Regional Assessment of Energy-Producing Metabolic Activity in the Endothelium of Donor Corneas

Abstract

Purpose.: We characterized mitochondrial respiration and glycolysis activity of human corneal endothelium, and compared metabolic activity between central and peripheral regions.

Methods.: Endothelial keratoplasty-suitable corneas were obtained from donors aged 50 to 75 years. The endothelium–Descemet membrane complex (EDM) was isolated, and 3-mm punches were obtained from central and peripheral regions. Endothelium–Descemet membrane punches were assayed for mitochondrial respiration (oxygen consumption) and glycolysis (extracellular acidification) using an extracellular flux analyzer. Enzymatic (citrate synthase, glucose hexokinase) and mitochondrial density (MitoTracker) assays also were performed.

Results.: Ten corneas were analyzed per assay. Metabolic activity for mitochondrial respiration and glycolysis showed expected changes to assay compounds (P < 0.01, all pairwise comparisons). Basal mitochondrial respiration and glycolysis activity did not differ between regions (P > 0.99). Similarly, central versus peripheral activity after assay compound treatment showed no significant differences (P > 0.99, all time points). The intracorneal coefficient of variation for basal readings between two and four peripheral punches was 18.5% of the mean. Although peripheral samples displayed greater enzymatic activity than central samples (P < 0.05), similar to extracellular flux results, mitochondrial density did not differ between regions (P = 0.78).

Conclusions.: Extracellular flux analysis of oxygen and pH is a valid technique for characterizing metabolic activity of human corneal endothelium. This technique demonstrates high reproducibility, allows quantification of metabolic parameters using small quantities of live cells, and permits estimation of overall metabolic output. Neither oxygen consumption nor extracellular acidification differed between central and peripheral regions of transplant suitable corneas in this series. Our results show that endothelial cell health can be quantified biochemically in transplant suitable corneas.

Improper fluid retention in the cornea causes swelling, disruption of the stromal collagen fiber matrix, and loss of transparency.¹ The ability of the cornea to maintain proper hydration is dependent on the activity of the corneal endothelium, which counters the passive leak of fluid into stroma with the active pumping of ions into the anterior chamber using a process that is dependent on metabolic energy.^2–5 To date, significant progress has been made in elucidating the various mediators, regulators, and molecular mechanisms that underlie this process,^5,6 including the sodium–potassium-dependent ATPase that is a principle component of the pump mechanism.⁷ In stark contrast, comparatively little attention has been given to molecular mechanisms used by the corneal endothelium to fuel the energetically expensive, adenosine triphosphate (ATP)–dependent activities of this tissue.⁸ Consequently, the energy-producing metabolic activity of human corneal endothelial cells has yet to be described and compared in healthy, diseased, or injured states. 

Mitochondrial respiration and glycolysis are the most important metabolic pathways for producing the ATP required for proper human corneal endothelial cell function.⁸ Therefore, a deeper understanding of the corneal endothelium's capacity for mitochondrial respiration and glycolysis is likely to provide significant insight into the influence of cellular energy status on endothelial pump function and cellular health. Recently, a novel, noncontact method has been described for measuring mitochondrial respiration and glycolysis in ex vivo cells and tissues using an extracellular flux analyzer (XFe24; Seahorse Bioscience, North Billerica, MA, USA).^9–12 Using this instrument, continuous real-time recordings of oxygen consumption (mitochondrial respiration) and extracellular acidification (glycolysis) are obtained while treating cells with stimulants and inhibitors of key components in the electron transport chain (ETC) and glycolysis pathways. To our knowledge, this technology has not yet been applied to the study of corneal endothelial cells or human corneal endothelial tissue. 

In this study, we sought to adapt mitochondrial respiration and glycolysis metabolic assay methods using extracellular flux analysis of oxygen and pH to the investigation of native human corneal endothelium and normal energy production metabolism. Herein, we describe a technique for mounting and testing punches of native endothelium bound to Descemet membrane. Using this technique, we characterized the oxygen consumption and extracellular acidification rates, at baseline and after mitochondrial stress and glycolysis stress testing, in endothelial cell punches prepared from transplant-suitable corneas. Recognizing that the corneal endothelium may not be uniform in structure or function,^13–16 we compared central and peripheral endothelial punches to characterize regional differences. Lastly, we examined the feasibility of testing individual enzymatic steps involved in mitochondrial respiration and glycolysis (citrate synthase and glucose hexokinase enzymatic assays, respectively) as well as mitochondrial density quantification by fluorescent microscopy as alternatives to extracellular flux analysis. Results of this study provide a foundation for investigating corneal endothelial health using quantifiable metrics based on the energy-producing metabolic activity of these cells. 

Methods

Donor Corneal Tissue

All experimental procedures conformed to the Declaration of Helsinki. Corneoscleral tissue was obtained, inspected and stored in Optisol GS (Bausch & Lomb, Irvine, CA, USA) at 4°C in accordance with Eye Bank Association of America (EBAA) compliant protocols at Iowa Lions Eye Bank (ILEB). Donor tissue characteristics, including donor age, death to preservation time, death to preparation time, and endothelial cell density (ECD), including endothelial cell hexagonality and coefficient of variation via noncontact specular microscopy (KeratoAnalyzer EKA-10; Konan Medical USA, Irvine, CA, USA) were noted, and all tissues were determined to be suitable for cornea transplantation. The donor's diabetic status, determined by standard eye bank medical records review and donor risk assessment interview,¹⁷ also was noted. All corneas were analyzed within 14 days of preservation. The Institutional Review Board at the University of Iowa determined that approval was not required for this study, and research consent was obtained for all donor corneas. 

Metabolic Assay Tissue Processing and Preparation

One day before analysis, the endothelium-Descemet membrane complex (EDM) was prestripped using a standardized technique for Descemet membrane endothelial keratoplasty (DMEK) graft preparation¹⁷ with minor modifications. Briefly, each cornea was mounted on a 9.5-mm vacuum trephine and a partial thickness trephination was made through the EDM into posterior stroma. The tissue was stained with VisionBlue (DORC International, Zuidland, The Netherlands) to visualize the edges and rinsed carefully with Optisol GS. After submerging the tissue in Optisol GS, tissue peripheral to the scoring groove was removed and the outer 3 mm of the EDM was stripped away from stroma using a nontoothed forceps, leaving the central portion attached. After stripping the entire circumference, the EDM was returned to its normal anatomical location and replaced in Optisol GS for storage overnight at 4°C. 

The day of the assay, prestripped corneas were warmed to 25°C for 1 hour, then mounted on a vacuum trephine. Under Optisol GS, one central and two to four peripheral partial thickness punches through EDM were made using a 3-mm diameter biopsy punch (Integra LifeSciences, Plainsboro, NJ, USA; Fig. 1). Each tissue punch was secured endothelium-side up on a 4-mm diameter punch of transparent backing membrane (Millicell Cell Culture Inserts, 1.0 μm polyethylene terephthalate; EMD Millipore, Billerica, MA, USA) using a 1:2 mixture of Matrigel extracellular matrix (Corning Incorporated, Tewksbury, MA, USA) diluted in XF Base Media (Seahorse Bioscience). Excess Matrigel was removed and the coupled tissue and backing were placed in the bottom of a XF24 cell culture microplate (Seahorse Bioscience) and submerged in Optisol GS. Tissue mounts were rinsed and maintained in assay medium for 1 hour at 37°C before assaying. The assay medium for mitochondrial respiration consisted of XF Base Media, 10 mM glucose, 2 mM sodium pyruvate, and 1 mM glutamine, pH 7.40 (± 0.05). The assay medium for the glycolysis consisted of XF Base Media and 2 mM glutamine, pH 7.40 (±0.05). 

Figure 1

 

Preparation of EDM tissue punches from prestripped corneas for extracellular flux analysis of oxygen and pH. (A) Partial thickness punches through the EDM are made using a 3-mm diameter biopsy punch. (B–D) Tissue punches are secured endothelium-side up on a 4-mm diameter punch of transparent backing membrane (arrow). (E) A 1:2 mixture of extracellular matrix diluted in assay medium is used to secure EDM tissue to the transparent backing membrane. (F) After removing excess fluid, the coupled tissue and backing are placed in the bottom of an XF24 cell culture microplate well and submerged in corneal storage medium.

Metabolic Assays

Measurements of mitochondrial respiration and glycolysis were performed on an extracellular flux analyzer (XFe24; Seahorse Bioscience) at 37°C following manufacturer suggested protocols using XF Cell Mito and Glycolysis Stress Test Kits, respectively. The assay kits use specific chemicals that target and allow quantification of individual metabolic components of the mitochondrial respiration and glycolysis pathways (Supplementary Fig. S1). Briefly, the mitochondrial stress assay consisted of a series of 12 oxygen concentration measurements, beginning with a set of three basal recordings in glucose-supplemented media, followed by three recordings each after adding 1 μM oligomycin (ATP synthase inhibitor), 0.5 μM carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP, ETC uncoupler), and 1 μM Rotenone (Rot, ETC inhibitor) with 1 μM Antimycin A (AA, ETC inhibitor). Mitochondrial respiration was quantified by measuring oxygen concentration change in the extracellular media and reported as the oxygen consumption rate (OCR; pmole/min). The glycolysis stress assay consisted of a series of 12 pH measurements, beginning with a set of three basal recordings in glucose-free media, followed by three recordings each after adding 10 mM glucose (glycolysis substrate), 1 μM oligomycin, and 10 mM 2-Deoxy-D-glucose (2-DG, glycolysis inhibitor). Glycolysis was quantified by measuring pH change in the extracellular media and reported as the extracellular acidification rate (ECAR; mpH/min). 

Metabolic Assay Post Processing

Following metabolic assays, tissue punches were fixed for 20 minutes using 4% paraformaldehyde in PBS, rinsed with tris-buffered saline (TBS), and labeled fluorescently for 20 minutes using 1:1000 Sytox Green Nucleic Acid Stain (Life Technologies, Grand Island, NY, USA) diluted in TBS. Tissues were rinsed and maintained in PBS for imaging on an Olympus IX-81 inverted microscope (Olympus America, Center Valley, PA, USA). Images were captured using a ×4 objective with 200 ms exposure and a FITC filter, with multiple images taken to cover the entire punch area. Montaged images were imported into ImageJ,¹⁸ the threshold function used to generate binary images labeling only the nuclei, and the analyze particles function used to count the number of individual nuclei (cells). Cell counts were used to normalize metabolic assay results to generate ECAR/cell and OCAR/cell values, respectively. 

Enzymatic Assays

To quantify citrate synthase (CS) and glucose hexokinase (GH) activity in corneal endothelium, EDM punches were prepared as described above with minor modifications. Briefly, following 9.5-mm partial thickness trephination and EDM prestripping, a second 5-mm inner trephination was made using a biopsy punch (Integra LifeSciences) in the central cornea, generating separate central and peripheral punch areas (5-mm circle and 2.25-mm wide ring, respectively). The decision to use a 5-mm central punch was based on preliminary experiments indicating a 5-mm section of EDM tissue was necessary to generate adequate protein concentration for the assays. Both EDM punches were removed and placed in separate tubes containing lysis buffer. For CS, samples were lysed in 30 μL CellLytic M Cell Lysis Reagent (Sigma-Aldrich Corp., St. Louis, MO, USA) supplemented with Complete Mini protease inhibitor (Roche, Brandford, CT, USA). For GH, samples were lysed in 120 μL HK Assay Buffer. After 20 minutes at 4°C, the resulting protein-containing lysate was normalized across samples based on protein concentration; 4 μg were used to quantify CS activity (μmole/mL/min) using the Citrate Synthase Assay Kit (Sigma-Aldrich Corp.) and 8 μg for GH activity (pmole/min/mL) using the Hexokinase Colorimetric Assay Kit (Sigma-Aldrich Corp.), following manufacturer suggested protocols. 

Mitochondrial Density Assay

To quantify mitochondrial density in the corneal endothelium, the EDM was prepared using the standardized DMEK graft preparation technique without modifications,¹⁷ leaving a peripheral zone of the 9.5-mm EDM attached to the stroma. Prestripped corneas were incubated at 37°C in mitochondrial respiration assay medium for 1 hour followed by a 45-minute incubation at 37°C in 500 nM MitoTracker Orange CM-H₂TMRos (Life Technologies) diluted in assay medium. Following mitochondrial labeling, corneas were rinsed in PBS, fixed for 30 seconds in 0.5% paraformaldehyde in PBS, and after submerging in PBS, the EDM was stripped completely from the stroma. After transferring the EDM to an 18-mm coverslip using nontoothed forceps, peripheral relaxing incisions (4) were made at 90° intervals using a 15° microsurgical knife (Katena Products, Denville, NJ, USA), the tissue was fixed again for 5 minutes in 4% paraformaldehyde in PBS, and nuclei were labeled by incubating in 1:25,000 4′,6-diamidino-2-phenylindole (DAPI) diluted in PBS. The samples were rinsed in PBS and mounted in Aqua-Mount (Lerner Laboratories, Pittsburgh, PA, USA) on Superfrost Plus slides (Fisher Scientific, Waltham, MA, USA). Mounted tissues were imaged on an Olympus IX-81 inverted microscope. Images were captured using a ×4 objective with 200 ms exposure using a tetramethylrhodamine isothiocyanate (TRITC) filter and 100 ms exposure using a DAPI filter, with multiple images taken to cover the entire tissue area. Displayed images were obtained using a Leica SP2 Confocal Microscope (Leica Microsystems, Buffalo Grove, IL, USA) and a ×60 objective. 

To quantify staining intensities in the central and peripheral regions of the tissue, MitoTracker fluorescent images were montaged and imported into ImageJ. Using the oval tool, 3-mm circular tracings were created centrally (1) and peripherally (4) between relaxing incisions. The histogram tool was used to obtain individual pixel intensity values within each tracing. To adjust for differences in cell density and dropout among regions, background pixel intensity values were measured in areas of Descemet membrane lacking endothelial cells. Average background intensity values were used as a threshold, above which was considered to be mitochondrial labeling and subtracted from calculations of average pixel intensity per region. 

Statistical Analysis

A linear mixed model analysis for repeated measures was used to test for overall differences in extracellular flux metabolic output, and for differences based on central versus peripheral corneal regions, at baseline and following compound injections for mitochondrial respiration and glycolysis assays at manufacturer-specified time points (for mitochondrial respiration, the third glucose-supplemented media measurement, minimum postoligomycin value, maximum post-FCCP value, and minimum post-Rot/AA value; for glycolysis, the third glucose-free media measurement, maximum postglucose value, maximum postoligomycin value, and minimum post–2-DG value). This was followed by Tukey's HSD test to assess differences in pairwise comparisons between mean values at baseline and following compound injection, and Bonferroni's post hoc test to compare mean central versus peripheral values at each specified time point. 

A linear mixed model analysis also was used to obtain estimates of intercorneal and intracorneal coefficients of variation using peripheral corneal samples for mitochondrial respiration and glycolysis metabolic assays. Intercorneal and intracorneal variances were used to obtain an intraclass correlation coefficient as a measure of repeatability based on random sampling for each metabolic assay. 

Paired t-tests were used to analyze central and peripheral regional differences in citrate synthase and hexokinase enzymatic activity, respectively. 

A linear mixed model analysis for repeated measures was used to test for differences in mitochondrial staining intensity and peripheral variability and repeatability using the same approach as described above for extracellular flux analysis. 

Statistical analyses were performed using SAS version 9.3 (SAS Institute, Inc., Cary, NC, USA). A value of P ≤ 0.05 was considered statistically significant. 

Results

We analyzed 50 corneas from 43 donors, consisting of 10 corneas from unique donors per assay (mitochondrial respiration, glycolysis, citrate synthase, glucose hexokinase, mitochondrial density). Tissue characteristics of donor corneas analyzed are included in the Table. The mean donor age was 62.8 (SEM 0.9) and no significant differences in donor age between assays were noted. A history of diabetes was noted in 15 corneas (30%), distributed among assays with a range of two to four diabetic tissues per assay. 

Table

 

Donor Characteristics of Corneal Tissue by Metabolic Test

 

Metabolic Assays, Response to Stimulus

As expected physiologically (Supplementary Fig. S1), mean OCR/cell and ECAR/cell estimates obtained at baseline and following exposure to assay compounds were statistically different (P < 0.01 for all pairwise tests in each assay; Fig. 2). All responses to assay compounds were prototypical (inhibition after oligomycin and Rot/AA, stimulation after FCCP, acidification after glucose and oligomycin, deacidification after 2-DG). 

Figure 2

 

Mean OCR (pmole/min) and ECAR (mpH/min) for transplant quality corneal endothelial tissue. (A) Mean OCR/cell and (B) ECAR/cell for all regions combined were significantly different in response to compounds influencing mitochondrial respiration and glycolysis. Asterisks indicate significant differences on pairwise testing (*P < 0.01). BR, basal respiration; O, oligomycin; A/R, antimycin a/rotenone; BA, basal acidification; G, glucose.

Metabolic Assays, Regional Analysis

To quantify the variability in metabolic activity among different regions of the endothelium, OCR/cell, and ECAR/cell means and estimates of variability were obtained and compared between the central and peripheral regions of the cornea. For mitochondrial respiration and glycolysis stress testing, basal measurements for central and peripheral mitochondrial respiration (central, 0.019 pmole/min/cell SEM 0.002; peripheral, 0.020 pmole/min/cell SEM 0.002) and glycolysis (central, 0.009 mpH/min/cell SEM 0.001; peripheral, 0.009 mpH/min/cell SEM 0.001) showed no significant differences (P > 0.99, respectively). Comparisons between central and peripheral regions at all other time points analyzed also showed no significant differences (P > 0.99 for each assay; Fig. 3). Comparisons of overall plot profiles for central versus peripheral tissues showed no significant difference in mitochondrial respiration (P = 0.48) or glycolysis (P = 0.48). The intracorneal coefficient of variation for basal measurements among peripheral punches for mitochondrial respiration and glycolysis (n = 20 corneas, assays analyzed separately) was 19.9% and 17.2% of the mean, with intraclass correlation coefficient estimates of 0.52 and 0.43, respectively. The intercorneal coefficient of variation for basal measurements among peripheral punches for mitochondrial respiration and glycolysis was 21.8% and 12.7% of the mean, respectively. 

Figure 3

 

Regional analysis of energy-producing metabolic activity in the endothelium of donor corneas. (A) Schematic representation of the regional orientation of 3-mm tissue punches (dashed circles) taken from donor corneas in central (C) and peripheral (P) locations. (B) Mitochondrial respiration and (C) glycolysis metabolic activity for central (light gray) and peripheral (dark gray) corneal tissue (n = 10, each assay). The OCR (pmole/min) per cell and ECAR (mpH/min) per cell did not differ between central and peripheral tissues (P > 0.99 for all time points tested), and plot profile comparisons of central versus peripheral tissues showed no significant differences (mitochondrial respiration, P = 0.48; ECAR, P = 0.48). Dashed lines in (B) and (C) indicate assay drug injections and abbreviations above them represent specific compounds injected.

Enzymatic Assays, Regional Analysis

Unlike metabolic assay measurements, the peripheral region of the cornea displayed 51% greater CS activity and 106% greater GH activity compared to the central region of the cornea (CS: central, 0.18 μmole/mL/min SEM 0.03; peripheral 0.28 μmole/mL/min SEM 0.02, and GH: central, 1.5 pmole/min/mL SEM 0.39; peripheral, 3.0 pmole/min/mL SEM 0.65; P < 0.05 for each assay; Fig. 4). 

Figure 4

 

Regional comparisons of enzymatic activity in the central and peripheral cornea. (A) Schematic representation of the regional orientation of the 5-mm tissue punch (dashed circle) used to generate central (C) and peripheral (P) regions for enzymatic analysis. Mean glucose hexokinase activity (B) and citrate synthase activity (C) are presented for central and peripheral regions of the cornea. Asterisks indicate significant differences between the central and peripheral regions (*P ≤ 0.05).

Mitochondrial Density Assay, Regional Analysis

Similar to metabolic assay findings, mean mitochondrial staining intensity for central and peripheral regions of the cornea showed no significant differences (central, 536 intensity units SEM 36.8; peripheral, 531 intensity units SEM 34.4; P = 0.78; Fig. 5). The intracorneal coefficient of variation for staining intensity among peripheral regions was 9.7% of the mean with an intraclass correlation coefficient estimate of 0.67. The intercorneal coefficient of variation for staining intensity among peripheral punches was 19.8% of the mean. 

Figure 5

 

Regional comparisons of mitochondrial density in the central and peripheral cornea. (A) Representative example of MitoTracker mitochondrial staining (red) counterstained with DAPI nuclear staining (blue) in endothelial cells imaged with a confocal microscope using a ×60 objective lens. (B) Average mitochondrial staining intensity quantified in central and peripheral regions of the cornea, presented using pixel intensity values.

Discussion

There is a critical need to better understand the energy-producing metabolic activity and relative metabolic health of the corneal endothelium for clinicians, scientists, and the transplant community. In 2013, 24,987 endothelial keratoplasty procedures and an additional 20,954 penetrating keratoplasty procedures were performed in the United States to treat disorders of the corneal endothelium and various disorders requiring healthy endothelium, respectively.¹⁹ Additionally, 3695 corneas recovered in 2013 with transplant intent were processed and released for surgery, but not transplanted because the tissue expired or was unable to be placed,¹⁹ representing a significant financial cost to eye banks and wasted donor tissue. Currently, the tissue evaluation process for donor corneas is based on visual inspection techniques and fueled by surgeon-driven perceptions of a tissue's viability for transplant (donor age, time since recovery, cell counts) rather than objective measures of cell function.²⁰ We hypothesize that tissue wastage and financial losses may be preventable using a data-based approach and testing strategies that can indicate a level of metabolic activity commensurate with adequate tissue function. 

Our study demonstrated, for the first time to our knowledge, that extracellular flux analysis of oxygen and pH can be used to quantify multiple metabolic parameters for native endothelium tissue samples obtained from transplant suitable donor corneas. Our pilot data indicated a high degree of reproducibility in the cellular response to reagents that influence mitochondrial respiration and glycolysis. These observations highlight the potential of this technique to establish a normative database of corneal endothelial cell energy metabolism. Such a normative database would be useful in efforts to develop practical tests that ensure donor cornea functionality for transplantation, evaluation of metabolic activity based on storage time and conditions, and investigations of known metabolic diseases characterized by oxidative stress damage including Fuchs' dystrophy^21,22 and diabetes mellitus.²³ Use of native and cultured endothelial cells on this high-throughput platform also may prove particularly useful in efforts to identify side effects of therapeutic agents^24,25 and improve corneal storage media to support corneal health before transplant. Once established, we anticipate that metabolic data and parameters characterizing native corneal endothelium will serve as a benchmark in the potency validation of cultured and bioengineered corneal endothelial cells and tissues.²⁶ 

Despite documented regional differences in cell density,¹³ regenerating pluripotent stem cell populations,^14,15 and endothelial damage in eyes with Fuchs' dystrophy,¹⁶ our results demonstrated that the peripheral endothelium on average is functionally equivalent on a per cell basis to the central endothelium and multiple peripheral samples are capable of generating an accurate representation of the central corneal metabolic activity. It is worth noting that oxygen concentrations vary significantly within the anterior chamber of the eye.²⁷ However, at present, we are unable to determine if oxygen concentrations are rate limiting in mitochondrial respiration output and influence in vivo metabolic output. Regardless, our findings support the feasibility of ex vivo testing to correlate and validate peripheral metabolic activity with corneal graft function clinically, and the development of a functional assay to evaluate donor cornea suitability before transplant surgery. However, given the variability in metabolic activity and mitochondrial density measurements among peripheral punches within a given cornea, additional investigation of the peripheral corneal endothelium is required to better define its functional orientation and devise an appropriate sampling methodology. 

Additional studies also are needed to help identify the tissue-specific differences influencing the variability in mitochondrial respiration and glycolysis activity among corneas observed in this study (between 12.7% and 21.8% of the mean). For example, age may represent a significant variable in determining endothelial cell function.^28–31 While the mean donor age in our metabolic assays was 64.5 years, donors ranged between 54 and 72 years, and an age-stratified analysis may identify age-related changes in metabolic output and reduce variability. Likewise, 30% of our cohort included tissue from diabetic donors. This proportion constitutes a representative sample of diabetes based on current estimates of the donor pool in the United States.^17,32 However, it is plausible that diabetic donor endothelial tissue may exhibit abnormal responses to metabolic stress testing, as diabetes induces biochemical changes in endothelial cells.^33–35 It will be important to quantify the influence of age and diabetes on the energy-producing metabolic activity of the corneal endothelium. This approach would help improve our understanding of these conditions and their impact on cornea transplantation, and may lead to new therapeutic avenues to preserve and restore endothelial cell function. 

Our results also demonstrated that alternate assays of mitochondrial respiration and glycolysis have important functional limitations in assessing corneal endothelial cell metabolism. We observed no quantifiable differences in mean mitochondrial respiration, glycolysis, or mitochondrial density between central and peripheral regions of the endothelium when normalized to cell counts, yet regional differences in enzymatic activity were detected. At present, we are uncertain if these particular enzymatic data have any functional significance for the corneal endothelium. However, enzymatic regulation elsewhere in a pathway may alter interpretation of enzyme activity,^36,37 a disadvantage of testing individual steps in metabolic pathways as a surrogate for total metabolic output, and may help explain these findings. It is possible that other metabolic tests, such as the MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide) assay more closely approximate total metabolic activity. Additionally, although mitochondrial density measurements closely reflected the mitochondrial respiration activity we observed in extracellular flux testing, the MitoTracker assay lacks high-throughput capacity and cannot account for glycolysis activity. Altogether, while the assays of enzymatic activity and mitochondrial density we tested may supplement metabolic assay data, extracellular flux assays have distinct advantages for measuring total metabolic output in the corneal endothelium. 

In conclusion, extracellular flux analysis of oxygen and pH is a valid technique for characterizing the energy-producing metabolic activity of native corneal endothelial tissue. This technique is favorable compared to other tested assays of energy metabolism based on its high reproducibility, ability to quantify metabolic parameters using small quantities of live cells, and capacity to estimate overall metabolic output. Mean oxygen consumption and extracellular acidification rates in mitochondrial and glycolysis stress testing did not vary on a per cell basis between central and peripheral regions of transplant quality corneas in this series. Although additional investigation is required, results of this study provided a foundation for investigating corneal endothelial cell health using quantifiable metrics based on the energy-producing metabolic activity. 

Acknowledgments

Supported by Lloyd and Betty Schermer, Research to Prevent Blindness, Saving Sight, Cleveland Eye Bank, Michigan Eye Bank, Old Dominion Eye Foundation, Illinois Eye Bank, Tennessee Donor Services, Lions Medical Eye Bank and Research Center of Eastern Virginia, Minnesota Lions Eye Bank, Lions Eye Bank of West Central Ohio, Sierra Donor Services, Iowa Lions Eye Bank, and Midwest Eye-Banks 2014-15 (MAG). The authors alone are responsible for the content and writing of the paper. 

Disclosure: M.A. Greiner, None; K.A. Burckart, None; M.D. Wagoner, None; G.A. Schmidt, None; C.R. Reed, None; C.A. Liaboe, None; M.J. Flamme-Wiese, None; M.B. Zimmerman, None; R.F. Mullins, None; R.H. Kardon, None; K.M. Goins, None; B.T. Aldrich, None