Human peripheral blood was obtained in 20 ml quantities with a 20 U/ml heparin-coated syringe and then mixed with Hank's balanced salt solution (HBSS, GIBCO). Neutrophils were separated from red cells and other lymphocytes using Ficoll-Paque (Pharmacia & Upjohn, Biotech AB Uppsala, Sweden). After that, neutrophils were mixed with dextran (SIGMA, St. Louis, Mo, USA) and then preserved in an ice bag for 45 minutes. The upper fluid was removed and the remaining red blood cells were destroyed with hypotonic solution (0.2% NaCl). Purified neutrophils were obtained by centrifugation of the solution at 4℃ and 1000 rpm for 10 minutes with 10% fetal bovine serum (FBS, GIBCO).

Using a 96-well plate, about 1 × 10⁶ of the purified neutrophils were placed into each well along with 50 ul of HBSS. Into each well was then placed one of the three types of HA, molecular weight (MW) 700,000 (Viscoat®, Alcon, USA), MW 2,000,000 (Healon®, Pharmacia & Upjohn, Sweden), and MW 4,000,000 (Healon GV®, Pharmacia & Upjohn, Sweden) at four different concentrations (0.1, 1, 2, 3 mg/ml). We also employed two types of control groups: viscoelastic free wells and 2% hydroxymethylcellulose (Ocucoat®, Storz, USA) wells. The 2% hydroxymethylcellulose wells did not contain HA but had similar viscoelasticity to the HA wells. Twenty microliters of a 10 ng/ml phorbol myristate acetate solution (PMA, SIGMA), a chemoactive peptide for neutrophils, was applied into each well. Fifty microliters of cytochrome C (SIGMA) was applied into each well and the plates were incubated for 90 minute in a CO₂ incubator. After incubation, the reduced form of cytochrome C was measured at 550 nm with a Microplate reader (MR 700, Dynatech laboratories, USA) with spectral filter. To avoid measurement of superoxide that was not derived from PMN, we prepared wells with superoxide dismutase (SOD, SIGMA). The inhibitory effect of viscoelastics was expressed as a percentage of inhibition as follows:

BCEC were prepared with modification as described by Crawford KM et al.5 Briefly, bovine eyes were obtained from a local abattoir, and the corneas were excised with an attached scleral ring and placed, endothelium side up, in a concave plastic cup. The endothelial surface was ringed and subsequently incubated at 37℃ for 90 minutes in a sterile solution of 20mM HEPES buffered Earle's balanced salt solution (EBSS,GIBCO). Endothelial cells were dislodged from Descemet's membrane by gentle scraping using a spatula with forceps. Dislodged cells were aspirated with pipette and added with RPMI 1640 media with 10% FBS. The cells were gently centrifuged (600g, 2 minutes), resuspended in 5 ml of medium supplement and incubated in 25-cm² flasks at 37℃ in 95% air-5%CO₂. Third-passage cells were inoculated in a 96- well plate, 1 × 10⁵ per well, and cultured for three additional days for confluencing. Into each well, we applied human PMN, 1 × 10⁶ per well, and then supplied one of the three types of HA, or methylcellulose, at the four different concentrations (0.1, 1, 2, 3 mg/ml). After PMN activation with PMA, PMNs were incubated for 90 minutes and then washed out with HBSS. The viability of the remaining BCECwas measured with an MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrasodium bromide, SIGMA) assay.6 The protective effect of viscoelastics was expressed as a percentage of protection as follows:

All of these procedures were repeated three times. The results were analyzed with one-way ANOVA according to molecular weight and concentration and Duncan's pairwise comparison test was performed. To analyze for the dual effects of molecular weight and concentration, we used two-way ANOVA. We calculated the above statistics using a PC-statistical analysis system program (SAS, version 6.12).

The degree of superoxide release decreased in a concentration-dependent manner in all three types of HA and methylcellulose (p < 0.0001) (Table 1). However, the comparison according to molecular weight at each concentration showed no significant difference (p value of 0.1mg/ml concentration was 0.8495, that of 1 mg/ml concentration was 0.0733) between types of viscoelastics except for methylcellulose and 4,000,000 MW HA, both over 2 mg/ml concentration: the percentage inhibition of 2mg/ml concentration of methylcellulose and 4,000,000 MW HA was 78.23 ± 2.08 and 86.69 ± 7.45, respectively (p = 0.0001), and of 3mg/ml concentration was 75.00 ± 2.08 and 78.43 ± 6.27, respectively (p = 0.0001) (Table 1). In two-way ANOVA analysis that considered both the effects of molecular weight and concentration of HA, both were found to influence the results. Specifically, HA inhibited superoxide release dose-dependently at concentrations over 0.1 mg/ml (Fig. 1). The mean values of cytochrome C reduction in SOD wells and viscoelastic-free wells were equal. This revealed that the main source of superoxide was PMN.

In all three types of HA and methylcellulose, the results of the MTT assay showed that the protective effect of viscoelastics on BCEC from activated PMN occurred in a concentration-dependent manner. However, there was no significant difference in the types of viscoelastics in the one-way ANOVA analysis: percentage protection of methylcellulose, and HA of MW 700,000, 2,000,000 and 4,000,000, was 112.22 ± 4.19, 113.2 ± 1.83, 108.7 ± 0.26 and 112.2 ± 1.05, respectively (p = 0.3663) (Table 2). In the two-way ANOVA, both type and concentration of viscoelastics influenced the results. In the concentration analysis, the BCEC survival rate was positively correlated with viscoelastic concentration: percentage protection of methylcellulose, and HA of MW 700,000, 2,000,000, 4,000,000, at 1 mg/ml, was 161.67 ± 10.12, 131.9 ± 3.23, 131.9 ± 3.67 and 135.1 ± 9.29, respectively (p = 0.0002), and at 2 mg/ml was 192.04 ± 24.90, 130.8 ± 4.86, 131.7 ± 7.35 and 132.5 ± 4.50, respectively (p = 0.0001) (Table 2). However, the curve of survival plateaued at concentrations over 1 mg/ml (Fig. 2). In the type analysis, there was no significant difference in BCEC viability except for methylcellulose wells in which more BCEC survived than in other wells (Fig. 2).