A total of eight nanofiber formulations ( Figure 1) were prepared by combining four different losartan potassium (Sigma-Aldrich, St. Louis, MO, United States) doses (0, 0.5, 5, and 50 mg/mL) and two PA molecules with opposite charge. Cationic PA molecules (K-PA) with the sequence palmitoyl-VVVAAAKKK and anionic PA molecules (E-PA) with the sequence palmitoyl-VVVAAAEEE were synthesized by solid phase peptide synthesis as described previously (Sather et al., 2021), and had a purity of more than 95% by liquid chromatography-mass spectroscopy (LC-MS). PA nanofiber assemblies at each losartan dose were prepared by first solubilizing K-PA or E-PA at 10 mg/mL in 150 mM NaCl and 3 mM KCl and pH adjusting to 7.4 using 1 M NaOH before adding various amounts of losartan powder. The solutions were vortex mixed, sonicated for 30 min, and then thermally annealed in a water bath at 80°C for 30 min followed by slow cooling to room temperature overnight. The annealed solutions were then frozen in liquid nitrogen and freeze-dried for 24 h, with the resulting lyophilized powders stored at 4°C until further use.

Samples for transmission electron microscopy (TEM) were resuspended at a PA concentration of 10 mg/mL and then diluted 10-fold–1 mg/mL immediately before 10 µL of sample solution was transferred to plasma-cleaned 300-mesh copper grids with lacey carbon support (Electron Microscopy Science). Samples were stained with 2% uranyl acetate. Imaging was performed using a FEI Spirit G2 TEM working at 120 kV accelerating voltage.

Nuclear magnetic resonance (NMR) spectroscopy was performed on a Bruker Neo 600 MHz system with QCI-F cryoprobe at 298 K. The lyophilized powders from the sample preparation protocol above were resuspended in D₂O at 10 mg/mL PA concentration and 5 mg/mL losartan concentration. This corresponds to 8.7 mM E-PA, 8.7 mM K-PA, and 10.8 mM losartan.

Absorbance spectroscopy was used to measure losartan release from PA nanofibers. Freeze dried PA-losartan powders were reconstituted in milli-Q H₂O at 10 mg/mL PA and 5 mg/mL losartan and divided into 100 μL aliquots in 1 mL Eppendorf tubes. PA nanofiber hydrogels were formed by adding 50 μL of either 1x phosphate buffered saline (PBS) to K-PA solutions or 25 mM CaCl₂ in 1x tris buffered saline (TBS) to E-PA solutions. Samples were incubated at 37°C for 1 h to allow for gel formation at the bottom of the tube. Then 400 μL of release buffer (1x PBS for K-PA gels and 10 mM CaCl₂ in 1x TBS for E-PA gels) was added to each tube and the samples were incubated at 37°C. Every 24 h, 400 μL of release buffer was removed from each tube for absorbance spectroscopy measurements and replaced with fresh release buffer. All release experiments were repeated in triplicate. Absorbance spectra were collected using an Ocean Optics USB4000-UV-VIS spectrometer, with data update rate = 100 μs, scans to average = 100, and boxcar width = 2. Release buffer samples at each time point were loaded into a quartz cuvette with 1 cm pathlength, and fresh release buffer solutions were used for background subtraction. The primary losartan peak absorbance at 239 nm was measured and compared to a calibration curve of known concentrations of losartan to determine the amount of losartan released at each time point. The absorbance in the release buffer at the t = 0 time point was used to calculate the drug loading efficiency.

The ATDC5 cell line is a continuous long-term culture line derived from mouse teratocarcinoma that is commonly used as a model for in vitro chondrocyte research (Yao and Wang, 2013). ATDC5 (Sigma-Aldrich, St. Louis, MO, United States) cells were plated at a density of 1 × 10⁴ cells/well on the bottom of a 24-well transwell plate (Corning, Glendale, AZ, United States) (Figure 2). These cells were cultured for 48 h in a “complete chondrogenic media” defined as Dulbecco’s Modified Eagle Medium (DMEM)/Nutrient F-12 Ham (Thermo Fisher, Winsor, NJ, United States), containing 5% fetal bovine serum (FBS) (Thermo Fisher, Winsor, NJ, United States), and 1% penicillin and streptomycin (Invitrogen, Waltham, MA, United States), followed by 1% insulin-transferrin-sodium selenite (ITS) (Sigma-Aldrich, St. Louis, MO, United States) to support chondrogenic differentiation as previously published (Chen et al., 2020). Total of 12 groups were divided by losartan dose (0, 0.5, 5, and 50 mg/mL) and presence/absence of positively charged K or negatively charged E nanofibers (Table 1). Each nanofiber was implanted on a 0.4 μm semipermeable membrane (Corning, Glendale, AZ, United States) (Figure 2). Losartan solution without nanofibers was prepared by dissolving powdered losartan, identical to the drug contained in the nanofibers, in culture medium. The culture was continued for 2 weeks at 5% CO₂, 21% O₂, and 37°C, with the nanofibers remaining untouched and the medium changed every 3 days from the start of nanofiber treatment.

Cell proliferation of ATDC5 cells in each group was evaluated by PrestoBlue assay (Invitrogen, Waltham, MA, United States) according to the manufacturer’s protocol (Boncler et al., 2014). This non-destructive assay can approximate the number of living cells present based on the mitochondrial reduction of resazurin dye (blue, non-fluorescent) to resorufin (red, highly fluorescent). Cell proliferation was semi-quantitated as a longitudinal fold change from control group at 24 h, 5 days, and 14 days after treatment follow a 1 h of reaction with PrestoBlue at 37°C. Fluorescence in each well is quantified by measuring absorbance at 570 nm using a microplate reader (Infinite 200 Pro, Tecan, Männedorf, Switzerland) and normalizing by the value at 600 nm (n = 8 per group).

Twenty four hours after plating ATDC cells (1 × 10⁴ cells/well) into a 24-well transwell plate, 10 ng/mL interleukin (IL)-1β (Peprotech, Cranbury, NJ, United States) was added to the complete chondrogenic media for 24 h to model OA in vitro as previously published (Deng et al., 2021). The cells were then washed twice with phosphate buffered saline (PBS) (Thermo Fisher, Winsor, NJ, United States) and cultured again for 24 h in complete chondrogenic media with or without losartan delivery.

GAG production was quantitatively evaluated by measuring the reaction of GAG with 1,9-dimethyl-methylene blue (DMMB) (Sigma-Aldrich, St. Louis, MO, United States) 14 days after initiation of treatment according to the well-defined method. Briefly, each well of a 96-well plate was filled with 15 μL of supernatant, and 150 μL of DMMB solution. GAGs production was measured at 575 mm with a microplate reader (Infinite 200 Pro, Tecan, Männedorf, Switzerland) using a standard curve obtained with shark chondroitin sulfate (Sigma-Aldrich, St. Louis, MO, United States). GAGs production was standardized by calculating GAGs production per cell unit using the number of cells estimated by PrestoBlue assay for each sample (n = 8 per group).

Total RNA was isolated from ATDC5 cells using TRizol Reagent (Thermo Fisher, Winsor, NJ, United States) and reverse transcribed according to the manufacturer’s protocol using iScript reverse transcription supermix cDNA synthesis kit (Bio-Rad, Hercules, CA, United States) 14 days after treatment. qRT-PCR was performed using the Applied Biosystems™ SYBR™ Green Assay Kit (Thermo Fisher, Winsor, NJ, United States) and Applied Biosystems StepOnePlus RT-PCR thermocycler (Applied Biosystems, San Francisco, CA, United States). Francisco, CA, United States). Primers were designed using PRIMER-Blast (NCBI) and the sequences are listed in Table 2. The evaluated mRNAs were the inflammation marker interleukin-6 (Il6), the hypertrophic chondrocyte marker type X collagen (colXa1), matrix metalloproteinase-13 (mmp13), and the chondrocyte markers type II collagen (colIIa1) and Aggrecan (acan). To evaluate the inhibitory effect of losartan on TGF-β1, mRNA expression of tgfβ1 was also evaluated. These genes of interest were normalized to the expression of the housekeeping gene glyceraldehyde 3-phosphate dehydrogenase (gapdh) expression. Relative gene expression was calculated compared to the control culture condition (no nanofibers or losartan) using the 2^−ΔΔCT method (n = 8 per group).

All experiments were repeated at least twice with a n = 3 replicates per experiment. All plots and statistical comparisons were made using GraphPad Prism version 9.4.0 (GraphPad Software, La Jolla, CA, United States). Bar plots represent the mean ± standard deviation with each individual replicate plotted as a dot on the bar graph. The groups were compared by analysis of variance (ANOVA) with statistical significance set at p < 0.05. For significantly different groups, Tukey’s Honestly Significant Different (HSD) multiple comparison post hoc testing was performed with statistical significance again set at p < 0.05.

Losartan encapsulated PA nanofibers were prepared by physical mixing of PA and losartan solutions followed by thermal annealing to drive co-assembly of losartan into the PA supramolecular structure (see Figure 1). We hypothesized that hydrophobic interactions between losartan and the aliphatic core of the nanofiber would initiate encapsulation and that opposing electrostatic effects on positively charged K-PA vs. negatively charged E-PA nanofibers may alter the binding affinity to the drug. In TEM imaging, we observed similar fiber length distributions (∼200–2,000 nm) for K-PA and E-PA nanofibers both with and without losartan (Figures 3A–D). We note that these nanofiber lengths are shorter than previously reported PA nanofibers from our group because the lyophilization step after thermal annealing can fracture the supramolecular fibers (Chen et al., 2018). However, for our target therapy of a liquid injection in the intra-articular space, we preferred the shorter nanostructures to minimize aggregation or gelation.

We first measured the drug loading efficiency and found that both K-PA and E-PA nanofibers were highly effective at encapsulating losartan (93% loading for K-PA and 84% loading for E-PA). Then, we measured the release of losartan from positively charged K-PA and negatively charged E-PA nanofibers by monitoring the losartan absorbance in a release buffer over the course of two weeks. We found that both nanofibers offered a sustained release profile but that negatively charged E-PA nanofibers released losartan at approximately twice as fast as the positively charged K-PA nanofibers (Figure 3E). At a moderate dose of losartan (5 mg/mL), all of the losartan was released from negatively charged E-PA after 1 week, while positively charged K-PA nanofibers still retained ∼55% of the losartan after 2 weeks.

We utilized ¹H NMR spectroscopy in solution to probe the binding of losartan to E− and K-PA nanofibers. Our analysis focused on the chemical shift region featuring the aromatic protons from losartan (Figure 4), in which E− and K-PA assemblies alone show no signal (Figures 4A, B). The spectrum of free losartan shows a set of sharp resonances that correspond to ortho- (purple) and para-disubstituted (orange) benzene moieties (Figure 4C). This is expected since losartan, a small molecule in the monomeric state, can undergo fast rotational tumbling in solution. In contrast, the ¹H NMR signals from losartan measured in the samples of the drug-PA co-assemblies are significantly broadened out. This is due to chemical exchange of the drug molecules among the free and PA nanofiber bound states (losartan _free ⇌ losartan _bound ), indicating encapsulation of losartan into the PA nanofiber assembly. It is expected that losartan’s bound state shows faster transverse relaxation rate (R ₂ ) in comparison to its free state (R _2,free < R _2,bound ), considering the larger size of the losartan-nanofiber assembly compared to monomeric losartan, which causes the excessive NMR line broadening of the losartan signals in the samples where PAs are present. Importantly, the positively charged K-PA/losartan spectrum shows signals shifted downfield relative to those measured for losartan alone and in the negatively charged E-PA/losartan co-assembly. This shift can be attributed to attractive electrostatic interactions between the tetrazolate segment of losartan and the lysine ammonium moieties (NH₃ ⁺···tetrazolate), which are present in positively charged K-PA, but not in negatively charged E-PA. Thus, in addition to the hydrophobic effect that drives the PA nanofiber/losartan association, unique charge complementarity present within the K-PA/losartan system enhances the persistency of K-PA/losartan relative to the E-PA/losartan co-assembly, which is in line with the losartan release experiments where K-PA releases losartan slower than E-PA.

Cytotoxicity and ATDC5 cell proliferation was evaluated at 24 h, 5 days, and 14 days in groups treated according to the schema shown in Table 1 and compared to a control group that was not treated with PA nanofibers or losartan (Figure 5). No cytotoxicity was observed in any of the treatment groups. We found that both positively and negatively charged PA nanofibers significantly increased cell numbers at all time points compared to no nanofibers at each losartan dose. The addition of losartan did not result in a significant difference in cell proliferation at any dose.

Bioactivity and the regenerative effect of the nanofibers with or without losartan was tested for our clinical application by using an established in vitro model of OA (Deng et al., 2021). We first confirmed that, in our hands, 24 h of IL-1β (10 ng/mL) leads to reduced GAG production (Figure 6A). Subsequently, GAG production in the OA-induced chondrocytes was evaluated following losartan treatment either with or without the PA nanofibers. Significant increase in GAG production was observed with the positively charged K-PA nanofibers alone or when dosed with low (0.5 mg/mL) or moderate (5 mg/mL) losartan compared to controls without losartan or nanofibers (Figure 6B, green bars). The negatively charged E-PA nanofibers had no significant impact on GAG production compared to controls without nanofibers and losartan (Figure 6B, blue bars).

Chondrocyte response to the losartan nanofibers was also evaluated by looking at relative changes in expression of canonical hyaline cartilage markers (col2a1, acan), the pro-inflammatory marker il6, and the chondrocyte hypertrophy and degradation markers (colXa1, mmp13). Induction of an osteoarthritis-like phenotype in the ATDC5 chondrocytes was confirmed by a significant upregulation of il6, mmp13, and colx10a1 (Figures 7A–C) and a decrease in col2a1 and acan (Figures 7D, E) expression following treatment with IL-1β. Since experimentation up to this point supports a more beneficial effect of the positively charged K-PA nanofibers over the negatively charged E-PA nanofibers, gene expression analysis subsequently focused only on the positively charged K-PA nanofibers at the low and medium doses of losartan. The positively charged nanofibers themselves significantly suppressed il6 expression, and this stayed low, but not significantly different with the addition of losartan (Figure 7A). Conversely, positively charged nanofibers did not significantly change mmp13 expression, but increasing doses of losartan resulted in a dose dependent decrease in expression of both the metalloproteinase (mmp13) and hypertrophy (col10a1) marker (Figures 7B, C). A pro-regenerative response was stimulated by the addition of the positively charged nanofibers, as evident by the significantly increased expression of col2a1 and acan (Figures 7D, E). Bioactivity of the losartan as a TGFβ antagonist was further confirmed by showing a dose-dependent decrease in tgfβ1 expression with increasing losartan delivery (Figure 7F).