Increased levels of proinflammatory cytokines and MMPs have been observed on the ocular surface of patients with KCS.
2 34 35 36 Furthermore, inflammation has been observed to develop in neurturin-deficient mice that have naturally occurring and permanent dry eye. These mice have been observed to have increased concentrations of IL-1β and MMP-9 in tear fluid washings and stimulated expression of IL-1β, TNF-α, macrophage inflammatory protein 2 (MIP-2), cytokine-induced neutrophil chemoattractant (KC), and MMP-9 mRNA by the corneal epithelia as the dry eye develops with age.
26 The present study provides convincing evidence that ocular surface inflammation develops in response to experimentally induced dry eye in mice. The concentrations of IL-1β and MMP-9 in the tear fluid washings and gelatinolytic activity on the ocular surface epithelia were observed to increase significantly in mice treated with subcutaneous injections of scopolamine and placement in a blower hood for 5 or 10 days
(Figs. 1 2 4) . The increased IL-1β concentration in the tear fluid washings could be due in part to a concentration effect caused by decreased tear volume. However, experimental evidence suggests that this is not the only cause. IL-1β was measured in tear fluid washings, rather than in pure collected tear fluid. IL-1β in the tear fluid was collected by placing a 1.5-μL drop of PBS on the ocular surface and collecting 1 μL. Therefore, all these samples were diluted tear washings, and the dilution factor was greater in the dry eyes. Using a fluorescein dilution method, we calculated the tear volume to be approximately 0.1 μL in normal eyes and 0.01 μL in dry eyes. Using these tear volumes, an approximate real tear concentration of IL-1β was calculated. In normal eyes, the tear fluid is diluted approximately 16 times (0.1 μL tears + 1.5 μL PBS). In contrast, the tear fluid in dry eyes was diluted approximately 151 times (0.01 μL tears + 1.5 μL PBS). The measured IL-1β concentrations of 20 and 80 pg/mL in the tear fluid washings of untreated mice and mice treated for 10 days, respectively
(Fig. 1) , multiplied by these dilution factors, yields an estimated true tear fluid concentration of 320 pg/mL in normal eyes and 12,080 pg/mL in dry eyes. This estimated 38-fold difference is much higher than the 10-fold difference that would be expected from the difference in tear volume. Therefore, it is likely that the increased IL-1β concentration in the tear washings of mice with dry eye is produced by the lacrimal glands and/or the stressed ocular surface epithelia. We also found the concentrations of IL-1β and TNF-α significantly increased in the corneal and conjunctival epithelia of the mice treated for 5 or 10 days
(Fig. 3) . In addition, both conventional RT-PCR and real-time PCR showed that the levels of IL-1β, TNF-α, and MMP-9 mRNA in the corneal and conjunctival epithelia noticeably increased in mice treated for 5 or 10 days, compared with untreated mice
(Figs. 5 7) . The gene array data further supported that dry eye induces inflammation at the transcriptional level by showing increased expression of inflammatory cytokines (IL-1β, TNF-α) and their receptors (IL-1R1, IL-1R2, and TNFR1) in the corneal epithelia of the mice treated for 5 days
(Fig. 8) . It is likely that elevated levels of these cytokines in the corneal epithelium were due to stimulated production by the corneal epithelial cells, because we did not detect any infiltration of inflammatory cells into the corneal epithelium over a 2-week dry eye treatment period. In contrast, slight infiltration of the conjunctival epithelium and stroma with inflammatory cells has been observed in this murine dry eye model, and these cells could be responsible for some of the increased levels of inflammatory cytokines. IL-1 is a potent inducer of other inflammatory cytokines such as IL-6 and TNF-α, and chemokines such as IL-8.
37 In mice, IL-1 and TNF-α stimulate the production of the key chemoattractants KC and MIP-2.
38 39 IL-1 and TNF-α also stimulate the production of MMPs by epithelial and inflammatory cells.
30 40 The gelatinase MMP-9 is one of the most important MMPs on the ocular surface. Overexpression of MMP-9 by the corneal epithelia has been reported to impede re-epithelialization of the cornea after experimental thermal injury in animal models and has been associated with sterile corneal ulceration in humans.
41 The concentration and activity of MMP-9 have been found to be significantly higher in the tear fluid of patients with KCS, with the highest levels observed in patients with sterile ulceration.
5 MMP-9 is also an efficient activator of latent precursor cytokines, including TGF-β1 and IL-1β.
42 We used multiple methods to demonstrate conclusively that ocular surface dryness induces the production of these key inflammatory factors and provides a model for studying their roles in the pathogenesis of KCS.