Hypersensitive Response to Oxidative Stress in Keratoconus Corneal Fibroblasts

Abstract

purpose. It is unclear whether the oxidative damage found in keratoconus (KC) corneas results from innate defects of corneal fibroblasts or is due to excessive environmental challenges encountered by the patient with KC. The purpose of this study was to explore whether KC cells have inherent, exaggerated hypersensitivity to oxidative stressors.

methods. Normal and KC corneal stromal fibroblasts were incubated in neutral or low-pH conditions, with or without hydrogen peroxide (H₂O₂). Reactive oxygen/nitrogen species (ROS/RNS) production was measured with 2′,7′-dichlorodihydrofluorescein diacetate dye. Caspase-9 and -12 activities were measured by fluorochrome inhibitor to caspase (FLICA) assays. Long-extension polymerase chain reaction (LX-PCR) was used to amplify mtDNA. RNA was extracted, full-length cDNA synthesized, and PCR performed for mitochondria-encoded genes. Mitochondrial membrane potential (ΔΨm) was measured by a cationic dye assay.

results. In neutral pH conditions, KC fibroblasts had increased ROS production (P = 0.047), higher RNA levels for cytochrome c oxidase (complex IV) subunit II (P < 0.05), and decreased cathepsin K RNA (P = 0.04) compared with levels in normal cultures. In low-pH conditions, KC fibroblasts had decreased ΔΨm (P = 0.015) and increased activation of caspase-9 (P = 0.013) and -12 (P = 0.01) compared with normal cultures. Changes in ΔΨm were independent of cathepsin inhibition. The combination of low-pH+H₂O₂ treatment degraded intact mtDNA and decreased the mtDNA-to-nuclear DNA ratio.

conclusions. Cultured KC fibroblasts have an inherent, hypersensitive response to oxidative stressors that involves mitochondrial dysfunction and mtDNA damage. KC fibroblast hypersensitivity may play a role in the development and progression of keratoconus.

Keratoconus (KC) is a corneal thinning disorder associated with irregular astigmatism along with decreased visual acuity and is a leading indication for corneal transplantation. Intact KC corneas have altered antioxidant enzymes, ^{1 2 3} accumulations of cytotoxic ROS/RNS, ⁴ and mitochondrial DNA (mtDNA) damage. ⁵ In addition, a genomic deletion in the superoxide dismutase 1 (SOD1) gene has been associated with KC. ⁶ More recently, abnormal oxidative stress-related properties were found in KC corneal cells in vitro. ⁷ In response to oxidative stress-related conditions, KC fibroblasts have increased ROS/RNS production, catalase activity, and caspase-3 activity along with a loss of cell viability. ⁷ Other studies have shown that oxidative stress elements can induce activation of degradative enzymes and degradation of tissue inhibitors of metalloproteinases (TIMPs). ⁸ These findings are significant, in that the extreme thinning of KC is related to increased proteinase activities ^{9 10 11 12 13} and decreased proteinase inhibitors. ^{14 15}  

Apoptosis is a noninflammatory programmed cell death that involves complex signaling cascades mediated by caspases, a family of cysteine proteases. Two of the initiator caspase pathways (caspase-9 and -12) are associated with cellular stress. In response to mitochondrial stress, caspase-9 is activated, which leads to cytochrome c release and caspase-3 activation and is involved with various pathologic conditions. ^{16 17 18} Cells subject to endoplasmic reticulum stress undergo caspase-12 activation, which causes cleavage of caspase-3 in a cytochrome c-independent manner. ¹⁹ Caspase-12 activation and endoplasmic reticulum stress have been implicated in neurodegenerative disorders. ²⁰ In intact KC corneas, apoptosis is associated with epithelial, endothelial, and stromal cells. ²¹ Although KC corneal fibroblast cultures have elevated caspase-3 activity compared with normal cultures, ⁷ it is not known whether the stress-related initiator caspase pathways are activated in these cultures. 

KC corneas have higher than normal levels of cathepsin B, -G, and -V/L2. ^{3 13 22 23} Cathepsin-B, -L, and -K are cysteine proteases, whereas cathepsin-D and -G are aspartic proteases and serine proteases, respectively. Cathepsins present in cellular lysosomes can activate the caspases that in turn initiate apoptosis. ^{24 25} Isolated lysosomal extracts can cause mitochondrial dysfunction and apoptosis. ²⁶ Cathepsin-D translocates to the cytosol and causes apoptosis, ²⁷ and cathepsin-B plays a role in TNF-α-induced apoptosis. ²⁸ The recently described cathepsin-K is involved in bone resorption. ^{29 30} To date, no study has examined the relationship of cathepsins to oxidative stress–related apoptosis in KC cells. 

The purpose of this study was to determine whether oxidative stress–related challenges cause mitochondrial dysfunction or activate the stress-related caspase pathways in corneal fibroblasts cultures. 

Materials and Methods

Corneal Fibroblast Cultures

Normal corneas (n = 11; mean age, 40.2 years; range, 24–59) were received within 24 hours after death from the National Disease Research Interchange (Philadelphia, PA). KC corneas (n = 11; mean age, 34.8 years; range, 25–46) were sent to the laboratory in MK (McCarey-Kaufman) medium on ice and were received within 24 hours after penetrating keratoplasty (Table 1) . The diagnosis of KC in the participating patients was based on the presence of more than one of the following criteria: Munson’s sign, Rizzuti phenomenon, slit-lamp findings of stromal thinning, Vogt’s striae, Fleischer ring, and epithelial or subepithelial scarring. Retroillumination signs included scissoring on retinoscopy and the oil droplet sign (Charleux); photokeratoscopy signs were compression of mires inferotemporally, inferiorly or centrally; and videokeratography signs were localized increased surface power and/or inferior superior dioptric asymmetry. Informed consents were obtained from participants and the study was performed according to the tenets of the Declaration of Helsinki for research involving human subjects and human tissue. Normal and KC stromal cells were isolated and cultured as described previously. ³¹ When corneal stromal cells are cultured in 10% fetal bovine serum, they differentiate into fibroblasts. ^{32 33} Our corneal cells stained with vimentin (a marker for fibroblasts) but did not stain with α-smooth muscle actin (a marker for myofibroblasts). In addition, the morphology was consistent with fibroblasts and not corneal keratocytes. For individual experiments, third-passage fibroblasts were used. The cells were plated overnight in either 24-well plates (1 × 10⁵) or 60-mm dishes (5 × 10⁵), rinsed in Dulbecco’s phosphate-buffered saline containing calcium and magnesium (D-PBS; Mediatech Inc, Herndon, VA), and then incubated for 1 hour in D-PBS, (pH 7.0 or 5.0), with or without H₂O₂ (200 or 400 μM; Sigma-Aldrich, St. Louis, MO). Cultures were rinsed in D-PBS and then harvested for assays. 

In this study, fibroblast cultures were stressed by two different mechanisms. The first stress challenge exposed the cultured fibroblasts to low-pH (pH 5.0) conditions for 1 hour, which lowered the extracellular pH causing passive H⁺ influx and decreased the intracellular pH. In response, cells must expend energy (ATP) to eliminate H⁺ through the Na⁺/H⁺ exchange, placing high metabolic demands on mitochondrial function. The second stress method exposed cells to H₂O₂, a major ROS element that induces an oxidative stress response in cells. 

Detection of ROS/RNS Production

ROS/RNS production was measured by using the fluorescent dye 2′,7′-dichlorodihydrofluorescein diacetate (H₂DCFDA; Invitrogen-Molecular Probes, Eugene, OR), ⁷ which detects hydrogen peroxide, peroxyl radicals, and peroxynitrite anions. 

Caspase-9 and -12 Assays

Cells were treated for 1 hour with or without H₂O₂, rinsed with PBS-EDTA, and incubated at 37°C for 1 hour in MEM with 10% FBS. ⁷ The fluorochrome inhibitor of caspases (red-LEHD-FMK/caspase-9 or FITC-ATAD-FMK/caspase-12) solutions (BioVision Research Products, Mountain View, CA) were added and incubated for an additional hour. Caspase activities were measured at an excitation of 490 or 532 nm and emission 520 or 580 nm, with a fluorescent imager (FMBIO III; Hitachi-MiraiBio Inc., Alameda, CA). 

Mitochondrial Membrane Potential (ΔΨm) Measurements

Loss of ΔΨm was measured with a kit (JC-1 Mitochondrial Membrane Potential Detection Kit; Cell Technology, Minneapolis, MN). ⁷ The cells were plated on 24-well plates and incubated for 1 hour at 37°C in D-PBS at either pH 7.0 or 5.0, with or without H₂O₂. Some cultures were incubated with inhibitors (20 μM cyclosporine, 100 μM pepstatin A, or 50 nM cystatin C) for 1 hour before incubation with H₂O₂. Cystatin C inhibits cathepsin B, pepstatin A inhibits cathepsin D, and cyclosporine inhibits cathepsins B and L and blocks the loss of ΨΔm. ^{34 35} Red fluorescence was measured at 532 nm excitation and 580 nm emission. Green fluorescence was measured at 488 nm excitation and 520 nm emission on a fluorescence imager (FMBIO III, Hitachi-MiraiBio Inc.). 

Mitochondria Isolation

Mitochondria from normal and KC cultures were isolated with a kit (Pierce Biotechnology, Rockford, IL). Protein concentrations were measured with a BCA protein assay reagent kit (Pierce Biotechnology) according to the manufacturer’s protocol. 

Western Blot Analysis

Western blot analysis were performed as described previously, ³ with the monoclonal porin antibody (1 μg/mL; Invitrogen, Carlsbad, CA). Alkaline phosphatase-conjugated goat anti-mouse IgG antibody (1:5000) was used as the secondary antibody. Blots were developed with substrate solution (Immun-Star AP; Bio-Rad, Hercules, CA). 

Extraction of Total DNA

DNA from the fibroblast cultures were extracted according to the method of Atilano et al. ⁵  

Long Extension Polymerase Chain Reaction (LX-PCR)

The LX-PCR was performed according to modified methods of Melov et al. ³⁶ and Jessie et al. ³⁷ LX-PCR was performed as described previously using forward primer: TGA GGC CAA ATA TCA TTC TGA GGG GC (15,148–15,174 bp); reverse primer: TTC ATC ATG CGG AGA TGT TGG ATGG (14,841–14,816 bp). ⁵  

Extraction of RNA

RNA from normal and KC cultures were isolated (RNeasy Extraction Kit; Qiagen, Inc., Valencia, CA), according to the manufacturer’s protocol. Quality and quantity of the RNA was then determined (2100 Bioanalyzer; Agilent Technologies, Palo Alto, CA). 

Full-Length cDNA Synthesis and PCR

RNA (500 ng) was reverse transcribed according to the manufacturer’s protocol (PowerScript; BD-Clontech, Mountain View, CA). cDNA samples were subjected to PCR with specific primers (Invitrogen; Table 2 ). 

Statistical Analysis

Data were subjected to statistical analysis with a commercial program (Prism ver. 3.0; GraphPad Software Inc., San Diego, CA). The data were analyzed by the unpaired two-tailed student t-test. Data are expressed as the mean ± SEM. The data from triplicate fibroblast cultures were combined for statistical analyses. P < 0.05 was considered statistically significant. 

Results

In neutral-pH conditions, the ROS/RNS production was significantly higher in KC fibroblast cultures than in normal fibroblast cultures (Fig. 1) . The KC cultures showed a 1.7-fold increased ROS/RNS production compared with normal cultures (P = 0.047). Treatment with 200 μM H₂O₂ decreased the ROS/RNS production in normal (P = 0.021) and KC cultures (P = 0.004). With 400 μM H₂O₂ treatment, the normal cultures showed a 2.0-fold decrease (P = 0.028) and KC fibroblasts had an 8.6-fold decline in ROS/RNS formation compared with untreated cultures (P = 0.0009). 

Caspase-3 activities were elevated in KC fibroblast cultures treated with low pH, ⁷ but it was not clear whether the stress-related caspase pathways were involved. In the present study, we examined the caspase-9 and -12 activity levels in low-pH–treated fibroblast cultures. KC fibroblasts treated with low pH had significantly higher caspase-9 activity than did normal cultures (P = 0.013; Fig. 2A ). Normal fibroblasts treated with low-pH had increased caspase-9 activity with 200- and 400-μM H₂O₂ treatments (P = 0.0005 and 0.003, respectively). KC cultures had no statistically significant change in caspase-9 activity after treatments with 200 or 400 μM H₂O₂. 

KC fibroblast cultures treated with low-pH also showed significantly higher caspase-12 activity than did normal cultures (P = 0.01; Fig. 2B ). Normal fibroblast cultures showed increased caspase-12 activities after treatment with 200 and 400 μM H₂O₂ compared with untreated cultures (P = 0.018 and P = 0.03, respectively). In contrast, in KC fibroblast cultures, 200 and 400 μM H₂O₂ treatments did not significantly increase the caspase-12 activities compared with the untreated cultures. The KC fibroblasts treated with low-pH had caspase-12 and -9 activities comparable to normal cultures treated with the low-pH+400 μM H₂O₂. 

Figure 3Ashows that at neutral pH, the ΨΔm in untreated normal and KC cultures were similar to each other (P = 0.6) and that culturing the normal cells with 200 or 400 μM H₂O₂ did not significantly change the ΨΔm. KC cells treated with 200 and 400 μM H₂O₂ showed a ΨΔm similar to untreated KC cultures. 

In low-pH stress conditions (Fig. 3B) , untreated KC cultures had a 2.7-fold decrease of ΨΔm compared with untreated normal fibroblast cultures (P = 0.015). Treatment with 200 or 400 μM H₂O₂ resulted in a nonsignificant decrease in ΨΔm compared with that in untreated normal cultures (P = 0.052 and 0.054, respectively). 

In normal and KC fibroblast cultures, mtDNA was damaged after low-pH+H₂O₂ treatment (Fig. 4) . The LX-PCR method allows amplification of 16,262 bp of the mitochondrial genome along with subgenomic-sized products that contain the mtDNA primer sites. When mtDNA contains deletions, substitutions, or rearrangements, LX-PCR yields smaller products. Normal and KC fibroblast cultures were incubated at neutral pH and low pH with or without H₂O₂, and mtDNA was amplified by LX-PCR. Figure 4Ashows a major band of ∼16.2 kb representing intact mtDNA in the C7 (control, pH 7), C5 (control, pH 5), and H7 (H₂O₂, pH 7) samples. After treatment with H₂O₂+low pH (H5), the mtDNA was degraded and smaller fragments appeared. This demonstrated that the low-pH treatment alone (C5) or H₂O₂ treatment alone (H7) did not significantly damage the full-sized mtDNA. As the mtDNA degraded (H5), nuclear DNA (nDNA), as represented by 18s, increased (Fig. 4B) . For the low-pH+H₂O₂ (H5) samples, the ratios of LX-PCR mtDNA to 18s were significantly decreased in both normal and KC fibroblast cultures compared with the untreated pH7 cultures (C7; P = 0.0003 and 0.002, respectively). 

Normal and KC fibroblast cultures had similar levels of porins in the mitochondrial fractions (Fig. 5) . Western blot analyses were performed for mitochondrial and cytosolic fractions of normal and KC fibroblast cultures. The 35.5-kDa band representing porins was found in the mitochondria fractions but not in the cytosolic fractions. As measured by scanning densitometry, the normal and KC porin bands had similar intensities (100 density units vs. 108 density units, respectively). 

In addition to genomic DNA analysis, we examined RNA expression patterns in normal and KC fibroblasts incubated at neutral pH and low-pH (Fig. 6) . The RNA expression levels for mitochondrial encoded cytochrome B (MTCYB) and mitochondrial cytochrome oxidase subunit 1 (MTCO1) in KC fibroblast cultures were similar to those in normal cultures. In contrast, the mitochondrial cytochrome oxidase subunit 2 (MTCO2) RNA levels were increased in the KC fibroblasts (KC7) cultured at pH 7 compared with the normal (NL7) cultures (P = 0.01). When exposed to low-pH conditions, the MTCO2 RNA levels were higher in the normal (NL5) fibroblasts (P = 0.001) and KC (KC5) fibroblasts (P = 0.0004) than in untreated neutral-pH cultures (NL7). 

RNA levels for both cathepsin-D and -K are expressed by corneal fibroblasts. At neutral pH, the cathepsin K RNA levels in KC cultures were initially threefold lower than normal fibroblasts (P = 0.04). In low-pH conditions, the KC cultures showed a threefold increase in cathepsin K RNA levels compared with neutral pH cultures (P = 0.03), whereas the normal fibroblast cultures did not change significantly. The RNA levels for cathepsin D were similar in both normal and KC cultures when incubated in neutral- or low-pH conditions. 

We measured the ΨΔm of cultures treated with known inhibitors for various cathepsins. We found that cystatin C, cyclosporine, and pepstatin A did not reverse the loss of ΨΔm that occurred after low-pH or H₂O₂ treatments (Fig. 7) . 

Discussion

In the present study, KC fibroblasts had increased activation of stress-related caspase pathways (caspase-9 and -12) compared with normal fibroblasts. In addition, mitochondrial dysfunction, represented by a loss of ΔΨm, was independent of various cathepsin pathways. The mtDNA was selectively degraded by the combination of a low-pH stress+H₂O₂. Furthermore, KC fibroblasts possessed characteristics similar to those found in normal fibroblasts treated with 200 and 400 μM H₂O₂. 

Human corneal fibroblasts have a different response to the H₂O₂ challenge if they are at neutral-pH (ROS/RNS levels decrease) or low-pH conditions (ROS/RNS levels increase). A H₂O₂ challenge to human corneal fibroblasts elicits elevated catalase activity, ⁷ which eliminates H₂O₂. Catalase upregulation is likely to be responsible for the decline in ROS/RNS levels in corneal fibroblast cultures treated with neutral pH+H₂O₂. As little as 50 μM H₂O₂ can upregulate catalase in human lens epithelial cells ³⁸ and chronic, sublethal levels of H₂O₂ can increase the catalase levels and modulate cells to become resistant to acute higher H₂O₂ levels. ³⁹ Once human corneal fibroblasts are incubated in low pH+H₂O₂, the ROS/RNS levels increase. ⁷ The H₂O₂ can damage cells directly or alternatively, it can combine with superoxide anions to produce hydroxyl radicals. This could increase nitric oxide availability, since the superoxides would then not be available to form peroxynitrite. These reactions involve SOD, the levels of which are altered in KC corneas, and may represent a mechanism by which H₂O₂ could modify ROS production. Our findings suggest that normal human corneal fibroblasts have the capacity to eliminate large amounts of H₂O₂, which is consistent with the known corneal function of protecting the inner eye from oxidative stress elements. In addition, fibroblasts from KC corneas have a significantly higher ROS/RNS burden that may contribute to high oxidative damage in these corneas. 

Significantly increased caspase-9 and -12 activities were found in KC fibroblast cultures compared with normal cultures. Both pathways are activated by stress: the caspase-9 via mitochondrial stress ^{35 40} and caspase-12 through endoplasmic reticulum stress. Studies have shown elevated caspase-3 activities in KC cultures compared with normal fibroblasts. ⁷ To our knowledge, activation of the caspase-12 pathway in corneal fibroblast cultures has not been reported. The latent form of caspase-12 is found at the cytosolic face of the endoplasmic reticulum; it is activated in response to stress which then triggers caspase-3 activation. ⁴¹ Activation of the stress-related caspase-9 and -12 pathways suggests that oxidative stress plays a role in KC pathogenesis. The involvement of specific caspase pathways may be dependent on the cell type and cause of stress. For example, ARPE-19 cells stressed with 7-ketocholesterol have activation of caspase-8 and -12 pathways but not caspase-9. ⁴² The interactions between the caspase-9 and -12 pathways in human corneal fibroblasts are unknown and will be investigated in future studies. 

Measuring mtDNA integrity is a valuable way to determine viability of cells. Mitochondria have unique DNA that is circular, lacks introns, and has high transcription rates and an inefficient repair system, making it particularly susceptible to damage. This susceptibility is important, because mtDNA encodes for 13 OXPHOS proteins, 22 transfer (t)RNAs, and 2 ribosomal (r)RNAs. When mitochondria are damaged, OXPHOS decreases and ROS/RNS formation increases, which lead to mitochondrial dysfunction, altered gene expression, apoptosis, and loss of cell viability. KC corneas have increased levels of smaller-sized LX-PCR mtDNA bands and an accumulation of somatic mtDNA damage. ⁵ In the present study, normal and KC corneal fibroblast mtDNAs were susceptible to degradation by a combination of low-pH stress and exogenous H₂O₂. Of note, low-pH or H₂O₂ stress alone did not lead to significant mtDNA damage within the period that we studied. This result is in contrast to other cell types that have shown significant mtDNA damage when stressed with H₂O₂. ^{43 44} We suspect that in our fibroblast cultures, the low-pH challenge increased the metabolic demand on cells as they adjusted to maintain their intracellular pH. The additional H₂O₂ stress may overwhelm the mitochondrial integrity and cause mtDNA damage. Our findings suggest that in vivo intact KC corneas may have both a metabolic defect and high levels of oxidants that contribute to damaged mtDNA found in the intact KC corneas. ⁵  

Mitochondrial porin is an outer membrane protein that forms channels between the cytosol and intermembrane space. These 3-nm channels allow the passage of molecules less than 10 kDa. Since porins are quite stable, they are used as a reference protein for mitochondria and can reflect levels of mitochondria within cells. In our study the normal and KC cultures had similar levels of the mitochondrial porins which agrees with studies showing that mitochondrial porins in the cytoplasm of epithelial, stromal, and endothelial cells were similar in normal and KC corneas. ⁵  

KC fibroblasts contain increased levels of MTCO2 RNA compared with normal cultures. MTCO1 and MTCO2 are subunits of respiratory complex IV that are encoded by mtDNA. MTCO1 protein is diminished in areas of active disease in KC corneas. ⁵ In the present study, the MTCO1 RNA showed an elevated, but nonsignificant, trend in KC cultures. However, at neutral pH, the MTCO2 RNA levels were significantly higher in the KC cultures than in the normal cultures and in fact were at levels similar to those in the normal fibroblasts exposed to low-pH conditions. It is likely that the metabolic stress generated by the low-pH conditions increases OXPHOS demand. The fact that the neutral pH KC cultures (KC7) showed such elevated MTCO2 RNA levels implies that these fibroblasts have a stress-related burden not found in normal cells. 

In our study, human corneal fibroblasts expressed RNA for cathepsin-K and -D. Cathepsin K protein levels are elevated in KC corneas. ⁴⁵ Of interest, in our study the RNA levels for cathepsin K were threefold lower in the neutral pH cultures compared with the normal control but increased significantly at low-pH conditions. Although cathepsin K is collagenolytic and implicated in rheumatoid arthritis, ²⁹ little is known about the regulation of cathepsin K and its role in KC. 

Cathepsins represent a caspase-independent pathway for apoptosis. Lysosomal rupture causes release of mitochondrial cytochrome c followed by apoptosis. ⁴⁶ Cathepsins act as mediators of apoptosis by triggering mitochondrial dysfunction, cleaving Bid, and releasing cytochrome c. ^{24 27 47 48} In addition, cathepsin K induces apoptosis by cleaving Bid at Arg65 or Arg71, ⁴⁸ and the addition of purified cathepsin D increases ROS/RNS production by isolated mitochondria. ⁴⁶ Pepstatin A, an inhibitor of cathepsin D, decreases caspase-9 and -3 activation in vitro. ⁴⁹ Therefore, since cathepsins are elevated in KC corneas and play a significant role in mitochondrial related apoptosis, we speculated that inhibiting cathepsins may decrease the ΨΔm, an early hallmark of apoptosis. In our study, the cathepsin inhibitors did not inhibit the loss of ΨΔm in the KC fibroblast cultures or the cultures treated with low pH+H₂O₂. These findings differ from studies showing that cyclosporine prevented loss of ΨΔm and stabilized the mitochondrial permeability transition pore (mtPTP) after treatment with TNF-α in heavy metal ion–deficient conditions. ³⁵ Our findings suggest that cathepsin-related pathways do not play a role in the loss of ΨΔm or the mtPTP changes associated with apoptosis. 

Our data showed that under identical conditions, KC fibroblasts had increased ROS/RNS production, activated caspase-9 and -12 pathways, and decreased ΨΔm compared with normal fibroblasts. Furthermore, mtDNA damage, similar to that found in the intact KC corneas, was induced by the combination of low-pH+H₂O₂. These findings support our hypothesis that oxidative stress contributes to the development and progression of KC. 

 Supported by National Eye Institute Grant EY06807, the Schoellerman Charitable Foundation, the Discovery Fund for Eye Research, the Research to Prevent Blindness Foundation, the Guenther Foundation, the Iris and B. Gerald Cantor Foundation, and the Skirball Molecular Ophthalmology Program.

 Submitted for publication March 3, 2008; revised April 24, 2008; accepted July 28, 2008.

 Disclosure: M. Chwa, None; S.R. Atilano, None; D. Hertzog, None; H. Zheng, None; J. Langberg, None; D.W. Kim, None; M.C. Kenney, 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.

 

Table 1.

 

Cornea Donor Characteristics

	Age	Sex	Race	COD/Diagnosis
Normal
1	59	NA	NA	Subarachnoid hemorrhage
2	25	F	C	Respiratory arrest
3	40	M	C	SI GSW head
4	56	F	C	Metastatic breast cancer
5	48	M	C	Spinal fracture
6	24	M	NA	NA
7	33	M	C	SI GSW head
8	33	M	C	SI GSW head
9	44	M	C	MVA
10	46	M	C	NA
11	34	M	C	Acute MI
Mean age	40.2
Keratoconus
1	34	M	NA	KC
2	35	F	NA	KC
3	43	F	NA	KC
4	27	F	NA	KC
5	46	F	NA	KC
6	30	M	NA	KC
7	43	M	NA	KC
8	32	M	B	KC
9	37	F	B	KC
10	25	M	NA	KC
11	31	F	NA	KC
Mean Age	34.8

 NA, not available; COD, cause of death; MVA, motor vehicle accident; MI, myocardial infarction; GSW, gun shot wound; SI, self inflicted.

Table 2.

 

Primers

Name	Forward Primer	Reverse Primer	Size (bp)	PCR Annealing Temp. (°C)	Number of Cycles	MgCl2 (mM)
β2MG	CTCGCGCTACTCTCTCTTTCTG	GCTTACATGTCTCGATCCCACTT	334	60	25	1.5
CAT-D	CGAGGTGCTCAAGAACTACA	AGCACGTTGTTGACGGAGAT	432	60	25	1.5
CAT-K	GGCCAACTCAAGAAGAAA	GTACCCTCTGCATTTAGC	225	55	30	1.5
MTCYB	GGGGCCACAGTAATTACAAA	GGGGGTTGTTTGATCCCGTTT	200	57	30	2
ND2	GCCCTAGAAATAAACATGCTA	GGGCTATTCCTAGTTTTATT	232	53	25	2
MTCO1	ACTGATGTTCGCCGACCGTT	GATTAGGACGGATCAGACGAAGA	573	57	25	1.5
MTCO2	CTGAACCTACGAGTACACCG	TTAATTCTAGGACGATGGGC	322	55	30	3
18s	ACGGACCAGAGCGAAAGCAT	GGACATCTAAGGGCATCACAGAC	531	53	25	2

Figure 1.

 

Increased ROS formation in KC fibroblast cultures. At neutral pH conditions, KC fibroblast cultures (n = 4) had a 1.7-fold increase in ROS production compared with that in NL cultures (n = 5; 34,940 ± 5,001 vs. 20,610 ± 3,564; *P = 0.047). The ROS levels were decreased in NL cultures after treatment with 200 μM H₂O₂ (20,610 ± 3,564 vs. 9,802 ±1,175, *P = 0.021) and 400 μM H₂O₂ (20,610 ± 3,564 vs. 10,120 ± 1,591, *P = 0.028). The KC fibroblasts treated with 200 μM H₂O₂ had a 3.3-fold decrease (34,940 ± 5,001 vs. 10,510 ± 1,970, **P = 0.004) and after 400 μM H₂O₂ treatment had a 8.6-fold decrease in ROS/RNS activity (34,940 ± 5,001 vs. 4,069 ± 1,073, ***P = 0.0009).

Figure 2.

 

KC cell cultures had increased caspase-9 and -12 activity compared with that in NL cell cultures. (A) Low-pH stressed KC fibroblasts (n = 4) had significantly higher caspase-9 activity compared with that in normal cultures (n = 4; 3983 ± 824 vs. 1019 ± 196; *P = 0.013). In normal cultures, the caspase-9 activity increased with the 200- and 400-μM H₂O₂ treatments (1019 ± 196 vs. 3257 ± 262, ***P = 0.0005 and 1019 ± 196 vs. 4070 ± 601, **P = 0.003). The KC cultures stressed with low pH, and 200 or 400 μM H₂O₂ had similar caspase-9 levels compared with untreated KC cultures. Note that low-pH, untreated KC fibroblasts had a caspase-9 level comparable to that in the low-pH+400 μM H₂O₂–treated normal fibroblast cultures. (B) Low-pH–treated KC fibroblast cultures showed significantly higher caspase-12 activity than the normal cultures (mean ± SEM; **P = 0.01). The caspase-12 activity increased with the 200 μM H₂O₂ (mean ± SEM; *P = 0.018) and 400 μM H₂O₂ (mean ± SEM; *P = 0.03) treatment. The untreated KC fibroblast cultures had caspase-12 activity higher than that in the low-pH+400 μM H₂O₂–treated normal cultures (7595 ± 1684 vs. 4939 ± 1242). *Statistically significant difference between designated groups.

Figure 3.

 

Decreased ΨΔm after low-pH+H₂O₂ stress conditions in normal and KC cultures. (A) At neutral-pH conditions, the untreated normal (n = 5) and KC (n = 5) cultures had similar levels of ΨΔm (3.84 ± 0.62 vs. 3.38 ± 0.58, P = 0.60). There was not a significant change in ΨΔm after treatment with 200 or 400 μM H₂O₂ in the normal (2.74 ± 0.55 or 2.58 ± 0.57) or KC (3.28 ± 0.53 or 4.84 ± 1.33) cultures. (B) In the low-pH–stressed cultures, the untreated KC fibroblasts (n = 4) showed a 2.7-fold decrease in ΨΔm compared with that in the normal (n = 3) untreated fibroblast cultures (1.275 ± 0.06 vs. 3.43 ± 0.71, *P = 0.015). At low-pH+200 μM H₂O₂, the normal fibroblast cultures showed a 2.3-fold decrease in ΨΔm compared with that in the untreated normal cultures (3.43 ± 0.71 vs. 1.47 ± 0.15, P = 0.052). The low-pH+400 μM H₂O₂ also had a 2.3-fold decrease compared with the control (3.43 ± 0.71 vs. 1.50 ± 0.16, P = 0.054). The ΨΔm in the KC fibroblast cultures incubated at low-pH+200 μM H₂O₂ (1.1 ± 0.07) and the low-pH+400 μM H₂O₂–treated samples (1.25 ± 0.09) were similar to that in the control (1.275 ± 0.06).

Figure 4.

 

mtDNA was damaged after low-pH+H₂O₂ treatment. (A) After treatment with low-pH+H₂O₂, the LX-PCR mtDNA was degraded and smaller fragments appeared. The low-pH treatment alone or H₂O₂ treatment alone did not damage the mtDNA. As the mtDNA was degraded (H5), the nDNA (18s) increased. NL, normal; C7, control, pH 7; C5, control, pH 5; H7, hydrogen peroxide–treated, pH 7; H5, hydrogen peroxide treated, pH 5. (B) Decreased mtDNA-to-nDNA ratios after low-pH+H₂O₂ treatment. The ratios of LX-PCR mtDNA and 18s (representing nuclear DNA) were measured by semiquantitative PCR. The mtDNA-to-nDNA ratios were decreased after treatment with low-pH+H₂O₂ stress conditions in both the normal (***P = 0.0003) and KC (**P = 0.002) fibroblast cultures.

Figure 5.

 

Mitochondrial porin levels are similar in KC and normal cultures. The normal (n = 4) and KC (n = 3) cultures were separated into the mitochondrial and cytosolic fractions and proteins subjected Western blot analyses with a porin-specific antibody. Shown are representative porin levels in mitochondrial and cytosolic fractions. In Western blot analysis, the 35.5-kDa band represents porins, which are a stable structural protein of the mitochondria. Note that the NL and KC porin bands are similar in intensity.

Figure 6.

 

Cathepsin-D and -K are expressed by normal and KC corneal fibroblasts in vitro. RT-PCR analysis of gene expression for cathepsin D and -K in normal and KC fibroblasts incubated for 1 hour at pH 7 and pH 5. In neutral pH conditions, the KC fibroblast cultures showed a threefold decrease of RNA levels for cathepsin K compared to normal fibroblasts (0.192 ± 0.082 vs. 0.577 ± 0.136, *P < 0.05). At the low-pH conditions, the cathepsin K RNA levels increased significantly in the KC cultures compared with the KC neutral-pH samples (0.598 ± 0.137 vs. 0.192 ± 0.082; *P = 0.03). The cathepsin D RNA levels were similar in normal and KC cultures when incubated at neutral- or low-pH conditions. RNA levels for MTCYB and MTCO1 were similar in KC (KC7 and KC5) and normal (NL7 and NL5) cultures. In neutral pH, the RNA levels for MTCO2 were higher in the KC fibroblasts (KC7) than in the normal cultures (NL7) (1.713 ± 0.162 vs. 0.768 ± 0.135, **P = 0.01). In low-pH conditions the MTCO2 RNA increased in the normal (1.71 ± 0.16, ***P = 0.001) and KC (1.86 ± 0.16, ***P = 0.0004) fibroblast cultures compared with neutral pH normal samples.

Figure 7.

 

Cyclosporine, pepstatin A, and cystatin C did not reverse the reduced mitochondrial membrane potentials (ΨΔm) found in normal or KC cultures after H₂O₂ treatment.