Beneficial Effects of Polydeoxyribonucleotide (PDRN) in an In Vitro Model of Fuchs Endothelial Corneal Dystrophy

Fuchs endothelial corneal dystrophy (FECD) is a bilateral, hereditary syndrome characterized by progressive irreversible injury in the corneal endothelium; it is the most frequent cause for corneal transplantation worldwide. Oxidative stress induces the apoptosis of corneal endothelial cells (CECs), and has a crucial function in FECD pathogenesis. The stimulation of the adenosine A2A receptor (A2Ar) inhibits oxidative stress, reduces inflammation and modulates apoptosis. Polydeoxyribonucleotide (PDRN) is a registered drug that acts through adenosine A2Ar. Thus, the goal of this study was to assess the effect of PDRN in an in vitro FECD model. Human Corneal Endothelial Cells (IHCE) were challenged with H2O2 (200 μM) alone or in combination with PDRN (100 μg/mL), PDRN plus ZM241385 (1 μM) as an A2Ar antagonist, and CGS21680 (1 μM) as a well-known A2Ar agonist. H2O2 reduced the cells’ viability and increased the expression of the pro-inflammatory markers NF-κB, IL-6, IL-1β, and TNF-α; by contrast, it decreased the expression of the anti-inflammatory IL-10. Moreover, the pro-apoptotic genes Bax, Caspase-3 and Caspase-8 were concurrently upregulated with a decrease of Bcl-2 expression. PDRN and CGS21680 reverted the negative effects of H2O2. Co-incubation with ZM241385 abolished the effects of PDRN, indicating that A2Ar is involved in the mode of action of PDRN. These data suggest that PDRN defends IHCE cells against H2O2-induced damage, potentially as a result of its antioxidant, anti-inflammatory and antiapoptotic properties, suggesting that PDRN could be used as an FECD therapy.


Introduction
Fuchs endothelial corneal dystrophy (FECD) is a frequent corneal dystrophy with an unmet therapeutic approach. FECD has been documented in approximately 4% of individuals above the age of 40 years. It is characterized by gradual endothelial cell reduction, the thickening of the Descemet membrane, and localized excrescences known as guttae that normally begin in the middle of the cornea and gradually spread to the periphery [1]. Corneal endothelial cells (CECs) are involved in maintaining corneal transparency, deturgescence, and controlling corneal hydration. The loss of endothelial cells during FECD is irreversible because the corneal endothelium (CE) does not divide in vivo [2], thus leading

The Effects of PDRN on Oxidative Stress
A significant increase in ROS levels was observed in IHCE cells as a consequence of H2O2 stimulation (p < 0.0001 vs. the control group; Figure 2A). PDRN and CGS treatment significantly reduced ROS levels compared to untreated IHCE cells, confirming that A2Ar stimulation is involved in the antioxidant effects of PDRN (p < 0.0001 vs. the H2O2 group; Figure 2A). The co-incubation with ZM241385 blunted the antioxidant effects of PDRN, confirming that PDRN acts through A2Ar modulation ( Figure 2A).
In order to better characterize the antioxidant effects of PDRN, the malondialdeyide (MDA) levels were measured in the IHCE cells. Low levels of MDA were detected in the control cells, whereas H2O2 challenge considerably increased the MDA levels (p < 0.0001 vs. the control group; Figure 2B). Both PDRN and CGS21680 decreased MDA production (p < 0.0001 vs. H2O2 group; Figure 2B), while co-incubation with ZM241385 counteracted the beneficial effects of PDRN ( Figure 2B).

The Effects of PDRN on Oxidative Stress
A significant increase in ROS levels was observed in IHCE cells as a consequence of H 2 O 2 stimulation (p < 0.0001 vs. the control group; Figure 2A). PDRN and CGS treatment significantly reduced ROS levels compared to untreated IHCE cells, confirming that A 2Ar stimulation is involved in the antioxidant effects of PDRN (p < 0.0001 vs. the H 2 O 2 group; Figure 2A). The co-incubation with ZM241385 blunted the antioxidant effects of PDRN, confirming that PDRN acts through A 2Ar modulation ( Figure 2A).
In order to better characterize the antioxidant effects of PDRN, the malondialdeyide (MDA) levels were measured in the IHCE cells. Low levels of MDA were detected in the control cells, whereas H 2 O 2 challenge considerably increased the MDA levels (p < 0.0001 vs. the control group; Figure 2B). Both PDRN and CGS21680 decreased MDA production (p < 0.0001 vs. H 2 O 2 group; Figure 2B), while co-incubation with ZM241385 counteracted the beneficial effects of PDRN ( Figure 2B).

The Effects of PDRN on Inflammatory Markers
The gene expression of TNF-α, IL-1β, IL-6 and IL-10 was studied in order t ate the anti-inflammatory effects of PDRN in this experimental model. H2O2 e caused a significant increase in TNF-α, IL-1β and IL-6 gene expression, as well nificant decrease of IL-10 expression compared to untreated cells (p < 0.0001 vs. trol group; Figure 3). PDRN treatment blunted TNF-α, IL-1β and IL-6 expressi upregulated IL-10 mRNA expression compared to the H2O2 group (p < 0.0001 H2O2 group; Figure 3). CGS21680, a selective A2Ar agonist, caused similar effects, ing that A2Ar is involved in inflammatory cascade modulation (p < 0.0001 vs. t group; Figure 3). The co-incubation with the A2Ar antagonist ZM241385 abrogate fects of PDRN, thus corroborating the involvement of A2Ar in PDRN's mechanis tion ( Figure 3).
In addition, the expression of mature proteins was evaluated in order to PDRN's anti-inflammatory effects. TNF-α, IL-1β, p-NF-κB and IL-6 protein exp were markedly increased in H2O2-stimulated cells compared to the untreated c 0.0001 vs. the control group; Figure 4). PDRN treatment dampened TNF-α, IL-1β κB and IL-6 increased expression following H2O2 incubation (p < 0.0001 vs. th group; Figure 4). Additionally, IL-10 protein expression was markedly reduced ing H2O2 incubation compared to the untreated cells (p < 0.0001 vs. the contro Figure 4). PDRN and CGS21680 treatments showed a significant increase of IL tein levels, confirming that PDRN anti-inflammatory effects were related to A2A tion (p < 0.0001 vs. the H2O2 group; Figure 4). The co-incubation with the A2Ar an ZM241385 reverted the effects of PDRN, indicating that A2Ar activation is invo PDRN's mechanism of action ( Figure 4).

The Effects of PDRN on Inflammatory Markers
The gene expression of TNF-α, IL-1β, IL-6 and IL-10 was studied in order to evaluate the anti-inflammatory effects of PDRN in this experimental model. H 2 O 2 exposure caused a significant increase in TNF-α, IL-1β and IL-6 gene expression, as well as a significant decrease of IL-10 expression compared to untreated cells (p < 0.0001 vs. the control group; Figure 3). PDRN treatment blunted TNF-α, IL-1β and IL-6 expression, and upregulated IL-10 mRNA expression compared to the H 2 O 2 group (p < 0.0001 vs. the H 2 O 2 group; Figure 3). CGS21680, a selective A 2Ar agonist, caused similar effects, indicating that A 2Ar is involved in inflammatory cascade modulation (p < 0.0001 vs. the H 2 O 2 group; Figure 3). The co-incubation with the A 2Ar antagonist ZM241385 abrogated the effects of PDRN, thus corroborating the involvement of A 2Ar in PDRN's mechanism of action ( Figure 3).
In addition, the expression of mature proteins was evaluated in order to confirm PDRN's anti-inflammatory effects. TNF-α, IL-1β, p-NF-κB and IL-6 protein expression were markedly increased in H 2 O 2 -stimulated cells compared to the untreated cells (p < 0.0001 vs. the control group; Figure 4). PDRN treatment dampened TNF-α, IL-1β, p-NF-κB and IL-6 increased expression following H 2 O 2 incubation (p < 0.0001 vs. the H 2 O 2 group; Figure 4). Additionally, IL-10 protein expression was markedly reduced following H 2 O 2 incubation compared to the untreated cells (p < 0.0001 vs. the control group; Figure 4). PDRN and CGS21680 treatments showed a significant increase of IL-10 protein levels, confirming that PDRN anti-inflammatory effects were related to A 2Ar activation (p < 0.0001 vs. the H 2 O 2 group; Figure 4). The co-incubation with the A 2Ar antagonist ZM241385 reverted the effects of PDRN, indicating that A 2Ar activation is involved in PDRN's mechanism of action ( Figure 4).

Effects of PDRN on Apoptosis
The gene expressions of Bcl-2, Bax, Caspase-3 and Caspase-8 were studied in order to evaluate the apoptosis process. IHCE cells challenged with H2O2 for 2 h exhibited a significant down-regulation of Bcl-2 mRNA levels with a simultaneous increase of Bax,

Effects of PDRN on Apoptosis
The gene expressions of Bcl-2, Bax, Caspase-3 and Caspase-8 were studied in order to evaluate the apoptosis process. IHCE cells challenged with H2O2 for 2 h exhibited a significant down-regulation of Bcl-2 mRNA levels with a simultaneous increase of Bax,

Effects of PDRN on Apoptosis
The gene expressions of Bcl-2, Bax, Caspase-3 and Caspase-8 were studied in order to evaluate the apoptosis process. IHCE cells challenged with H 2 O 2 for 2 h exhibited a significant down-regulation of Bcl-2 mRNA levels with a simultaneous increase of Bax, Caspase-3 and Caspase-8 compared to the untreated cells (p < 0.0001 vs. the control group; Figure 5). On the other hand, PDRN treatment for 24 h completely inverted the gene expression of Bcl-2, Bax, Caspase-3 and Caspase-8 (p < 0.0001 vs. the H 2 O 2 group; Figure 5). CGS21680, a selective A 2Ar agonist, showed similar effects, thus confirming the role of A 2Ar activation in the modulation of the apoptotic process (p < 0.0001 vs. H 2 O 2 group; Figure 5). The A 2Ar antagonist ZM241385 counteracted the effects of PDRN, highlighting the role of A 2Ar in PDRN's mode of action ( Figure 5). In order to confirm the antiapoptotic effect of PDRN, we also evaluated the mature protein. A significant decrease of the Bcl-2 protein levels together with a concurrent increase of Bax, Caspase-3 and Caspase-8 were detected following H 2 O 2 challenge compared to the untreated cells (p < 0.0001 vs. the control group; Figure 6). PDRN or CGS21680 treatment for 24 h caused a significant increase of Bcl-2 protein expression with a concomitant decrease of Bax, Caspase-3 and Caspase-8 protein levels (p < 0.0001 vs. the H 2 O 2 group; Figure 6). ZM241385 blocked the effects of PDRN, again confirming the role of A 2Ar activation in apoptosis modulation ( Figure 6).
Caspase-3 and Caspase-8 compared to the untreated cells (p < 0.0001 vs. the control group; Figure 5). On the other hand, PDRN treatment for 24 h completely inverted the gene expression of Bcl-2, Bax, Caspase-3 and Caspase-8 (p < 0.0001 vs. the H2O2 group; Figure 5). CGS21680, a selective A2Ar agonist, showed similar effects, thus confirming the role of A2Ar activation in the modulation of the apoptotic process (p < 0.0001 vs. H2O2 group; Figure 5). The A2Ar antagonist ZM241385 counteracted the effects of PDRN, highlighting the role of A2Ar in PDRN's mode of action ( Figure 5).
In order to confirm the antiapoptotic effect of PDRN, we also evaluated the mature protein. A significant decrease of the Bcl-2 protein levels together with a concurrent increase of Bax, Caspase-3 and Caspase-8 were detected following H2O2 challenge compared to the untreated cells (p < 0.0001 vs. the control group; Figure 6). PDRN or CGS21680 treatment for 24 h caused a significant increase of Bcl-2 protein expression with a concomitant decrease of Bax, Caspase-3 and Caspase-8 protein levels (p < 0.0001 vs. the H2O2 group; Figure 6). ZM241385 blocked the effects of PDRN, again confirming the role of A2Ar activation in apoptosis modulation ( Figure 6).

Discussion
FECD is a bilateral heterogeneous disease characterized by the progressive loss of CECs, the production of localized excrescences known as guttae in the Descemet membrane, corneal oedema and impaired vision [3]. The loss of CECs observed in FECD is irreversible, as CE does not divide in vivo. Corneal transplantation is the only therapy option for the restoration of impaired vision, making FECD the most common cause of corneal-endothelial transplants worldwide [22]. Oxidative stress plays a key role in a variety of ocular disorders; in particular, chronic oxidative stress is likely to contribute to

Discussion
FECD is a bilateral heterogeneous disease characterized by the progressive loss of CECs, the production of localized excrescences known as guttae in the Descemet membrane, corneal oedema and impaired vision [3]. The loss of CECs observed in FECD is irreversible, as CE does not divide in vivo. Corneal transplantation is the only therapy option for the restoration of impaired vision, making FECD the most common cause of cornealendothelial transplants worldwide [22]. Oxidative stress plays a key role in a variety of ocular disorders; in particular, chronic oxidative stress is likely to contribute to cellular and molecular damage in CECs, inducing apoptosis that causes CE degeneration, thus leading to FECD [6,23]. Previous papers have demonstrated that adenosine A 2Ar activation prevents ROS generation, also reducing inflammation and the apoptotic process [24,25]. PDRN, which is recognized as an A 2Ar agonist, showed several properties-including antioxidant, anti-apoptotic and anti-inflammatory ones-in several pre-clinical in vivo disease models, such as airway inflammation, cerebral ischemia, interstitial cystitis, acute lung injury, and ischemic colitis [26][27][28][29][30]. Moreover, a previous clinical trial demonstrated that PDRN eye drops stimulated corneal epithelium regeneration, thus supporting the hypothesis that it could be used for the treatment of corneal diseases [15,18]. Therefore, in the present study, the efficacy of PDRN was evaluated in an in vitro FECD model induced by H 2 O 2 stimulation. Previous studies have demonstrated that hydrogen peroxide stimulation may promote the functional and structural changes that occur in corneal endothelial cells in FECD [6,31]. Furthermore, in this experimental setting, CECs challenged with H 2 O 2 showed an increase of oxidative stress, inflammatory and apoptotic processes, in accordance with previous papers [32,33]. PDRN treatment drastically reduced oxidative stress markers such as ROS content and MDA levels compared to the H 2 O 2 group. By contrast, these effects were abolished when the A 2Ar antagonist ZM241385 was used together with PDRN, thus demonstrating that its antioxidant effects occurred through adenosine A 2Ar stimulation. Moreover, CECs challenged with H 2 O 2 and treated with the specific A 2Ar agonist CGS21680 showed a marked reduction of either ROS or MDA levels, thus confirming the involvement of A 2Ar in oxidative stress suppression in these experimental conditions. These results are consistent with previous articles that showed the protective effects of PDRN against ROS-related diseases, such as IBD [13], supporting the idea that the implementation of an antioxidant system through adenosine A 2Ar activation may also play a role in the treatment of diseases characterized by oxidative stress, such as FECD.
Moreover, H 2 O 2 stimulation activated the transcriptional factor NF-κB, which in turn enhanced the expression of the pro-inflammatory cytokines TNF-α, IL-1β and IL-6, with a concomitant downregulation of the anti-inflammatory cytokine IL-10. PDRN significantly reduced pro-inflammatory cytokine expression and increased IL-10 levels compared to the untreated H 2 O 2 group of cells. CECs treated with CGS21680 showed similar results, while ZM241385 co-incubation inhibited the beneficial effects of PDRN, demonstrating again that PDRN activity occurs through A 2Ar stimulation. These findings are consistent with earlier data, which have characterized well the anti-inflammatory effects mediated by A 2Ar in several preclinical inflammatory models [34,35], and with recent papers that pointed out the inactivation of the NF-κB pathway mediated by PDRN [36,37].
The activation of the oxidative stress mechanism may be also responsible for the induction of cell death processes; in fact, H 2 O 2 stimulus induced apoptosis in CECs. PDRN treatment decreased the expression of the proteins involved in apoptosis-such as Bax, Caspase-3 and Caspase-8-whereas it increased Bcl-2 levels compared to untreated H 2 O 2stimulated cells. Furthermore, cells treated with CGS21680 showed a modulation of the apoptotic process; conversely, by adding ZM241385 to PDRN-treated cells, the positive effects observed when the compound was used alone were abolished, thus confirming that PDRN acts through modulation of adenosine A 2Ar . As mentioned, these results are in line with the data obtained from previous studies, and indicate that adenosine A 2Ar modulation might prevent the worsening of oxidative stress-related diseases in eye models [38,39].
These preliminary findings are intriguing, and point to new therapeutic alternatives. However, a major limitation of the study is that the effects of PDRN were evaluated in an in vitro FECD model; hence, additional in vivo research would be required in order to characterize its efficacy. Instead, PDRN might be deemed to be a novel treatment for FECD, because this adenosine A 2Ar agonist is already on the market for a variety of uses and it could be readily available for a clinical trial in FECD patients. Moreover, previous works demonstrated that the half-life of PDRN is around 12-17 h, indicating that it may be suitable for once-daily administration in ordinary clinical practice [13]. Finally, it should be pointed out that PDRN is already commercially available, it is well tolerated, and it showed a very good safety profile across several clinical trials and in a variety of therapeutic applications [40][41][42]. These intriguing preclinical findings, in light of PDRN's strong translational potential, should be confirmed in a FECD clinical context.

Cell Cultures
Human Corneal Endothelial Cells (IHCE) were purchased by Creative Biolabs neuroS London, UK. The IHCE cells were cultured in PriNeu I medium enriched with 10% Fetal Bovine Serum (FBS), recombinant human epidermal growth factor (5 ng/mL), a 5 µg/mL insulin plus 1% antibiotic mixture (penicillin/streptomycin), and G418 antibiotic in a humidified incubator at 37 • C and with a percentage of 5% CO 2 .

Cell Treatments
The IHCE cells were cultured in 6-well plates at a density of 3 × 10 6

FDA/PI Staining
The cells were plated at a density of 5 × 10 5 cells/well in a 24-well plate, and were incubated with H 2 O 2 (200 µM) for 2 h; then, the cells were treated with PDRN (100 µg/mL), CGS21680 (1 µM) and PDRN + ZM241385 (1 µM) for 24 h. At the end of the treatment period, the cells were washed in sterile PBS and stained with the FDA/PI staining solution, with the addition of 3.2 µL FDA and 20 µL PI at the concentrations of 5mg/mL and 2 mg/mL, respectively, in 2 mL culture medium without FBS, for each well, for 5 min at room temperature in the dark. The viable cells were observed with a fluorescence microscope. ImageJ software for Windows was used to calculate the number of positive cells (Softonic, Barcelona, Spain).

ROS Measurement
In order to evaluate the effects of PDRN on oxidative stress, the production of total Reactive Oxygen Species (ROS) in the IHCE cells was measured using an assay kit (Thermo Fisher, Carlsbad, CA, USA), as previously described in detail [31].

Malondialdehyde Assay
The antioxidant effect of PDRN in IHCE cells was evaluated by measuring the malondialdehyde (MDA) levels, as previously described in detail [31].

Western Blot Analysis
After 24 h of treatment, the IHCE cells were collected and protein extraction was performed to evaluate pNF-κB, Bax, Bcl-2, Caspase-3 and Caspase-8 expression by Western Blot analysis, as previously described in detail [46].

Statistical Analysis
The data presented are the results of at least five experiments, and are expressed as the mean ± SD. In order to guarantee repeatability, all of the assays were carried out in duplicate. The differences between the groups were evaluated and analysed using one-way ANOVA with the Tukey post-test. A p value of less than 0.05 was considered significant. SPSS Statistics for Windows v22.0 was used for the statistical analysis (SPSS, Inc, Chicago, IL, USA), and GraphPad Prism was used to create the graphs (Version 8.0 for macOS, San Diego, CA, USA).

Conclusions
In conclusion, we demonstrated for the first time that PDRN through A 2Ar stimulation defends IHCE cells against H 2 O 2 -induced damage, potentially as a result of its antioxidant, anti-inflammatory and antiapoptotic properties, suggesting that PDRN could be used as an FECD therapy.