Normal pooled adult Human Epidermal Keratinocytes (NHK) were obtained from PromoCell (cat # C-12006) maintained and passaged in serum free Keratinocyte complete growth medium (PromoCell, cat # C-20111). Early passage cells (P4-P6) were used for all our studies. To induce senescence and asses the effect of AT compounds on senescence phenotype we treated cells either with 200 nM Mitomycin (cat # 10107409001) for 24 h, 25 mM of D-glucose to mimic high glucose conditions and 10 µM of H₂O₂ for 48 h or 400 µM for 1 h followed by washout and consequently cells were analyzed for senescence markers at day 10. In treatment groups, AT compounds (AT-001, AT-003, AT-007) were added at a concentration of 0.2 nM 10 min prior to senescence induction. For all compounds, 1 mM stock solutions were prepared in DMSO, consequent dilutions to 0.2 µM stocks were prepared in PBS under sterile conditions. Vehicle also referred to as Veh which was added to control cells were prepared accordingly. The dose used is calculated based on EC50 for all AT compounds (Shendelman, 2018; Wasmuth et al., 2013; Wasmuth et al., 2014).

Colorimetric assessment of human AR activity was performed using AR Activity Kit (BioVision, Inc. cat # K369) as per manufacturer’s instructions. Absorbance was measured at 340 nm for a duration of 60 min at a regular time interval of 90 s using TECAN Infinity Pro 200 plate reader. Calculations were performed as per manufacturer’s instructions. The resultant data was normalized to total protein, which was assessed using Pierce™ BCA Protein Assay Kit (Thermo Scientific™ cat # 23225).

NHK cells were plated in standard six well plates. SA-β-gal staining was performed using Cellular Senescence Staining Kit obtained from Cell Biolabs, Inc. (cat# CBA230) as per manufacturer guidelines. In brief, cells were washed and fixed by the fixative provided by the kit. Cells were incubated with SA-β-gal staining solution for 16 h. Cells were counter stained with Syto-13 (ThermoFisher Scientific cat# S7575) dye to obtain accurate cell count. Alternatively, florescent based approach was used to measure cellular senescence using CellEvent™ Senescence Green Detection Kit from ThermoFisher Scientific (Cat# C10851). The experiment was performed as per manufacturer’s guidelines. Cells were counter stained with DAPI to obtain accurate cell count. Images were taken at ×10 magnification using Ziess Axio-observer light and florescent microscope. For CellEvent™ Senescence Green Detection Kit, images were taking at 488 nm filter set. Cell count was performed using NIH ImageJ software.

NHK cells were plated in 12 wells plates. Post respective treatments, cells were stained with live cell superoxide indicators. For detecting cytosolic superoxide, cells were incubated with 5 µM dihydroethidium (ThermoFisher Scientific Inc. cat # D1168) for 30 min. To detect mitochondrial superoxide, cells were incubated with 5 µM MitoSOX dye (ThermoFisher Scientific, cat# M36008) for 10 min. Cells were later washed with PBS to remove excess dye and fixed with 3.7% paraformaldehyde (PFA). Imaging was performed at ×10 magnification using EVOS XL Core Imaging System. Relative intensity for images was calculated using NIH ImageJ software.

RNA from treated cells was extracted using RNeasy Plus Mini Kit (QIAGEN, cat#74134) and cDNA was synthesized using qScript™ cDNA SuperMix (Quanta BioSciences, Inc. cat # 95048). RT-qPCR was performed using Taqman gene expression assays. 20 ng of RNA input per reaction was used for all qPCR studies. Amplification reactions were performed by Applied Biosystems 7500 Fast Real-Time PCR System. Relative fold change in expression for desired genes was performed using the ΔCT method of analysis. All gene expression data were normalized to housekeeping gene PPIA. Genes used to determine expression are listed in Table 1.

ELISA was performed to determine cytokine release into Condition medium. Condition medium at day 10 was collected from skin cells treated H₂O₂ as described above. ELISA was performed for cytokines released in condition medium for IL1β (Abcam, cat # ab214025), TNFα (Abcam, cat # ab181421), MCP1 (Abcam, cat # ab179886) and IL6 (Abcam, cat # ab178013) as per manufacturers protocols. All data was normalized to total protein content assessed by BCA assay. Absorbance as described in the protocol was measured using TECAN infinity pro 200 plate reader.

To measure DNA damage in NHK cells treated with long term H₂O₂, Gamma H2A.X Staining Kit (ABCAM; cat # ab242296) was used. Cells were stained according to company protocol followed by florescent imaging using EVOS XL Core Imaging System at ×10 magnification. Relative intensity and cell count for images was calculated using NIH ImageJ software.

All data was analyzed using GraphPad PRISM 9 software (GraphPad, LaJolla, CA). n represents biological replicates obtained from two consecutive batches performed at two different time points with cells originating from separate source. Each batch comprised of n = 2 obtained from two separate culture dishes or wells. All data are expressed as mean ± SEM. Unpaired t-test was performed to compare data between two groups and Ordinary one-way ANOVA with Tukey’s multiple comparisons test was implemented for comparison between more than two groups. p < 0.05 was considered statistically significant between groups.

Previous studies have demonstrated the deleterious effect of the activity of the polyol pathway in accelerating keratinocyte senescence (Kang et al., 2011). To determine if senescence is linked to changes in AR activity, we induced senescence by treating human keratinocytes either with mitomycin C (MMC), H₂O₂ or high glucose (HG) and measured AR activity. We observed a 3-fold increase in AR activity upon treatment with 10 µM H₂O₂ or 25 mM D-glucose compared to controls (Figures 1A, B) and a 2.5-fold increase upon treatment with 200 nM MMC (Figure 1C). These data link increases in activity of polyol pathway enzyme AR in keratinocyte senescence.

The main focus of our study is to test the preventive role of three different, novel and highly-selective ARIs, AT-001, AT-003 and AT-007 in induced keratinocyte senescence. We initially tested if AT-001, which is the parent drug for AT-003 and AT-007 can inhibit AR activity in normal keratinocytes and our data clearly demonstrated more than 75% reduction in AR activity upon AT-001 treatment compared to vehicle treatment (Figure 1D). SA-β-gal staining, gene expression of CDKN1A, TP53 and SERPINE and both cytosolic and mitochondrial (mito) ROS superoxide production are well-established gold standard markers for cellular senescence. To test our hypothesis, we measured these key markers to confirm the preventive role of ARIs on human keratinocyte senescence. As it was evident that all three senescence-inducing agents significantly increased AR activity, we next wanted to investigate if ARIs prevented HG induced senescence. Treatment with HG for 48 h induced senescence by 30% as demonstrated by SA-β-gal staining (Figure 2A). Treatment with all three compounds 10 min prior to addition of HG significantly prevented HG induced senescence, of which AT-001 and AT-003 demonstrated maximum inhibition. Morphology observed by cell tracker dye Syto-13^® displayed fewer flattened cells (typical morphology of senescent cells) in ARI treatment groups compared to the control vehicle-treated HG group (Figure 2A). We next confirmed significant downregulation of HG increased gene expression CDKN1A, TP53 and SERPINE by all three ARIs (Figure 2B). Furthermore, HG induced cytosolic and mito superoxide generation, which was measured by DHE and MitoSOX dye respectively was also strongly inhibited by AT-001, AT-003 and AT-007 (Figures 2C, D). Our data above provides strong evidence and emphasizes the importance of these compounds in preventing HG induced cellular senescence.

H₂O₂ is a well-known chemical agent, which causes senescence and increases reactive oxygen species generation in cells (Ramis et al., 1989). We next tested the efficacy of ARIs in preventing H₂O₂ induced senescence phenotype. Exposure to 10 µM H₂O₂ for 48 h significantly upregulated all the key markers of senescence tested with a 3-fold increase in SA-β-gal positive keratinocytes (Figure 3A). Treatment with the ARI compounds 10 min prior to addition of H₂O₂ significantly inhibited H₂O₂ induced senescence phenotype of which AT-001 completely abolished H₂O₂ induced SA-β-gal positive staining (Figure 3A). H₂O₂ induced DHE and MitoSOX staining was significantly inhibited by AT-001, AT-003 and AT-007 (Figures 3B, C). Gene expression of CDKN1A, TP53 and SERPINE was consistently downregulated by all three ARIs (Figure 3D). AT-001 is the parent drug for other derivatives AT-003 and AT-007. We next wanted to investigate the efficacy of AT-001 on long term treatment of H₂O_2. As described in the methods, keratinocytes were initially treated with 400 µM H₂O₂ for 1 h with or without AT-001 and the cells were allowed to be incubated in AT-001 for 10 days. At day 10 we observed a strong increase in fluorescence signal in cells treated only with H₂O₂ compared to vehicle treated group for γ-H2AX stain which is a marker for DNA damage and impaired cell proliferation. The fluorescence signal was abolished in cells co-treated with AT-001 and H₂O_2. (Figure 3E). Alternatively, we have used florescent based SA-β-galactosidase staining approach to detect senescence in these conditions. As observed in Figure 3F, AT-001 prevented florescent senescence signal induced by long term H₂O₂ treatment. We then assessed the SASP release into condition medium in keratinocytes upon H₂O₂ treatment. We observed significantly higher release of SASP IL1β, IL6, MCP1 and TNFα in H₂O₂ treated cells and that treatment with AT-001 reduced the increases in SASP (Figure 3G). Long term exposure of H₂O₂ significantly increased gene expression of CDKN1A, TP53, SERPINE, CDKN2A and well-established keratinocyte senescence marker LMNB1 ( Figure 3H), AT-001 treatment blocked the increases in all these senescent marker genes (Figure 3H).

To strengthen our hypothesis, we have used a third approach which is drug induced accelerated senescence approach. Here, we treated with 200 µM of MMC for 24 h which induced approximately 35% SA-β-gal positive keratinocytes (Figure 4A). Consistent with our above finding, all three compounds significantly prevented MMC induced senescence (Figure 4A). In addition, all three compounds significantly downregulated MMC induced gene expression of CDKN1A, TP53 and SERPINE. MMC treatment induced cytosolic and mito superoxide production was consistently prevented with AT-001, AT-003 and AT-007 (Figures 4B–D). Taken together our data demonstrates that all three ARI compounds were effective in attenuating human keratinocyte senescence.