Association of PDGF-BB–Induced Thrombomodulin with the Regulation of Inflammation in the Corneal and Scleral Stroma

Abstract

Purpose.: The responses of corneal and scleral stromal cells to platelet-derived growth factor (PDGF)-BB were assessed and inflammatory reactions in the cornea and sclera were investigated.

Methods.: Primary cultures of cells obtained from human subjects and strains derived from human corneal or scleral stromal cells (Cs3 and Sc1, respectively) were used. Changes in gene expression after 24 hours of exposure to 10 ng/mL PDGF-BB were analyzed with an Sc1 DNA microarray. The upregulation of several genes in Cs3 and Sc1 was confirmed by reverse transcription–polymerase chain reaction (RT-PCR) and Western blot analysis. The expression of bioactive factors was detected immunohistochemically in nine different clinical specimens.

Results.: DNA microarray analysis revealed that the gene encoding thrombomodulin (TM) was induced in Sc1 by PDGF-BB. RT-PCR confirmed that TM expression at the mRNA level was increased in both corneal and scleral stromal cells. At the protein level, TM expression was increased in scleral stromal cells, but not in corneal cells, and TM was detected in both the membrane and cytoplasmic compartments. TM was detected immunohistochemically in inflamed scleral and several corneal specimens. After TM stimulation, interleukin (IL)-18 transcription was increased in Sc1.

Conclusions.: PDGF-BB induced TM mRNA expression in scleral and corneal stromal cells, but Western blot analysis revealed the increase in TM expression only in the scleral cells. TM induced IL-18 in scleral stromal cells. A cascade involving these biologically active factors may regulate scleral and corneal inflammation. The results also reveal differences in the biological response of scleral and corneal stromal cells.

The cornea and sclera help maintain the shape of the eyeball against internal and external forces and protect against infection. Both are rich in extracellular matrix (ECM), and interactions between their cellular components and the ECM are mediated by cytokines ^{1 –3} ; however, their structures and biochemical properties differ. ^4,5  

The stroma of the cornea and sclera is composed mainly of type I collagen. The cornea is transparent and refracts light as it passes into the intraocular space. In contrast, the opaque sclera blocks the entry of unneeded light. This difference in function is a result of the differing structure of the collagen fibrils in the two regions. ⁶  

As the outer layers of the eyeball, the cornea and sclera are at risk of inflammation due to infection, ^7,8 surgery, or noninfectious diseases, including marginal corneal ulcers and scleritis. ^{9 –11} Although corneal and scleral stromal cells are thought to play important roles in association with the ECM during wound healing after injury or surgery, these functions are unproven. ^12,13 To investigate the functional similarities and differences between these cells in compromised situations, we devised experiments using cells derived from human tissues. 

The supply of cells derived from human ocular tissues depends on eyeball enucleation or excision. As these operations are rare and given that the amount of extracted tissues is limited, the supply of cells is unreliable. In addition, corneal and scleral stromal primary culture cells are extremely difficult to proliferate past the fifth passage. Thus, cell strains that can thrive long-term in culture are needed. In this study, we established corneal stromal (Cs3) and scleral stromal (Sc1) cell strains by transfection with E6E7 from human papilloma virus 16, ¹⁴ and investigated their biological characters. 

In investigating the characteristics of Cs3 and Sc1, we sought to elucidate the molecular mechanism of inflammation as a biological reaction to infection. Platelet-derived growth factor (PDGF) is secreted by various cells, including platelets, and has been shown to accelerate wound healing by stimulating cellular proliferation, migration, and ECM production. ^15,16 To uncover the interactions between stromal cells and the ECM, we investigated PDGF function at the molecular level. 

PDGF is a 35-kDa cysteine knot–containing dimer composed of A and B chains that can form three different dimers: PDGF-AA, -AB, and -BB. Both PDGF receptors (α [PDGFRα] and β [PDGFRβ]) have been shown to possess tyrosine kinase activity. PDGFRα binds all PDGF isoforms, whereas PDGFβ binds PDGF-BB with high affinity and PDGF-AB with low affinity; it does not bind PDGF-AA appreciably. In the human cornea, the presence of PDGFRβ at the surface of epithelial cells, stromal fibroblasts, and endothelial cells is more pronounced than that of PDGFRα. ^15,17  

Thrombomodulin (TM), which is associated with anticoagulation, anti-inflammation, and cellular proliferation, is a glycoprotein with five distinct domains that is expressed on the surface of various cells. Epidermal growth factor (EGF)-like repeats 3, 4, 5, and 6 have been studied in detail and are essential for the activation of protein C (PC) by thrombin. Activated PC (APC) inactivates factors Va and VIIIa and the production of thrombin. It also has anti-inflammatory effects and suppresses NFκB activation; moreover, it inhibits the expression of tumor necrosis factor (TNF)-α in monocytes and macrophages. ¹⁸ As shown in Figure 1, the N-terminal lectin-like domain of TM also suppresses inflammation. ^19,20 In the human eye, TM has been observed in the corneal epithelium and endothelium, forniceal conjunctiva, trabecular meshwork, canal of Schlemm, and nonpigmented ciliary body epithelium ²¹ ; however, its function in ocular tissues is unknown. 

Figure 1.

 

TM possesses five domains. Among its epidermal growth factor (EGF)-like repeats, one alters thrombin (TH) substrate specificity and promotes the conversion of PC to APC, which has anticoagulant and anti-inflammatory effects. The N-terminal lectin-like domain of TM also suppresses inflammation.

In this study, we examined the responses of corneal and scleral stromal cells to PDGF-BB and investigated inflammatory reactions in the cornea and sclera. 

Methods

Cell Culture

We used two cell strains (Cs3 and Sc1) established in our laboratory and their primary culture cells without immortalization by gene transfection. The cells used for culture were extracted from the enucleated eye tissues of a 65-year-old woman with choroidal malignant melanoma. Informed consent was obtained from the patient in accordance with the Declaration of Helsinki, and the study was approved by the Yamagata University Faculty of Medicine. The introduction of an immortalizing gene by transfection with E6E7 from human papilloma virus 16 enabled the cultured cells to proliferate to about the 25th passage. The strains and primary cells were grown in Dulbecco's modified Eagle's medium (DMEM; Invitrogen, San Diego, CA) supplemented with 10% fetal bovine serum (FBS; Nichirei, Tokyo, Japan) at 37°C under 5% CO₂. 

Ocular Tissue

Nine eyes of nine patients (four males and five females) were examined. Of these, three eyes had choroidal malignant melanoma, five had endophthalmitis, and one had herpetic keratitis (Table 1). Because the cornea and sclera in an eye with choroidal malignant melanoma are uninvolved in the disease, they were compared with eyes showing inflammation as controls. Samples were harvested from human tissues with approval from the Medical Ethics Committee of the Yamagata University School of Medicine for immunohistologic evaluation and were managed in compliance with the Declaration of Helsinki. 

Table 1.

 

Immunohistochemistry of PDGF-BB in Human Ocular Tissue Samples

Case	Age	Sex	Disease	Cell Infiltration (−)				Cell Infiltration (+)
				Cornea			Sclera	Cornea			Sclera
				Epi.	Str.	End.		Epi.	Str.	End.
1	65	F	L, choroidal malignant melanoma				+
2	78	F	L, choroidal malignant melanoma	+	+	+	+
3	59	M	L, choroidal malignant melanoma	+	+	+
4	78	F	R, endophthalmitis				+	+	+		+
5	71	M	L, endophthalmitis				+		+	+	+
6	70	M	L, herpetic keratitis	+	+	−		+	+
7	85	F	L, endophthalmitis	+	+		+	+	+		+
8	81	M	R, endophthalmitis	+	+		+	+	+		+
9	60	F	R, endophthalmitis	+	+		+	+	+		+

 L, left eye; R, right eye; Epi., epithelium; Str., stroma; End., endothelium; +, positive; −, negative; blank cell, was not observed or estimated.

DNA Microarray

Cells at 50% to 60% confluence from the scleral stromal cell strain were cultured in DMEM supplemented with 1% FBS for 24 hours, after which 10 ng/mL recombinant PDGF-BB (R&D Systems, Minneapolis, MN) was added to the medium, the cells were cultured for an additional 24 hours, and total cellular RNA was extracted (RiboPure RNA Isolation Kit; Applied Biosystems [ABI], Foster City, CA). The mRNA expression profile was produced by Bio Matrix Research (Chiba, Japan) with a gene microarray technology (Human Genome U133 Plus 2.0 Array; Affymetrix, Santa Clara, CA). Data analysis was performed with array technology software (Affymetrix), and the change ratios between the hybridization intensities of PDGF-BB in the treated and control samples were determined. 

Reverse Transcription–Polymerase Chain Reaction

Corneal and scleral stromal cell strains and their primary culture cells were cultured as in our DNA microarray experiment with 10 ng/mL recombinant PDGF-BB. Recombinant human (rh)TM (0.1, 1, 10, or 100 ng/mL; R&D Systems) was added to the scleral stromal cell strain medium, after which the cells were cultured for 24 hours before the extraction of total cellular RNA. A total of 2 μg of RNA was reverse transcribed with 200 U reverse transcriptase (Promega, Madison, WI), 0.5 μg Oligo dT₁₆, and 20 U RNase inhibitor (Takara-Bio, Shiga, Japan) for 60 minutes. PCR was performed in a 25-μL volume containing 1.5 μL of cDNA, 200 μM dNTPs, 1 μM primers, and 1 U DNA polymerase (KOD-Plus kit; Toyobo, Osaka, Japan). The sequences of the primers used are given in Table 2. Amplification was achieved as follows: 2 minutes at 94°C followed by 20 (for β-actin) or 25 to 37 (depending on the primer; see Table 2) cycles of 15 seconds at 94°C and 30 seconds at 60°C (β-actin) or 50 to 64°C (depending on the primer; see Table 2). A final extension for 30 seconds or 1.5 minutes at 68°C was added. 

Table 2.

 

Primer Sequences for RT-PCR and PCR

Primer Name	Primer Sequences 5′–3′	PCR (°C/cycles)	PCR Product (bp)
β-Actin	Fw: CCCATGCCATCCTGCGTCTG	60/20	577
	Rv: CGTCATACTCCTGCTTGCTG
TM	Fw: CTAGCTGTGAGTGCCCTGAAG	60/33	273
	Rv: GCCTATGAGCAAGCCCGAATG
IL-6	Fw: GTGTGAAAGCAGCAAAGAGGC	58/25	159
	Rv: CTGGAGGTACTCTAGGTATAC
IL-18	Fw: GCTTGAATCTAAATTATCAGTC	57/33	342
	Rv: GAAGATTCAAATTGCATCTTAT
TGF-β1	Fw: AACACATCAGAGCTCCGAGAA	62/29	500
	Rv: GTCAATGTACAGCTGCCGCAC
TNF-α	Fw: CTGTAGCCCATGTTGTAGC	64/34	431
	Rv: CAATGATCCCAAAGTAGACCT

 Fw, forward; Rv, reverse.

The products were separated by 2% agarose gel electrophoresis (Iwai Chemicals, Tokyo, Japan), visualized by ethidium bromide staining, and quantified (CS Analyzer; ATTO Corp., Tokyo, Japan). 

Western Blot Analysis

Cells were incubated with or without 10 ng/mL PDGF-BB, as in our DNA microarray experiment, for 48 hours and then collected and lysed in sample buffer (50 mM Tris-HCl [pH 8.0], 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% SDS, 0.1% sodium deoxycholate, and 1 mM phenylmethylsulfonyl fluoride [PMSF]; Wako Pure Chemical Industries, Osaka, Japan). After sonication in ice water, the crude lysates were cleared by centrifugation at 22,000g for 30 minutes at 4°C. The total protein concentration of the lysates was measured by the Bradford assay. In our TM localization experiment, the cells were lysed in sample buffer (50 mM Tris-HCl [pH 8.0], 150 mM NaCl, 1 mM EDTA, and 1 mM PMSF), and the lysates were centrifuged at 50,000g for 30 minutes at 4°C. We defined the supernatants as lysates of intracellular TM. The precipitates, which were defined as lysates of cell membrane TM, were lysed in buffer (50 mM Tris-HCl [pH 8.0], 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% SDS, 0.1% sodium deoxycholate, and 1 mM PMSF). 

Equal amounts of protein (20 μg) were subjected to 10% SDS-PAGE. The separated proteins were then electrotransferred to PVDF membranes (Immun-Blot; Bio-Rad, Hercules, CA). After electrotransfer, the blots were incubated for 60 minutes in blocking solution (0.3% dried low-fat milk/TTBS) on an orbital shaker. The primary (0.4 μg/mL mouse anti-TM [Santa Cruz Biotechnology, Santa Cruz, CA] or 20 ng/mL mouse anti-β-actin [Sigma-Aldrich, St. Louis, MO]) and secondary (1 μg/mL horseradish peroxidase–labeled mouse anti-IgG [Vector Laboratories, Burlingame, CA]) antibodies were prepared in blocking solution. Detection and quantification were achieved with a chemiluminescence kit (ECL Plus Western Blot Detection Kit; GE Healthcare, Buckinghamshire, UK) with a chemiluminescence imaging system (Light-Capture; ATTO Corp.). 

Immunohistochemistry

Enucleated eye tissues were fixed in 10% formalin for 24 hours, then embedded in paraffin at the Department of Pathology, Yamagata University School of Medicine, and at the Applied Medical Research Laboratory (Osaka, Japan). The embedded sections were cut 8 μm thick and stained with 0.2 μg/mL anti–PDGF-BB (Abcam, Cambridge, UK), 4 μg/mL anti-TM (Santa Cruz Biotechnology), 1 μg/mL anti-IL-18 (MBL, Nagoya, Japan), or 25 μg/mL anti-IL-18 receptor α-chain antibody (R&D Systems). Biotin-labeled anti-IgG (1.5 μg/mL anti-mouse or -rabbit; Vector Laboratories) was used as the secondary antibody. Detection was achieved with the avidin-biotin-peroxidase complex (ABC) system (Vector Laboratories). 

Immunofluorescence

The cells were incubated with or without 10 ng/mL PDGF-BB for 48 hours and then fixed in 4% paraformaldehyde phosphate buffer solution for 20 minutes at room temperature. After the addition of 5% horse serum to the cells with incubation for 30 minutes at room temperature, antigen-antibody reactions were initiated. The primary antibody was 2 μg/mL anti-TM (mouse; Santa Cruz Biotechnology), whereas the secondary antibody was 1 μg/mL biotin-labeled anti-IgG (anti-mouse; Vector Laboratories). For immunofluorescence, we used 4 μg/mL Alexa 488 (BD Biosciences, Franklin Lakes, NJ) and mounted the cells with fluorescent mounting medium; (Dako Glostrup, Denmark). A confocal laser scanning microscope (LSM510META; Carl Zeiss, Oberkochen, Germany) was used for detection. 

Enzyme-Linked Immunosorbent Assay

After incubation of the cells for 48 hours with 10 ng/mL PDGF-BB, the medium was collected and enriched 50-fold (Amicon Ultra; Millipore, Billerica, MA). ELISAs were run on the collected cell medium by Mitsubishi Chemical Medicine Corp. (Tokyo, Japan). 

Statistics

The PDGF-BB– and TM-treated groups were compared with the control groups by Mann-Whitney U test. Statistical significance was set at P < 0.05. 

Results

PDGF-BB Expression in Human Ocular Tissue Samples

PDGF-BB was expressed in the corneal epithelium, stroma, endothelium, and sclera, regardless of the presence of inflammation (Table 1, Figs. 2d, 2i, 3d, 3i). 

Figure 2.

 

Hematoxylin and eosin (HE) staining and immunohistochemical analyses of PDGF-BB and TM were performed on corneal and scleral tissues taken from cases of choroidal malignant melanoma. (a–e) Case 3 cornea: (a) HE, (b) PDGF-BB negative control, (c) TM negative control, (d) PDGF-BB, and (e) TM. (f–j) Case 1 sclera: (f) HE; (g) PDGF-BB negative control; (h) TM negative control; (i) PDGF-BB; and (j) TM. PDGF-BB was expressed in the corneal and scleral stroma, whereas TM was not. Bar, 50 μm.

Figure 3.

 

Hematoxylin and eosin (HE) staining and immunohistochemical analyses of PDGF-BB and TM were performed on inflamed corneal and scleral tissues. (a–e) Case 8 cornea: (a) HE; (b) PDGF-BB negative control; (c) TM negative control; (d) PDGF-BB; and (e) TM. (f–j) Case 9 sclera: (f) HE; (g) PDGF-BB negative control; (h) TM negative control; (i) PDGF-BB; and (j) TM. PDGF-BB was expressed in the corneal and scleral stroma, whereas TM was expressed in the scleral, but not the corneal, stroma. Bar, 50 μm.

Effects of PDGF-BB on Gene Expression in Corneal and Scleral Stromal Cells

To evaluate the biological effects of PDGF-BB, we performed Sc1 DNA microarray analysis. A two-fold or greater increase in expression was noted for some genes on the addition of PDGF-BB (Table 3). 

Table 3.

 

Genes Increased by Addition of PDGF-BB in DNA Microarray

Gene	GenBank*	Change Ratio
Tenascin C (hexabrachion)	NM_002160	3.00
Synaptotagmin-like 4 (granuphilin-a)	AI167292	3.01
YTH domain containing 1	BF592058	3.04
Chromosome 21 open reading frame 7	NM_020152	3.09
Homo sapiens, clone IMAGE:4285253, mRNA	BC026304	3.09
Ubiquitin-conjugating enzyme E2N-like	AL109622	3.09
Family with sequence similarity 111, member B	AA960844	3.10
Dynein, cytoplasmic 2, light intermediate chain 1	AA947051	3.11
Suppressor of fused homolog (Drosophila)	AF222345	3.17
Lamin B1	NM_005573	3.22
F-box protein 5	NM_012177	3.23
CUB domain containing protein 1	AK026028	3.30
Signal peptide, CUB domain, EGF-like 3	AI733234	3.33
Leucine rich repeat containing 17	NM_005824	3.54
Protease, serine, 3 (mesotrypsin)	NM_002771	3.57
Establishment of cohesion 1 homolog 2 (S. cerevisiae)	AL120674	3.58
CDNA: FLJ21037 fis, clone CAE10055	AK024690	3.67
Methylenetetrahydrofolate dehydrogenase (NADP+ dependent) 2-like	BF185922	3.72
Histone cluster 1, H4c	NM_003542	3.73
E2F transcription factor 8	NM_024680	3.75
Protease, serine, 3 (mesotrypsin)	AW007273	3.77
Stanniocalcin 1	AI300520	3.79
Transmembrane protein 158	BF062629	3.96
Stanniocalcin 1	U46768	4.00
Apolipoprotein L, 6	AW026509	4.01
MRNA; cDNA DKFZp547O0210 (from clone DKFZp547O0210)	AL831884	4.30
Hepatocyte growth factor (hepapoietin A; scatter factor)	M77227	4.33
Growth associated protein 43	NM_002045	4.71
BCL2-related protein A1	NM_004049	4.90
Full-length cDNA clone CS0DF020YD11 of Fetal brain of Homo sapiens	AL536899	5.12
Thrombomodulin	NM_000361	5.47
Tissue factor pathway inhibitor 2	AL574096	5.54
Fibroblast growth factor 13	NM_004114	6.10
Thrombomodulin	NM_000361	6.34
Phosphatidylinositol-specific phospholipase C, X domain containing 3	AI694325	11.67

*  GenBank Accession numbers: http://www.ncbi.nlm.nih.gov/Genbank; provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD.

An increase in the mRNA expression of TM after PDGF-BB treatment in Cs3, Sc1, and their primary culture cells was confirmed by RT-PCR. TM expression increased significantly in all cells (Fig. 4A; Cs3 1.48 ± 0.36-fold, Sc1 2.19 ± 0.52-fold, corneal stromal primary cells 1.78 ± 0.17-fold, and scleral stromal primary cells 1.78 ± 0.23-fold). 

Figure 4.

 

(A) Changes in TM expression in Cs3 and Sc1 and their primary culture cells resulting from PDGF-BB stimulation as shown by RT-PCR. TM expression in the PDGF-BB–treated groups was significantly increased (P < 0.05, Mann-Whitney U test; Cs3: n = 6; Sc1: n = 8; Cs. pri.: n = 5; Sc. pri.: n = 3) compared with the control level. (B) Changes in TM expression in Cs3 and Sc1 and their primary culture cells resulting from PDGF-BB stimulation as shown by Western blot analysis. TM expression in the PDGF-BB–treated scleral cells was significantly increased (P < 0.05, Mann-Whitney U test, Sc1: n = 6; Sc. pri.: n = 3), whereas that in the corneal cells was not (Mann-Whitney U test, Cs3: n = 6; Cs. pri.: n = 3) Cs. pri., primary corneal stromal cells; Sc. pri., primary scleral stromal cells.

Changes in TM Protein Expression in Corneal and Scleral Stromal Cells

TM expression in the PDGF-BB–treated cell groups was compared with that in the control cells by Western blot analysis. The treated group/control ratios were 2.60 ± 0.48 for Sc1 (P < 0.05) and 3.82 ± 1.36 for the scleral stromal primary cells (P < 0.05). TM expression was not increased in strain Cs3 (1.32 ± 0.29) or its primary cells (1.05 ± 0.55; Fig. 4B). 

Localization of TM in Corneal and Scleral Stromal Cells

Western blot analysis of intracellular and cell membrane TM (Figs. 5A, 5B) revealed that PDGF-BB increased TM expression in both the intracellular environment and cell membrane of strain Sc1 (treated group/control: 1.73 ± 0.43 intracellular, P < 0.05; 1.60 ± 0.20 cell membrane, P < 0.05) but not Cs3 (treated group/control: 0.77 ± 0.15 intracellular, P < 0.05; 0.95 ± 0.39 cell membrane, P < 0.05). 

Figure 5.

 

(A) Changes in TM expression in the cytoplasm of Cs3 and Sc1 cells as shown by Western blot analysis. TM expression in PDGF-BB–treated Sc1 cells was significantly increased (P < 0.05, Mann-Whitney U test, n = 3) but was significantly decreased in PDGF-BB–treated Cs3 cells (P < 0.05, Mann-Whitney U test, n = 3). (B) Changes in TM expression in the cell membrane of Cs3 and Sc1 cells were investigated by Western blot analysis. TM expression in the PDGF-BB–treated groups was significantly increased for Sc1 (P < 0.05, Mann-Whitney U test, n = 3) but not for Cs3 (Mann-Whitney U test, n = 3). (C) ELISA was used to examine 50-fold enriched Cs3 and Sc1 media. PDGF-BB stimulation did not alter the amount of TM released into the media (Mann-Whitney U test, Cs3: n = 3; Sc1: n = 3).

Immunofluorescence revealed that TM localization in the cytoplasm and cell membrane was significantly increased in strain Sc1. Although the increase in TM in strain Cs3 was less than that in Sc1, TM expression in Cs3 was greater than that in Sc1 before PDGF-BB treatment (Fig. 6). 

Figure 6.

 

Immunofluorescence analysis of Cs3 and Sc1 cells. (A) TM was increased in the Sc1 cytoplasm and cell membrane (Ad–f). (B) The increase in TM in strain Cs3 cytoplasm and cell membrane (Bd–f) was less than that in strain Sc1 but greater than that in strain Sc1 before PDGF-BB treatment. Nuclear, stained with propidium iodide.

Changes in TM in the Corneal and Scleral Stromal Cell Media

To evaluate changes in the amount of TM released into the ECM, we used ELISA to examine the media from strains Cs3 and Sc1. TM was not detected in either medium (data not shown). We next enriched the media 50-fold and re-examined them by ELISA. The amount of TM released was unchanged by PDGF-BB stimulation (Fig. 5C; treated group/control: cornea, 1.15 ± 0.20; sclera, 1.04 ± 0.17). 

TM Expression in Ocular Tissue Samples

Immunohistochemical analysis of TM expression was performed on nine enucleated tissues (Table 4). TM expression was observed in the inflamed scleral stroma (Fig. 3j); in contrast, the inflamed corneal stroma included both TM-expressive and nonexpressive samples (Fig. 3e). TM expression was not observed in malignant melanoma tissues (Figs. 2e, 2j). 

Table 4.

 

Immunohistochemistry of TM in Human Ocular Tissue Samples

Case	Cell Infiltration (−)				Cell Infiltration (+)
	Cornea			Sclera	Cornea			Sclera
	Epi.	Str.	End.		Epi.	Str.	End.
1				−
2	−	−	+	−
3	−	−	+
4				+	+	+		+
5				−		−	+	+
6	+	−	+		+	+
7	+	−		−	+	−		+
8	+	−		−	+	+		+
9	+	−		−	+	−		+

 Abbreviations are as defined in Table 1.

Effects of TM on Cytokine Production in Scleral Stromal Cells

To evaluate the effects of the extracellular domain of TM on the scleral stroma, we investigated the changes in IL-1β, IL-6, IL-18, transforming growth factor (TGF)-β1, and tumor necrosis factor (TNF)-α expression after the addition of rhTM to the Sc1 medium by RT-PCR. IL-1β, IL-6, TGF-β1, and TNF-α expression did not change significantly. Only IL-18 expression was increased by the addition of TM (Fig. 7, Table 5). 

Figure 7.

 

Changes in IL-1β, IL-6, IL-18, TGF-β1, and TNF-α expression in Sc1 cells due to recombinant TM stimulation as shown by RT-PCR (IL-1β: n = 4; IL-6: n = 4; IL-18: n = 4; TGF-β1: n = 4; and TNF-α: n = 4). Only the expression of IL-18 increased (P < 0.05, Mann-Whitney U test).

Table 5.

 

Changes in Levels of Cytokines by TM Stimulation in Scleral Stromal Cells, Determined by RT-PCR

rhTM Concentration (ng/mL)	IL-1β	IL-6	IL-18	TGF-β1	TNF-α
0.1	0.81 ± 0.38	1.10 ± 0.41	2.34 ± 0.56*	0.93 ± 0.33	0.93 ± 0.29
1	0.75 ± 0.31	1.10 ± 0.42	2.10 ± 0.45*	0.85 ± 0.18	0.62 ± 0.07
10	0.83 ± 0.45	1.01 ± 0.42	1.95 ± 0.42*	0.75 ± 0.25	0.68 ± 0.32
100	0.87 ± 0.60	0.85 ± 0.30	2.08 ± 0.46*	0.75 ± 0.28	0.86 ± 0.51

*   P < 0.05.

Expression of IL-18 and Its Receptor in Ocular Tissue Samples

We evaluated the expression of IL-18 and its receptor in the corneal and scleral stroma in inflamed tissues by immunohistochemistry. The expression of IL-18 and its receptor was observed in inflamed corneal and scleral samples. In inflamed corneal stromal tissue, both IL-18 receptor expressive and nonexpressive samples were identified (Table 6, Fig. 8). 

Table 6.

 

Immunohistochemistry of IL-18 and IL-18 Receptor in Human Ocular Tissue Samples

Case	IL-18				IL-18 Receptor
	Cornea			Sclera	Cornea			Sclera
	Epi.	Str.	End.		Epi.	Str.	End.
4	+	+		+	+	+		+
5		+	+	+		−	+	+
6	+	+			+	+
7	+	+		+	+	−		+
8	+	+		+	+	+		+
9	+	+		+	+	−		+

 Abbreviations are as in Table 1.

Figure 8.

 

Hematoxylin and eosin (HE) staining and immunohistochemical analysis of IL-18 and its receptor in inflamed corneal and scleral tissues. (a–e) Case 6 cornea: (a) HE; (b) IL-18 negative control; (c) IL-18 receptor negative control; (d) IL-18; and (e) IL-18 receptor. (f–j) Case 9 sclera: (f) HE; (g) IL-18 negative control; (h) IL-18 receptor negative control; (i) IL-18; and (j) IL-18 receptor. IL-18 was expressed in the corneal and scleral stroma. IL-18 receptor was detected in both expressive and nonexpressive samples of inflamed corneal stroma. Bar, 50 μm.

Discussion

We investigated inflammatory reactions in the cornea and sclera. TM was included in this study, because it has both direct and indirect anti-inflammatory effects. 

We found that PDGF-BB was expressed in the cornea and sclera, regardless of the presence of inflammation. Previously, PDGF expression was reported in the corneal epithelium, stroma, endothelium, and sclera. ^2,22 More PDGF-BB may be produced in inflamed than in noninflamed tissues, and this may affect surrounding tissues in a dose-dependent manner. 

According to our DNA microarray results, TM expression was increased in strain Sc1. RT-PCR confirmed the increase in TM mRNA expression in both Cs3 and Sc1 and in their primary culture cells. However, Western blot analysis revealed the increase in TM expression only in the scleral cells. Although scleral and corneal cells are similar in shape, we observed differences in properties between the cells that may be based on their localization. Because TM is an angiogenic factor, ²³ the control of its expression in the avascular cornea could be different from that in the sclera. Our use of cell strains established by the transfection of E6E7 from human papilloma virus 16 raises the possibility that changes in the characteristics of the cells affected the results of our biological experiments. However, our strains showed the same tendency as the primary cells. Hence, we believe that these strains can be used in place of primary cells in experimental research. 

Immunofluorescence and Western blot analysis showed that TM expression was increased in the Sc1 cytoplasm and cell membrane, suggesting that PDGF-BB stimulates the anticoagulant and anti-inflammatory effects of the TM extracellular domain. 

Analyses of enriched cell media by ELISA showed that PDGF-BB stimulation did not change the amount of TM released. This result suggests that little TM is released and that little of it is released into the ECM. 

To confirm TM expression in the corneal and scleral stroma, we performed an immunohistochemical analysis of TM. TM expression was observed in the inflamed scleral stroma, whereas in inflamed corneal stroma, both TM-expressing and nonexpressing samples were identified. TM was not expressed in malignant melanoma tissue. These results suggest that an increase in PDGF-BB accompanies inflammation, with a greater increase in TM in the sclera than in the corneal stroma. However, TM expression has also been reported in the corneal stroma ²¹ ; thus, further study is necessary. As just mentioned, TM has direct and indirect anti-inflammatory effects. ¹⁸ PDGF-BB may be associated with TM and the control of inflammation. The difference in TM expression revealed by our histologic and biological investigations may have been caused by thresholds, the PDGF-BB concentration used in this study, effects from existing tissues, or other factors. 

Because TM is located on the cell surface, it may be involved in cell–cell interactions. In addition, soluble TM is released into the blood in heart disease or sepsis and is considered an index of endothelial injury. ^24,25 Released TM also reverses inflammation. Cellular injury may also occur in inflamed ocular tissues, and TM may be released into the extracellular space. Our PCR results showed that the addition of TM to the scleral stromal cell strain increased IL-18 mRNA expression. IL-18-enhanced T-cell and natural-killer cell (NK cell) cytotoxicity directly induces NK cell IFN-γ production and promotes IFN-γ, TNF-α, and GM-CSF production by Th1 clones. ^26,27 Immunohistochemical analysis of IL-18 and its receptor in inflamed tissues showed that both proteins were expressed in the cornea and sclera. These results suggest that TM is not limited to anti-inflammatory effects, but also controls inflammation (Fig. 9). 

Figure 9.

 

Various cell types released PDGF-BB under inflammatory conditions. TM stimulated by PDGF-BB inactivates inflammation, with or without thrombin and protein C. However, IL-18 induced by TM had inflammatory effects, suggesting that TM has various effects on the control of inflammation. Bar, 50 μm.

Further evaluation of the factors controlling TM expression and the factors involved in the inflammatory control of TM is needed, to support the development of novel anti-inflammatory therapies in ophthalmology. 

Although corneal and scleral stromal ECM structures have been investigated, ²⁸ stromal cell characteristics have rarely been compared. This study demonstrates the possibility that fibroblasts are tissue-specific and shows the effects of TM in ocular tissues. 

Footnotes

 Disclosure: H. Namba, None; Y. Kashiwagi, None; K. Nishitsuka, None; H. Takamura, None; T. Yamamoto, None; H. Yamashita, None