Injectable, Drug-Eluting Nanocrystals Prevent Fibrosis and Stricture Formation In Vivo

Abstract

BACKGROUND & AIMS:

Tissue fibrosis results from uncontrolled healing responses leading to excessive mesenchymal cell activation and collagen and other extracellular matrix deposition. In the gastrointestinal tract, fibrosis leads to narrowing of the lumen and stricture formation. A drug treatment to prevent fibrosis and strictures in the gastrointestinal tract would be transformational for patient care. We aimed to develop a stricture treatment with the following characteristics and components: a small molecule with strong antifibrotic effects that is delivered locally at the site of the stricture to ensure correct lesional targeting while protecting the systemic circulation, and that is formulated with sustained-release properties to act throughout the wound healing processes.

METHODS:

A high-throughput drug screening was performed to identify small molecules with antifibrotic properties. Next, we formulated an antifibrotic small molecule for sustained release and tested its antifibrotic potential in 3 animal models of fibrosis.

RESULTS:

Sulconazole, a US Food and Drug Administration–approved drug for fungal infections, was found to have strong antifibrotic properties. Sulconazole was formulated as sulconazole nanocrystals for sustained release. We found that sulconazole nanocrystals provided superior or equivalent fibrosis prevention with less frequent dosing in mouse models of skin and intestinal tissue fibrosis. In a patient-like swine model of bowel stricture, a single injection of sulconazole nanocrystals prevented stricture formation.

CONCLUSIONS:

The current data lay the foundation for further studies to improve the management of a range of diseases and conditions characterized by tissue fibrosis.

Graphical Abstract

Normal tissue repair involves immune, as well as mesenchymal, cell activation; extracellular matrix (ECM) deposition; and tissue remodeling.¹ However, severe or repetitive injury, such as in the case of chronic inflammation, can lead to excessive ECM accumulation and formation of fibrotic tissue.¹ Fibrosis can affect any organ and lead to disruption of normal tissue structure and organ dysfunction or failure.² Myofibroblasts display high expression of α–smooth muscle actin (α-SMA), are capable of contractility, and play central roles in fibrosis and scar formation.¹ In addition, chronic inflammation causes impaired mucosal tissue repair and myofibroblast proliferation.^3,4 Hence, targeting the myofibroblast proliferation pathway has been proposed as a strategy for the development of effective antifibrotic drugs.^2,5,6 To date, however, there are only 2 US Food and Drug Administration–approved antifibrotic drugs in general—pirfenidone and nintedanib—both are for the treatment of idiopathic pulmonary fibrosis.⁷ Except for idiopathic pulmonary fibrosis, there are no antifibrotic drugs for any other indication, including for intestinal fibrosis. Thus, effective antifibrotic therapies represent a critically important unmet need.

In the gastrointestinal (GI) tract, inflammatory bowel diseases (IBDs) are the main cause of intestinal fibrosis and the significant morbidity derived from intestinal fibrosis.⁸ IBDs can be separated into 2 major sub-types—Crohn’s disease (CD) and ulcerative colitis—and affect more than 3 million people in the United States.⁹ Fibrosis is the underlying mechanism for the development of intestinal strictures,^1,10 a narrowing of the intestine that occurs in 27%–54% of patients with CD and 1.5%–11.2% of patients with ulcerative colitis.¹¹ Such strictures often require surgical or endoscopic treatment, and almost invariably recur due to ongoing fibrosis.¹ Recurrent fibrotic strictures typically lead to multiple surgeries, short gut, stomas, poor quality of life, and high cost of health care.¹² Because profibrotic mechanisms in IBDs are triggered by inflammation, there was significant hope that potent anti-inflammatory therapies developed over the past 20 years would decrease the incidence of stricturing and fibrosing complications. This expectation, however, did not materialize.^13–15 A dedicated antifibrotic treatment, developed and formulated for fibrotic strictures in the GI tract, is urgently needed.

Intestinal strictures offer the opportunity for administering localized, targeted therapy that would reduce the likelihood of systemic complications. We anticipate that local injection of antifibrotic drugs at the time of endoscopic dilation has the potential to modulate wound healing and fibrotic processes with reduced risk of systemic adverse effects. Of note, prior efforts involved single injections of immediate-release drugs, such as corticosteroids and mitomycin C, and were largely ineffective.^16,17 This lack of effectiveness is likely due, in part, to the short duration of action before being cleared from the body. Here we proposed that a sustained-release formulation of an effective antifibrotic is optimal for impacting tissue remodeling processes like fibrosis that occur over days and weeks.

Here we report that sulconazole, a topical antifungal agent, has potent antifibrotic activity. We hypothesized that providing sustained, effective concentrations of sulconazole intralesionally could modulate the acute healing response and interrupt pathologic fibrotic tissue remodeling processes. We demonstrate that a sulconazole-nanocrystal (Sul-NC) formulation was highly effective in preventing fibrosis in rodent models of skin and intestinal fibrosis, as well as a novel swine model of esophageal stricture.¹⁸ Finally, we demonstrate that Sul-NC was well-tolerated and safe. Our data build a foundation to support further preclinical studies leading to development of Sul-NC as an antifibrotic drug.

Materials and Methods

Development of Sulconazole Nanocrystals

Sul-NC was formulated using a wet-bead–nanomilling method described previously.¹⁹ The wet-bead milling was carried out using a laboratory-scale tissue homogenizer (TissueLyser LT, Qiagen Inc, Germantown, MD). Various stabilizers were employed to determine optimal sulconazole formulation, that is, a formulation that resulted in particles that were relatively uniform in size, stable at room temperature, and injectable through a small-gauge endoscopy needle. These stabilizers included polyvinyl alcohol, hyaluronic acid, carboxymethylcellulose, hydroxypropyl methylcellulose, and Pluronic F127 (Supplementary Table 1). The final formulation approach for animal dosing incorporated 500 mg sulconazole, 2.0 g of 0.5-mm zirconium oxide beads, and 1 mL of 2% (w/v) Pluronic F127 solution in a 2-mL Eppendorf tube. The contents were milled for 10 hours at a speed of 3000 oscillations/min in a 4°C cold room. The mixture was then passed through a 100-μm cell strainer to isolate the milling beads. Milling time was optimized to obtain particles approximately 200 nm, which was 5 hours for the final formulation. Particle size, polydispersity index, and surface charge (ζ-potential) of the Sul-NC were measured using a Malvern Zetasizer Nano ZS (173-degree scattering angle) (Malvern, Westborough, MA). For the particle size and polydispersity index measurement, Sul-NC were diluted 1:100 in ultrapure water and for the ζ-potential measurement, Sul-NC were diluted 1:40 in 10 mM NaCl (pH 7).

Animal Welfare Statement

All animal studies were approved by and conducted in accordance with the guidelines of the Institutional Animal Care and Use Committee at the Johns Hopkins University. All procedures were performed in compliance with the Guide for the Care and Use of Laboratory Animals and the Animal Welfare Act at an Association for Assessment and Accreditation of Laboratory Animal Care–accredited facility. C57BL/6J male and female mice (5–8 weeks old) were obtained from Jackson Laboratory (Bar Harbor, ME). Yorkshire pigs (Sus scrofa domestica, female, 35–50 kg) were purchased from Archer Farms (Darlington, MD).

Esophagus Stricture Treatment With Local Sulconazole-Nanocrystal Injection Via Esophagogastroduodenoscopy

Fourteen days after the argon plasma coagulation procedure, pigs were sedated and esophagogastroduodenoscopy with fluoroscopy was applied for internal evaluation of the stricture formation and measurement of the lumen diameters. Of the 30 ablated regions in the 10 pigs, 23 strictures were <6 mm and were included in further analyses. Pigs were randomly assigned to either the vehicle control (n = 5 pigs) or treatment (n = 5 pigs) groups, such that the 23 strictures included in the study were split into 12 in the control group and 9 in the Sul-NC group. Strictures were then balloon-dilated to 10 mm as described previously¹⁸ and injected with either vehicle (2% F127) or Sul-NC. In the first 2 pigs (n = 3 strictures) that received Sul-NC, 5 mL at 500 mg/mL concentration was split into 4 × 1.25-mL injections circumferentially around each stricture site (total dose 150 mg/kg). We found that this concentration was challenging to push through the endoscopy needle, and the NCs clogged the needle multiple times. In addition, there was immediate fluid leakage from the tissue post injection. Thus, we experimented with different dilutions in vitro to determine a maximum concentration that we could easily administer through an endoscopy needle. We found that diluting the Sul-NC formulation with an additional 2% F127 to 300 mg/mL eliminated the needle-clogging effect. Thus, the remaining 3 pigs (n = 6 strictures) were injected with 300 mg/mL Sul-NC with a total volume of 3 mL split into 6 × 0.5-mL injections circumferentially around each stricture site (total dose 54 mg/kg). For the control pigs, 2% F127 was split into 6 × 0.5-mL injections circumferentially around each stricture site. Fourteen days after the treatment, pigs were sedated, the endoscope was positioned in the esophagus just proximal to the stricture, iodinated contrast solution Omnipaque 240 was then injected through a catheter advanced through the scope, and fluoroscopy images were obtained for evaluation of the stricture reformation. Subsequently, esophagogastroduodenoscopy with fluoroscopy was applied to internally evaluate the stricture reformation and measure the lumen diameters. Some strictures were so narrow that the esophagogastroduodenoscopy could not pass through, therefore, pigs were euthanized and the esophagi were taken out to measure the stricture diameters ex vivo. In addition, healthy tissue specimens were collected from uninvolved regions of the esophagus.

Statistical Analysis

Data are shown as mean ± SD or mean ± SEM for each graphical representation. Graphs were generated with Graph-Pad Prism 9. A 2-tailed Student t test (for comparison of 2 groups) or 1-way analysis of variance test (for comparison of more than 2 groups) was used to determine statistical significance. Statistical significance was assumed if P < .05, P < .01, or P < .001.

The Supplementary Material includes further methodological details.

Results

Sulconazole Is a Potent Antifibrotic

To identify small molecules with antifibrotic actions, we used a high-throughput microscope-based drug-screening approach using human primary colonic fibroblasts (CCD-18Co) activated with transforming growth factor-β (TGF-β). We then screened a panel of 1586 US Food and Drug Administration–approved, small-molecule drugs for their ability to reduce expression of α-SMA and type 1 collagen (COL1A1), as assessed by fluorescent immunocytochemistry. Based on its known antifibrotic effects, pirfenidone was included in this screen as a positive control, and found effective at a dose of 4 mM, as reported previously²⁰ (Figure 1A). Notably, sulconazole provided a similar or better decrease in the staining of α-SMA and Collagen I at a much lower dose of 10 μM (Figure 1A). We then used both activated CCD-18Co cells and human hepatic stellate cells (liver fibroblasts, LX2) in a confirmatory Western blot analysis that showed sulconazole significantly reduced the protein production of α-SMA and Collagen I (Figure 1B and C). We also analyzed α-SMA expression by reverse transcription polymerase chain reaction in both activated CCD-18Co cells and LX2 cells after treatment with several concentrations of antifibrotic drugs. Sulconazole at concentrations as low as 1 μM and 5 μM significantly reduced ACTA2 expression in activated CCD-18Co and LX-2 cells, respectively, whereas pirfenidone showed no effect until reaching 4 mM concentration (Figure 1D). These in vitro results indicated that sulconazole is a potent antifibrotic and acts at much lower concentration than pirfenidone. To quantify the impact of sulconazole on the production of ECM, we treated CCD-18Co, as well as LX2 cells, with increasing concentrations of sulconazole. As shown in Figure 1E, the ECM staining intensity decreased in a concentration-dependent manner in both cell lines. The 50% inhibitory concentration (IC50) for the effect of sulconazole on ECM staining intensity was 0.9 μM in LX2 cells and 6.7 μM in CCD-18Co cells (Figure 1F).

Figure 1.

Sulconazole inhibits fibrotic responses in CCD-18Co and LX2 cells activated with TGF-β. (A) Inactive CCD-18Co colon fibroblasts were stimulated with TGF-β (activated, in the figure), and then treated with pirfenidone (4 mM) or sulconazole (10 μM) before staining for α-SMA, red), type I collagen (Collagen-I, green), and 4′,6-diamidino-2-phenylindole (DAPI) (cell nuclei, blue). Scale bar: 100 μm. (B) Western blot analysis displayed reduced expression of α-SMA and Collagen-1 in inactive as well as activated CCD-18Co and LX2 cells treated with 10 μM sulconazole (sul) compared with activated control cells. (C) Quantification of Western blot images presented in panel B (n = 3–7). Data presented as mean ± SEM. *P < .05, **P < .01 compared with activated control cells. (D) Reverse transcription polymerase chain reaction demonstrated that sulconazole decreased expression of α-SMA (ACTA2) at much lower drug concentrations (1–10 μM) compared with pirfenidone (4 mM) in CCD-18Co (n = 3–5) and LX2 (n = 3) cells. Data presented as mean ± SEM. *P < .05 compared with activated control cells. The “inactive” cells are fibroblasts that were deprived with fetal bovine serum and not stimulated with TGF-β. (E) Sulconazole inhibits production of ECM in a dose-dependent fashion. Scale bar: 100 μm. (F) Quantification of sulconazole effects on ECM production and the IC50 for the effect of sulconazole on ECM production (n = 3–10). Data presented as mean ± SEM. **P < .01, ***P < .001 compared with activated cells.

Sulconazole Effects on Fibroblast Proliferation and Viability

To investigate the effect of sulconazole on fibroblast cell growth, we treated activated CCD-18Co and LX2 fibroblast cells with increasing concentrations of sulconazole, as well as pirfenidone, respectively (Figure 2A). We found that cells activated with TGF-β and subsequently treated with pirfenidone grow at rates similar to inactive cells, but slower than untreated, activated cells, in accordance with previous publications.²⁰ Similarly, we found that higher concentrations of sulconazole slow cell growth similar to inactive cell levels (≥7.5 μM LX2, ≥10 μM CCD-18Co) (Figure 2A). We then assessed the effect of sulconazole, as well as pirfenidone, on cell viability in vitro. Based on live/dead staining (Figure 2B), as well as based on lactate dehydrogenase (LDH) activity (Figure 2C), there was no effect on cell viability. Similarly, flow cytometry analysis of cultured cells found that the highest tested dose of sulconazole (10 μm) did not induce cell apoptosis in activated CCD-18Co and LX2 cells (Figure 2D and F). Similar to what was observed based on cell growth rate, Ki67 staining supported that the highest dose of sulconazole inhibited activated cell proliferation to levels that were similar to inactive cells in vitro (Figure 2E and G). Indeed, sulconazole increased the proportion of cells arrested in the G0-G1 phase, reducing the percentage of cells in the S phase (Figure 2H and I). Overall, sulconazole appeared to reduce cell proliferation without causing overt cell death, similar to what has been described previously for pirfenidone.²⁰

Figure 2.

Sulconazole decreases fibroblast cell growth but does not cause cell death in vitro. (A) Growth curves for CCD-18Co and LX2 cells treated with increasing concentrations of sulconazole, as well as pirfenidone (n = 6). (B) Sulconazole and pirfenidone did not affect relative cell viability (n = 5–6) or (C) relative LDH activity in CCD-18Co and LX2 cells (n = 5–6). (D) Flow cytometry plot of apoptosis in sulconazole-treated CCD-18Co and LX2 cells. (E) Quantification of flow cytometry data in (D) demonstrated that apoptosis is not increased by sulconazole in CCD-18Co or LX2 cells (n = 3–4). (F) Ki67 immunofluorescence staining of CCD-18Co and LX2 cells treated with sulconazole. Scale bar: 100 μm. (G) Quantification of immunofluorescence image in (F) showed proliferation is decreased by sulconazole compared with activated fibroblasts (n = 3–7). (H) Flow cytometry histogram image of cell cycle analysis in CCD-18Co or LX2 cells treated with sulconazole. (I) Quantification of cell cycle analysis in (H) showed sulconazole induces G0-G1 arrest, which explains the decreased proliferation (n = 3). Data were presented as mean ± SD. *P < .05, **P < .01, ***P < .001 compared with activated control.

Sulconazole Effects on CCD-18co Cell Morphology and Gene Expression

We investigated the effects of increasing doses of sulconazole on fibroblast cell morphology as assessed by α-SMA staining (Figure 3A). We observed both a decrease in the intensity of α-SMA and the overall cell surface area with increasing doses of sulconazole (Figure 3B), similar to what has been reported for itraconazole-treated fibroblasts.⁵ When calculating the IC50 for the effect of sulconazole on the intensity and area of α-SMA expression, close correlation was observed (Figure 3C). To characterize pathways and specific genes affected by sulconazole, we performed transcriptomic analyses on activated CCD-18co cells treated with sulconazole vs vehicle control. As shown in Figure 3D, genes from the ontology groups ECM Remodeling Enzymes and Anti-Fibrotic were statistically significantly differentially expressed in a manner supportive of a reduction in fibrosis-related mechanisms in cells treated with sulconazole. Some of the genes in the epithelial to mesenchymal transition (EMT), TGF-β Superfamily Members, and ECM Remodeling Enzymes ontology groups were also differentially expressed supportive of reduced fibrotic response. More specifically, a volcano plot of relative expression of fibrosis gene showed that several genes known to be overexpressed in fibrosis were found to be down-regulated by sulconazole treatment (Figure 3E).

Figure 3.

Sulconazole decreases α-SMA, fibroblast surface area, and key profibrotic genes in CCD-18Co cells. (A) Sulconazole decreases the protein level of α-SMA in parallel with decreasing the surface area of fibroblasts. Scale bar: 100 μm. DAPI, 4′,6-diamidino-2-phenylindole. (B) Quantification of decreased α-SMA as well as cell surface area indicative of fibroblast inactivation with sulconazole treatment (n = 4–6). Data presented as mean ± SD. ***P < .001 compared with activated control. (C) The IC50 of sulconazole based on the relative expression of α-SMA and relative fibroblast cell surface area. (D) Transcriptomic analysis identified ontology groups of fibrosis-related genes affected by treatment with sulconazole. (E) Volcano plot showing relative expression of fibrosis-related genes that were affected by treatment with sulconazole.

Sulconazole Was Successfully Formulated as Sustained-Release Nanocrystals

We tested an array of generally regarded as safe stabilizers at a range of concentrations for use in nanomilling of sulconazole. The resulting nanocrystals varied in size and polydispersity, and generally larger molecular-weight polymer stabilizers, such as hyaluronic acid, resulted in larger and more polydisperse particle sizes (Supplementary Table 1). Incorporation of Pluronic F127 generally led to smaller and more uniform particle sizes, even as the concentration of sulconazole was increased from 10 mg/mL to 500 mg/mL (Supplementary Table 2). Although 2%–6% Pluronic F127 resulted in similar particle sizes, the 2% was chosen in the final formulation due to decreased viscosity, which is preferred for injection through an endoscopy needle. X-ray diffraction of raw and processed Sul-NC (500 mg/mL) revealed similar peak patterns and area under the curve, reflective of polycrystalline material that was largely unaltered by the nanomilling process (Supplementary Figure 1). Similarly, differential scanning calorimetry analysis indicated that the sulconazole and F127 in the 500 mg/mL Sul-NC were unaltered by the processing (Supplementary Figure 2). Electron microscopy of the 500 mg/mL Sul-NC showed nonspherical particles with cuboidal edges, reflective of the crystalline nature of the drug (Figure 4A). The particle size was consistent across 9 batches produced on different days, ranging in size from 196 to 261 nm (mean ± SD across batches, 232 ± 19 nm) (Figure 4B). Using a rapid equilibrium dialysis system for in vitro drug release under accelerated conditions, we confirmed that free sulconazole was not impeded by the dialysis membrane, and that the Sul-NC dissolved over a period of 7–8 days (Figure 4C). Of note, this in vitro experimental system underestimates the duration of drug elution in vivo. For example, in other work with intraocular injection, we have observed that a 7-day release in vitro corresponded to more than 3 months in vivo.²¹ The 500 mg/mL Sul-NC was also stable over 112 days of storage at room temperature or at 4°C, as assessed by measuring the particle size (Figure 4D).

Figure 4.

Sul-NC characterization. (A) Transmission electron microscopy (TEM) image of 500 mg/mL Sul-NC. Scale bar represents 500 nm. (B) Bar graph showing 500 mg/mL Sul-NC size over 9 batches with the overall average (Ave.) across batches shown as mean ± SD. (C) Sulconazole release under accelerated in vitro conditions in a rapid equilibrium dialysis device for Sul-NC compared with free sulconazole. Data shown as mean ± SD. (D) Stability of the 500 mg/mL Sul-NC as assessed by particle size during storage at room temperature or at 4°C for 112 days. Data shown as mean ± SD.

The Maximum Tolerated Dose of Sulconazole Nanocrystals Was 25-Fold Higher Than Free Sulconazole

In order to evaluate the safety of Sul-NC, as well as inform the dose for efficacy experiments, we determined the MTD for both free sulconazole and Sul-NC. First, 3 groups of 5 mice each received an intraperitoneal (IP) injection with a single dose of 15 mg/kg, 30 mg/kg, or 40 mg/kg free sulconazole. All mice survived (Supplementary Table 3). However, increasing the IP dose to 50 mg/kg resulted in a survival rate of only 3 of 9 mice (33%) (Supplementary Table 3). Similarly, dosing sulconazole at 50 mg/kg via the subcutaneous (SC) route resulted in survival of only 2 of 4 mice (50%) (Supplementary Table 3). All mice that did not survive died within 1–4 days of injection (not shown).

We then investigated the effect of nanoformulation on the tolerability of sulconazole injected SC. Having previously observed that injection volume can affect the rate and extent of drug uptake from crystalline formulations,²² we also tested different injection volumes (10–100 μL) and Sul-NC concentrations (100–500 mg/mL). We found that a combination of higher injection volume (50–100 μL) and lower sulconazole concentration (100–200 mg/mL) showed higher mortality (Supplementary Table 4). For example, only 2 of 3 mice (66%) survived after SC injection of 100 μL of the 100 mg/mL formulation (500 mg/kg), whereas there was 100% survival when injecting 50 μL of the 500 mg/mL formulation (1250 mg/kg) (Supplementary Table 4). This suggested that increasing the formulation concentration and using a smaller injection volume aided in slowing the rate of systemic drug absorption. However, as the Sul-NC dose was increased to 1875 mg/kg, only 2 of 4 mice survived (50%), and at 2500 mg/kg, 0 of 4 mice survived (0%). It is possible that a smaller injection volume (<75 μL) would have resulted in improved tolerability, but the concentration was the limiting factor at such high doses. Overall, Sul-NC doses as high as 1250 mg/kg were well-tolerated, whereas a free sulconazole dose of only 50 mg/kg resulted in reduced survival.

Sulconazole Nanocrystals Showed No Signs of Systemic Toxicity in Mice and Pigs In Vivo

We aimed to investigate whether the sustained local drug had any signs of systemic exposure associated with sulconazole. First, we injected C57Bl6 mice with Sul-NC SC at the concentration of 150 mg/kg and 300 mg/kg, respectively, and obtained peripheral blood 6 hours later. As shown in Supplementary Table 5, all measures of kidney and liver function were normal. Similarly, Sul-NC at doses of 450 mg/kg and 2500 mg/kg did not cause any notable histologic changes in the liver, kidney, lung, or spleen at 7 days after dose (Supplementary Figure 3). In addition, we also harvested major organs (ie, liver, lung, kidney, heart, and pancreas) in Sul-NC (150 mg/kg)–treated and untreated pigs. As shown in Supplementary Figure 4, there were no notable histologic findings in animals treated with Sul-NC vs control. Of note, we found 2 Sul-NC–treated pigs and 1 phosphate-buffered saline–treated control pig with mild liver steatosis, likely due to decreased food intake, as reported previously.²³

Sulconazole Nanocrystals Are an Effective Antifibrotic in Rodent Models of Skin and Bowel Fibrosis

There is a paucity of animal models of GI tract fibrosis or strictures.¹⁸ To obtain preliminary evidence regarding efficacy of Sul-NC as antifibrotic in vivo, we used a well-described and reproducible bleomycin-induced rodent skin fibrosis model.²⁴ To induce fibrosis, we injected bleomycin in mouse skin at 5 spots within a 1 × 1-cm region on the back every other day for 4 weeks (Supplementary Figure 5), as described previously.²⁴ Five groups of mice (n = 5 each) were treated with either free sulconazole at 10 mg/kg IP every other day, Sul-NC at 50 mg/kg or 150 mg/kg SC once per week, phosphate-buffered saline (vehicle) SC once per week, or daily oral pirfenidone at 100 mg/kg. Examining tissue sections of the skin samples dissected from regions injected with bleomycin by H&E staining or Masson’s trichrome staining (Figure 5A and B), we found that all treatments significantly reduced the thickness of the dermis compared with vehicle (Figure 5C). In addition, the Masson’s trichrome staining confirmed that the thickness of the dermis was due to deposition of collagen and other ECM proteins (blue staining in Figure 5). We further assessed expression of key genes involved in fibrosis in skin samples, including α-SMA (ACTA2), collagen type 1 (COL1A1), collagen type 3 (COL3A1), matrix metalloproteinase (MMP)-2, and MMP9. Both the 50 and 150 mg/kg doses of Sul-NC significantly reduced the expression of all genes assessed compared with vehicle (Figure 5D–H). In contrast, oral pirfenidone treatment significantly reduced expression of COL1A1, COL3A1, and MMP2 compared with vehicle, but not ACTA2 and MMP9 (Figure 5D–H). Similarly, the free sulconazole treatment significantly reduced the expression of only ACTA2 and COL1A1 (Figure 5D–H). Of note, there was a striking difference between the dose and frequency of administration when comparing pirfenidone (daily dosing, total dose 2800 mg/kg) vs Sul-NC (once weekly, total dose 200 mg/kg).

Figure 5.

Once-weekly injection of sul-NC reduced dermal thickening and reduced expression of key fibrosis-related genes in a bleomycin-induced mouse skin fibrosis model. (A) Representative tissue sections from mice with bleomycin-induced skin fibrosis (n = 5) after once-weekly injection of sul-NC (50 mg/kg or 150 mg/kg), IP injection of sulconazole (free sul, 10 mg/kg) every other day (every 2d IP), or daily oral administration of pirfenidone (100 mg/kg). Mice not induced with bleomycin shown as Sham, and bleomycin with vehicle injections shown as Vehicle. Dermis areas shown by double-sided arrows). (B) Mason’s trichrome staining of tissue sections shows the thickness and collagen layer of samples corresponding to each treatment group. Scale bar: 100 μm. (C) Quantification of dermal thickness from the bleomycin-induced skin fibrosis model. (D–H) Expression of key fibrosis-related genes in a bleomycin-induced skin fibrosis model after various treatments: (n = 3–5) after once-weekly injection of Sul-NC (50 mg/kg or 150 mg/kg), IP injection of sulconazole (free sul, 10 mg/kg) every other day (every 2d IP), or daily oral administration of pirfenidone (100 mg/kg). Mice not induced with bleomycin shown as Sham, and bleomycin with vehicle injections shown as Vehicle. Data shown as mean ± SD. *P < .05 compared with vehicle. (D) α-SMA (ACTA2), (E) collagen type I (COL1A1), (F) collagen type 3 (COL3A1), (G) MMP2, and (H) MMP9.

We then employed an intestine transplantation model (Supplementary Figure 6) that was previously used to validate the antifibrotic efficacy of oral pirfenidone.²⁵ Sul-NC at a dose of 50 mg/kg or 150 mg/kg were given as a single injection at the site of the transplantation and compared with oral pirfenidone at the dose previously described to be effective in the model—3 times daily oral dosing with 100 mg/kg. Tissue sections stained by Mason’s trichrome showed that Sul-NC significantly reduced collagen deposition (collagen layer thickness) at both 50 mg/kg (9.9 ± 0.6 μm) and 150 mg/kg (10.3 ± 0.2 μm) compared with vehicle-injected animals (13.2 ± 1.1 μm), and was similar to the 3 times daily oral pirfenidone (9.1 ± 0.3 μm) (Figure 6). Again, there was a striking difference between the dose and frequency of administration when comparing pirfenidone (3 times daily dosing, total dose 2100 mg/kg) vs Sul-NC (single dose, 50 mg/kg).

Figure 6.

Sul-NC reduces collagen deposition in a mouse model of intestinal tissue fibrosis. Representative Masson’s trichrome–stained intestinal tissue sections from mice (n = 4–5) treated with (A) phosphate-buffered saline (vehicle), (B) 3 times daily oral gavage with pirfenidone at 100 mg/kg for 7 days, (C) a single injection of Sul-NC at 50 mg/kg, (D) a single injection of Sul-NC at 150 mg/kg or (yello warrows indicate collagen capsule). Scale bar: 100 μm. (E) The collagen layer thickness in small intestine grafts was decreased significantly with the single Sul-NC injections and the daily oral pirfenidone treatment. Data shown as mean ± SEM. *P < .05 compared with vehicle control.

Sulconazole Nanocrystals Are Effective in a Swine Model of Luminal Gastrointestinal Tract Strictures

We recently established a clinically relevant pig esophagus stricture model that includes endoscopic balloon dilation and subsequent stricture reformation.¹⁸ Two weeks after induction, strictures with <6 mm diameter were first endoscopically balloon-dilated to 10 mm and then injected with either vehicle or Sul-NC. Two weeks post treatment, we used endoscopy (Figure 7A) and x-ray with contrast (Figure 7B) to examine the stricture sites, and observed significantly larger luminal openings in the pigs treated with Sul-NC (6.3 ± 1.9 mm) compared with vehicle (1.2 ± 0.9 mm) (Figure 7C). To evaluate the extent of fibrosis, we harvested tissue sections from treated stricture sites and stained them with H&E, Masson’s trichrome, and Sirius Red (Figure 7D). As expected, histopathologic analysis of stricture sites revealed fibrosis of the esophageal lamina propria and submucosa. Quantification of the collagen layer thickness (fibrosis) showed that Sul-NC treatment significantly reduced extent of fibrosis compared with vehicle (collagen layer thickness = 1.8 ± 0.2 mm [Sul-NC] vs 2.93 ± 0.7 mm [vehicle]) (Figure 7E). Furthermore, the Sul-NC–treated tissues showed a folded, flexible overlying epithelium more similar to healthy tissue from uninvolved regions of the esophagus (Figure 7D). These results confirmed the anti-fibrotic effect of the Sul-NC formulation in pigs.

Figure 7.

Sul-NC is effective in preventing fibrosis in a pig esophagus stricture model. (A) Representative endoscopic images of the esophageal lumen at the stricture sites (30, 40, and 50 cm from the incisors) in pigs treated with vehicle (2% F127) or Sul-NC. (B) X-ray images of strictures with contrast after treatment with vehicle or Sul-NC (yellow arrows point to esophageal lumen). (C) The luminal diameter was measured from tissue sections from pigs treated with vehicle (n = 12 strictures from 5 pigs) or Sul-NC (n = 9 strictures from 5 pigs). Data shown as mean ± SEM. *P < .05 compared with vehicle. images of stained tissue sections from the esophageal lumen of pigs treated with vehicle or Sul-NC at the site of the stricture compared with healthy, uninvolved regions of the esophagus. Scale bar represents 5 mm. (E) The collagen layer thickness in Sul-NC–treated esophagus was significantly decreased compared with control. Data shown as mean ± SD. *P < .01 compared with vehicle control.

Discussion

One of the major unmet clinical needs in the management of IBD is the management of treatment for patients with CD with anastomotic strictures. Within 20 years of diagnosis, up to 54% of patients with CD will develop fibrosis strictures.¹¹ Endoscopic dilation is possible only in select cases, whereas the majority require surgery with ensuing loss of bowel, morbidity, and mortality. However, strictures almost invariably recur, resulting in further surgery, abdominal adhesions, loss of bowel, short gut, and/or other complications.⁸ The seemingly immovable clinical evolution of these patients results, at least in part, from the lack of effective antifibrotic medications.

Recently, a number of antifibrotic candidates, including itraconazole (chemically modified to CBR-096–4),⁵ haloperidol,²⁶ and several other azole antifungal agents (oxiconazole, clotrimazole, and butoconazole),²⁷ have been identified through high-throughput screens. However, to our knowledge, there has not been validation of antifibrotic drugs in the GI tract, nor has there been a focus on formulation for local treatment. Although further work is needed to completely elucidate antifibrotic mechanisms downstream of sulconazole, here we report that sulconazole impacts the growth of activated fibroblasts through a relative arrest in the G0-G1 phase of the cell cycle. The end result is a slower growth rate of activated fibroblasts. Of note, in vitro studies found no increase in cell apoptosis, death, or toxicity, which is important for clinical translation. In addition, in vivo studies further demonstrated no toxicity, based on biochemical assays, as well as major organ histology. A second identified the mechanism of action for sulconazole as inactivation of fibroblasts, as evidenced by the decrease in the cell surface area and staining for α-SMA. Furthermore, we found that sulconazole decreases the secretion of ECM. Last, a number of fibrotic gene pathways were modulated, including ontology groups ECM Structural Constituents and Anti-Fibrotics.

Sulconazole inhibits a number of known profibrotic pathways (Figure 3E). For example, sulconazole induced a 3-fold down-regulation of Integrin α1 (ITGA1). Integrins are cell adhesion molecules that interact with proteins of the ECM, such as fibronectin and collagen.²⁸ Snail 1 (Snai1)²⁹ drives EMT in response to TGF-β, and blocking Snai1 signaling can alleviate fibrosis.³⁰ We found that sulconazole treatment resulted in a 4.5-fold inhibition of Snai1. Tissue inhibitor of metalloproteinases 1 (TIMP1, down-regulated 2-fold by sulconazole) participates in the maintenance of homeostasis balance of ECM by inhibition of MMPs.³¹ TIMP1 was found to increase in parallel with development of fibrosis in a number of organs, such as skin, kidney, liver, and others.³² In addition, key fibrosis genes Collagen I and III, as well as ACTA2 were also down-regulated by treatment with sulconazole by a factor of 3, 5, and 4, respectively. Among transcripts that were up-regulated was interleukin-10 (IL10, 7-fold up-regulation). IL10 has pleiotropic effects, and is generally considered to be an anti-inflammatory cytokine, as well as a gatekeeper for fibrotic processes.³³ IL-13 receptor α2 (IL13RA2, 8-fold up-regulation) has been previously shown to act as a nonsignaling decoy receptor.³⁴ This receptor binds IL-13 with high affinity and inhibits the IL-13 induction of collagen and TGF-β.³⁴ Bone morphogenic protein 7 (BMP7, 2.5-fold upregulation) inhibits fibrosis in kidneys,³⁵ and also inhibits scar formation.³⁶ Serine protease inhibitors (Serpins) are a family of protease inhibitors.³⁷ Serpina1 (up-regulated by treatment with sulconazole) encodes antitrypsin, and Serpine1 (inhibited by treatment with sulconazole) encodes the plasminogen activator inhibitor I.³⁷ Their roles in tissue fibrosis remains to be elucidated.

Some early efforts toward localized treatment to prevent fibrosis in the GI tract involved injection of anti-inflammatory agents, such as triamcinolone¹⁷ or anti-fibrotic nucleic acids,³⁸ directly into the inflamed tissue. However, these attempts did not result in clinically meaningful results, likely because a single injection would have a short duration of action before being cleared from the body. In contrast, tissue remodeling processes like fibrosis occur over days to weeks, necessitating the development of a sustained-release formulation. As sulconazole has low water solubility, it is amenable to various approaches for sustained drug release, including encapsulation into polymer matrices. When considering the limitations on the volume that can be injected directly into intestinal tissue and the need for loading the highest amount of drug possible to achieve the longest duration of therapeutic effect, formulation as particulates with pure drug cores can achieve higher drug loadings than encapsulation.^22,39 For example, the 500 mg/mL Sul-NC formulation described here would contain approximately 96% drug loading by weight. Reduced amounts of excipients may also be advantageous for minimizing the potential for injection site reactions or buildup of materials that can also drive a fibrotic response. Furthermore, we observed that increasing the drug concentration and reducing the injection volume led to an increase in the MTD for Sul-NC (approximately 1250 mg/kg compared with 50 mg/kg for free sulconazole) with subcutaneous injection in mice, likely due to the decreased surface area for absorption and decreased dissolution rate.²² Future work to increase the crystallinity of the formulation either by using bottom-up formulation approaches or by fully crystallizing the sulconazole before processing using top-down formulation approaches could further prolong the duration of drug release.

Although more in-depth pharmacokinetic and safety studies are required, there is a rationale for repurposing a drug typically employed with topical administration as an injectable for the GI tract. Prior pharmacokinetic studies showed that systemic absorption of topically applied sulconazole was much higher than other azoles, in the range of 8.7%–11.3% of a 9-g dose (two 4.5-g doses 12 hours apart).⁴⁰ Furthermore, elimination via the feces was one of the primary clearance routes, suggesting that topical sulconazole dosing is already associated with relatively high levels of drug exposure in the intestines.⁴⁰ Furthermore, the potential toxicity of sulconazole is likely associated with its accumulation in the liver, which is going to be much higher with systemic dosing.

This study represents the first demonstration of successful nanocrystal formulation of sulconazole for local, sustained-release, antifibrotic effect in the GI tract. One potential limitation of the study is that animal models of GI tract fibrosis are few. In addition, fibrosis in the GI tract is initiated by a variety of stimuli. However, to alleviate some of these concerns, the efficacy of Sul-NC was tested here in 3 animal models in which fibrosis was induced by different inciting factors, that is, chemical (the skin bleomycin model), ischemia (bowel transplantation model), and thermal (swine GI stricture model). Although none of these models are a perfect representation of strictures in the GI tract in patients, the fact that Sul-NC was effective across these models that included different organ systems and inciting factors is reassuring and creates the foundation for further studies that are aimed at specific etiologies for fibrosis in the GI tract. In addition, although we have performed MTD experiments in rodents, further work should be focused on similar experiments after injection into the GI tract, because the systemic absorption may be different vs injection in the subcutaneous space. Therefore, a scaled-up systemic preclinical study of its pharmacokinetics, toxicity, and MTD is warranted as future work.

WHAT YOU NEED TO KNOW.

BACKGROUND AND CONTEXT

Fibrosis and strictures in the gastrointestinal tract are major contributors to morbidity and health care costs, for which, currently, there are no US Food and Drug Administration–approved antifibrotic medications.

NEW FINDINGS

A sustained-release nanocrystal formulation of sulconazole is effective as an antifibrotic and can prevent stricture formation in preclinical models of fibrosis.

LIMITATIONS

Further preclinical studies are needed before human clinical trials.

CLINICAL RESEARCH RELEVANCE

The current study lays the foundation for advancing sustained-release formulations of sulconazole to human clinical trials for fibrotic indications. In addition, the general methodology can be used to discover and validate other antifibrotic small molecules.

BASIC RESEARCH RELEVANCE

The current study uncovers antifibrotic mechanisms of action and gene pathways for sulconazole. The principles of nanocrystal formulation lay the foundation for other sustained-release, locally administered antifibrotics.

Abbreviations used in this paper:

α-SMA

α–smooth muscle actin

CD

Crohn’s disease

ECM

extracellular matrix

GI

gastrointestinal tract

IBD

inflammatory bowel disease

IL

interleukin

IP

intraperitoneal

MMP

matrix metalloproteinase

MTD

maximum tolerated dose

SC

subcutaneous

Sul-NC

sulconazole nanocrystal

TGF-β

transforming growth factor–β

Data Availability

The main data supporting the findings of this study are available within the article and its Supplementary Material. The associated raw data are available from the corresponding author on reasonable request.