The Presence of Guttae in Fuchs Endothelial Corneal Dystrophy Explants Correlates With Cellular Markers of Disease Progression

Abstract

Purpose: Fuchs endothelial corneal dystrophy (FECD) is characterized by an accelerated depletion of corneal endothelial cells. There is growing evidence that mitochondrial exhaustion is central in the pathology. Indeed, endothelial cells loss in FECD forces the remaining cells to increase their mitochondrial activity, leading to mitochondrial exhaustion. This generates oxidation, mitochondrial damage, and apoptosis, fueling a vicious cycle of cells’ depletion. This depletion ultimately causes corneal edema and irreversible loss of transparency and vision. Concurrently to endothelial cells loss, the formation of extracellular mass called guttae on the Descemet's membrane, is a hallmark of FECD. The pathology origins at the center of the cornea and progress outward, like the appearance of guttae.

Methods: Using corneal endothelial explants from patients with late-stage FECD at the time of their corneal transplantation, we correlated mitochondrial markers (mitochondrial mass, potential, and calcium) and the level of oxidative stress and apoptotic cells, with the area taken by guttae. The different markers have been analyzed using fluorescent-specific probes and microscopic analysis.

Results: We observed a positive correlation between the presence of guttae and the level of mitochondrial calcium and apoptotic cells. We found a negative correlation between the presence of guttae and the level of mitochondrial mass, membrane potential, and oxidative stress.

Conclusions: Taken together, these results show that the presence of guttae is correlated with negative outcome in the mitochondrial health, oxidative status, and survival of nearby endothelial cells. This study provides insight on FECD etiology that could lead to treatment targeting mitochondrial stress and guttae.

The corneal endothelium is the most posterior cellular layers of the cornea.¹ The main role of this endothelium is to keep the corneal stroma partially dehydrated,² which is essential for the proper alignment of the collagen fibrils necessary for corneal transparency.³ The corneal endothelial cells (CECs) pumps ions out of the stroma and to the anterior chamber using Na⁺/K⁺ ATPase pump necessitating a high mitochondrial contribution. Fuchs endothelial corneal dystrophy (FECD) is a corneal pathology characterized by an accelerated depletion of CECs.⁴ This depletion jeopardizes the role of the endothelium causing infiltration of liquid in the corneal stroma, leading to a loss of corneal transparency and a loss of vision.⁵ Concurrently to the loss of endothelial cells, the formation of extracellular mass called guttae on the endothelium basement membrane, the Descemet's membrane (DM), can be observed in FECD endothelium.⁴ FECD affects about 4% of the US population over 40 years old and the only curative treatment available is corneal transplantation, making it one of the leading uses of cornea from donors.^6–8 

Although some mutations are linked to early-onset (COL8A2) and late-onset (TCF4) FECD,⁹ the etiology of most FECD cases remains unknown and have been linked to oxidative stress and damaged mitochondria.^10–12 Studies bring evidence that these characteristics are part of a vicious cycle fueled by an overuse of mitochondria.^13,14 Indeed, as we have previously shown, using the same markers as the one in this study, that the accelerated loss of endothelial cells in patients with FECD forces the remaining cells to increase the amount of ions they pump thus increasing their energy demand and consequent mitochondrial contribution and reactive oxygen species (ROS) generation.¹³ This generates oxidative stress, leading to mitochondrial damage and ultimately to apoptosis, fueling a vicious cycle of cells depletion.¹³ 

The pathology originates at the center of the cornea andes progress outward, much like the appearance of guttae.¹⁵ The size of the guttae varies between 10 and 40 µm and their presence is accepted as a marker of FECD progression.^16,17 It has been hypothesized that guttae are the results of cell secretions, weak points in the DM, or intracellular leftover of a dying cell fusing with the DM.¹⁸ Few studies have investigated the link between guttae and markers of FECD progression in a native setting.^13,19 In this study, we used markers of the mitochondrial burnout¹⁴ and correlated them with the guttae. We used late-stage FECD explants, in which we have previously shown that each cell is not at the same progression level in the pathology, making it a good model to study FECD progression. Using this model, we correlated the level of mitochondrial mass, potential and calcium, and the level of oxidative stress and apoptotic cells, with the area taken by guttae. 

Materials and Methods

All experiments in this study were performed in accordance with the Declaration of Helsinki, and the research protocol received approval by the CIUSSS-EMTL (Montréal) and the CHU de Québec-Université Laval (Québec) institutional ethics committees for the protection of human subjects with written informed patient consent for study participation. FECD explants (corneal endothelium and DM) were collected from 14 consenting patients (age = 53–90 years, median = 75, SD ± 11.5) with late stage FECD at the time of their corneal transplantation (Centre universitaire d'ophtalmologie [CUO], Hôpital Maisonneuve-Rosemont [HMR], and CUO – Hôpital du Saint-Sacrement, Québec, Canada). We received FECD explants in one piece and every cellular portion of each explant was analyzed and with no discrimination between central and peripheral portion of the explant. Age, sex, surgical method of removal, and pseudophakia status of all donors used for this study are listed in Supplementary Material (Table S1). 

Markers Used on Explants

Corneal endothelium on their DM were kept overnight at 37°C in growth medium, as previously described.¹⁴ We received the explants from different location and at different periods of the day. Keeping them overnight helps standardizing the method and insures they are all treated the same. Explants were then washed with Opti-Mem-I (for all markers but CM-H2DCFDA; Invitrogen, Burlington, ON, Canada) or PBS (for CM-H2DCFDA). The different markers have been performed as previously described.¹⁴ Briefly, JC-1 (Invitrogen; 2.5 µM) was used to measure mitochondrial membrane potential. Rhod-2 AM (Invitrogen; 3 µM) was used to measure mitochondrial calcium. CellEvent Caspase-3/7 Green Detection (Invitrogen; 5 µM) was used to measure caspase-3/7 activity marker. CM-H2DCFDA (Invitrogen; 2.5 µM) was used to measure ROS. Mitotracker Deep Red FM (Invitrogen; 80 nM) was used in combination with JC-1, CellEvent Caspase-3/7 Green Detection Reagent, and CM-H2DCFDA. Mitotracker Green FM (Invitrogen; 80 nM) was used with Rhod-2 AM. The explants were put in staining solution of PBS (for CM-H2DCFDA), PBS with 5% fetal bovine serum (for CellEvent caspase-3/7 marker) or Opti-Mem-I (for JC-1 and Rhod-2) containing the markers with their appropriate mitochondrial probe. The explants were then incubated for 30 minutes at 37°C and 8% CO2, washed with Opti-Mem-I or PBS, put on a microscope slide. Explants were observed with corresponding channel for the marker and we used Blue FP (Ex: 380 nm / Em: 440 nm) to image guttae. Solvents for each marker (PBS, PBS with 5% fetal bovine serum [FBS], or Opti-Mem-1) was determined according to the manufacturer's recommendations. 

Image Analysis

The signal produced by the different markers was quantified using AxioVision 4.8.2 software (Zeiss, Oberkochen, Baden-Württemberg, Germany). The background was measured and removed from the signal of the different markers. The percentage of the area occupied by guttae in a field was measured using ImageJ. 

Statistical Analysis

Linear regressions were performed for each marker against the total area of guttae in each field. Simple linear regression t-tests were performed to evaluate the relationship between the guttae area in the fields and the different markers. A P value ≤ 0.05 was defined as a statistically significant linear relationship between guttae and the markers. 

Results

Caspase 3/7 activity was measured in conjunction with guttae in FECD explants. Results show that Caspase 3/7 positive cells are concentrated near guttae (Fig. 1A). Imaged fields were plotted for Caspase 3/7 positive cells against guttae (Fig. 1B). A linear regression was drawn and shown to be statistically significant, indicating that increased guttae in an area translates in more apoptotic CECs. The guttae area accounts for 48.5% of the variance in apoptotic cells ratio. 

Figure 1.

 

Presence of apoptotic cells positively correlates with guttae. (A) A marker of apoptosis (caspase 3/7 activity; green) was measured in conjunction with guttae auto-fluorescence (pale blue) using Blue FP filter. A marker of mitochondrial mass (Mitotracker Deep Red; red) was used to label mitochondria. The blue arrow points to a cell near guttae and red arrow points to a cell far from guttae. (B) Ratio of number of caspase 3/7 positive cells on total cells was plotted against guttae area in each microscopic field. Each dot represents a microscopic field. A positive correlation was found between apoptosis and guttae area (R² = 0.4850; P < 0.01). Scale bar = 40 µm. Experiments were performed with four different explants from patients with FECD (15 fields analyzed, N = 4).

Signal of mitochondrial mass was correlated with the guttae area in FECD explants. As seen in Figure 2A, the mitotracker signal is lower near the guttae and higher further away. The mitochondrial mass signal was plotted against guttae area and a linear regression was drawn (Fig. 2B). This regression is statistically significant and shows that increased guttae in an area translates in less mitochondrial mass in CECs. The guttae area accounts for 78% of the variance in mitochondrial mass. 

Figure 2.

 

Mitochondrial mass negatively correlates with guttae. (A) A marker of mitochondrial mass (Mitotracker Deep Red; red) was measured in conjunction with guttae auto-fluorescence (pale blue) using Blue FP filter. The blue arrow points to a cell near guttae and red arrow point to a cell far from guttae. (B) Mitotracker signal per cell was plotted against guttae area in each microscopic field. Each dot represents a microscopic field. A strong negative correlation was found between mitochondrial mass and guttae area (R² = 0.7801; P < 0.0001). Scale bar = 40 µm. Experiments were performed with three different explants from patients with FECD (17 fields analyzed, N = 3).

Signal of mitochondrial membrane potential was correlated with guttae area in FECD explants. We found that the JC-1 signal is lower near guttae and higher away from them (Fig. 3A). JC-1 signal per cell was plotted against the area of guttae in the same microscopic field (Fig. 3B). A linear regression was performed and was statistically significant. There is a negative correlation between the area guttae takes and the mitochondrial membrane potential signal in CECs. The guttae area accounts for 73.6% of the variance in mitochondrial membrane potential. 

Figure 3.

 

Mitochondrial potential negatively correlates with guttae. (A) Markers of mitochondrial mass (Mitotracker Deep Red; blue) and of mitochondrial potential (JC-1; red) were measured in conjunction with guttae auto-fluorescence (pale blue) using Blue FP filter. The blue arrow points to a cell near guttae and the red arrow points to a cell far from guttae. (B) Ratio of JC-1 on mitotracker signal was plotted against guttae area in each microscopic field. Each dot represents a microscopic field. A strong negative correlation was found between mitochondrial potential and guttae area (R² = 0.6082; P < 0.0001). Scale bar = 40 µm. Experiments were performed with three different explants from patients with FECD (15 fields analyzed, N = 3).

The signal for mitochondrial calcium was correlated with guttae area in FECD explants. Results show that the Rhod-2 signal is higher near guttae (Fig. 4A). The Rhod-2 signal per cell was plotted against the area taken by guttae in the microscopic field (Fig. 4B). A linear regression was performed and was statistically significant. There is a positive correlation between the presence of guttae and mitochondrial calcium content in CECs. The guttae area accounts for 86.6% of the variance in mitochondrial calcium. 

Figure 4.

 

Intro-mitochondrial calcium positively correlated with guttae. (A) A marker of mitochondrial calcium (Rhod-2; red) was measured in conjunction with Rhod-2 marked guttae (⁎). A marker of mitochondrial mass (Mitotracker Green; green) was also used. The blue arrow points to a cell near guttae and the red arrow points to a cell far from guttae. (B) Rhod-2 per cell was plotted against guttae area in each microscopic field. Each dot represents a microscopic field. A strong positive correlation was found between mitochondrial calcium and guttae (R² = 0.8663; P < 0.0001). Scale bar = 40 µm. Experiments were performed with three different explants from patients with FECD (17 fields analyzed, N = 3).

The signal of oxidative stress was correlated with guttae area in FECD explants. We found the CM-H₂DCFDA signal to be lower near guttae (Fig. 5A). The CM-H₂DCFDA signal per cell was plotted against the area occupied by guttae (Fig. 5B). A linear regression was performed and was statistically significant. There is a positive correlation between oxidative stress and the presence of guttae in CECs. The guttae area accounts for 29.3% of the variance in oxidative stress. 

Figure 5.

 

Oxidative stress negatively correlates with guttae. (A) A marker of oxidative stress (CM-H₂DCFDA; green) was measured in conjunction with guttae auto-fluorescence (pale blue) using Blue FP filter. A marker of mitochondrial mass (Mitotracker Deep Red; red) was also used. The blue arrow points to a cell near guttae and the red arrow points to a cell far from guttae. (B) CM-H₂DCFDA signal per cell was plotted against guttae area in each microscopic field. Each dot represents a microscopic field. A negative correlation was found between oxidative stress and guttae area (R² = 0.2928; P < 0.01). Scale bar = 40 µm. Experiments were performed with four different explants from patients with FECD (22 fields analyzed, N = 4).

Discussion

In a previous study, we have determined the mitochondrial-related biomarkers of FECD is a vicious cycle.¹⁴ Here, we show that those markers correlate with the presence of guttae. In this previous study, we have shown that FECD CECs in an explant have variable levels of mitochondrial mass, which is an indicator of pathology progression at the cellular level.¹⁴ We also found that FECD CECs have a lower mitochondrial membrane potential and higher mitochondrial calcium than non-pathologic cells. Other studies have shown that FECD CECs are in an oxidative stress status and have increased apoptotic cells number.^13,14 In this study, we aimed to determine the potential correlation between the abundance of guttae and the mitochondrial manifestation of FECD. At first, we found a direct correlation between apoptotic cells and guttae (see Fig. 1). This indicates that guttae are toxic for CEC and/or that apoptotic cells are linked to the formation of guttae. 

FECD cells were shown to have increased variation in mitochondrial mass levels compared to healthy cells despite having similar median mass.¹⁴ We hypothesized that the FECD cells with higher mitochondrial mass were compensating for the fewer number of CECs and those with lower mitochondrial mass were dying from the unbearable energic stress (mitochondrial burnout).¹⁴ We found that cells in regions with higher density of guttae have a lower mitochondrial mass than those with lower density of guttae (see Fig. 2). Because the apoptotic status has been linked to lower mitochondrial mass in FECD CECs,^14,20 we can make a direct link among apoptosis, mitochondrial mass, and the presence of guttae. 

Previous studies have shown that FECD CECs’ mitochondria have a lower mitochondrial membrane potential than healthy CECs.^14,21,22 Here, we show that CECs in regions with higher density of guttae have lower mitochondrial membrane potential than region with fewer guttae (see Fig. 3). CECs with mitochondria harboring lower membrane potential have a reduced potential for ATP production, which affects their capacity to perform their function of pumping ions out of the corneal stroma. These CECs most likely drive the energetic stress to other regions. Because loss of mitochondrial membrane potential is a trigger for mitochondrial recycling via mitophagy,²³ this could explain the decreased mitochondrial mass in the regions with higher guttae density. The loss of the mitochondrial membrane potential could also be a contributing factor in the increased apoptosis in guttae rich regions.²⁴ 

As we have previously shown, intra-mitochondrial calcium is increased in FECD CECs.¹⁴ Here, we show CECs in regions with higher guttae density show increased intra-mitochondrial calcium levels (see Fig. 4). Because increasing intra-mitochondrial calcium is the last-ditch effort by the cell to increase its energy production,^25,26 the increase in intra-mitochondrial calcium found in guttae-rich regions is concordant with the low mitochondrial level and membrane potential found in these regions. Intra-mitochondrial calcium leads to an increase in ATP production. However, if the concentration reaches a certain threshold, it can also contribute to the loss of mitochondrial membrane potential and consequent apoptosis observed in guttae-rich regions via the opening of the mitochondrial permeability transition pore (MPTP).²⁶ In regions with lower density of guttae, the amount of healthy mitochondria is sufficient to avoid the need for mitochondrial calcium internalization. 

Oxidative stress has been described as an important hallmark of FECD.^10,27 Our results show a negative correlation between oxidative stress and the presence of guttae (see Fig. 5). As we have shown, the CECs in guttae-rich regions have characteristics that suggest they have a reduced pumping ability (i.e. higher apoptosis, lower mitochondrial mass and membrane potential, and higher intra-mitochondrial calcium). Consequently, ROS production is lowered in these cells due to a decrease in overall mitochondrial activity. This would force the CECs in guttae-poor regions to increase their ATP production to compensate. This increase in ATP production would increases ROS production and leads to oxidative stress.²⁸ 

Taken together, our results clearly show that guttae negatively affect neighboring CECs by affecting their mitochondria to a point that CECs are undergoing apoptosis. This created an energetic pression on CECs in guttae-poor regions, which translate in an oxidative stress. Whether guttae are toxic for CECs and are thus responsible for the changes observed or the consequence of CECs poor health conditions remains unknown. It is most likely a combination of both, where guttae affect CECs and affected CECs secrete extracellular matrix proteins found in guttae and modify their immediate environment. This suggests that removal of guttae in a patient with FECD could lead to an improvement in CECs health and/or that finding a way to prevent mitochondrial decline health could prevent or reduce guttae formation. Descemet stripping only (DSO) is a novel treatment alternative for patients with FECD.²⁹ DSO is performed in the same manner as Descemet's membrane endothelial keratoplasty (DMEK) but a smaller area of the central area of DM is removed (4-5 mm) and there is no transplantation of donor endothelium.³⁰ The central area being the most affected, the removal of the central most guttae-rich region of the endothelium allows the peripheral CEC to grow/migrate and restore the endothelium. Our results are perfectly consistent with the DSO studies and the removal of guttae may indeed represent a very promising avenue of treatment in view of our cellular and molecular analysis. 

Acknowledgments

The authors are grateful to the Banque d'Yeux du Centre Universitaire d'Ophtalmologie (Québec City, Canada) and to Hema-Québec's employees for ongoing collaboration for the procurement of research-grade healthy corneas, and to Drs Johanna Choremis, Marie-Eve Légaré, Michèle Mabon, Ralph Kyrillos, Mathieu Mercier, and Julia Talajic and the CUO-HMR operating room nurses for their collaboration in obtaining the FECD Descemetorhexis specimens. 

Supported by a grant from the Canadian Institutes of Health Research (CIHR) to I.B., S.P., and P.J.R. P.J.R. and S.P. are research scholars from the Fonds de Recherche du Québec - Santé (FRQ-S). Procurement of eyes and corneas for research from the CUO Eye bank was possible thanks to an infrastructure from the Vision Health Research Network. 

Author Contributions: Study conceptualization: S.M. and P.J.R. Data curation: S.M. and P.J.R. Formal analysis: S.M. and P.J.R. Funding acquisition: P.J.R., I.B., and S.P. Investigation: S.M. and P.J.R. Methodology: S.M. Project administration: P.J.R. Resources: P.J.R. Software: N/A. Supervision: P.J.R. Validation: S.M. and P.J.R. Visualization: S.M. and P.J.R. Writing of the original draft: S.M. Manuscript writing, review, and editing: S.M., P.J.R., and S.P. 

Disclosure: S. Méthot, None; S. Proulx, None; I. Brunette, None; P.J. Rochette, None