Super Carbonate Apatite-miR-497a-5p Complex Is a Promising Therapeutic Option against Inflammatory Bowel Disease

The incidence of inflammatory bowel disease (IBD) is increasing worldwide. It is reported that TGF-β/Smad signal pathway is inactivated in patients with Crohn’s disease by overexpression of Smad 7. With expectation of multiple molecular targeting by microRNAs (miRNAs), we currently attempted to identify certain miRNAs that activate TGF-β/Smad signal pathway and aimed to prove in vivo therapeutic efficacy in mouse model. Through Smad binding element (SBE) reporter assays, we focused on miR-497a-5p. This miRNA is common between mouse and human species and enhanced the activity of TGF-β/Smad signal pathway, decreased Smad 7 and/or increased phosphorylated Smad 3 expression in non-tumor cell line HEK293, colorectal cancer cell line HCT116 and mouse macrophage J774a.1 cells. MiR-497a-5p also suppressed the production of inflammatory cytokines TNF-α, IL-12p40, a subunit of IL-23, and IL-6 when J774a.1 cells were stimulated by lipopolysaccharides (LPS). In a long-term therapeutic model for mouse dextran sodium sulfate (DSS)-induced colitis, systemic delivery of miR-497a-5p load on super carbonate apatite (sCA) nanoparticle as a vehicle restored epithelial structure of the colonic mucosa and suppressed bowel inflammation compared with negative control miRNA treatment. Our data suggest that sCA-miR-497a-5p may potentially have a therapeutic ability against IBD although further investigation is essential.


Introduction
Inflammatory bowel disease (IBD) such as ulcerative colitis (UC) and Crohn's disease (CD) is an intractable chronic inflammatory disease, and the number of patients is increasing in the world year by year [1][2][3]. Medical treatments such as 5-aminosalicylic acid (5-ASA), corticosteroids, and anti-tumor necrosis factor-α (TNF-α) antibody are first line-therapies against IBD, but remissions and relapses are often repeated [4,5]. In recent years, antiinterleukin 12/23 antibody, JAK inhibitors, and anti-α4β7 integrin antibody emerged as new molecular-targeted drugs [6][7][8][9], but they carry the risk of immunocompromise, allergy and other side effects and they still cannot cure IBD. Therefore, continuous effort to develop novel therapy is required against IBD.
Although the cause of IBD has not been fully clarified, involvement of genetic factors and environmental factors is suggested [10][11][12]. When the barrier mechanism of the intestinal mucosa is destroyed, food residues and intestinal bacteria are phagocytosed by antigen presenting dendritic cell which present antigen to Naïve T cells and induce differentiation into regulatory T lymphocytes (Treg) and inflammatory T lymphocytes (Th17) [13,14]. In IBD patients, Th17 becomes dominant and Treg declines, so that inflammatory cytokines, TNF-α, and interferon-γ (IFN-γ) increase, and an anti-inflammatory cytokine transforming growth factor-β (TGF-β) decreases [15]. It is reported that TGF-β/Smad signal pathway is suppressed in IBD patients [16][17][18]. Smads involved in this pathway are classified into three types: Inhibitory Smads (I-Smad: Smad 6/7) that inhibit the signal pathway, Common mediator Smad (Co-Smad: Smad 4) that forms a complex with Smad 2/3, and Receptor-regulated Smads (R-Smads: Smad 2/3 and others) that activate the signal pathway [19,20]. It is reported that Smad 7 was highly expressed in mononuclear cells at intestinal lamina propria in patients with IBD [17,18,21]. Intestinal macrophages also play an important role in IBD [22][23][24]. It is reported that intestinal-specific macrophages subset CD14 + macrophages produce a large amount of inflammatory cytokines IL-23, TNF-α and IL-6, leading to chronic inflammation in Crohn's disease [25].
MicroRNA (miRNA) is a single-stranded non-cording RNA of 21 to 25 bases MiRNA that binds to the 3 UTR of the target mRNA to suppress translation, or control gene expression by cleaving mRNA [26,27]. Although limited numbers of siRNA-and miRNAbased therapeutic options have advanced to clinical stages [28][29][30][31][32][33][34][35], venous infusion of nucleic acid medicine is expected as a powerful therapeutic option especially against severe IBD at acute exacerbation. Using IBD models considerable efforts have been made for systemic delivery of various miRNAs [36][37][38][39][40][41][42][43], but it still remains an unsolved clinical challenge mainly due to lack of suitable delivery system. Thus, miRNA and siRNA are rapidly degraded when administered to the blood stream, which made it difficult to supply sufficient amount of nucleic acid to target lesions.
A phase II clinical trial showed that oral Smad 7 antisense oligonucleotides improved clinical symptoms in patients with Crohn's disease [21], but the phase III clinical study was unfortunately discontinued [56]. Some reports suspect insufficient quality of nucleic acid prepared in the phase III study [56][57][58][59]. Unlike antisense oligonucleotides, miRNA can bind to and regulate multiple genes [26,27]. Instead of single molecule targeting, we currently attempted to identify certain miRNAs based on TGF-β/Smad signal activity, which should exert multiple function. Finally, we investigated therapeutic efficacy of miR-497a-5p in mouse dextran sodium sulfate (DSS)-induced colitis using super carbonate apatite as a systemic delivery vehicle.  Figure S1, Point 1). Among 18 miRNAs we chose 13 miRNAs which conserve identical sequences also in human species (Supplementary Figure S1, Point 2). Potential binding combination between 3 UTR mRNA of the negative regulators and mmu miRNAs are summarized in Table 1. To find certain miRNAs that activate TGF-β/Smad signal pathway, SBE reporter assays were performed using the two different systems (Supplementary Figure S1, Point 3,4). Supplementary Figure S2 shows the principle of this reporter assay. Thus, once Smad 3/Smad 4 binds to SBE together with various transcriptional factors, luciferase signal comes out.
In the second screening, we employed a dual luciferase assay system in which SBE activity is normalized by expression of co-transfected Renilla luciferase vector, thus providing more accurate data. Seven miRNAs selected in the first-round screening were transfected 24 h prior to transfection of the plasmids. Then cells were exposed in the assay medium containing 0.5 ng/mL TGF-β for 24 h (Figure 2A). As results, we found that 3 miRNAs (miR-497a-5p, miR-186-5p, miR-195a-5p) again significantly activated the SBE activity when compared with miR-NC (* p < 0.05) ( Figure 2B). Because miR-195a-5p had already been reported as a potential treatment option for IBD by promoting intestinal barrier integrity and restoration of the intestinal epithelium [62,63], we focused on miR-479a-5p and miR-186-5p in the subsequent experiments. In the second screening, we employed a dual luciferase assay system in which SBE activity is normalized by expression of co-transfected Renilla luciferase vector, thus providing more accurate data. Seven miRNAs selected in the first-round screening were transfected 24 h prior to transfection of the plasmids. Then cells were exposed in the assay medium containing 0.5 ng/mL TGF-b for 24 h ( Figure 2A). As results, we found that 3 miRNAs (miR-497a-5p, miR-186-5p, miR-195a-5p) again significantly activated the SBE activity when compared with miR-NC (* p < 0.05) ( Figure 2B). Because miR-195a-5p had already been reported as a potential treatment option for IBD by promoting intestinal barrier integrity and restoration of the intestinal epithelium [62,63], we focused on miR-479a-5p and miR-186-5p in the subsequent experiments. , 3miRNAs (miR-497a-5p, miR-186-5p, miR-195a-5p) significantly activated SBE reporter activity when compared to miR-NC (** p < 0.01, * p < 0.05, miR-NC vs. miR-497a-5p, p = 0.016; miR-NC vs. miR-186-5p, p = 0.001; miR-NC vs. miR-195a-5p, p = 0.001).

Smad 7 Is a Direct Target of miR-497a-5p
Based on the findings of western blots, we preferentially focused on miR-497a-5p. It is reported that miR-497-5p indirectly activated latent TGF-β via reversion-inducible cysteine- rich protein (Reck) in lung fibrosis model [67]. Here we show that miR-497a-5p directly inhibit Smad 7 expression. In silico survey showed that mouse Smad 7 mRNA has the binding site of miR-497a-5p in its 3 UTR ( Figure 4A). Seed sequence of human miR-497-5p and its binding site in 3 UTR of human Smad 7 mRNA are both well conserved between mouse and human species (Supplementary Figure S3). We constructed a luciferase reporter plasmid containing the miR-497a-5p binding sites in the 3 UTR of Smad 7 ( Figure 4B). When luciferase assay was performed using HCT116 cells, it was revealed that miR-497a-5p significantly suppressed luciferase activity compared with miR-NC (p < 0.05), indicating the direct binding between miR-497a-5p and the 3 UTR of Smad 7 ( Figure 4C).   It is reported that Smad 7 was highly expressed in mononuclear cells in lamina propria of intestinal mucosa in patients with IBD [16][17][18]21]. A part of mononuclear cells turns into macrophages which produce a large number of inflammatory cytokines such as IL-  It is reported that Smad 7 was highly expressed in mononuclear cells in lamina propria of intestinal mucosa in patients with IBD [16][17][18]21]. A part of mononuclear cells turns into macrophages which produce a large number of inflammatory cytokines such as IL-23, TNF-α, and IL-6 by stimulation of intestinal bacteria, leading to chronic inflammation [23][24][25]. Co-culture of macrophages and intestinal epithelial cells is also used as a colitis model in vitro [68]. Therefore, we examined whether miR-497a-5p would suppress the production of inflammatory cytokines TNF-a, IL-6, and IL-12p40 (a subunit of IL-23), when lipopolysaccharides (LPS) at 100 ng/mL was added to mouse macrophage cell line J774a.1 according to the time schedule shown in Figure 5A. qRT-PCR assays showed that miR-497a-5p suppressed the production of TNF-α, IL-6, and IL-12p40 compared with miR-NC at the indicated time points with asterisks (* p < 0.05, Figure 5B).  Figure 5A. qRT-PCR assays showed that miR-497a-5p suppressed the production of TNF-α, IL-6, and IL-12p40 compared with miR-NC at the indicated time points with asterisks (* p < 0.05, Figure 5B).

sCA Delivered miRNA to Macrophages in Colonic Mucosa
In our previous study, we showed that sCA incorporating miR-NC tagged with Alexa Fluor 647 was largely co-localized with CD11c + dendritic cells in the inflamed colon [46]. In this study, we performed in vivo uptake test of miRNA into macrophages. To visualize the extent and localization of miRNA in the normal and inflamed colon, sCA incorporating miR-NC tagged with Alexa Fluor 647 was administered via tail vein, and the colon was excised 4 h after administration. Fluorescence microscopy showed that the red fluorescence of the Alexa 647 conjugate miR-NC was present in the mucosa and submucosa of the colonic epithelium. Immunostaining of macrophages with the anti-F4/80 antibody showed that co-localization of miRNA with the F4/80 positive macrophages was often found ( Figure 6A) and the percentage of uptake of miRNA in macrophages was 47.12 ± 8.27 in inflamed colon and 38.23 ± 2.79 in normal mucosa, respectively ( Figure 6B). There was no significant difference between the two groups.
Anti-TNF-a antibodies such as infliximab and adalimumab are already used in the treatment of IBD [69]. Because miR-497a-5p was able to suppress IL-6 and IL-12p40 in addition to TNF-α in J774a.1, sCA-miR-497a-5p complex targeting macrophages at inflamed colon may have a clinical benefit.

Therapeutic Efficacy of Systemic Administration of sCA-miR-497a-5p on Mouse DSS-Induced Colitis
Mice were treated with 1.5% DSS in drinking water for 16 days. sCA-miR complexes were injected to tail vein 8 times on days 9, 11, 13, 15, 17, 19, 21, and 23. On day 24, mice were sacrificed ( Figure 7A). Here we attempted a long-term experiment to evaluate the therapeutic efficacy of miR-497a-5p; 1.5% DSS for 16 days followed by therapeutic treatments from day 9 to day 23 every two days. Because most studies were performed to assess preventive effect of drugs or gene manipulation in DSS-induced colitis [69][70][71][72][73], we are not aware of any reports that assessed the therapeutic effect of miRNA in DSS-induced colitis especially in such a long-term schedule. As a result, a drastic inflammatory change was noted as early as on day 5 in the inflamed rectum and colon (Supplementary Figure S4). Compared with normal colon epithelium, DSS treatment alone or DSS and sCA-miR-NC destroyed normal epithelial structures, and numerous inflammatory cells infiltrated into the lamina propria of colonic mucosa ( Figure 7B). By contrast, DSS and sCA-miR-497a-5p treatment restored epithelial structures of the colonic mucosa and infiltration of inflammatory cells rather decreased ( Figure 7B). The colon length was significantly longer in mice treated with DSS and sCA-miR-497a-5p as compared to those treated with DSS alone or DSS and sCA-miR-NC (* p < 0.05, Figure 7C). There was no significant difference in body weight loss among the DSS-treated groups ( Figure 7D). Significantly worse histological scores in mice treated with DSS alone or DSS and sCA-miR-NC were noted, whereas sCA-miR-497a-5p treatment significantly improved the histological damages ( Figure 7E, * p < 0.05). here. Normal mice (n = 3), DSS-treated mice (n = 3), DSS and sCA-miR-NC-treated mice (n = 4), DSS and sCA-miR-497a-5p-treated mice (n = 3). (B) H&E staining of distal colon in each group. The mucosal structure was destroyed and many inflammatory cells were noted in DSS-treated mice or DSS and sCA-miR-NC-treated mice. By contrast, DSS and sCA-miR-497a-5p-treated mice had the notable therapeutic effect. Scale bars, 100 µm for each. (C) The colon length was significantly longer in mice treated with DSS and sCA-miR-497a-5p as compared to those treated with DSS alone or DSS and sCA-miR-NC (** p < 0.01, * p < 0.05, DSS alone vs. DSS and miR-497a-5p, p = 0.002; DSS and miR-NC vs DSS and miR-497a-5p, p = 0.022). (D) There was no significant difference in body weight loss among the DSS-treated groups. (E) Significantly worse histological scores in mice treated with DSS alone or DSS and sCA-miR-NC were noted, whereas sCA-miR-497a-5p treatment significantly improved the histological damages (** p < 0.01, * p < 0.05, DSS alone vs DSS and miR-497a-5p, p = 0.046, DSS and miR-NC vs DSS and miR-497a-5p p = 0.034).

Therapeutic Efficacy of Systemic Administration of sCA-miR-186-5p on Mouse DSS-Induced Colitis
Finally, we compared the in vivo efficacy of miR-186-5p and miR-497a-5p loaded on sCA. Studies have shown anti-tumor effect of miR-186-5p in carcinomas of colon, breast, bladder, prostate, and osteosarcoma through maintaining NK cell stability and suppressing epithelial-mesenchymal transition (EMT) [74][75][76][77][78][79], but its role in IBD has not been investigated. A shorter time course study, where 2% DSS in drinking water was given for 8 days and sCA-miRNAs were injected to tail vein 6 times ( Figure 8A), indicated that miR-186-5p had similar therapeutic efficacy to miR-497a-5p in terms of histological score ( Figure 8B-E). Our current data with regard to selected three miRNAs acting at activation of TGF-β/Smad signal pathway support the notion that this pathway is an important factor to suppress IBD.

Therapeutic Efficacy of Systemic Administration of sCA-miR-186-5p on Mouse DSS-Induced Colitis
Finally, we compared the in vivo efficacy of miR-186-5p and miR-497a-5p loaded on sCA. Studies have shown anti-tumor effect of miR-186-5p in carcinomas of colon, breast, bladder, prostate, and osteosarcoma through maintaining NK cell stability and suppressing epithelial-mesenchymal transition (EMT) [74][75][76][77][78][79], but its role in IBD has not been investigated. A shorter time course study, where 2% DSS in drinking water was given for 8 days and sCA-miRNAs were injected to tail vein 6 times ( Figure 8A), indicated that miR-186-5p had similar therapeutic efficacy to miR-497a-5p in terms of histological score (Figure 8B-E). Our current data with regard to selected three miRNAs acting at activation of TGF-b/Smad signal pathway support the notion that this pathway is an important factor to suppress IBD.

Limitation and Future Perspective
There are several limitations in this study. (i) TGF-β activation and production of cytokines from mouse macrophages had not been examined in the in vivo model yet. (ii) It remains to be clarified how miR-186-5p acts against IBD. (iii) In vivo experiments for miR-186-5p should be repeated although in vivo efficacy of miR-497a-5p was confirmed by two different experiments. During preparation of this manuscript, Zhang M et al. demonstrated a preventive role of miR-497 in DSS-induced colitis using knockout mice and inhibition of Wnt/β-catenin pathway was suggested as one possible mechanism [80]. Collectively it is considered that miR-497 exerts multiple functions such as activation of TGF-β signaling pathway through targeting Smad 7 and inhibition of Wnt/β-catenin pathway. Our study proved therapeutic efficacy of miR-497a-5p using sCA as a delivery tool. Recent review articles introduce sCA nanoparticle as a hopeful non-viral systemic strategy [81][82][83][84].
The specific miRNAs (mmu miR-497a-5p) and the negative control miRNA-1 (NC-miR-1) were used in in vivo experiments. The specific miRNAs and NC-miR-1 were purchased from Gene Design Inc. (Ibaragi, Osaka, Japan) and NC-miR-2 was purchased from Sigma-Aldrich. The sequences of miRNAs used are listed in Supplementary Table S1.

TGF-β Pathway-Responsive Reporter Assays
The first round screening was performed using HEK293 cells where SBE reporter plasmid was introduced (BPS Bioscience, Cat. No. 60653, Court West, Suite E San Diego, CA, USA). The cells were maintained with 400 µg/mL of Geneticin (Invitrogen, Cat. No. 10131035, Carlsbad, CA, USA). Cells were seeded in 96-well plates at a density of 2.5 × 10 4 per well and transfected with miR-NC and candidate miRNAs at a final concentration of 50 nM. The second round screening was performed using SBE Reporter Kit (BPS Bioscience, Cat. No. 60654). The kit contains transfection-ready SBE luciferase reporter vector. This reporter contains a firefly luciferase gene under the control of multimerized SBE responsive element located upstream of a minimal promoter. The SBE reporter is premixed with constitutively expressing Renilla-Sea Pansy luciferase vector that serves as internal control for transfection efficiency. Luciferase assay was performed using Dual-Luciferase ® Reporter Assay System (Promega, Cat, No. E1910, Madison, WI, USA) and luminescence was measured by a luminometer (TriStar 2 LB942).

Western Blotting
Cells were seeded in six-well plates at a density of 1 × 10 5 -2 × 10 5 per well and transfected with miR-NC, miR-497a-5p and miR-186-5p at a final concentration of 50 nM.  The sequences of inserts and vectors were confirmed by Sanger sequencing.
Cells were seeded in 96-well plates at a density of 1 × 10 4 cells per well and were co-transfected with 50 ng pmirGLO plasmid vector containing the insert and either miRnegative control (5 pmol) or miR-497a-5p (5 pmol). At 24 h after transfection, firefly and Renilla luciferase activities were measured using the Dual-Luciferase Reporter Assay System (Promega, Cat, No. E1910). All experiments were conducted in triplicate.

qRT-PCR
Total RNA was extracted using TRIzol TM Reagent (Invitrogen, Cat, No. 15596018). RNA quality was assessed with a NanoDrop ONE spectrophotometer (Thermo Fisher Scientific, Wilmington, DE, USA). About 2 µg of RNA was reverse transcribed with the high-capacity RNA to cDNA Kit (Applied Biosystems, Cat, No. 4388950, Foster City, CA, USA). qPCR analysis was performed using THUNDERBIRD SYBR qPCR Mix (TOYOBO LIFE SCIENCE, Cat, No. QPS-201). The qPCR was performed on the LightCycler ® 480 real-time PCR system (Roche Diagnostics, Basel, Switzerland). The qPCR conditions were as follows: 95 • C for 30 s; followed by 40 cycles of 95 • C for 10 s, 60 • C for 10 s and 72 • C for 30 s. The expression of the target gene was normalized to endogenous GAPDH expression. Relative expression was quantified by the 2 −∆∆Cq method.

Therapeutic Model for DSS-Induced Mouse Colitis
Eight-week-old BALB/c mice (female) which retain intact immune system were purchased from CLEA (Tokyo, Japan). DSS (MW 36,000-50,000) was purchased from MP Biomedicals (Cat, No. 9011-18-1, Santa Ana, CA, USA). For producing therapeutic model of DSS-induced colitis, drinking water at a concentration of 1.5% DSS was given to mice for 16 days with reference to previous studies [46,81]. MiR-497a-5p loaded on super carbonate apatite nanoparticle was injected eight times on the tail vein from day 9 to day 23 every two days. Mice were sacrificed on day 24. For a comparative therapeutic study between sCA-miR-497a-5p and sCA-miR-186-5p, 2% DSS in drinking water was given to mice for 8 days [82][83][84]. MiR-497a-5p or miR-186-5p loaded on super carbonate apatite nanoparticle was injected on days 8,9,10,12,13

Fluorescent Immunostaining of Macrophages at Propria Muscularis of Colon Mucosa
DSS-induced colitis was produced by free drinking of 2% DSS for 7 days in female mice (n = 2). Non-treated mice (n = 2) served as a comparative reference. The Alexia 647-tagged NC-miRNA (25 µg) encapsulated in sCA was injected into the tail vein and the distal colon was collected 4 h later, and frozen in OCT compound. About 8 µm sections were cut and fixed in 4% paraformaldehyde. The frozen sections (n = 6 per mouse) were incubated overnight with rat anti-mouse F4/80 antibody (BIO RAD, Cat, No. MCA497G, Hercules, CA, USA) at a concentration of 1:100. As a secondary antibody, FITC-conjugated goat antirat IgG was used (Jackson ImmunoResearch, Cat, No. 112-095-167, West Grove, PA, USA). The nuclei were stained with ProLong Gold anti-fade reagent with DAPI (Invitrogen, Cat, No. #8961). Sections were observed using a fluorescence microscope (BZ-X 700, Keyence Corporation, Osaka, Japan).

Statistics
F-test was performed to find out if there were equal variances between the two groups. Statistical significance of the difference between two groups was then calculated by Student's t-test or Welch's t-test, and data are presented as means ± standard deviations unless specifically otherwise indicated. When more than two groups were compared, one-way ANOVA was used followed by Bonferroni correction to determine the statistical significance of the differences. Statistical analyses were performed using the JMP13 program (SAS Institute, Cary, NC, USA). Differences with p < 0.05 were considered significant (File S1).

Conclusions
In conclusion, we have demonstrated that sCA-miR-497a-5p complex exerts a potent anti-inflammatory effect through activation of TGF-β/Smad signal pathway and inhibition of secretion of inflammatory cytokines from macrophages in IBD therapeutic mice model. These results suggest that sCA-miR-497a-5p may potentially have a therapeutic ability against IBD although further investigation is essential.