The following reagents and consumables have been used for this study: hyaluronic acid (HA) MW 850 kDa (Novozymes, Denmark), sodium hydroxide (NaOH, VWR, Radnor, PA), iodoacetic acid (ThermoScientific, Waltham, MA), isopropanol (VWR, Radnor, PA), Whatman filter paper (Fisher, Waltham, MA), hydrochloric acid 6N (HCl, Fisher, Waltham, MA), Slide-A-Lyzer Dialysis Cassettes, 3500 MWCO (ThermoScientific, Waltham, MA), deuterated water (D₂O, ThermoScientific, Waltham, MA), d-Methionine (D-Met, Waltham, MA), 2-(N-morpholino)ethanesulfonic acid buffer (MES, ThermoScientific, Waltham, MA), 1-ethyl-3-(3- dimethylaminopropyl) carbodiimide (EDC, ThermoScientific, Waltham, MA), lithium citrate buffer (FUJIFILM Wako, Osaka, Japan), Agilent PL Aquagel-OH Mixed H larger molecular weight SEC column (Agilent Technologies, Santa Clara, CA), Agilent PL Aquagel-OH SEC column (Agilent Technologies, Santa Clara, CA).

The following cell lines, cell culture reagents and assay kits have been used for this study: House Ear Institute-Organ of Corti (HEI-OC1) and Stria Vascularis k1 (SV-k1) (Kalinec Lab, UCLA, Los Angeles, CA), three-dimensional (3D) in vitro tissue models, Epi-Airway day 7 (MatTek, Ashland, MA), fetal bovine serum (FBS, Corning, Corning, NY), Dulbecco’s Modified Eagle Medium (DMEM, Corning, Corning, NY), Cell-Titer 96 Aqueous One Solution Cell Proliferation Assay (Promega, Madison, WI), CyQUANT NF Cell Proliferation Assay Kit (Molecular Probes, Eugene, OR), phosphate buffered-saline (PBS, Corning, Corning, NY), hydrogen peroxide (H₂O₂, Millipore Sigma, St. Louis, MO), chloromethyl derivative of 2′,7′-dichlorodihydrofluorescein diacetate (CM-H2DCFDA, Invitrogen, Waltham, MA), CellEvent Caspase-3/7 Green Detection Reagent (ThermoFisher, Waltham, MA), Trypsin-ethylenediaminetetraacetic acid (trypsin-EDTA, Corning, Corning, NY), NADP/NADPH Assay Kit (Abcam, Waltham, MA), 10 kDa column (Biovision, Milpitas, CA), fluorometric methionine assay kit (Biovision, CA), Ibidi μ -Slide 8 well (Ibidi, Fitchburg, WI), boron-dipyrromethene-labeled HA (HA-BODIPY, Echelon Biosciences, Salt Lake City, UT), MitoTracker Red (Invitrogen, Waltham, MA), LysoTracker Red (Invitrogen, Waltham, MA), bovine serum albumin (BSA, Corning, Corning, NY), formaldehyde solution (VWR, Radnor, PA), 4′,6-diamidino-2-phenylindole nuclear stain (DAPI, ThermoFisher, Waltham, MA).

The following instruments have been used for this study: Varian VNMRS 500 MHz nuclear magnetic resonance spectrometer (Agilent Technologies, Palo Alto, CA), High Speed Amino Acid Analyzer L-8900 (Hitachi, Tokyo, Japan), FilterMax F5 multimode microplate reader (Molecular Devices, San Jose, CA), Leica Stellaris 5 Confocal Microscope (Leica, Germany), Agilent High Pressure Liquid Chromatography instrument with a UV-VIS detector (Agilent, Santa Clara, CA) additionally equipped with an OptiLab differential refractive index (RI) detector (Wyatt Technologies, Santa Barbara, CA), and a miniDawn 3-angle/multi-angle light scattering (MALS) detector (Wyatt Technologies, Santa Barbara, CA).

HA was derivatized with carboxymethyl functionalities to increase the conjugation sites available for the covalent attachment of molecules containing primary amines. Briefly, 2 g of HA, MW 850 kDa were added to 20 ml of 45% w/v NaOH and allowed to activate at room temperature (RT) for 2 h. In parallel, 2 g of iodoacetic acid were dissolved in 50 ml of isopropanol. The viscous HA solution was added to 150 ml of isopropanol, then the iodoacetic acid solution was added to the activated HA/isopropanol solution. The reaction mix was allowed to react for 2 h and was then filtered using a Buchner funnel with a #2 Whatman filter paper. The white filter cake obtained after filtration was dissolved in 200 ml of deionized water (diH₂O) and the pH of the resulting solution was neutralized using 6N HCl. The resulting CMHA solution was then loaded into 3500 MWCO Slide-A-Lyzer Dialysis Cassettes and dialyzed for 72 h with a minimum of five water changes per 24 h to remove residual reagent and salts. After dialysis the CMHA solution was removed from the cassettes, frozen by placement in a −80°C freezer for a minimum of 4 h and subsequently lyophilized. The reaction yielded 1.580 g of CMHA. The carboxymethylation was confirmed by proton nuclear magnetic resonance (¹H-NMR).

D-Met was chemically conjugated to CMHA using carbodiimide chemistry. Briefly, 1 g of CMHA was dissolved into 200 ml of MES buffer in a 450 ml beaker covered with parafilm. The solution was stirred for approximately 25 min until CMHA was fully dissolved. Next, 4.4 g of D-Met were added to the solution and allowed to dissolve (∼5 min). For the covalent conjugation of the antioxidant to CMHA, 2 g of EDC, a zero-length crosslinker, was added to the solution and allowed to react at RT for 20 h. Subsequently, the reaction was neutralized to a pH of 7.0 with an NaOH solution. The reaction mix was then transferred to 3500 MWCO Slide-A-Lyzer Dialysis Cassettes and dialyzed for 72 h with a minimum of five water changes per 24 h to remove residual D-Met. After dialysis the HAM solution was removed from the cassettes, frozen by placement in a −80 °C freezer for a minimum of 4 h and subsequently lyophilized. The reaction yielded 1.214 g of HAM. The conjugation was confirmed by ¹H-NMR.

To confirm the structures and purity of CMHA and HAM, 1H-NMR spectral data were obtained using a Varian VNMRS 500 MHz at 20°C. For the analyses, samples were dissolved in deuterium oxide at a concentration of 5 mg/ml and all spectra were referenced to the residual solvent peak (H2O) at δ =   4 . 65  ppm .

HEI-OC1 and SV-k1 were seeded in a 96-well plate with a seeding density of 1.5 × 10⁵ cells/mL in 100 μ L of DMEM media with 2% FBS and incubated at 33°C/10% CO₂ overnight (Kalinec et al., 2016). Subsequently the media was aspirated, replaced with sterile filtered media containing HAM conjugate and controls (CMHA only, D-Met only and CMHA/D-Met blend, all at concentrations equivalent to HAM) and plates were incubated overnight. The viability and proliferation of the cells was quantified with a Cell-Titer 96 Aqueous One Solution Cell Proliferation Assay and the absorbance at 450 nm was measured with a FilterMax F5 multimode microplate reader.

HEI-OC1 and SV-k1 cells were used to assess the proliferation of cells treated with the HAM conjugate and controls. Cells were seeded in a 96-well plate with a seeding density of 1.5 × 10⁵ cells/mL in 100   μ L of DMEM media with 2% FBS and incubated at 33°C/10% CO₂. The cells were allowed to adhere overnight, and the media was aspirated and replaced with sterile filtered media containing HAM conjugate and controls then incubated overnight. The proliferation of cells was quantified with a CyQUANT NF Cell Proliferation Assay Kit and the florescence (ex/em 485/530 nm) measured using the FilterMax F5 multimode microplate reader.

HEI-OC1 and SV-k1 cells, respectively, were seeded in a 96-well plate at a seeding density of 2 × 10⁵ cells/mL in DMEM media with 10% FBS and incubated at 33°C/10% CO₂. The cells were allowed to adhere 24 h, then the media was aspirated and replaced with sterile filtered media containing HAM conjugates and controls, followed by incubation for 24 h. The media was then aspirated and replaced with 1   μ M CM-H2DCFDA in PBS and incubated for and incubated for 30 min. The experimental conditions used for the ROS assays associated with this project were experimentally determined by our group to yield appropriate stress responses without causing cell detachment from the culture plates. The reagent was then aspirated and replaced with either PBS for control or 0.5 mM H₂O₂ in PBS and incubated for 30 min. Media was then aspirated, and wells were washed 2 times with PBS. The amount of ROS was quantified using florescence (ex/em 495/530 nm) with the FilterMax F5 multimode microplate reader.

HEI-OC1 and SV-k1, respectively, were seeded in a 96-well plate with a seeding density of 2 × 10⁵ cells/mL in DMEM media with 10% FBS and incubated at 33°C/10% CO₂. The cells were allowed to adhere 24 h, then the media was aspirated and replaced with sterile filtered media containing HAM conjugate and controls and then incubated for 24 h. The media was then aspirated and replaced with either PBS for control or 0.5 mM H₂O₂ in PBS and incubated for 30 min. Media was then aspirated, and wells were washed 2 times with PBS. DMEM with 10% FBS was added to each well and the plate was incubated for 48 h at 33°C/10% CO₂. The amount of apoptosis was assessed with 4 μ M CellEvent Caspase-3/7 Green Detection Reagent and the florescence (ex/em 495/530 nm) was read using the FilterMax F5 multimode microplate reader.

HEI-OC1 and SV-k1, respectively, were seeded in a 75 cm² flask with a seeding density of 1 × 10⁶ cells/mL in DMEM media with 10% FBS and incubated at 33°C/10% CO₂ for 48 h. Media was then aspirated and replaced with sterile filtered media containing HAM conjugate and controls and incubated for 24 h. Treatment solutions were aspirated, and flasks were washed with 1x PBS two times. Cells were then stressed with fresh 0.5 mM H₂O₂ or PBS only for controls for 30 min. Stressing solution was then aspirated and flasks were washed two times with 1x PBS. Cells were detached using 0.25% trypsin-EDTA. Cells and and trypsin solution was centrifuged at 300 RCF for 5 min and cells were counted using Countess Cell Counter. Cell numbers for each group were normalized to 4 × 10⁶ cells. Cells were washed two times with 1x PBS to remove any remaining trypsin then processed according to the NADP/NADPH Assay Kit manufacture’s protocol for cellular lysates. Sample quantification was performed via absorbance detection at 450 nm with a microplate reader.

A fluorometric methionine assay kit was used for this. HEI-OC1 and SV-k1, respectively, were seeded in a 12 well plate with 0.5 ml of cell solution seeded at 1 × 10⁶ cells/mL in DMEM media with 10% FBS and incubated at 33°C/10% CO₂ for 48 h. Media was then aspirated and replaced with sterile filtered media containing HAM conjugate and controls and incubated for 24 h. Cells were detached with 0.25% trypsin-EDTA and pelleted via centrifugation at 300 RCF. The cell pellets were resuspended in 1x PBS two times to remove any traces of trypsin. Cells were then counted with Countess Cell Counter, and each treatment group was normalized to 2 × 10⁶ cells. Subsequently, 500   μ L of assay kit buffer were added to each treatment group and vortexed. The solutions were then centrifuged at 17,000 RCF for 15 min. The supernatants were transferred to 10 kDa column and centrifuged for 10 min at 10,000 RCF. The ultrafiltrates were collected and further processed per the manufacturer’s protocol using cell lysates. Samples were analyzed via detection at fluorescence (ex/em 535/595 nm) on a microplate reader.

HEI-OC1 and SV-k1, respectively, were plated at 5 × 10⁴ cells/mL with 200   μ L into each well of an Ibidi μ -Slide 8 well. Cells were incubated for 48 h at 33°C/10% CO₂. Media was aspirated and 200 μL of 50 μM HA BODIPY, 50 nM MitoTracker Red, 50 nM LysoTracker Red and 1x PBS with 0.1% BSA were added to respective wells for 2.5 h and incubated at 33°C/10% CO_2. Following incubation, organelle stains and HA were aspirated, and wells were washed two times with 1x PBS with 0.1% BSA. Cells were then fixed with 10% formaldehyde solution for 1 h and incubated at RT. The formaldehyde was then removed, and wells were washed two times with 1x PBS and 0.1% BSA, followed by the addition of 1 μ M DAPI solution to each well and incubation at RT for 5 min. Subsequently, DAPI was removed, wells were washed two times with 1x PBS, and 200 μ L of 1x PBS were added to every well for imaging with a Leica Stellaris 5 confocal microscope, by using the ×40 objective.

An Epi-Airway Day 7 3D in vitro tissue model was used as mimetic of the round window membrane (Rask-Andersen et al., 2015; Liebesny et al., 2018). HAM conjugate solution at 50 mg/ml and individual controls diluted in 1x PBS, respectively, were added to the top of Epi-Airway Day 7 3D tissues in transwell plates, with 500   μ L of 1x PBS in the receiver compartment of a 12 well plate. The plate and tissues were incubated at 37 °C/5% CO₂ for 24 h. After 24 h the receiver solution was collected and filtered with a 0 . 1   μ m filter prior to size exclusion chromatography (SEC) analyses.

The filtered samples were immediately frozen until ready to use. For analyses, 50   μ L of filtered sample were run through an Agilent PL Aquagel-OH Mixed H SEC column for higher molecular weights (detection range: 6 kDa - 10 MDa) for HAM or CMHA detection and analysis, and collections of each sample run were conducted between 8 and 11 min of the 22-min total run time. These collections were then frozen until later use. After all samples were run through the SEC column for larger MW, 50   μ L of each collection were run through an Agilent PL Aquagel-OH SEC column for lower molecular weights (detection range: 100 Da–20 kDa) to detect for D-Met (MW = 149 Da). The concentrations of HAM and CMHA, respectively, that permeated through the tissue models of the larger MW SEC column where quantified with the RI detector. Additionally, tissue permeants were also analyzed with the UV-VIS detector which specifically detects D-Met at 235 nm. Standard curves of HAM, CMHA and D-Met, respectively, were created via area under the curve at 235 nm.

All experiments were run in triplicates or greater and all data were presented as mean ± standard deviation of the mean. Student t-tests were performed for 2-group comparisons while one-way ANOVA with multiple comparisons were performed for more than 2-group comparisons with confidence limits of 95% considered significant.

For the conjugate syntheses, D-Met was selected as the antioxidant of choice based on its reported cochlea-specific antioxidant protection and chemical structure. Considering HA’s chemical structure (Figure 1A), our strategy for chemical conjugation was to use carbodiimide chemistry to create a covalent bond between the carboxyl functionalities existing on the HA backbone and the primary amine groups of D-Met. This synthetic scheme is well characterized, uses a zero-length crosslinker that directly couples the two compounds of interest without adding any additional bulk, and requires mild reaction conditions. The HA backbone, that consists of repeating, identical disaccharide units, intrinsically contains one carboxyl functionality per repeat (Figure 1A). To increase the amount of D-Met that could be covalently attached to the HA backbone, in a first step the HA backbone was enriched in carboxyl functionalities. This was achieved via the addition of carboxymethyl groups to the primary hydroxyl functionalities of the disaccharide units to yield CMHA. The successful structural change was confirmed by ¹H-NMR (Supplementary Figure S1), and based on the spectra, the extent of the functionalization was assessed to be approximately 58% (Supplementary Figure S2). The CMHA was then used as a substrate for D-Met attachment according to the reaction scheme illustrated in Figure 1B. Three different HA to D-Met stoichiometries were used (2X, 10X, 20X molar excess of D-Met to HA disaccharide unit), and the amount of conjugated D-Met was quantified by amino acid analysis.

Our data (Supplementary Table S1) indicated that the amount of conjugated D-Met was higher for the 10X reaction stoichiometry compared to the 2X and was comparable with the 20X reaction mix (within the detection method variability range). We therefore conducted all our subsequent experimental data with the 10X reaction product, which was determined to contain ∼427 nmol/mg of D-Met. As controls, 10X HAM equivalent amounts of CMHA and D-Met were used either alone or as blend (Supplementary Table S2).

To evaluate the potential applicability of HAM as a prophylactic or therapeutic against cochlear oxidative cell damage, typically associated with hearing loss, HEI-OC1 and SV-k1 cochlear cell lines were treated with HAM, CMHA, D-Met and CMHA + D-Met blend, respectively, and assayed for cell viability. Two different assays were used for this. The first one was a colorimetric methyl tetrazolium salt (MTS)-based assay that quantifies cell viability as a function of the cellular enzymatic reduction of the MTS which results in the development of a red color in the assay solution. The intensity of the color typically directly correlates with the number of viable cells. The second assay used was a more sensitive quantification method that relies on the binding of a cell-permeant fluorescent dye to the nucleic acid of live cells. The colorimetric assay results (Figure 2A) appear to indicate increased cell numbers/proliferation in cells treated with HAM, compared to those treated with CMHA, D-Met and CMHA + D-Met, or control, media-treated ones. Specifically, compared to the control group, the number of HAM treated cells appeared significantly higher, while CMHA and CMHA + D-Met, treated cells exhibited slightly lower viability. Only D-Met treatment did not seem to affect cell viability/number compared to the control. The colorimetric assay, however, is based on the enzymatical reduction of the assay reagent, particularly, it reflects the activity of mitochondrial enzymes. A validation of the MTS assay-observed trends was expected through the CyQuant assay. Interestingly, the fluorescent CyQuant assay results appear to indicate no statistically significant differences in cell numbers/viability for any of the treatment groups compared to the control, media-treated ones (Figure 2B). Cumulatively, these results seem to indicate that HAM is cytocompatible, and when added to cells it impacts the activity of mitochondrial enzymes but does not induce cell proliferation in HEI-OC1 or SV-k1 cells.

With an understanding of HAM cytocompatibility and cellular compartmentalization, we next sought to assess its potential prophylactic and/or therapeutic effect on oxidatively stressed cochlear cells. As oxidative stress is one of the major players in sensorineural hearing loss, these experiments are expected to serve as initial indicators of the potential prophylactic and/or therapeutic value of HAM against hearing loss. Both HEI-OC1 and SV-k1 cells were pre-treated with HAM, D-Met, or just tissue culture medium (control) and the amount of reactive oxygen species (ROS) present in each treatment group was quantified. As anticipated, stressing of the cells with H₂O₂ resulted in a significant increase in ROS in both cell types, with significantly higher levels detected in SV-k1 cells (Figure 3). The pre-treatment of cells with HAM yielded cellular ROS levels comparable to the unstressed controls for both HEI-OC1 and SV-k1. All the other pre-treatment conditions (D-Met or CMHA + D-Met blend) significantly reduced the amount of ROS produced by the cells, but to a significantly lesser extent than the HAM pre-treatment, and the cell line-specific differences observed in the stressed cells remained present, with SV-k1 cells displaying higher ROS levels than their HEI-OC1 counterparts. Post-stress treatment of cells with HAM, D-Met or CMHA + D-Met, did not result in any significant protective effects indicating that HAM could be potentially used as a prophylactic but not as a therapeutic (Supplementary Figure S3A). Considering that ROS are produced in the mitochondria and, as our MTS assay data indicated an increase in mitochondrial enzymatic activity in HAM-treated cells, we also sought to confirm that the observed mitochondrial enzymatic upregulation would not result in ROS production. In agreement with our assumption, our data indicated no change in ROS levels in unstressed cells, when pre-treated with HAM, D-Met or CMHA + D-Met (Supplementary Figure S3B).

As oxidative stress typically leads to cell death via apoptosis, we next evaluated the effect of HAM treatment on oxidatively-induced apoptosis in HEI-OC1 and SV-k1 cells via caspase 3/7 detection. In agreement with our ROS results, the data indicated cell type dependent levels of caspase 3 and 7 (Figure 4). Interestingly, even though SV-k1 cells expressed higher levels of ROS in response to oxidative stress, the apoptosis marker levels were lower than those in HEI-OC1. All treatment conditions appeared to reduce the amount of apoptosis markers in HEI-OC1 cells, but in SV-k1 cells, CMHA, D-Met or CMHA + D-Met blend treatments appeared to slightly increase the apoptotic marker levels. It is important to note here that overall SV-k1 cells did not appear to be particularly susceptible to oxidative stress-induced apoptosis.

Methionine has been extensively demonstrated to have antioxidant properties by directly targeting ROS and undergoing oxidation to methionine sulfoxide and methionine sulfone respectively (Figure 5A). Therefore, we sought to further investigate the insignificant D-Met protective effects observed in our ROS and apoptosis assays in both HEI-OC1 and SV-k1 cells. To assess if the enantiomeric form or oxidative status of the methionine were important, cells were treated with D-Met, L-Met, L-Met sulfoxide, L-Met sulfone and D, L-Met sulfone (Figure 5B). The results revealed several different findings. First, in agreement with the ROS data, there was no substantial reduction in the ROS levels as a result of D-Met treatment. L-Met treatment had the same effect as D-Met treatment, indicating no enantiomeric specificity in the cellular responses. Interestingly, pre-treatment of cells with methionine sulfoxide or sulfone, appeared to increase the amount of ROS detected in the cells. We hypothesize that this might be caused by the oxidative interaction of the assay reagent with the sulfoxide and sulfone. These data indicate that treatment with unconjugated D-Met or L-Met is ineffective against ROS and that methionine’s oxidative state does not have a positive impact on cellular ROS. Cumulatively, these findings appear to indicate that D-Met might not be internalized enough to trigger protective cellular effects.

The above-mentioned results appeared to indicate that the covalent conjugation of D-Met to CMHA was crucial for protective cellular effects, as CMHA or D-Met treatments alone or as a blend (CMHA + D-Met) did not induce similar changes in the two cochlear cell lines. HA is known to be able to internalize intracellularly via receptor-mediated endocytosis in various tissues. With that consideration, our assumption was that conjugation of D-Met to HA might be able to more effectively internalize the antioxidant to induce the observed changes in mitochondrial enzymatic activity and trigger cell protective effects. To test this, first, the potential internalization of unconjugated HA in the two cochlear cell lines was interrogated, via the use of fluorescently labeled-HA (HA-BODIPY) and microscopical visualization of the compound’s cellular distribution. Our data indicate that in both cell types, HA was able to relocate inside the cells and appeared to co-localize with lysosomes (stained with LysoTracker) and mitochondria (stained with MitoTracker), although its exact organelle-specific location could not be pinpointed at this resolution (Figure 6A). In the context of these observations, however, a mitochondrial localization would be in line with our observed MTS assay results.

Second, we sought to assess the amount of D-Met internalized when cells were treated with conjugated versus unconjugated antioxidant. For this HEI-OC1 and SV-k1 cells were treated with HAM, D-Met or CMHA + D-Met and the content of intracellular methionine was quantified. The obtained data indicate that treatment of cells with D-Met alone or as blend with CMHA resulted in significantly less intracellular D-Met than in the cell lysates treated with HAM (Figure 6B). Interestingly, the data also seem to indicate that in the HEI-OC1 cells the blend of D-Met and CMHA did increase the intracellular concentration of D-Met compared to the D-Met alone treatment (Figure 6B inset). Nevertheless, these changes were insignificant compared to the effects of HAM treatment on the amount of intracellular D-Met. Overall, these results further highlight the impact of the chemical conjugation between D-Met to CMHA in the context of the development of a novel compound targeting hearing loss.

With the D-Met cellular internalization confirmed, the next step was to interrogate the potential mechanism of protection employed by the internalized antioxidant. One such mechanism could involve the modulation of the NADP/NADPH ratio in the pentose phosphate pathway (PPP). To investigate this potential mechanism, HEI-OC1 and SV-k1 cells were treated with HAM, D-Met or CMHA + D-Met and the NADP/NADPH ratio was quantified. Both cell lines showed no significant difference in the NADP/NADPH ratio in any of the treatment groups or control groups (Figure 7). These results suggest that mechanisms other than upregulation of the NADP/NADPH ratio in the PPP pathway are responsible for the observed protective effects; however, further mechanistic investigations were out of the scope of this study.

The original intent of this study was to develop a topically applicable prophylactic or therapeutic against hearing loss. Therefore, the next step in HAM’s characterization process was to assess its ability to permeate through biological membranes typical for ear structures. Specifically, a therapeutic targeting the inner ear would be expected to permeate at a minimum the round window membrane (RWM). To test this, we used a three-dimensional tissue model of RWM, consisting of underdeveloped airway cells.

Therefore, HAM and controls were then placed on top of the tissues (donor side) and the amount that traveled through the tissue (receiver side) was quantified (Figure 8A). The receiver side solutions were then analyzed with two SEC columns: one for higher molecular weights (6 kDa–10 MDa) capable of detecting HAM, and CMHA and one for lower molecular weights (100 Da–20 kDa) capable of detecting free D-Met (Figure 8B). Our data indicate that HAM and CMHA were both able to permeate the RWM model, at 2.93% and 0.77%, respectively, of the donor side-applied amount (Figure 8C). Interestingly, the CMHA amount that permeated the model when applied to the donor side as blend, was slightly higher, at 0.90%, than when applied individually. For D-Met detection a standard curve based on antioxidant-specific absorbance A235 values were created to ensure proper chromatographic detection and quantification of the antioxidant (Supplementary Figure S4). However, all treatment conditions resulted in no detectable levels of D-Met in the receiver solution. These results could indicate that the amount of antioxidant able to penetrate the RWM model was either insignificant or below the detection threshold of our instrument (0.01 mg/ml) in the D-Met only or CMHA + D-Met treated samples, and that D-Met was most likely still covalently attached to the CMHA backbone in the HAM treated sample (Supplementary Figure S5). To further assess the latter, we used absorbance A235 detection on the large molecular weight range SEC column and quantified the area under the curve (AUC) for HAM and CMHA, before and after permeation through the tissue models. The analyses indicated a significantly higher AUC of 958.419 for HAM compared to the CMHA only treated group, that had an AUC of 12.945. These values were similar to those obtained for control HAM and CMHA that have not been permeated through the tissue models (Supplementary Table S3). Overall, our data indicate that a significantly higher amount of HAM permeated the tissue compared to CMHA only or CMHA + D-Met blend, and it appears that D-Met remains covalently attached to the CMHA backbone throughout the RWM model permeation.