Release of Soluble Tumor Necrosis Factor Receptor 1 from Corneal Epithelium by TNF-α–Converting Enzyme-Dependent Ectodomain Shedding

Abstract

purpose. An involvement of tumor necrosis factor-α–converting enzyme (TACE)–dependent ectodomain shedding in the release of soluble tumor necrosis factor receptor 1 (sTNFR1) from corneal epithelium was evaluated.

methods. In vitro experiments were performed using the human SV40-transformed human corneal epithelial cell (HCEC) line. Ectodomain shedding was stimulated by phorbol myristate acetate (PMA, 3 μM) or peptidoglycan (PGN, 100 μg/mL), with or without TACE inhibition, using TNF-α processing inhibitor-1 (TAPI-1, 250 μg/mL) or tissue inhibitor of metalloproteinase-3 (TIMP-3, 2 μg/mL) by addition to the HCEC culture medium. The concentrations of sTNFR1 in culture medium were analyzed by enzyme-linked immunosorbent assay. To induce an inflammatory response in the ocular surface, corneal alkali burn of BALB/c mice was made from a filter paper dipped in 1 N NaOH solution. TNFR1 expression in corneal and conjunctival epithelia was evaluated by immunohistochemistry 28 days after wounding.

results. In HCEC culture medium, sTNFR1 release was significantly increased by the addition of PMA (t-test, P < 0.01) or PGN (P < 0.01). The increased release of sTNFR1 was significantly inhibited by the addition of TAPI-1 or TIMP-3, indicating the possibility of TACE-dependent ectodomain shedding of TNFR1. In the corneal alkali burn model, TNFR1 was expressed in corneal and conjunctival epithelia.

conclusions. TACE-dependent ectodomain shedding of sTNFR1 was recognized in corneal epithelium. In the inflamed ocular surface, TNFR1 was expressed in the corneal and conjunctival epithelia after alkali burn treatment. sTNFR1, released from the ocular surface, may play an anti-inflammatory role in the inflammatory condition.

Tumor necrosis factor (TNF)-α plays a key role in inflammation, immunity, and apoptosis. Although TNF-α can transmit its signal through two receptors, TNF receptor 1 (TNFR1) and 2 (TNFR2), most of the biological effects of TNF-α are mediated by TNFR1 signaling. ¹ TNFR1 is expressed ubiquitously on almost all cell types, whereas TNFR2 expression is limited to hematopoietic and endothelial cells. TNFR1 mediates various activities of TNF-α, including cytotoxicity, proliferation, and apoptosis. ² Therefore, TNFR1 is considered to play an important role in the regulation of various inflammatory conditions. 

TNFR1 can be released from cells to the extracellular compartment by proteolytic cleavage of the TNFR1 extracellular domain. This process, called ectodomain shedding, generates soluble TNF-binding protein that inhibits TNF bioactivity. ^{3 4} The resultant decrease in the number of receptor molecules on the cell surface occurs as a consequence of ectodomain shedding and serves to desensitize the cells to TNF action. ⁵ In addition, the pool of soluble TNFR1 in the extracellular compartment could function as physiological attenuators of TNF bioactivity, which blocks the ligand to bind to the cell surface receptor. In humans, soluble TNFR1 is constitutively released in the general circulation, ⁶ and its level increases in the course of various diseases. ^{7 8} Elevated plasma level of soluble TNFR1 is suggested to be a biomarker for morbidity and mortality in patients with acute inflammatory disease. ⁹ Additionally, impaired TNFR1 ectodomain shedding from the cell surface has recently been proposed to be responsible for an autoinflammatory disease, termed TNFR1-associated periodic syndromes (TRAPS). ¹⁰ Therefore, these evidences suggest the pivotal role of ectodomain shedding of TNFR1 in inflammatory conditions. 

TNF-α–converting enzyme (TACE, ADAM17), a member of a disintegrin and metalloproteinase (ADAM) family, has been proposed to be a TNFR1 sheddase. ^{3 4} TACE is proposed to play a central role in ectodomain shedding, and its substrates are TNF-α, TNFR1, TNFR2, epidermal growth factor receptor ligand family, IL-6 receptor, CD40, p75NTR, TrkA, and Notch. ^{11 12 13 14 15 16 17}  

Previously, we showed that TACE and TNFR1 were expressed in corneal stromal cells during corneal wound healing and that soluble TNFR1 was generated from both fibroblast and macrophage by TACE-dependent ectodomain shedding. ¹⁸ In this study, we focused on the corneal epithelium for its ability to produce soluble TNFR1 and demonstrated that cultured corneal epithelium cells were able to produce soluble TNFR1 by TACE-dependent ectodomain shedding. 

Materials and Methods

Animal experiments were performed under the recommendation of the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the Animal Care and Use Committee at Nihon University School of Medicine. We followed the tenets of the Declaration of Helsinki for procedures involving human tissue. 

Soluble TNFR1 in Cultured Corneal Epithelial Cells

To elucidate TACE-dependent ectodomain shedding of TNFR1 in cultured corneal epithelial cells, the SV40-transformed human corneal epithelial cell (HCEC) line was purchased from American Type Culture Collection (ATCC, Manassas, VA). This cell line was seeded into 24-well plates and was cultured in DMEM (Invitrogen, Carlsbad, CA) supplemented with 10% heat-inactivated fetal bovine serum, 2 mM l-glutamine, and antibiotics at 37C°, in 5% CO₂ incubator (CPE-1201; Hirasawa Works, Tokyo, Japan). After reaching 80% subconfluence, each well was pretreated for 4 hours with phorbol myristate acetate (PMA, 3 μM; Sigma-Aldrich, St. Louis, MO) or for 24 hours with peptidoglycan (PGN, 100 μg/mL; Sigma-Aldrich) dissolved in serum-free DMEM. In experiments using TACE inhibitor, TNF-α–processing inhibitor-1 (TAPI-1, 250 ng/mL; EMD Chemicals, Darmstadt, Germany; 30 minutes) or tissue inhibitor of metalloproteinase-3 (TIMP-3, 2 μg/mL, 4 hours; R&D Systems, Minneapolis, MN) was dissolved in serum-free DMEM and added 30 minutes before the addition of PMA or PGN as described. After treatment, the culture medium from each well was sampled and analyzed by enzyme-linked immunosorbent assay (ELISA) for soluble TNFR1 (R&D Systems). Those experiments were repeated three times, and statistical analysis was performed with the Student’s t-test. 

Surgical Procedure

Ten eyes of five BALB/c male mice were examined in the following experiments. The mice were anesthetized with intramuscular injection of xylazine (Selactar; Bayer HealthCare, Leverkusen, Germany). Oxybuprocaine 0.4% eyedrop (Benoxil, Santen Pharmaceutical, Osaka, Japan) was used for local anesthesia. The mice were euthanatized by cervical dislocation. Corneal alkali burns were made by exposing a filter paper of 1.5-mm diameter for 60 seconds. The filter was presoaked in 1 N NaOH. After the filter paper was removed from the cornea, the eye was rinsed by sterilized saline and an antibiotic ointment (Tarivid; Santen Pharmaceutical) was instilled. After the treatment, eye globe was enucleated on day 28 (three eyes, three mice). Nontreated eyes (seven eyes, two mice) were also enucleated and used as controls. 

Immunohistochemistry

Enucleated eyes were embedded (Tissue-Tek; Sakura Finetek, Tokyo, Japan), and a 7-μm-thick section was made using a cryostat (HM505; Microm, Walldorf, Germany). These thin sections were stained by anti–TNFR1 antibody (dilution 1:200; Santa Cruz Biotechnology, Santa Cruz, CA) and then stained by the indirect enzyme-antibody method (avidin-biotin complex). The section was examined by light microscopy (BH-2; Olympus, Tokyo, Japan). 

Results

TACE-Dependent Ectodomain Shedding of TNFR1 in Cultured Corneal Epithelial Cells

We examined the release of soluble TNFR1 from cultured corneal epithelial cells by TACE-dependent ectodomain shedding in the presence of the metalloproteinase stimulator PMA or PGN, with or without the TACE inhibitor TAPI-1 or TIMP-3. PMA is reported to upregulate the activity of ectodomain shedding through ADAM family activation in vitro, ^{19 20} and PGN is reported to upregulate the activity of metalloproteinase. ²¹ TAPI-1 is reported to inhibit ADAM activity, and TIMP-3 is the internal inhibitor of TACE. ⁴  

In the culture medium of HCECs, soluble TNFR1 release was significantly increased by the addition of PMA (P < 0.01). This increased release of soluble TNFR1 was significantly inhibited by the addition of TAPI-1 (P < 0.01) or TIMP-3 (P < 0.01) with PMA (Fig. 1) . Furthermore, soluble TNFR1 release was significantly increased by the addition of PGN (P < 0.01) and was inhibited by a mixture of TAPI-1 (P < 0.01) or TIMP-3 (P < 0.01) with PGN (Fig. 2) . These results indicate the possible involvement of TACE in this soluble TNFR1-releasing process. 

TNFR1 Expression in Alkali-Burned Cornea

Twenty-eight days after the alkali treatment, TNFR1 expression was detected in corneal and conjunctival epithelial layers. In addition, positive immunoreactivity was noted in corneal and conjunctival stroma. In nontreated eyes, no immunoreactivity of TNFR1 was noted (Fig. 3) . 

Discussion

In this study, we showed that soluble TNFR1 release was significantly increased in HCEC culture medium by the addition of the metalloproteinase stimulator PMA or PGN. The increased release of soluble TNFR1 was inhibited by the addition of the TACE inhibitor TAPI-1 or TIMP-3. These results indicate the possibility that ectodomain shedding of TNFR1 is caused by TACE. In a mouse model, TNFR1 was expressed in corneal and conjunctival epithelia after alkali burn. Although the presence of TNFR1 in corneal and conjunctival epithelia has been described, ^{22 23} our results showed TNFR1 ectodomain shedding in corneal epithelial cells. 

In our in vitro experiments, soluble TNFR1 release was significantly upregulated by adding PMA to the culture medium. PMA is frequently used as an activator of ADAM-dependent ectodomain shedding. Soluble TNFR1 concentration in HCEC culture medium was significantly upregulated by the addition of PMA, and this upregulation was inhibited by the addition of TAPI-1 or TIMP-3. As mentioned, TAPI-1 and TIMP-3 are frequently used as TACE inhibitors. Therefore, it can be speculated that soluble TNFR1 release to HCEC culture medium is caused by TACE-dependent ectodomain shedding. Additionally, we confirmed that PGN holds its ability to activate TACE-dependent ectodomain shedding. Although its ability to activate metalloproteinase has been reported, ²¹ we were able to link PGN with ectodomain shedding. This evidence is interesting because PGN—a cell wall component of Gram-positive bacteria—is well known not only as a virulent factor that causes inflammation but also as an eliciting factor in most clinical manifestations of Gram-positive infection. ^{24 25} Therefore, the relationship between PGN and corneal epithelium regarding ectodomain shedding of various proteins may play an important role in corneal inflammation and infection. 

In the alkali-burned eye, TNFR1 expression was upregulated in corneal and conjunctival epithelia, indicating that soluble TNFR1 is produced during inflammatory conditions in the ocular surface. Combined with HCEC-culture experiments, corneal epithelium is capable of producing soluble TNFR1. Other investigators showed that soluble TNFR1 could be released from conjunctival epithelium, ²² and we confirmed their results by using a human conjunctival epithelial cell line (CCL-20.2; ATCC). A significant increase of soluble TNFR1 was noted after the addition of PGN (P < 0.01), and this upregulation was inhibited by the mixture of TAPI-1 (P < 0.01) or TIMP-3 (P < 0.05) with PGN (data not shown). We hypothesize that soluble TNFR1 is produced to stabilize inflammatory conditions in the ocular surface. In our previous study on corneal alkali burn model, increased TACE expression was detected in corneal stroma after the acute phase of wound healing. ¹⁸ In the present study, TNFR1 ectodomain shedding was promoted by PGN, which is a constituent of bacterium. Therefore, TNFR1/TACE expression in chronic inflammation or infectious conditions of ocular surface should be investigated. Furthermore, investigations of soluble TNFR1 in various corneal diseases, including tear analysis, are needed to elucidate its role in the pathophysiology of ocular surface disorders. 

Our study was limited in that finding were not confirmed by the activated form of TACE, which is active originally on the cellular membrane. Given that TACE is primarily a membrane-type metalloproteinase, it was technically difficult to use an activated form that is active on the cellular membrane. Therefore, several studies were used for TACE knockout cell, ²⁶ TACE-transfected cell, ²⁷ or dominant negative-TACE cell. ²⁸ However, many studies were performed using PMA, lipopolysaccharide, or another metalloproteinase stimulator to stimulate TACE and TAPI or another metalloproteinase inhibitor to inhibit TACE. ^{12 15 19 20 29 30 31 32} Therefore, it cannot be concluded that TACE was the sole candidate proteinase to cause TNFR1 ectodomain shedding in the present study. It can be said, however, that TACE plays an important role in ectodomain shedding. Regarding TACE stimulation and inhibition, our experiments were performed using single concentrations based on our preliminary study. TACE stimulation and inhibition should be performed in various concentrations in further studies. In addition, SV40-transformed HCEC was the only cell line used in this study. Additional studies using the primary cell line should be performed, as suggested by Cook et al. ²²  

The ADAM family was originally identified in cellular kinetics of cell fusion and proteolysis. The discovery of TACE (ADAM17) and other members of the ADAM family elucidated their roles in ectodomain shedding of various cytokines, growth factors, and receptors that are involved in inflammation, cell proliferation, and cell death. ^{3 4 10 11 12 13 14 15 16 17} Therefore, it is no longer thought that the ADAM family is composed of matrix-degrading proteins; rather, ectodomain shedding caused by the ADAM family should be taken into account as an important posttranscriptional mechanism with the capability of regulating a variety of biological process. Although substrates of members of the ADAM family overlap, the representative substrates of TACE (ADAM17) are TNF-α, TNFR1, TNFR2, IL-6R, and EGF receptor ligand family, and the substrates of ADAM10 are heparin-binding EGF, IL-6R, TNF- α, CXCL16, and CD23. Within these proteins, some substrates—such as TNFR2 (soluble TNFR2, etanercept), ³³ IL-6R (anti-IL-6R antibody, tocilizumab), ³⁴ and CD23 (anti-CD23 antibody, lumiliximab)—also have attracted attention in the field of molecular target therapy. ³⁵ Therefore, it could be said that the comprehension or regulation of ectodomain shedding in various disorders might be the key factor for the pathophysiological process. 

 Supported by Japan Society for the Promotion of Science, Grant-in-Aid for Scientific Research (C) 18591937 (MS), and The Ministry of Education, Culture, Sports, Science and Technology, Grant-in-Aid for Young Scientists (B) 20791284 (TS).

 Submitted for publication August 4, 2008; revised December 12, 2008, and March 10, 2009; accepted June 25, 2009.

 Disclosure: T. Sakimoto, None; A. Yamada, None; M. Sawa, None

 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.

 

Figure 1.

 

Relationship between PMA and TNFR1 ectodomain shedding through TACE activation in vitro. Ectodomain shedding of TNFR1 is stimulated by PMA, activator of ectodomain shedding, with or without TACE inhibitor, TAPI-1, or TIMP-3. PMA stimulation significantly upregulates soluble TNFR1 release in HCEC culture medium. This upregulated release of soluble TNFR1 is significantly inhibited by TAPI-1 or TIMP-3. Those experiments were repeated three times, and statistical analysis was performed with the use of the unpaired Student’s t-test. *P < 0.01.

Figure 2.

 

PGN induces ectodomain shedding of TNFR1 through TACE activation in vitro. Ectodomain shedding of TNFR1 is stimulated by PGN in the presence or absence of TAPI-1 or TIMP-3. PGN stimulation significantly upregulates soluble TNFR1 release in HCEC culture medium. This upregulated release of soluble TNFR1 is significantly inhibited by TAPI-1 or TIMP-3. *P < 0.01.

Figure 3.

 

TNFR1 expression in alkali-burned cornea. TNFR1 expression in alkali-burned corneal (A) and conjunctival (B) epithelial layer. TNFR1 staining is not observed in the nontreated eye. TNFR1 immunoreactivity is observed at day 28 in corneal (C) and conjunctival (D) epithelium. Scale bars, 40 μm.

The authors thank Akiko Ishimori (Department of Ophthalmology, Nihon University School of Medicine) for expert technical assistance.