Neprilysin inhibition promotes corneal wound healing

Neprilysin (NEP), an ectoenzyme that modulates inflammation by degrading neuropeptides, was recently identified in the human corneal epithelium. The cornea expresses many NEP substrates, but the function of NEP in homeostatic maintenance and wound healing of the cornea is unknown. We therefore investigated the role of this enzyme under naive and injured conditions using NEP-deficient (NEP−/−) and wild type (WT) control mice. In vivo ocular surface imaging and histological analysis of corneal tissue showed no differences in limbal vasculature or corneal anatomy between naive NEP−/− and WT mice. Histological examination revealed increased corneal innervation in NEP−/− mice. In an alkali burn model of corneal injury, corneal wound healing was significantly accelerated in NEP−/− mice compared to WT controls 3 days after injury. Daily intraperitoneal administration of the NEP inhibitor thiorphan also accelerated corneal wound healing after alkali injury in WT mice. Collectively, our data identify a previously unknown role of NEP in the cornea, in which pharmacologic inhibition of its activity may provide a novel therapeutic option for patients with corneal injury.


Supplementary
. Effect of thiorphan on NEP activity in the cornea and trigeminal ganglia. NEP activity in whole cornea and trigeminal ganglia isolated from uninjured WT mice at 1 and 6 h after a single intraperitoneal injection of 15 mg/kg thiorphan. Presented as % NEP activity of control, vehicle-treated tissue. In both tissues, NEP activity was significantly lower than in vehicle-treated tissue at 1 h after administration, but not at 6 h after administration of i.p. thiorphan (mean ± SEM; * P = 0.0226; ** P = 0.002; one-way ANOVA with Bonferroni correction).

Supplementary Figure 4. Corneal perforation in alkali-injured WT mice.
Incidence of corneal perforation by day 7 post-injury. Veh, vehicle; 5, 5 mg/kg thiorphan; 15, 15 mg/kg thiorphan. Figure 5. Thiorphan administration does not significantly affect CD45, CD31, or α-smooth muscle actin expression in alkali-injured WT corneas at one week. ( a ) Immunoblots showing expression of CD45, CD31, and α-smooth muscle actin in whole cornea lysates from WT mice at one week post-alkali or sham injury. Samples from each experimental group were randomized among three gels that were processed simultaneously. ( b ) Amido black stain for total protein on membranes corresponding to those shown in ( a ), utilized as a loading control. ( c ) Band intensities normalized to total protein and intermembrane reference protein. Each point represents expression in lysate from a single cornea ( mean ± SEM; one-way ANOVA with Bonferroni correction). SMA, α-smooth muscle actin ; veh, vehicle; thior, thiorphan (15 mg/kg). Figure 6. Expression of NEP protein in TKE2 cell line. Immunoblot of NEP (100 kDa) in lysates from confluent TKE2 cultures, with glyceraldehyde 3-phosphate dehydrogenase (GAPDH, 37 kDa) run as a loading control. Membrane was cut at 50 kDa.

Supplementary Figure 7. Thiorphan does not affect migration of TKE2 cells after wounding in vitro .
( a ) Representative images from in vitro scratch assays demonstrating that cell migration into the cell-free wound region does not vary with thiorphan treatment. ( b ) Summary graph of the course of wound healing over the duration of the assay. ( c ) Quantification of wound closure at the final timepoint (n = 5 wells per condition; mean ± SEM; two-tailed t test).

Supplementary Methods
NEP enzyme activity following thiorphan administration. Uninjured WT mice were administered a single dose of i.p. 15 mg/kg thiorphan or vehicle (5% ethanol in saline) with a 28 gauge U-100 insulin syringe (Becton-Dickinson, Franklin Lake, NJ, USA). Mice were decapitated under deep isoflurane anesthesia 1 or 6 h after treatment for collection of bilateral trigeminal ganglia and corneas. Tissue was frozen in liquid nitrogen and stored at -80° C.
Frozen corneas were homogenized in 100 μL of 0.5% NP-40 lysis buffer in two 15 s bursts with 30 s on ice between bursts. Frozen trigeminal ganglia were homogenized in 200 μL of buffer for 10 s.
Aprotinin (5 μg/mL; Thermo Scientific) and phenylmethylsulfonyl fluoride (200 μM; Thermo Scientific) were added to the buffer to prevent protein degradation. Samples were centrifuged at 10,000 g for 5 min at 4°C, and pellets were discarded. Total protein in each supernatant was determined by BCA assay according to manufacturer's instructions (Thermo Scientific).

Immunoblot analysis of WT corneas after alkali injury.
Anesthetized WT mice were unilaterally injured with a 5 μL drop of 0.5 M NaOH on the left ocular surface for 30 s. Saline served as a sham injury. Subcutaneous meloxicam (2 mg/kg) was administered in sham and injured groups immediately after injury and 24 h after injury. Mice received i.p. 15 mg/kg thiorphan or vehicle (5% ethanol in saline) within 1 h after corneal injury and each day thereafter until euthanasia at the end of one week. In order to avoid any effects of manipulation on the ocular surface, mice used for immunoblot analysis were not imaged.
Freshly enucleated eyes were trimmed at the sclerocorneal limbus. The left cornea from each mouse was homogenized (VWR 200 Homogenizer) in 100 μL of cold RIPA buffer containing protease/phosphatase inhibitors. Three gels were required to represent all experimental groups.
Corneal lysates containing equal amounts of protein (9.5 μg), determined by BCA assay, were randomized among the three 7.5% Mini-PROTEAN TGX precast polyacrylamide gels (BioRad), separated by electrophoresis, and transferred to nitrocellulose membranes with the BioRad Trans-Blot Turbo Transfer system. An equal concentration of intermembrane (IM) reference sample (pooled TKE2 lysate) was run on each gel in the same location. Membranes were cut at 50 and 150 kDa and blocked in 5% skim milk in TBST buffer at RT for 1 h. The upper third of each membrane (> 150 kDa) was immunoblotted in rabbit anti-CD45 (ab208022; Abcam) at 1:1000, the middle third (50 -150 kDa) in rabbit anti-CD31 (ab28364; Abcam) at 1:250, and the bottom third (< 50 kDa) in mouse anti-alpha smooth muscle actin (SMA; ab7817; Abcam) at 1:500 overnight at 4°C. The IM reference lane was incubated in rabbit anti-neurokinin-1 receptor (NK1R; ab183713; Abcam) at 1:1000. After washing in TBST, all membranes were incubated with HRP-conjugated secondary antibodies (Abcam) at 1:5000 for 1 h at RT and developed with SuperSignal West Femto Maximum Sensitivity Substrate.
Chemiluminescence was detected using a BioRad ChemiDoc XRS+ system with Image Lab software.
Following chemiluminescent detection, membranes were stained for total protein using amido black according to a published protocol 60 . In brief, membranes were washed in dH2O and submerged in 0.01% (w/v) amido black 10B (Abcam) in 10% acetic acid for 1 min. Membranes were then destained in 5% acetic acid, washed in dH2O, and dried at RT. Dried membranes were imaged using the Epi-White illumination setting with identical exposure times on the ChemiDoc XRS+ system.
For relative quantification, the integrated optical density value (defined as the sum of each background-subtracted pixel value) was determined for equal-sized boxes drawn around bands for each antibody in Image Studio Lite software (LI-COR Biosciences, Lincoln, NE, USA). Median background was calculated from perimeter values around each band of interest. As a loading control, the total amido black signal was determined for equal-sized strips centered in each lane, rather than a rectangle encompassing the entire stained area, to minimize errors due to lane bending 61 . The integrated optical density for each protein of interest was first normalized to the total amido black signal in its corresponding lane, and then to the normalized integrated optical density of the IM control protein, NK1R, from its corresponding membrane to control for intermembrane variability. NK1R bands were similarly normalized to the total amido black signal in corresponding lanes.
Samples were re-randomized and immunoblots were repeated with statistically similar results.
Normalized data from the second set of immunoblots are presented.
Immunoblot analysis of TKE2 cells. TKE2 cells (ECACC 11033107; Sigma-Aldrich, St. Louis, MO, USA), a line of mouse corneal epithelial progenitors isolated from a CD-1 female mouse, were harvested at passage 3, incubated in 150 μL cold RIPA buffer with protease and phosphatase inhibitors (Halt Cocktail; Thermo Scientific), and sonicated on ice. After centrifugation at 13,000 g for 5 min at 4°C, the supernatant was collected and protein concentration was determined by BCA assay according to manufacturer's instructions.
Lysates containing equal amounts of protein (20 μg) were separated on a 10% Mini-PROTEAN TGX precast polyacrylamide gel (BioRad) and transferred to a nitrocellulose membrane with the BioRad Trans-Blot Turbo Transfer system. For detection of NEP and the loading control glyceraldehyde 3-phosphate dehydrogenase (GAPDH), membranes were cut at 70 kDa and blocked in 5% milk in TBST buffer at RT for 1.5 h. The upper half of the membrane (> 70 kDa) was incubated in goat anti-CD10 (NEP; PA5-47075; Invitrogen) at 1:1000, and the lower half (< 70 kDa) was incubated in mouse anti-GAPDH (MAB374; EMD Millipore) at 1:1000, both overnight at 4° C. After washing in TBST, membranes were incubated with HRP-conjugated secondary antibodies (Abcam) at 1:5000 for 1 h at RT and developed with SuperSignal West Femto Maximum Sensitivity Substrate (Thermo Scientific).