal cycler (PCR Express, HyBaid, Inc., Ashford, UK), chilled on ice, and subjected to reverse transcription (RT) in a reaction mixture containing 200 U of reverse transcriptase (SuperScript; Life Technologies), 200 μM each of deoxynucleoside triphosphate, 25 mM MgCl₂, and 10 mM dithiothreitol. The reaction mixture was incubated at room temperature for 10 minutes, at 42°C for 50 minutes, and at 80°C for 10 minutes and then stored at −20°C until PCR was performed. 

Polymerase Chain Reaction

The oligonucleotide primers used in this experiment are shown in Table 2 . Primers for the constitutively expressed gene of glyceraldehyde-3-phosphate dehydrogenase (G3PDH) served as the internal RT-PCR control. First-strand cDNA was incubated in a final volume of 25 μL containing PCR buffer (Promega, Madison, WI), 200 μM of each deoxynucleoside triphosphate, 1 μM of each primer, and 2 U Taq polymerase (Promega). The amplifications of α1(XVIII) and G3PDH were performed as follows: 4 minutes at 94°C (pre-heating), 1 minute at 94°C (denaturation), 1 minute at 55°C (annealing), and 1 minute at 72°C (extension) for 30 cycles. After completion of the PCR amplification, 15 μL of each PCR reaction was analyzed on a 2% agarose gel. Specific-sized PCR products were recovered from the gel with a gel extraction kit (Qiaex; Qiagen, Hilden, Germany) and used for templates of nucleotide sequencing. 

Western Blot Analysis with Anti-collagen XVIII Antibodies

Normal and wounded corneas (7 days after surgery) were dissected from eyes under a surgical microscope, three corneas per group. Each sample was homogenized and lysed with extraction buffer containing 50 mM Tris-HCl (pH 8.5), 150 mM NaCl, 1% NP-40, and a combination of protease inhibitors (Sigma Chemical Co., St. Louis, MO). Samples were electrophoresed on 4% to 20% SDS-polyacrylamide gels and transferred onto membranes (Immobilon P; Millipore, Bedford, MA). Membranes were incubated with anti-collagen XVIII antibody. After three washes of 5 minutes each in TBST, immunoblots were developed with enhanced chemiluminescence reagent. 

RNA Probes for In Situ Hybridization

The template for the type XVIII collagen used in the in situ hybridization was constructed by inserting a PCR product of mouse type XVIII collagen into a TOPO-II vector (900 base pairs in the 3′ end of the NC1 fragment; Promega). The integrity of the template was verified by DNA sequencing and restriction enzyme digestion. Digoxigenin (DIG)-labeled RNA probes were generated by in vitro transcription, with a DIG-RNA labeling kit (Roche Molecular Biochemicals, Indianapolis, IN). The antisense probe of type XVIII collagen was transcribed by SP6 RNA polymerase on an XhoI linearized DNA template. The sense probe was transcribed by T7 RNA polymerase on a KpnI linearized DNA template. The RNA transcripts were labeled with DIG-11-uridine triphosphate, according to the labeling kit protocol. The length (approximately 200 base pairs) and integrity of the synthesized riboprobes were quantified by gel electrophoresis. The concentrations were estimated by the dot-spot test with DIG-labeled control RNA. 

In Situ Hybridization

In situ hybridization was performed by using a previously described technique with modifications. ¹³ Briefly, 6-μm cryosections were air dried at room temperature for 30 minutes and fixed with 4% paraformaldehyde in PBS for 15 minutes. The slides were washed in 2× SSC (0.3 M sodium chloride, 0.03 M sodium citrate), incubated in 2 μg/mL proteinase K in TE (10 mM Tris, 1 mM EDTA) at 37°C for 20 minutes, and then washed in 0.1 M glycine/PBS for 5 minutes. After postfixation in 4% paraformaldehyde-PBS for 10 minutes, sections were immersed in 0.25% acetic anhydride for 10 minutes and 2× SSC for 15 minutes. After denaturing the probes for 5 minutes at 80°C, hybridization was performed in 50% formamide, 10% dextran sulfate, 1× Denhardt’s solution, and 5 mM EDTA. Fifty microliters of hybridization mixture containing 25 ng of labeled probe was applied to each section and incubated at 45°C for 18 hours. Slides were washed in 2× SSC twice for 10 minutes each at 40°C, followed by another wash in RNase buffer (0.5 M NaCl, 10 mM Tris-HCl, and 1 mM EDTA) and treated with 20 μg/mL RNase (Roche) at 37°C for 30 minutes. After two washes in 2× SSC, the slides were blocked with 1% bovine serum albumin and reacted with an alkaline phosphatase-conjugated antibody against DIG (Roche) for 1 hour. The sections were incubated with the substrate nitro-blue tetrazolium/5-bromo-4-chloro-3-indolyl-phosphate (NBT/BCIP; Roche) for 2 hours in the dark to develop color. All experiments with antisense RNA probes were compared with matched control experiments with the appropriate sense RNA probe, to determine signal specificity. 

Competitive RT-PCR

To quantify the expression of specific genes, DNA competitors were prepared with a competitive DNA construction kit (Takara, Tokyo, Japan), according to the manufacturer’s instructions. The DNA competitor competes with sample cDNA for the same target gene primers. After PCR amplification, gels were stained with ethidium bromide and photographed (665 instant film; Polaroid, Cambridge, MA). The photographs were scanned and digitized and NIH image software was used to invert the image and to determine the optical density of the bands (NIH Image; W. Rasband, National Institutes of Health; available by ftp from zippy.nimh.nih.gov or on floppy disk from NTIS, Springfield, VA, catalog number PB95-500195GEI). Competitor stock solution was diluted to make a set of serial (10^0.5-fold) dilutions. The relative mRNA concentration was determined by comparison with the calculated regression line. Each experiment was repeated three times. 

Results

Expression of Type XVIII Collagen in Corneal Tissue and after Excimer Keratectomy

In the mouse cornea, type XVIII collagen was immunolocalized primarily to the corneal epithelium (Fig. 1) staining of the epithelial basement membrane was also evident. Negative control experiments with specific peptide preincubation (Figs. 1M 1N 1O) and preimmune serum (not shown) confirmed the specificity of the collagen XVIII immunolocalization. The stromal keratocytes showed evidence of minimal staining. The distribution of type XVIII collagen in the stromal keratocytes was unchanged after excimer laser keratectomy at 7 days and 4 weeks (Figs. 1D 1E 1F 1G 1H 1I 1J 1K 1L) . However, the immunolocalization to the epithelial basement membrane appeared to be decreased in the early stages of wound healing (Figs. 1D 1E 1F 1G 1H 1I) . 

We used RT-PCR to determine the expression of collagen XVIII mRNA in the three corneal layers. Transcripts of type XVIII collagen were expressed in the corneal epithelial cell layer (Fig. 1P , lane 2), stroma (lane 3), and endothelial cell layer (lane 4). The predicted size band (440 bp) was present only when reverse transcriptase was in the incubation mixture, indicating that RNA, not genomic DNA, served as an amplification template. Nucleotide sequences of PCR products were identical with those of reported mouse type XVIII collagen cDNA. 

To localize cells synthesizing collagen XVIII in the cornea, in situ hybridization was performed with a DIG-labeled collagen XVIII probe. In normal mouse cornea, the expression of collagen XVIII mRNA was detected at the epithelial, keratocyte, and endothelial cell layers (Fig. 1Q) . These results were consistent with the RT-PCR data. No hybridization signal was detected in control sections hybridized with the sense probe (Fig. 1R) . 

Immunolocalization of Type XVIII Collagen in Mouse Ocular Tissue by Immunofluorescence Microscopy

The specificity of the anti-hinge antibody was confirmed before immunolocalization studies by preabsorption of the antibody, with or without its cognate peptide. In addition to the immunofluorescence for type XVIII collagen in the corneal epithelium, the lens capsule, ciliary epithelium, retinal inner limiting membrane (ILM), and Bruch’s membrane showed intense immunolocalization. The ciliary body stained for type XVIII collagen, most notably in the nonpigmented ciliary epithelium (ciliary body basement membrane; Fig. 2C ). In the retina, intense linear staining appeared in the ILM, Bruch’s membrane, and vascular basement membrane (Fig. 2D) . The lens capsule displayed a distinct staining pattern only in the outer side of the capsule (Fig. 2E) . 

Expression of Type XVIII Collagen in Healing Corneal Wounds

On postoperative day 3, when reepithelialization was complete, the expression of type XVIII collagen expression was visible at the wound margin, especially in the epithelial and stromal layer of the cornea (Figs. 3A 3B) . One week after wounding, an increase in cellularity within the wound was evident (Figs. 3E 3F) , associated with intense expression of type XVIII collagen (Figs. 3D 3E) . Two and 4 weeks after wounding, the staining for type XVIII collagen in the anterior stroma became less intense (Figs. 3G 3H 3I 3J 3K 3L) . 

Western Immunoblot Analysis

Western blot analysis of normal corneal extracts and corneal extracts from 7-day wounded corneas showed bands at 180- to 200-kDa (Fig. 4) . Additional bands of lower molecular mass, presumably representing degradation products of type XVIII collagen, were also noted. Preabsorption of the anti-endostatin antibody confirmed the specificity of the antibody for Western immunoblot analysis. 

Competitive RT-PCR

Competitive RT-PCR was used to determine whether type XVIII collagen gene transcription in the cornea after incisional wounds differs from that in the unwounded control. Approximately 9.72 × 10⁶ and 1.38 × 10⁷ copies/5 μg mRNA were identified in the unwounded and wounded corneas, respectively. The increase in type XVIII collagen gene transcriptions after incisional wounds was not statistically significant (Fig. 5) . 

In Situ Hybridization in Healing Incisional Corneal Wounds

In situ hybridization showed that the induction of expression of type XVIII collagen during corneal wound healing was also evident at the mRNA level. At 3 days after surgery, the induction of expression of type XVIII collagen mRNA was already evident around the wound’s edge (Fig. 6A) . Maximum expression occurred at 1 week after surgery (Fig. 6B) . At 2 weeks after surgery, keratocytes continued to migrate into the wound, and a strong reaction for collagen XVIII mRNA was localized to these cells at the wound’s edge (Fig. 6C) . At 4 weeks after surgery, the expression of type XVIII collagen was still observable in the stroma, although considerably diminished, (Fig. 6D) . 

Discussion

We have described the pattern of type XVIII collagen immunostaining in mouse ocular tissues. It is present in the corneal epithelium and epithelial basement membrane, lens capsule, ciliary epithelium, retinal ILM, and Bruch’s membrane. Our data obtained by confocal microscopy in unwounded mouse corneas suggest that the main localization of type XVIII collagen is in the epithelial cell layer. In situ hybridization of unwounded mouse corneas with the RNA probes indicate that the source of type XVIII collagen is not only the corneal epithelium but also the keratocytes and the endothelium. This observation is consistent with the RT-PCR studies, showing that all these cell types synthesize type XVIII collagen mRNA. 

Type XVIII collagen has been shown to be present in the skin, brain, vascular basement membrane, and placenta. ^{14 15} In ocular tissues, Halfter et al. ¹ were the first to demonstrate the presence of type XVIII collagen in avian retina. A mutation in the COL18A1 gene causes Knobloch syndrome, an autosomal recessive disorder defined by occipital encephalocele and vitreoretinal degeneration. ¹⁶ The ILM is believed to play an important role in mediating vitreoretinal adhesion and in maintaining retinal structure. ¹⁷ Therefore, the absence of type XVIII collagen may cause vitreoretinal degeneration in Knobloch syndrome, which is consistent with our finding of intense linear staining of type XVIII collagen in the ILM. 

A surprising observation was the strong expression of type XVIII collagen in the ciliary epithelium. The ciliary epithelium is known to synthesize many protein components of the aqueous humor as well as the vitreous cavity. The vitreous gel contains a dilute network of extracellular matrix (ECM) molecules (e.g., types II and IX collagen and opticin) that are essential to maintain its structure. ¹⁸ Current evidence suggests that the ciliary body is assumed to be the source of type IX collagen and opticin in the vitreous cavity. ^{18 19 20} Halfter et al. ¹ also demonstrated that the chick vitreous body contains type XVIII collagen. Thus, it is possible that type XVIII collagen is secreted by the ciliary epithelium to form the vitreous gel. It also may play a role in excluding vessels from the vitreous. Obviously, further studies are needed to clarify the significance of type XVIII collagen production in the ciliary epithelium. 

The function of type XVIII collagen in the cornea during wound healing remains uncertain. We have demonstrated, for the first time, an apparent increase in the expression of type XVIII collagen during healing after incisional corneal wounding, as evidenced by immunohistochemistry and in situ hybridization. The ECM has been thought to function solely as an inert scaffolding serving to stabilize the structure of tissues, but recent studies suggest that the ECM plays a far more active and complex role in regulating cell behavior. At the protein level, the expression of type XVIII collagen after incisional wounding occurred by day 3 and was still observable in the 4-week wound. This is in contrast to excimer keratectomy wounds, in which the subepithelial region showed comparatively limited cellular proliferation in the early stages of wound healing. Confocal images with nuclear staining demonstrated that type XVIII collagen was expressed in the connective tissue produced to fill the wound gap where many migrating cells were observed. Type XVIII collagen may regulate cell migration in the subepithelial region. Kuo et al. ²¹ have suggested a mechanism for type XVIII collagen regulation of cellular motility, in which the endostatin-containing NC1 domain of the collagen undergoes oligomerization, resulting in altered motility and morphogenesis. 

The increase in the high-molecular-weight bands (representing type XVIII collagen) after incisional wounding seemed to be substantially less than the changes noted by immunofluorescence confocal microscopy and in situ hybridization. The negative control provided evidence that the changes that we observed with these histologic techniques after incisional wounding were not nonspecific. The area around the incision in which these changes occurred was relatively small, and thus the impact of these changes in the wounded cornea would be diluted during Western blot analysis by the relatively large area of unwounded cornea surrounding the incisions. 

A similar discrepancy was apparent when comparing the results of in situ hybridization results with those of competitive PCR. Possible explanations include differences in the sensitivity of these techniques and in the analytical methods used. The competitive PCR results indicated that the increase in expression of collagen XVIII after incisional wounding was not a statistically significant increase. The in situ hybridization results illustrated a strong, but localized, response in the area surrounding the incisional wounds. That this area represents a relatively small fraction of the cornea may also explain the apparent discrepancy between the results of PCR and in situ hybridization. 

An intriguing aspect of type XVIII collagen is that it is a heparan sulfate proteoglycan. Type XVIII collagen has Ser-Gly–containing sequences that conform to consensus sequences for glycosaminoglycan attachment sites in proteoglycan core proteins. ⁹ Type XVIII collagen has been shown to be a proteoglycan that contains heparan sulfate side chains. ¹ In fact, this molecule is the first member of the collagen family in which heparan sulfate side chains have been identified. It is indicated that heparan sulfate proteoglycans are capable of binding a wide variety of biologically active proteins, including ECM molecules and growth factors. ²² The structural features of type XVIII collagen suggest that it may affect cell–ECM interactions and take part in the regulation of tissue regeneration as a growth factor–binding molecule. Whether these binding properties of type XVIII collagen also affect corneal wound healing remains to be seen. Further studies involving biochemical, physiological, and molecular genetic analyses in transgenic and knockout mice are needed to elucidate the precise function of type XVIII collagen in corneal wound healing. 

 Supported by National Eye Institute Grant EY10101 (DTA), the New England Corneal Transplant Research Fund, the Massachusetts Lions Eye Research Fund (DTA), and a Research to Prevent Blindness Lew R. Wasserman Merit Award (DTA), a Bausch & Lomb/Massachusetts Eye and Ear Infirmary Research Fellowship Award (TK), and the Japan Eye Bank Association (TK).

 Submitted for publication December 19, 2001; revised June 19, 2002; accepted June 26, 2002.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Antibodies and Peptides Used in This Study

Antibodies	Peptides Used as Immunogens	Domain in Collagen XVIII
Anti-hinge antibody	DDILANPRLPDRQPYPGVPHH	Hinge domain
Anti-endostatin antibody	RRADRGSVPIVNLKDEVLSPSWD	Endostatin domain

Table 2.

 

Primers Used for PCR and Competitive RT-PCR Analysis

PCR Primers	Expected Size (bp)
Collagen XVIII
5′-CTCATCAGGACTTTCAGCCAGT (3398–3419)
5′-CTGACCTGTAGCCCCAGTAGTT (3837–3816)	440
G3PDH
5′-CGTATTGGGCGCCTGGTCAC (31–50)
5′-CCACCTTCTTGATGTCATC (778–760)	748
Collagen XVIII competitor
5′-CTCATCAGGACTTTCAGCCAGTGTACGGTCATCATCTGACAC-3′
5′-CTGACCTGTAGCCCCAGTAGTTCGCCATCCTGGGAAGACTCC-3′	344

Figure 1.

 

(A–L) Immunoconfocal microscopy of type XVIII collagen, with anti-NC1-hinge antibodies, in unwounded mouse cornea (A–C, M–O) and at multiple time points after excimer laser keratectomy (3 days: D–F; 1 week: G–I; and 4 weeks: J–L). (A, D, G, J, M) Immunolocalization, (C, F, I, L, O) propidium iodide, (B, E, H, K, N) combined images. Type XVIII collagen immunolocalized to the epithelial layer, with faint localization to the epithelial basement membrane in the unwounded cornea and 4 weeks after wounding (A, J, arrows) but not during the early stages of wound healing (D, G, arrowheads). (A, D, G, J, double arrows) Immunolocalization of Descemet’s membrane. (M–O) Negative control. Primary antibody was preincubated with blocking peptide before immunolocalization. (P) RT-PCR analysis of type XVIII collagen mRNA in the corneal epithelial layer (lane 2), stroma (lane 3), endothelial layer (lane 4), and negative control (reverse transcription performed without reverse transcriptase; lane 5). RT-PCR products were fractionated by agarose gel electrophoresis and stained with ethidium bromide. G3PDH mRNA served as an internal control. Molecular weight standard (lane 1). (Q, R) Expression of type XVIII collagen mRNA by in situ hybridization with a DIG-labeled RNA probe. Normal cornea (Q) showing positive staining in epithelial cells, keratocytes, and endothelium with antisense probes and (R) showing no significant staining with sense probes. Bar, 50 μm.

Figure 2.

 

Type XVIII collagen distribution in normal mouse ocular tissues. Confocal images show positive staining for type XVIII collagen in ciliary epithelium (A–C), retina (A, B, D), and lens capsule (E, F). Primary antibody (anti-NC1-hinge antibody) was used without preabsorption in (A), (C), (D), and (E). (B, F): Negative control. Primary antibody was preincubated with blocking antibody (B) or replaced with preimmune serum (F). (A) Note the staining of the ciliary body basement membrane (arrow) and the ILM (arrowhead). Higher magnification showing immunolocalization to (C) the ciliary epithelium and basement membrane and (D) the ILM, Bruch’s membrane, and the vascular basement membrane. Staining of lens capsule (E) and counterstaining of the epithelial cells within the lens capsule (E, F). Nuclei are stained red with propidium iodide. ILM, inner limiting membrane; INL, inner nuclear layer; ONL, outer nuclear layer. Bars, 50 μm.

Figure 3.

 

Confocal analysis of type XVIII collagen during corneal wound healing. (A, D, G, J) Immunolocalization of type XVIII collagen, (C, F, I, L) nuclear staining with propidium iodide, and (B, E, H, K) combined images at 3 days (A–C), 1 week (D–F), 2 weeks (G–I), and 4 weeks (J–L) after wounding. Type XVIII collagen was immunolocalized to the epithelium, epithelial basement membrane (A, D, G, J, arrows), and Descemet’s membrane. It is also immunolocalized to the stromal keratocytes in the wound-healing zone surrounding the corneal incisions 1 and 2 weeks after surgery. Strongest staining occurred around the wound 7 days after the linear incisions (D). (B, E, H, K, arrowheads) wound locations. Bar, 50 μm.

Figure 4.

 

Western blot analysis of type XVIII collagen in normal and wounded cornea, with anti-endostatin antibody. (A) The specificity of anti-endostatin antibody is assayed by Western blot analysis of the recombinant NC1 fragment of collagen XVIII in the absence (lane 1) or presence (lane 2) of blocking peptide. (B) Western blot analysis of type XVIII collagen in normal (lane 3) and wounded (lane 4; 7 days after surgery) cornea. Each sample was electrophoresed under reducing conditions in SDS polyacrylamide gel, transferred to a nitrocellulose membrane, and visualized with enhanced chemiluminescence. Arrows: Characteristic 180- to 200- kDa bands of collagen XVIII immunostaining.

Figure 5.

 

Competitive PCR analysis using known quantities of a cloned gene fragment. Ethidium bromide–stained gel shows PCR products from a competitive assay of a representative (A) unwounded and (B) wounded corneal sample. Sample cDNA was added to a PCR reaction containing 10^0.5-fold serial dilution of type XVIII collagen competitor. Lanes 1–5: 10⁸, 10^7.5, 10⁷, 10^6.5, and 10⁶ copies of type XVIII collagen competitor, respectively. Arrows: positions of the 440-bp type XVIII collagen target and the 344-bp type XVIII collagen competitor PCR products.

Figure 6.

 

Expression of type XVIII collagen mRNA during healing of corneal incisions. Induction of type XVIII collagen mRNA was clearly evident around the incisional wound (A) 3 days, (B) 1 week, and (C) 2 weeks after wounding. (C) Intense signal of type XVIII collagen mRNA was noted in fibroblast-like cells adjacent to the wound (arrowheads). Epi, epithelium; St, stroma. (D) Expression of mRNA in the stroma was still observable 4 weeks after wounding. (E) No significant signal was visible around the incisional wound at 3 days with DIG-labeled sense probes. Bars, 50 μm.

The authors thank Naomi Fukai for the type XVIII collagen plasmids and for helpful discussions.