Local and Sustained Baricitinib Delivery to the Skin through Injectable Hydrogels Containing Reversible Thioimidate Adducts

Abstract

Janus kinase (JAK) inhibitors are approved for many dermatologic disorders, but their use is limited by systemic toxicities including serious cardiovascular events and malignancy. To overcome these limitations, injectable hydrogels are engineered for the local and sustained delivery of baricitinib, a representative JAK inhibitor. Hydrogels are formed via disulfide crosslinking of thiolated hyaluronic acid macromers. Dynamic thioimidate bonds are introduced between the thiolated hyaluronic acid and nitrile-containing baricitinib for drug tethering, which is confirmed with ¹H and ¹³C nuclear magnetic resonance (NMR). Release of baricitinib is tunable over six weeks in vitro and active in inhibiting JAK signaling in a cell line containing a luciferase reporter reflecting interferon signaling. For in vivo activity, baricitinib hydrogels or controls are injected intradermally into an imiquimod-induced mouse model of psoriasis. Imiquimod increases epidermal thickness in mice, which is unaffected when treated with baricitinib or hydrogel alone. Treatment with baricitinib hydrogels suppresses the increased epidermal thickness in mice treated with imiquimod, suggesting that the sustained and local release of baricitinib is important for a therapeutic outcome. This study is the first to utilize a thioimidate chemistry to deliver JAK inhibitors to the skin through injectable hydrogels, which has translational potential for treating inflammatory disorders.

Keywords: JAK‐STAT; dermatology; hydrogel; psoriasis; thioimidate.

Figure 1.. Baricitinib and cysteamine thioimidate formation.

A) Baricitinib nitrile (red) interacts with thiols (blue) to form thioimidates (purple). The reaction forms through a modified Pinner reaction in which a free, nucleophilic thiol attacks the electrophilic nitrile on baricitinib. B) Liquid chromatography-mass spectrometry demonstrates product formation between baricitinib (373 Da), cysteamine (77 Da), and baricitinib-cysteamine (449 Da), resolving with three distinct retention time peaks of 0.3 s, 1.5 s, 1.4 s, respectively, which reflect their differences in polarity.

Figure 2.. Polymer design, thioimidate bonding, and hydrogel disulfide crosslinking.

A) Hyaluronic acid (100 kDa) is amidated at its carboxyl position with cysteamine to yield a thiolated product where x is the fraction of disaccharides with thiol modification and 1-x is the fraction of unmodified disaccharides. B) Baricitinib nitrile (red) interacts with the thiolated hyaluronic acid polymer (blue) to yield thioimidates (purple), where y represents the fraction of disaccharides with bound baricitinib. Simultaneously, unmodified thiols oxidize and form disulfide crosslinks between polymer chains under physiologic conditions to yield a dynamically crosslinked hydrogel. The degree of unbound thiols which form disulfides is represented by x-y and unmodified disaccharides are represented as 1-x.

Figure 3.. Quantification of hyaluronic acid thiolation and consumption of thiols by baricitinib.

A) We validated the presence of thiols on hyaluronic acid using a commercially available assay. Fluorescence excitation at 490 and emission at 520 nm was performed using glutathione (GSH) standards and compared to thiolated hyaluronic acid (10 µM, yellow bar), n=3 per group. B) Thiolated hyaluronic acid (25 µM) was incubated with varying doses of baricitinib (0–400 µM), n=3 per group. Free thiols were subsequently quantified using the thiol quantification assay demonstrating consumption of free thiols on hyaluronic acid with increasing concentrations of baricitinib. *p<0.05 compared to [CN]/[SH] of 1, n=3 per group, mean ± SD. Comparison of nitrile to thiol ratio demonstrates incomplete bond formation and reversibility. C) K_eq measurements at tested concentrations of [CN]/[SH] of 1, 2, 4, 8, and 16.

Figure 4.. 13C and 1H NMR validate thioimidate formation between baricitinib and thiolated hyaluronic acid.

A) ¹³C NMR of baricitinib, thiolated hyaluronic acid, and the product baricitinib-thiolated hyaluronic acid demonstrate the unique spectra of each individual reactant and a new peak at 170 ppm corresponding to the predicted chemical shift of the thioimidate bond (R-N=C(SR)R). B) ¹H NMR of baricitinib, thiolated hyaluronic acid, and the product baricitinib-thiolated hyaluronic acid demonstrate the unique spectra of each individual reactant with distinct aromatic protons in the baricitinib-hyaluronic acid group that is chemically shifted from the aromatic protons of baricitinib alone, suggesting the presence of a modified product in solution consistent with the thioimidate adduct.

Figure 5.. Hydrogel formation and shear oscillatory rheometry.

A) Hydrogels are formed between thiolated hyaluronic acid and baricitinib, where thioimidate adducts and disulfide crosslinks form simultaneously. Upon material deposition, thioimidate bonds reverse to release baricitinib into the surrounding environment. B) Shear oscillatory rheometry over 24 hours demonstrating crosslinking of polymers into hydrogels (~100 Pa) with and without baricitinib which occur over the course of 24 hours. Frequency sweeps demonstrate classic viscoelastic behavior of materials. C) Mechanical properties of hydrogels and baricitinib hydrogels measured by shear oscillatory rheometry (0.5% strain, 10 Hz) demonstrate comparable properties, n=3 per group, mean ± SD.

Figure 6.. Baricitinib release and ultraviolet spectrometry.

A) Quantification of baricitinib aromatic absorption in ultraviolet A and B range (280–400 nm) demonstrating highest absorption in the ultraviolet B range. Subsequent plotting of absorbance against dose demonstrates linear absorption at 300 nm leads to linear equation (y = 0.01344x + 0.4769, R² = 0.9992). B) Measuring baricitinib release at 300 nm against a standard curve allows quantification of baricitinib release from hydrogel formulations. Tunable release is exhibited over the course of six weeks with release dependent on material formulation. *p<0.05 between groups at each timepoint, n=3 per timepoint, mean ± SD. C) Release curves for hydrogels assembled at 0.2 mg/mL or 2 mg/mL baricitinib loaded (2 wt% hydrogel, 30% thiol modification).

Figure 7.. Validation of baricitinib activity on JAK/STAT signaling using a HEK293 luciferase reporter.

A) HEK293 cell line with the firefly luciferase gene under the control of Interferon Stimulated Response Element (ISRE) where type I interferon-induced JAK/STAT signaling pathway in the target cells can be monitored by measuring luciferase activity. IFNα (EC₅₀ = 15 U/mL) leads to induction of luminescence, n=3 per group, mean ± SD. B) After stimulation with 100 U/mL IFNα, baricitinib leads to near complete silencing of luciferase activity, n=3 per group, mean ± SD. C) Schematic of experimental design demonstrating hydrogel incubation in saline with collection and replacement of releasate performed at regular intervals and then added directly to cells. D) Incubation of cells with 5 µL releasates demonstrates sustained JAK/STAT inhibition but only from the 2 mg/mL formulation, n=3 per group, mean ± SD, *p<0.05 between 0.2 mg/mL and 2 mg/mL. Incubation of cells with 20 µL of releasates demonstrates near complete inhibition of JAK/STAT signaling at all tested timepoints, n=3 per group, mean ± SD. Luminescence is normalized to signals from cells with IFNα.

Figure 8.. In vivo injections of baricitinib hydrogel in an imiquimod model of psoriasiform dermatitis.

A) Two 25 uL injections of baricitinib hydrogel or PBS, hydrogel alone, or baricitinib alone controls are injected into four spots in dorsal mouse skin. After injection, imiquimod is applied daily for five days to induce psoriasiform inflammation before sacrifice. B) Images of skin changes at five days and seven days. C) H&E sections and pSTAT3 demonstrating skin thickening and pSTAT3 nuclear expression in response to imiquimod application with reduced thickening and pSTAT3 expression seen in treatment groups. Scale = 100 µm. Epidermal keratinocytes are counterstained with keratin 14 (K14) to highlight the hyperplastic epidermis. D) Quantification of epidermal thickness at seven days by quantifying epidermal ROI divided by length. *p<0.05 by one-way ANOVA, n=6 for PBS, hydrogel, and baricitinib hydrogel groups, n=5 for control and baricitinib groups, mean ± SD.