Comparison of Aqueous Levels of Inflammatory Mediators Between Toxic Anterior Segment Syndrome and Endotoxin-Induced Uveitis Animal Models

Abstract

Purpose.: To compare clinical findings and the aqueous levels of inflammatory mediators between toxic anterior segment syndrome (TASS) and endotoxin-induced uveitis (EIU) animal models and to evaluate the efficacy of systemic steroid pretreatment in both animal models.

Methods.: Rats were used in this study. Ortho-phthalaldehyde solution was injected into the anterior chamber to produce TASS (n = 30), and lipopolysaccharide was injected into one hind footpad to produce EIU (n = 30). Clinical findings were evaluated under slit-lamp examination, and the aqueous levels of prostaglandin E2 (PGE2) and tumor necrosis factor-α (TNF-α) were measured 24 hours after these procedures. Twelve of the rats in each animal model were pretreated with intraperitoneal injection of dexamethasone for 4 days before the development of TASS and EIU.

Results.: Corneal haze scores were significantly higher for TASS than EIU, but clinical scores for anterior uveitis were not different between the two animal models. Although aqueous levels of PGE2 were markedly increased in both animal models, PGE2 levels were significantly higher for TASS than for EIU. However, an increase in aqueous levels of TNF-α was observed only in EIU. Dexamethasone pretreatment reduced the corneal haze score and clinical score for anterior uveitis in both animal models and inhibited the increase in aqueous levels of PGE2 and TNF-α.

Conclusions.: Prostaglandin E2 and TNF-α in aqueous humor seem to be regulated differently in animal models of TASS and EIU. However, dexamethasone pretreatment improved the clinical findings and inhibited the increases in PGE2 and TNF-α in both animal models.

Introduction

Toxic anterior segment syndrome (TASS) is characterized by sterile inflammation following anterior segment surgery due to entry of a noninfectious substance into the anterior segment.¹ The inflammatory process of TASS begins within 24 hours after cataract surgery following toxic cellular and extracellular damage of intraocular tissue caused by the cytotoxic chemicals within the anterior segment. 

Endotoxin-induced uveitis (EIU) is a well-known animal model for evaluating the efficacy of treatment for ocular inflammation.^2–5 Prostaglandin E2 (PGE2) and tumor necrosis factor-alpha (TNF-α) play essential roles in EIU.^6–8 Previous studies have shown the effect of systemic steroids on reduction of the concentration of inflammatory mediators in the EIU animal model.^3,9 

In contrast with EIU,¹⁰ the ocular inflammation in TASS probably results from a direct effect of the cytotoxic chemicals on intraocular tissue, because dispersive viscoelastics were shown to partially protect corneal endothelial cells from toxic disinfectant damage in a previous TASS animal model study.¹¹ However, the subsequent inflammatory response and the role of related proinflammatory cytokines are still not well understood. Sagoo et al.¹² showed that proinflammatory cytokines induce apoptosis of corneal endothelial cells through nitric oxide. Thus, this study raised the following questions: (1) Are inflammatory mediators such as PGE2 and TNF-α expressed in the TASS animal model? (2) What is the difference in the levels of PGE2 and TNF-α in the aqueous humor between TASS and EIU animal models? (3) Could systemic dexamethasone (Dexa) pretreatment improve clinical findings and inhibit the increase of these inflammatory mediators in these animal models? To answer these questions, this study compared the clinical findings and the aqueous levels of PGE2 and TNF-α in the TASS and EIU animal models and evaluated the effects of systemic steroid pretreatment on the clinical findings and the aqueous levels of these inflammatory mediators in both animal models. 

Materials and Methods

Seventy-eight inbred male Lewis rats, each weighing 150 to 200 g (6–8 weeks old), were used in this study. Rats were randomly divided into five groups (see below). All procedures conformed to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research.¹³ Institutional Review Board approval was obtained from Korea University Guro Hospital, Seoul, South Korea, for this study. 

Animal Models and Subgroups

Two animal models, TASS and EIU, were established, and each animal model was divided into two groups according to whether rats were pretreated with systemic Dexa or not: a TASS group (n = 18), in which TASS rats did not receive Dexa pretreatment; a TASS-Dexa group (n = 12), in which TASS rats were pretreated with Dexa; an EIU group (n = 18), in which EIU rats did not receive Dexa pretreatment; and an EIU-Dexa (n = 12) group, in which EIU rats were pretreated with Dexa. Naïve, untreated rats were used as a control group (n = 18). 

Xylazine hydrochloride (Rompun 2%, 1 mg/100 g body weight; Bayer, Leverkusen, Germany) and Zoletil (30 mg/100 g body weight; Virbac Laboratories, Carros, France) were injected intramuscularly for general anesthesia. Additional topical anesthesia was administered with 0.5% proparacaine hydrochloride (Alcaine; Alcon Laboratories, Fort Worth, TX, USA). 

To establish an animal model for TASS, a solution of 0.55% ortho-phthalaldehyde (Cidex OPA; Advanced Sterilization Products, Irvine, CA, USA) was diluted to 0.14% with balanced salt solution,¹¹ and then 0.01 mL was injected into the anterior chamber of the right eyes of rats in the TASS and TASS-Dexa groups under general and topical anesthesia using a 30-gauge needle and a 1-mL syringe. For the establishment of an animal model for EIU, 200 μg lipopolysaccharide (LPS, Salmonella typhimurium; Sigma Chemical Company, St. Louis, MO, USA) diluted in 0.1 mL phosphate-buffered saline (PBS) was injected into one hind footpad of rats in the EIU and EIU-Dexa groups.^2,14 The right eyes of rats in the EIU, EIU-Dexa, and control groups were used for this study. 

Prior to the development of TASS and EIU, animals in the TASS-Dexa and EIU-Dexa groups had received once-daily intraperitoneal injection of Dexa (5 mg/kg body weight) for 4 days in order to evaluate the effect of steroid pretreatment on the clinical findings and the expression of PGE2 and TNF-α in aqueous humor. 

Infiltrating Cell Counting and Aqueous Humor Cytology

After general and topical anesthesia, six aqueous humor samples (10 μL) were collected from the TASS, EIU, and control groups by inserting a 30-gauge needle attached to a 1-mL syringe into the anterior chamber and gently aspirating under a microscope 24 hours after the injection of either ortho-phthalaldehyde solution or lipopolysaccharide. Of 10 μL aqueous samples, 8 μL was used for conducting cytology and 2 μL was used for counting infiltrating cells in the aqueous humor. For cell counting, the aqueous humor sample was suspended in 5 μL trypan blue stain solution, and the cells were manually counted using a hemocytometer (Marienfeld-Superior, Lauda-Königshofen, Germany) under a light microscope. The number of cells per field was obtained by averaging the results of four fields from each sample. 

For cytology, the aqueous humor sample on ice was immediately sent to the department of laboratory medicine. As the amount of one aqueous humor sample was not enough to conduct cytology, two samples were mixed together with 100 μL PBS. These three diluted samples for each group were cytocentrifuged at 263g for 5 minutes using the Cytospin (Cellspin; Hanil Science Industrial Co., Incheon, Korea). Cellular morphology in each slide was evaluated under a high-power field by a laboratory medicine doctor (Young E. Koh, Department of Laboratory Medicine, Korea University College of Medicine). Distribution of each cell type was expressed as a percentage. Cell distribution analysis was not conducted in the control group because only rare cells were observed in the aqueous humor sample from the control group. 

Clinical Evaluation of TASS and EIU Models

The corneal haze score and clinical score for anterior uveitis were evaluated under slit-lamp examination 24 hours after establishing each animal model. Corneal haze score was graded according to the Fantes classification (grades 0–4)¹⁵: grade 0, totally clear cornea; grade 0.5, trace haze cornea; grade 1, minimal haze cornea; grade 2, mild haze easily visible with direct focal slit illumination; grade 3, moderate opacity obscuring iris details; and grade 4, severe opacity blocking the view of anterior chamber structures. The clinical score for anterior uveitis was graded according to a previously defined scoring system (grades 0–4)^16–18: grade 0, no inflammatory reaction; grade 1, discrete inflammation of the iris and conjunctival vessels; grade 2, dilation of the iris and conjunctival vessels with moderate flare in the anterior chamber; grade 3, hyperemia in the iris associated with the Tyndall effect in the anterior chamber; and grade 4, same clinical signs as for 3 plus the presence of fibrin or synechiae. 

Enzyme-Linked Immunosorbent Assay (ELISA)

After clinical evaluation of both animal models under general and topical anesthesia, an aqueous humor sample (10 μL) was collected from each right eye of all rats using a 30-gauge needle attached to a 1-mL syringe under a microscope. 

Inflammatory mediators in the aqueous humor sample (10 μL) of five groups were assessed with the rat Biotrak PGE2 Enzyme Immunoassay system (EIA; GE Healthcare Ltd., Little Chalfont, UK) and rat TNF-α Immunoassay (R&D Systems, Minneapolis, MN, USA). All measurements were conducted according to the manufacturer's protocol using a microplate spectrophotometer (Spectramax Plus 384; Molecular Devices, Sunnyvale, CA, USA). 

Statistical Analyses

Statistical analyses were performed using the Kruskal-Wallis test and Mann-Whitney U test in SPSS version 12.0 (SPSS, Inc., Chicago, IL, USA). Values were expressed as median and interquartile range (IQR). Results were considered statistically significant if the P value was less than 0.05. Bonferroni's corrected post hoc Mann-Whitney test for three-group comparisons was employed (P < 0.017 significant). 

Results

Infiltrating Cell Counting and Aqueous Humor Cytology

The median cell count in the aqueous humor of the TASS group (IQR) was 36.5 × 10⁵/mL (29.3 × 10⁵–58.0 × 10⁵) and that of the EIU group was 19.5 × 10⁵/mL (11.5 × 10⁵–40.5 × 10⁵). Infiltrating cell counts of the TASS and EIU groups were both significantly greater than that of the control group (P = 0.002 and P = 0.009, respectively). However, there was no significant difference in the cell counts between the TASS and EIU groups (P = 0.093) (Table 1). 

Table 1

 

Comparison of the Inflammatory Cell Counts in Aqueous Humor Among Control, Toxic Anterior Segment Syndrome, and Endotoxin-Induced Uveitis Animal Models

Parameter	Infiltrating Cells/mL (1 × 105)
Groups
TASS	36.5 (29.3–58.0)
EIU	19.5 (11.5–40.5)
Control	0.0 (0.0–4.3)
P value
TASS vs. EIU vs. control	0.003*
TASS vs. EIU	0.093†
TASS vs. control	0.002†
EIU vs. control	0.009†

 TASS rats received an injection of 0.01 mL 0.14% ortho-phthalaldehyde into the anterior chamber of the right eye. Endotoxin-induced uveitis rats received an injection of 0.1 mL lipopolysaccharide (200 μg/0.1 mL) into one hind footpad. The infiltrating cells in the aqueous humor of each group were manually counted using a hemocytometer 24 hours after establishing each animal model. Data are expressed as median and interquartile range of six rats.

*   Kruskal-Wallis test comparing three groups.

†   Bonferroni corrected post hoc Mann-Whitney U tests for between-group comparisons (P < 0.017 significant).

In the aqueous humor cytology, polymorphonuclear neutrophil was the most abundant (78.7%–90.9% in the TASS group and 78.7%–93.3% in the EIU group) type, followed by lymphocytes (4.5%–12.9% in the TASS group and 6.7%–7.9% in the EIU group) and monocytes (0.0%–8.4% in the TASS group and 0.0%–8.8% in the EIU group). The distribution of cell types was similar between the two groups (Fig. 1). 

Figure 1

 

Aqueous humor cytology shows an abundance of polymorphonuclear neutrophils in aqueous humor of both toxic anterior segment syndrome (TASS) and endotoxin-induced uveitis (EIU) animal models. (A) TASS animal model. (B) EIU animal model.

Clinical Evaluation of TASS and EIU Models

The medians of both corneal haze score and clinical score for anterior uveitis in the TASS and EIU groups were significantly higher than those of the control group. The median corneal haze score of the TASS group (IQR) was 3.5 (2.8–4.0), and that of the EIU group was 0.5 (0.5–1.0). There was a significant difference in the corneal haze score between the TASS and EIU groups (P = 0.002). However, there was no significant difference in the clinical score for anterior uveitis between the TASS and EIU groups (P = 0.127) (3.0 [2.5–4.0] in the TASS group; 2.0 [2.0–3.0] in the EIU group; Fig. 2; Table 2). 

Figure 2

 

Anterior segment photographs of a rat cornea obtained by stereomicroscopy 24 hours after establishing each animal model. (A) Normal control. (B) Toxic anterior segment syndrome. (C) Endotoxin-induced uveitis.

Table 2

 

Comparison of Corneal Haze Score and Clinical Score for Anterior Uveitis Among Control, Toxic Anterior Segment Syndrome, and Endotoxin-Induced Uveitis Animal Models

Parameter	Corneal Haze Score	Clinical Score for Anterior Uveitis
Group
TASS	3.5 (2.8–4.0)	3.0 (2.5–4.0)
EIU	0.5 (0.5–1.0)	2.0 (2.0–3.0)
Control	0.0 (0.0–0.0)	0.0 (0.0–0.0)
P value
TASS vs. EIU vs. control	<0.001*	0.001*
TASS vs. EIU	0.003†	0.127†
TASS vs. control	0.002†	0.002†
EIU vs. control	0.002†	0.002†

 Corneal haze score and clinical score for anterior uveitis were evaluated under slit-lamp examination 24 hours after establishing each animal model. Corneal haze score was graded according to the Fantes classification (grades 0–4). Clinical score for anterior uveitis was graded from 0 to 4. Data are expressed as median and interquartile range of six rats. Experimental groups are the same as in Table 1.

*   Kruskal-Wallis test comparing three groups.

†   Bonferroni corrected post hoc Mann-Whitney U tests for between-group comparisons (P < 0.017 significant).

Corneal haze score and clinical score for anterior uveitis of the TASS-Dexa group (1.5 [1.0–2.3] and 1.5 [1.0–2.0], respectively) were significantly lower than those of the TASS group (P = 0.013 and P = 0.025, respectively). Corneal haze score and clinical score for anterior uveitis of the EIU-Dexa group (0.3 [0.0–0.5] and 1.0 [1.0–2.0], respectively) were also significantly decreased compared to those of the EIU group (P = 0.030 and P = 0.014, respectively; Table 3). 

Table 3

 

The Effects of Intraperitoneal Dexamethasone Injection on Corneal Haze Score and Clinical Score for Anterior Uveitis in Toxic Anterior Segment Syndrome and Endotoxin-Induced Uveitis Animal Models

	TASS	TASS-Dexa	Value*	EIU	EIU-Dexa	Value†
Corneal haze score	3.5 (2.8–4.0)	1.5 (1.0–2.3)	0.013	0.5 (0.5–1.0)	0.3 (0.0–0.5)	0.030
Clinical score for anterior uveitis	3.0 (2.5–4.0)	1.5 (1.0–2.0)	0.025	2.0 (2.0–3.0)	1.0 (1.0–2.0)	0.014

 TASS-Dexa and EIU-Dexa animals received once-daily intraperitoneal injection of dexamethasone (5 mg/kg body weight) for 4 days before the establishment of each animal model. Data are expressed as median and interquartile range of six rats. Experimental groups are the same as in Table 1.

*   Mann-Whitney U test between the TASS and TASS-Dexa groups.

†   Mann-Whitney U test between the EIU and EIU-Dexa groups.

Enzyme-Linked Immunosorbent Assay

The median aqueous PGE2 concentration in the TASS group (IQR), 210.9 ng/mL (157.0–271.1), and the EIU group, 26.9 ng/mL (19.2–45.2), was significantly higher than that of the control group, 0.0 ng/mL (0.0–6.2). In addition, PGE2 level was significantly higher in the TASS group than in the EIU group (P = 0.004). The median aqueous TNF-α concentration in the EIU group (IQR), 87.4 pg/mL (39.0–129.6), was significantly higher than that of the control group, 0.0 pg/mL (0.0–2.8). However, there was no significant difference in the aqueous TNF-α concentration between the TASS and control groups (Table 4). 

Table 4

 

Comparison of the Expression of Prostaglandin E2 and Tumor Necrosis Factor-Alpha in Aqueous Humor Among Control, Toxic Anterior Segment Syndrome, and Endotoxin-Induced Uveitis Animal Models

Parameter	PGE2, ng/mL	TNF-α, pg/mL
Groups
TASS	210.9 (157.0–271.1)	0.0 (0.0–19.9)
EIU	26.9 (19.2–45.2)	87.4 (39.0–129.6)
Control	0.0 (0.0–6.2)	0.0 (0.0–2.8)
P value
TASS vs. EIU vs. control	<0.001*	0.002*
TASS vs. EIU	0.004†	0.003†
TASS vs. control	0.003†	0.703†
EIU vs. control	0.003†	0.003†

 The levels of PGE2 and TNF-α were identified in aqueous humor samples using a commercially available ELISA kit 24 hours after establishing each animal model. Data are expressed as median and interquartile range of six rats. Experimental groups are the same as in Table 1.

*  Kruskal-Wallis test comparing three groups.

†   Bonferroni corrected post hoc Mann-Whitney U tests for between-group comparisons (P < 0.017 significant).

The TASS-Dexa (13.9 ng/mL [0.0–17.4]) and EIU-Dexa groups (4.2 ng/mL [0.0–12.3]) had significantly lower PGE2 levels in their aqueous samples compared to the TASS and EIU groups, respectively (P = 0.004 and P = 0.004, respectively). The EIU-Dexa group (4.5 pg/mL [0.0–15.4]) had significantly lower TNF-α levels in their aqueous samples compared to the EIU group (P = 0.005; Fig. 3). 

Figure 3

 

The effect of intraperitoneal dexamethasone injection on the expression of prostaglandin E2 (PGE2) and tumor necrosis factor-alpha (TNF-α) in aqueous humor in TASS and EIU animal models. TASS-Dexa and EIU-Dexa animals received once-daily intraperitoneal injection of dexamethasone (5 mg/kg body weight) for 4 days before the establishment of each animal model. (A, B) Expression of PGE2 in aqueous humor. (C, D) Expression of TNF-α in aqueous humor. Asterisk indicates a P value < 0.05 by the Mann-Whitney U test.

Discussion

Typical clinical findings for TASS are diffuse corneal edema and severe anterior chamber inflammation, which often results in hypopyon or fibrin formation.¹⁹ Although many different causes such as preservatives, detergents, and denatured ophthalmic viscosurgical devices have been reported to induce TASS, the main treatment is limited to use of intense topical corticosteroid to reduce further damage from secondary inflammatory response.^1,19–22 However, the subsequent inflammatory response and the role of related proinflammatory cytokines in TASS are still not understood. The EIU animal model is a well-established model to evaluate the inflammatory response in the anterior chamber.^2–5 Therefore, we thought that although the mechanisms of anterior chamber inflammation are different between the TASS and EIU animal models, comparison between the two animal models would be a good way to understand the secondary inflammatory response and related proinflammatory cytokines in the TASS animal model. 

In this study, we counted the number of infiltrating cells in the aqueous humor of both animal models. The cell counts of the TASS and EIU groups were significantly greater than that of the control group. However, there was no significant difference between the TASS and EIU groups. We also performed aqueous humor cytology to evaluate the infiltrating cell types in the anterior chamber of both groups. Polymorphonuclear neutrophils were the most abundant type, followed by lymphocytes and monocytes in the anterior chamber of both TASS and EIU animal models. The distribution of cell types was similar between the two groups. These findings indicated that Dexa pretreatment was effective not only in the EIU animal model, but also in the TASS animal model. Dexa pretreatment reduced corneal haze score and clinical score for anterior uveitis in the TASS animal model. Although these values were not restored to the level of normal control, steroid treatment seemed to prevent secondary damage from the inflammatory response in the anterior chamber. 

This study compared the levels of PGE2 and TNF-α in aqueous humor in the TASS and EIU animal models and evaluated the effect of Dexa on the levels of these inflammatory mediators. The results demonstrated that aqueous PGE2 was elevated in both the TASS and EIU animal models, but TNF-α was elevated only in the EIU animal model. In addition, the aqueous PGE2 level was shown to be significantly higher in the TASS animal model than in the EIU animal model. Dexa not only blocked the increase in PGE2 level in both animal models, but also attenuated the increase in TNF-α in the EIU animal model. 

TASS is characterized by limbus-to-limbus corneal edema and anterior segment inflammation that appear to be caused by widespread corneal endothelial damage and breakdown of the blood–aqueous barrier, respectively.¹ In a previous TASS animal model study, Cidex OPA disinfectant solution caused devastating corneal endothelial damage, and dispersive viscoelastics had a partial protective effect on corneal endothelial cell damage induced by the disinfectant.¹¹ Thus, direct cytotoxic effects caused by disinfectant are thought to be the main mechanism of corneal endothelial damage. In addition to the direct cytotoxic effect of disinfectant, damage of the anterior segment appears to be mediated by PGE2, an inflammatory cytokine induced by the breakdown of the blood–aqueous barrier. In support of this possibility, a previous study demonstrated that PGE2 is involved in corneal injury in an anterior segment ischemia animal model.²³ The present study additionally demonstrated that Dexa had an inhibitory effect on the expression of inflammatory mediators, thereby improving the corneal haze score and clinical score for anterior uveitis. It seems that Dexa blunted the increase in PGE2 level in the aqueous humor, and this resulted in reduction of further inflammatory damage of intraocular tissues. 

In the EIU animal model, the levels of PGE2 and TNF-α in aqueous humor were elevated, and Dexa ameliorated the increases in these inflammatory mediators. In line with the present study, previous studies have also shown that both PGE2 and TNF-α levels are increased in aqueous humor in the EIU animal model.^5,6,8,14 In the EIU animal model, endotoxin induces breakdown of the blood–aqueous barrier, resulting in cellular infiltration and protein extravasation in the anterior segment.¹⁶ Endotoxin activates resident monocytes and macrophages that induce genes encoding inflammatory mediators, which in turn leads to subsequent infiltration of neutrophils and mononuclear cells.^24,25 As a result, PGE2 and TNF-α have been shown to be early mediators of intraocular inflammation in the EIU animal model.^26–28 As PGE2 and TNF-α could be responsible for disruption of the blood–aqueous barrier, it is likely that reduced cell infiltration of the aqueous humor after Dexa injection could result in blockaded elevation of PGE2 and TNF-α induced by LPS.^3,9,29 

In contrast to the elevation of both PGE2 and TNF-α in aqueous humor in the EIU animal model, only the PGE2 level was elevated in aqueous humor in the TASS animal model. Stable levels of TNF-α in the TASS animal model might be explained by the dualistic role of PGE2 in the regulation of TNF-α: A low concentration of PGE2 stimulates the expression of TNF-α, whereas a high concentration of PGE2 inversely inhibits TNF-α expression. Congruent with this possibility, Renz et al.³⁰ showed that a low dose of PGE2 (0.1–10 ng/mL) stimulated the synthesis of TNF-α, but a high dose (especially more than 100 ng/mL) suppressed TNF-α release from rat resident peritoneal macrophages. In this study, the median aqueous PGE2 levels were 26.9 ng/mL in the EIU group and 210.9 ng/mL in the TASS group, and the median aqueous TNF-α levels were 87.4 pg/mL in the EIU group and 0.0 pg/mL in the TASS group. These results are similar to those of Renz et al.³⁰ Although, unlike the present study, the study by Renz et al.³⁰ was conducted on an in vitro model of the Lewis rat using peritoneal macrophages, it is probable that there is a similar dualistic role of PGE2 in the regulation of TNF-α in the TASS and EIU animal models. 

To our knowledge, this is the first study to evaluate the levels of PGE2 and TNF-α in the aqueous humor of the TASS animal model. The TASS animal model used in this study will be helpful to investigate related inflammatory mediators and to evaluate the efficacy of treatment for ocular inflammation. 

In conclusion, PGE2 in aqueous humor was significantly increased in both the TASS and EIU animal models, whereas an increase in TNF-α in aqueous humor was induced only by EIU. Intraperitoneal Dexa injection attenuated the increase in PGE2 induced in the TASS and EIU animal models. A similar inhibition of TNF-α induction by Dexa was observed in the EIU animal model. Based on these results, aqueous PGE2 level may be more suitable for evaluating the efficacy of treatment for TASS. 

Acknowledgments

The authors thank Young Eun Koh, MB, Department of Laboratory Medicine, Korea University College of Medicine, for conducting cytology. 

Supported by a grant from the Korean Health Technology R&D Project, Ministry of Health & Welfare, Republic of Korea (HI13C0055). 

Disclosure: Y. Eom, None; D.Y. Lee, None; B.-R. Kang, None; J.-H. Heo, None; K.-H. Shin, None; H.M. Kim, None; J.S. Song, None