Oral delivery of liquid mRNA therapeutics by engineered capsule for treatment of preclinical intestinal disease

Abstract

Oral delivery of messenger RNA (mRNA) therapeutics could offer non-invasive and self-administered treatments and vaccinations. However, the development of oral mRNA therapeutics remains challenging due to the degradative conditions of the gastrointestinal (GI) tract. Here, we engineered a capsule-based device, named RNACap, designed for oral delivery of liquid mRNA nanoparticle (NP) therapeutics to the intestines. RNACap protects mRNA from the acidic stomach environment while allowing rapid release into the intestines in response to intestinal neutral pH, the pressure release due to the dissolution of capsule cap and natural intestinal contractions (peristalsis). This process enables NP-mediated delivery of mRNA into intestinal cells for in vivo transfection. We optimized a NP formulation for rapid intestinal mRNA delivery. In rat and porcine models, we confirmed that the RNACap remains intact in the stomach but releases its contents within the intestines. The release of mRNA-NPs led to the expression of multiple mRNAs. The therapeutic effect of the RNACap was demonstrated by acute and delayed treatment in two rat colitis models. Orally-administered RNACap loaded with mRNA encoding interleukin 10 (IL-10 mRNA-NP) reduced proinflammatory cytokine concentrations in both blood and tissues, ultimately alleviating colitis. Furthermore, using a large animal model of swine, we showed that RNACap remained intact in the stomach, disassembled in the intestine, and resulted in robust mRNA expression just 8.5 h after administration. RNACap represents a promising platform for the oral delivery of liquid mRNA therapeutics to the GI tract for treating challenging intestinal diseases and potentially many other conditions.

One Sentence Summary:

RNACap can orally deliver liquid mRNA therapeutics to the intestines in response to intestinal pH, the resulting pressure release and contractions.

INTRODUCTION

Building on the remarkable success of mRNA vaccines in combating Covid-19 (1-3), recent years have witnessed the rapid expansion of messenger RNA (mRNA)-based therapeutics for treating and preventing a range of diseases (4-7), with many potential therapeutics now in advanced clinical trials (1). Although current mRNA therapeutics are primarily administered through needle injections, the reliance on a substantial number of trained professionals and the necessity for patients to visit clinics or pharmacies for drug administration have posed limitations on the widespread and rapid deployment of mRNA vaccines during pandemics (8). Additionally, this approach may not be practical for patients with chronic diseases requiring repeated dosing. Non-invasive and self-administered mRNA therapeutics, such as oral mRNA therapies, offer a promising solution to these challenges (9).

Oral delivery stands as the preferred method for drug administration for both patients and physicians (10-13). Unlike injection-based delivery, oral administration offers a painless, convenient, and self-administered approach, enhancing patient compliance and adherence, and extending population coverage (14), a critical aspect in rapidly controlling the spread of viruses during pandemics. In the context of infectious diseases, oral vaccines have the capacity to induce robust systemic and mucosal immunity (15). The latter is particularly crucial for safeguarding against infections at mucosal surfaces, an outcome not achievable with injection-based vaccines (16, 17). Moreover, oral administration also holds the potential to target cells within the gastrointestinal (GI) tract, a region that is challenging to reach through injection-based methods (18). Oral drug administration has also proven to be a highly effective method in treating intestinal diseases, for example, inflammatory bowel disease (IBD) and colon cancer.

Despite these convincing advantages, effective oral delivery of fragile macromolecules like mRNA remains challenging due to the physiological barriers within the GI tract (18, 19). Nucleic acid drugs (20) face rapid degradation by enzymes, such as pepsin, and the highly acidic conditions of the stomach (21, 22). Upon reaching the intestines, these drugs encounter another dual barrier of enzymatic degradation by nucleases and a mucus barrier composed of a layer of viscous, hydrophilic, high-molecular-weight glycoproteins that impede their access to the epithelial surface (15, 23). These challenges have been partially addressed by nanotechnology. For instance, both polymeric nanoparticles (NPs) (24) and lipid NPs (25) have been employed to facilitate the oral delivery of mRNA vaccines and mRNA therapeutics to the intestines for immune activation and treatment of ulcerative colitis, respectively. Oral administration of mRNA vaccines formulated with lipid NPs to tumor-bearing mice demonstrated suppression of tumor growth (26). In another study, the oral delivery of lipid NPs containing mRNA resulted in higher gene expression in the upper GI tract than in the intestines (27). However, these systems expose their mRNA-loaded NPs directly to the hostile stomach environment, potentially leading to NP destabilization and damage to the mRNA cargos, resulting in suboptimal efficacy. In another study, oral mRNA delivery to the stomach (28) was achieved by injecting a dry mRNA formulation into the relatively thick gastric wall (about 10 mm (15)) using a well-established ingestible milli-injector capsule platform (29, 30). However, this system may not be suitable for oral delivery of mRNA to the intestines, which have a relatively thin wall (about 2 mm (15)), due to the risk of perforation. Therefore, an oral mRNA delivery platform that can protect the mRNA formulations from the harsh stomach environment and enable mRNA entry into the intestinal cells for efficient transfection is highly desirable.

Here, we present a capsule device, RNACap, designed for oral delivery of liquid mRNA therapeutics to the intestines. mRNA is carried by a polymer-lipid hybrid NP, which is further encapsulated in RNACap to aid in GI transit. The lipid-based NPs used are specifically designed and optimized to facilitate the delivery of mRNA to intestinal cells for efficient transfection. RNACap can accommodate liquid mRNA formulations, eliminating the need for costly and time-consuming lyophilization processes (31), which can reduce efficacy even with the use of cryoprotectants (32). Unlike dry formulations, liquid formulations are readily absorbable, making them ideal for the rapid delivery of vulnerable mRNA therapeutics to intestines. All currently approved mRNA vaccines in the clinic are provided in liquid (aqueous) form (5). We first evaluated the capability of RNACap to deliver reporter mRNA, such as mRNA encoding enhanced green fluorescence protein (EGFP), and therapeutic mRNA, such as mRNA encoding interleukin 10 (IL-10), to the intestines of rats by oral administration. Then, we assessed the therapeutic efficacy of IL-10-mRNA-RNACap in two dextran sodium sulfate (DSS)-induced rat colitis models. Furthermore, we evaluated the capability of RNACap to deliver multiple mRNAs to the intestines in a large swine model, whose GI tract size more closely resembles that of humans, thereby offering greater translational potential. The RNACap device could address the critical challenges for oral and specific delivery of liquid mRNA therapeutics to the intestines, holding the promise of advancing the development of non-invasive and self-administered oral mRNA therapeutics.

RESULTS

Design and fabrication of oral RNACap

Our device, RNACap, is designed to deliver liquid mRNA formulations instead of dry mRNA formulations to the intestines, avoiding the additional expensive and time-consuming lyophilization process (31), which often results in a loss of efficacy (32). The RNACap is a self-administered, multi-compartment capsule-based device that protects liquid mRNA formulations from the acidic stomach environment and allows both diffusion- and contraction-driven releases in the intestines in response to neutral pH, the resulting pressure release and intestinal contractions (peristalsis), enabling NP-mediated mRNA entry into the intestinal cells for efficient transfection (Fig. 1, A and B). To fabricate the RNACap, we employed an FDA-approved gelatin capsule (size 5 for in vitro), which consists of a body and a cap and can rapidly dissolve in an aqueous solution (Fig. 1B and fig. S1, A and B). Here, rhodamine B (RB) solution was used as a surrogate due to its bright color and fluorescence. We first coated the inner surface of the capsule body with polymeric membranes to prevent direct contact between the aqueous contents and the capsules, thus avoiding the dissolution of the capsules. To optimize the inner coating membrane, we tested two different surface coatings: ductile membrane A and inductile membrane B. The inner surface of the capsule body was coated with either membrane A or membrane B before loading RB solutions (Fig. 1C). Subsequently, a sealing membrane (membrane B) was applied to seal the capsule body, preventing cargo leakage. At this stage, membrane B was used as the sealing membrane to complete the construction of RNACap; optimization of the sealing membrane will be carried out in the next step. The final sealing effect was achieved by pressing the capsule cap over the sealing membrane, exerting force to seal it (Fig. 1B). Both the resulting capsule-(A)B and capsule-(B)B successfully encapsulated the RB solution without damaging the capsules (Fig. 1D). However, capsule-(A)B accommodated a higher volume of contents than capsule-(B)B, attributable to the firm attachment of ductile membrane A to the inner surface of the capsule body (Fig. 1E). Consequently, ductile membrane A, allowing maximal loading volume, was chosen as the inner coating membrane for further studies.

Fig. 1. Design and fabrication of RNACap.

(A) Schematic illustration of RNACap-mediated oral delivery of liquid mRNA therapeutics to intestines. (B) Construction of RNACap. (C) Construction of capsule-(A)B and capsule-(B)B. (D) Images of capsule-(A)B and capsule-(B)B loaded with rhodamine B (RB) solution. (E) Maximal cargo loading volume of capsule-(A)B and capsule-(B)B. (F) Release of RB from capsule-(A)B (cap removed) and capsule-(A)A (cap removed) after 20 min of incubation in PBS at room temperature. (G) Release profile of RB-loaded capsule-(A)B (cap removed) and capsule-(A)A (cap removed). Trend line is nonlinear curve fitting to calculate the time for 50% release. Dotted line indicates 50% release. (H) Images of the empty capsule, RB-loaded RNACap (RB-RNACap) and pH-sensitive polymer L100-55-coated RNACap (L100-55-RB-RNACap). (I) Visualization of the release of RB from uncoated and L100-55-coated capsule at pH 1.2 and 7.4 solutions. (J to L) Release profile in pH 1.2 and 7.4 solutions for L100-55-RB-RNACap (J), RB-RNACap (K), and L100-RB-RNACap (L). Trend line is nonlinear curve fitting to calculate the time for 50% release. Dotted line indicates 50% release. (M) Images of RB-RNACap at different time points at rt. Data are presented as mean ± S.D. In (E) dots represent individual sample replicates. Statistical significance was evaluated by unpaired two-tailed Student’s t-test in (E). ****P < 0.0001. Data in (D to M) are representative of n = 3 independent experiments. All the schematic illustrations were created using Adobe Illustrator.

We then proceeded to determine the optimal material for a sealing membrane. We employed both membrane A and B separately to seal the capsule body (fig. S2). According to our design, the sealing mechanism relies on the force applied by the capsule cap, which dissolves rapidly in the intestines, thereby releasing pressure exerted on the sealing membrane, leading to rapid sealing membrane departure to release contents. To evaluate this capability, the caps of both capsule-(A)B and capsule-(A)A were removed before they were exposed to an aqueous solution (pH 7.4). The sealing membrane of the capsule-(A)B detached immediately and released 50% of the RB into the solution after only 3.5 min of incubation, as demonstrated by the red fluorescence of RB (Fig. 1, F and G and fig. S3). In contrast, the sealing membrane of the capsule-(A)A adhered tightly to the capsule’s body, and no obvious release of RB was observed even after 20 min of incubation. Thus, inductile membrane B, allowing rapid cargo release, was chosen as the sealing membrane for further studies.

Having established that RNACap can successfully accommodate an RB solution and rapidly release it in an aqueous buffer, we further coated the RNACap with pH-sensitive copolymers (enteric coating) to protect RNACap from the acidic environment of the stomach. The enteric coating was achieved by dipping the RNACap into the polymer solutions and allowing it to air-dry. Coating RNACaps with the copolymer Eudragit L100-55, which dissolves at a pH > 5.5, did not result in any leakage of RB cargo (Fig. 1H). More importantly, the L100-55-coated RNACaps remained intact in a solution with an acidic gastric pH (pH 1.2) for at least 2 h (Fig. 1, I and J). In contrast, in a solution with a neutral intestinal pH (pH 7.4), 50% of the RB cargo was released after 49 min of incubation. As expected, without enteric coating, RNACap rapidly dissolved and released its cargo with both pH conditions (Fig. 1K and fig. S4). Coating RNACaps with another copolymer, Eudragit L100, which dissolves in the pH range of 6-7, also protected RNACaps at acidic pH. However, the release of cargo in a neutral pH environment was slightly slower compared with the L100-55-coated RNACaps (50% release in 66 min versus 49 min) (Fig. 1L and fig. S5, A and B). The RB-loaded RNACaps (capsule-(A)B) remained intact at room temperature (RT)for at least 60 h without any leakage of cargo (Fig. 1M and fig. S6).

Optimized mRNA NPs efficiently deliver different mRNAs to multiple cell lines for transfection

After establishing that RNACap can protect liquid cargo from the stomach’s acidic pH and efficiently release them in response to the neutral intestinal pH, our next goal was to develop an mRNA NP formulation capable of bypassing mucosal barriers and transporting mRNAs into intestinal cells. The PEGylation of NPs has been demonstrated to enhance their mucus-penetration efficiency (33). Thus, a PEGylated NP system (34, 35) comprising a cationic lipid–like material G0-C14, poly(DL-lactide-co-glycolide) (PLGA), and PEG-lipid (DSPE-PEG) was chosen (Fig. 2A). This system was originally designed for the intravenous administration of anticancer therapeutics. Given that the administration route can affect the transfection efficacy of mRNA NPs (36), we further optimized the NPs for oral administration. We began by formulating EGFP-mRNAs with the NPs and assessing their transfection efficiency in a human intestinal epithelial cell line (Caco-2). We first optimized the PLGA:mRNA ratio. Compared with the original PLGA:mRNA ratio of 120:1 (37), an optimal ratio of 140:1 yielded higher transfection efficiency, evident from increased fluorescence intensity (Fig. 2B). We had previously identified DSPE-PEG as a suitable PEGylation lipid with a prolonged dissociation half-life of 24.76 h. In contrast, DMPE-PEG and Ceramide (Ce)-PEG served as effective de-PEGylation lipids, lipids that readily dissociate from the surface of the particles, with much shorter dissociation half-lives of 1.02 h and 0.66 h, respectively (fig. S7) (38). We hypothesized that incorporating de-PEGylation lipids into DSPE-PEG-based NPs could expedite mRNA release from the NPs, leading to enhanced transfection efficiency. Indeed, introducing 5.0% of DMPE-PEG (in all lipid-PEGs) to the NPs improved transfection efficiency over the original formulation (Fig. 2C). Introducing varying ratios of Ce-PEG into the NPs did not lead to an increase in transfection efficiency (Fig. 2D). The size and PDI of lead NP formulations from DSPE-PEG, DMPE-PEG, and Ce-PEG groups ranged from 175 nm to 189 nm and 0.335 to 0.375, respectively (fig. S8, A and B). Consequently, an optimized NP formulation with 5.0% DMPE-PEG was selected for further studies (Fig. 2E).

Fig. 2. Optimized mRNA NPs efficiently deliver different mRNAs to multiple cell lines for transfection.

(A) Schematic illustration of mRNA NP formation. (B to D) In vitro transfection of mRNA NPs containing different amounts of PLGA (B), DMPEG-PEG (C), or Ce-PEG (D) in intestinal cells (Caco-2). (E) Formulation of top-performing mRNA NPs. (F) Transmission electron microscope (TEM) images of mRNA NPs. Scale bars, 200 nm and 50 nm. (G to I) Diameter (G), polydispersity index (PDI) (H), and zeta potential (I) of mRNA NPs analyzed by dynamic light scattering (DLS) and Malvern Zetasizer Pro. n = 6 independent experiments. (J) Uptake of Cy5-labeled mRNA NPs by intestinal Caco-2 cells in the presence of different endocytosis inhibitors. EIPA, ethylisopropylamiloride; CPZ, chloropromazine. (K) Confocal microscopy images of Caco-2 cells treated for 8 hours with Cy5-mRNA NPs (Cy5-mRNA: 750 ng/mL), with Cy5-labeled mRNA (red); cytoskeleton (Phalloidin, green); nuclei (Hoechst,blue). Scale bar, 30 μm. (L) Bioluminescence imaging of cultured intestinal cells (Caco-2) treated with FLuc-mRNA NPs containing varying amounts of FLuc-mRNA. Color scale, luminescence. (M) Flow cytometry analysis of MC-38 cells treated with EGFP-mRNA NPs containing different amounts of EGFP-mRNA. MFI, mean fluorescence intensity. The presented values were normalized to the signal intensity of cells treated with PBS (Normalized MFI). (N) Fluorescence microscopy images of Caco-2, Raw 264.7 and MC-38 cells treated with EGFP-mRNA NPs (750 ng/mL). Scale bar, 200 μm. Data are presented as mean ± S.D. Dots represent individual sample replicates. Statistical significance was evaluated by one-way ANOVA with Tukey’s post hoc analysis in (B-D), (J) and (M). *P < 0.05, **P < 0.01, ****P < 0.0001. Data in (B, C, D, F, J, K, L, M, N) are representative of n = 3 independent experiments. All the schematic illustrations were created using Adobe Illustrator.

Transmission electron microscopy (TEM) TEM images revealed that the lead mRNA NPs exhibited a solid spherical morphology with a size of about 100 nm (Fig. 2F). Dynamic light scattering (DLS) measurements indicated a diameter of 176 nm, a polydispersity index (PDI) of 0.337, and a zeta potential of −7.8 mV (Fig. 2, G to I). The size measured by DLS is consistent with the size observed in TEM, with the slightly larger value from DLS attributed to the measurement of hydrodynamic diameter, whereas TEM captures the size of the mRNA NPs in their dry state.

To investigate the endocytosis mechanism of the mRNA NPs, Caco-2 cells were pretreated for 30 min with endocytosis inhibitors that block the clathrin-mediated pathway, caveolae-mediated pathway, or macropinocytosis, respectively, followed by incubation with Cy5-labelled mRNA (encoding firefly luciferase, FLuc)-NPs for 1 h. The results revealed that the endocytosis of Cy5-mRNA NPs by Caco-2 cells was predominantly mediated through macropinocytosis (Fig. 2J). We further assessed the cellular uptake of these NPs in Caco-2 cells using a fluorescent mRNA (Cy5-mRNA). The efficient uptake of Cy5-mRNA-NPs was confirmed by strong red fluorescent signals (Cy5.5 channel) (Fig. 2K and fig. S9). The rapid and efficient uptake of Cy5-mRNA NPs was also validated through flow cytometry (fig. S10, A and B). Next, we examined the endosomal escape of Cy5-mRNA-NPs in cells. The endosomal escape mechanism of our nanoparticles containing lipid G0-C14 has been previously demonstrated to involve a proton-sponge effect, which may at least partially contribute to the cytosolic release of mRNA cargo (34). Indeed, after a 6-hour incubation, most of the red fluorescence representing Cy5-mRNA did not colocalize with the green fluorescence representing the endosomes or lysosomes (fig. S11), indicating successful endosomal escape of the Cy5-mRNA-NPs. We then examined the capability of the NPs to deliver reporter mRNAs encoding FLuc or EGFP, FLuc-mRNA and EGFP-mRNA, to cells for transfection. The successful delivery of FLuc-mRNAs by NPs into Caco-2 cells was confirmed by robust dose-dependent bioluminescence signals (Fig. 2L and fig. S12). EGFP-mRNA NPs transfected intestinal cells at the low dose of 23.4 ng/mL (P = 0.0028 and P < 0.0001, respectively, compared to blank-NPs, Fig. 2M and fig. S13). Furthermore, the NPs demonstrated the capacity to deliver EGFP-mRNAs to various intestinal cell lines and macrophages, including Caco-2 cells, murine macrophage cells (RAW264.7), and murine colon adenocarcinoma cells (MC-38) (Fig. 2N), as evidenced by high green fluorescence intensity.

IL-10-mRNA NPs enable efficient in vitro transfection and exhibit anti-inflammatory effects

IL-10, a regulatory cytokine known for its potent anti-inflammatory properties, has been employed in the treatment of various inflammatory conditions, including IBD (39, 40). Thus, we selected IL-10-mRNA encoding murine IL-10 as the therapeutic mRNA aiming to address IBD. To this end, we first evaluated the anti-inflammatory effect of IL-10-mRNA NPs in vitro. Intestinal cells Caco-2 and MC-38, as well as macrophages, were exposed to IL-10-mRNA NPs for 24 h, resulting in dose-dependent expression of IL-10. This demonstrated the NPs’ ability to deliver therapeutic mRNAs to both intestinal cells and macrophages (Fig. 3A). Western blot analysis further confirmed that only IL-10-mRNA NP-treated cells elicited IL-10 expression, whereas free IL-10-mRNA or blank NPs did not (Fig. 3B). IL-10 expression in IL-10-mRNA NP-treated cells was additionally confirmed through immunofluorescence staining, showing a strong red fluorescence indicative of IL-10 (Fig. 3C). We then investigated whether pre-treatment with IL-10-mRNA NPs could alter the polarization of mouse RAW 264.7 macrophages to a proinflammatory M1-like macrophage in the presence of lipopolysaccharide (LPS) (Fig. 3D). Indeed, treatment of the cultured macrophages with LPS, an inflammatory stimulator, led to lower releases of the inflammatory cytokine tumor necrosis factor-alpha (TNF-α) and IL-6 in IL-10-mRNA NP-treated RAW264.7 cells than untreated RAW264.7 cells, due to expression of IL-10 (Fig. 3E). This indicated that IL-10 may partially inhibit macrophage polarization. We also explored the possibility of repolarizing a proinflammatory M1-like macrophage into an anti-inflammatory M2-like macrophage through IL-10 expressed by intestinal cells rather than the macrophages themselves. For this purpose, RAW264.7 cells were initially exposed to LPS for 24 h, followed by treatment with supernatants from IL-10-mRNA NP-treated or -untreated intestinal Caco-2 cells (Fig. 3F). The enzyme-linked immunosorbent assay (ELISA) results revealed that murine RAW264.7 cells treated with supernatants from mouse IL-10-mRNA NP-treated Caco-2 cells exhibited higher IL-10 concentration and lower TNF-α concentration than those treated with supernatants from untreated Caco-2 cells, indicating an IL-10-mediated repolarization effect on the macrophages (Fig. 3G).

Fig. 3. IL-10-mRNA NPs exhibit anti-inflammatory effects in vitro, and ex vivo imaging reveals in vivo distribution of RNACap.

(A) Enzyme-linked immunosorbent assay (ELISA) analysis of IL-10 concentration in supernatant from Caco-2, MC-38 and Raw 264.7 cells treated with various amounts of PBS, IL-10-mRNA or IL-10-mRNA NPs for 24 h. (B) Western blot analysis of IL-10 expression in MC-38 cells treated with PBS, IL-10-mRNA, blank NPs or IL-10-mRNA NPs for 12 h. IL-10-mRNA 750 ng/mL. (C) Confocal microscopy images of immunofluorescence staining of IL-10 expression in MC-38 cells treated with free IL-10-mRNA or IL-10-mRNA NPs. Hoechst (blue) was used to stain the cell nuclei. An Alexa Fluor 647-labeled antibody was used to stain IL-10. (D) Schematic illustration of how IL-10-mRNA NPs directly transfect macrophages, inhibiting the polarization of macrophages to proinflammatory M1 phenotype in the presence of lipopolysaccharide (LPS), an inflammatory stimulator. (E) ELISA analysis of anti-inflammatory cytokine IL-10 and proinflammatory cytokine TNF-α and IL-6 in supernatant from RAW 264.7 cells treated with PBS or IL-10-mRNA NPs (w/wo LPS stimulation) following procedures illustrated in (D). (F) Schematic illustration of how IL-10-mRNA NPs first transfect intestinal epithelial cells, then indirectly induce the repolarization of macrophages from proinflammatory M1 phenotype to anti-proinflammatory M2 phenotype. (G) ELISA analysis of IL-10 and TNF-α and IL-6 in supernatant from RAW 264.7 cells treated with supernatant from Caco-2 cells incubated with PBS or IL-10-mRNA NPs following procedures illustrated in (F). (H) Ex vivo images of excised rat gastrointestinal (GI) tract at various time points from 15 min to 6 h post-administration. Rats were orally administered six Cy5-mRNA-RNACaps (L100-55 coated). The left panel shows enlarged images of RNACaps at 1 h and 2 h (images 1–4). Cy5-mRNA: 50 μg per rat. Color scale, 0-255 gray value. (I) Confocal microscopy images of intestine sections from rats treated with Cy5-mRNA-RNACaps (purple) for 4 h. Hoechst (blue) was used to stain the cell nuclei. Scale bar, 100 μm. (J) Ex vivo images of excised intestines at 6 h post-administration. Rats were orally administered six Cy5-mRNA-RNACaps (L100 coated). Color scale, 0-255 gray value. Data are presented as mean ± S.D. Dots represent individual sample replicates. Statistical significance was evaluated by one-way ANOVA with Tukey’s post hoc analysis in (E) and (G). **P < 0.01, ***P < 0.001, ****P < 0.0001. Data in (A, B, C, E, G, H, I, J) are representative of n = 3 independent experiments. All the schematic illustrations were created using Adobe Illustrator.

mRNA NPs retain most of their activity in intestinal fluids but not in gastric fluids

To assess the stability of mRNA NPs under the environment of stomach, we incubated the EGFP-mRNA NPs in acidic gastric fluid (laboratory-prepared simulated gastric fluid) (41) containing gastric enzymes for 1 h at 37 °C before using them for transfection in MC-38 cells (fig. S14A). Cells treated with gastric fluid-pretreated EGFP-mRNA NPs-treated exhibited significantly lower green fluorescence (10%) compared with those treated with untreated EGFP-mRNA NPs (P < 0.0001) (fig. S14, B and C). This result indicates that even a brief 1 h exposure to gastric fluid can substantially reduce the activity of mRNA NPs.

Next, we investigated the stability of mRNA NPs in laboratory-prepared simulated fasted and fed intestinal fluids(41). To this end, the EGFP-mRNA NPs were incubated in fasted or fed intestinal fluids for 3 h at 37 °C before transfection (fig. S15A). Cells treated with fasted and fed intestinal fluid-pretreated EGFP-mRNA NPs exhibited reduced green fluorescence (78% and 70%, respectively) compared with those treated with untreated EGFP-mRNA NPs (fig. S15, B and C). Although 22% and 30% of the mRNA dose were lost after 3 h of incubation in the intestinal fluid, at least 70% of the mRNA activity remained due to the protective effect of the NPs. The size of the NPs after incubation with different fluids was measured by DLS. The sizes of these NPs showed no significant differences, indicating that the changes in translation were not attributed to size variations (fig. S16, A and B).

Ex vivo imaging reveals in vivo distribution of RNACap

To evaluate whether RNACap could effectively deliver these liquid mRNA NPs to the intestines in vivo, six RNACaps loaded with fluorescent Cy5-mRNA NPs were orally administered to each rat (Fig. 3H). Ex vivo images of the excised GI tract 15 min post-administration indicated that all the administered RNACaps remained in the stomach. Furthermore, all six RNACaps were found to be intact after the dissection of the stomach, showing that RNACaps could withstand the hostile stomach environment. Although all the RNACaps were still in the stomach at 1 h post-administration, a portion of them entered the intestine by 2 h post-administration. Visual inspection of the GI tract at the 2-h stage revealed different RNACap statuses following oral administration, including intact status in the stomach, partially dissolved status in the intestine, a squeezed status with evident shape change in the intestine, and a completely dissolved status with the substantial release of Cy5-mRNA NPs into the intestines (Fig. 3H). At 4 h and 6 h post-administration, no intact RNACaps were observed in either the stomach or intestines, indicating that all RNACaps entered the intestines and subsequently dissolved, releasing Cy5-mRNA NPs (Fig. 3H). Additionally, no fluorescence was detected in other major organs of the rats (fig. S17). To investigate in vivo absorption of Cy5-mRNA NPs, the intestines of untreated or Cy5-mRNA-RNACap-treated rats were sectioned and imaged. The strong fluorescent signal of Cy5 in sections of Cy5-mRNA-RNACap-treated rats indicated that NPs were able to deliver Cy5-mRNA to the intestines (Fig. 3I). We also explored the in vivo distribution of L100-coated RNACaps; however, given the slower degradation of this formulation, there were undissolved RNACaps remaining in the intestines even 6 h after administration, making them unsuitable for rapid mRNA delivery to intestines (Fig. 3J). Thus, L100-55-coated RNACaps were selected for further studies.

RNACap mediates intestinal mRNA expression in rats

Next, we sought to confirm that the mRNAs delivered to the intestines by RNACap can retain their biological activity. For this purpose, EGFP-mRNA, a reporter mRNA, was utilized. Oral administration of EGFP-mRNA-RNACaps resulted in substantial EGFP expression in the small intestines and colons of rats, as demonstrated by strong green fluorescence (figs. S18, A and B and S19, A and B). Hematoxylin and eosin (H&E) staining of intestines showed that no obvious toxicity or damage was induced by oral administration of EGFP-mRNA-RNACaps (fig. S20). We next characterized the type of cells transfected in the intestine and found that epithelial cells were primarily transfected. Additionally, cells in gut crypts, including gut stem cells (gut stem cell marker SOX9 was used), were also transfected (fig. S21).

We then investigated the oral delivery of therapeutic IL-10-mRNAs by RNACap in rats. IL-10-mRNA-RNACap-treated rats displayed higher blood IL-10 concentration than untreated rats or rats treated with empty-RNACap or EGFP-RNACap at 24 h and 48 h post-oral administration (Fig. 4, A and B). The IL-10 concentration of the EGFP-mRNA-treated groups were not significantly different from those in the untreated or empty-RNACap-treated groups, indicating that the IL-10 elevations in blood do not reflect a response to non-specific inflammation induced by the NP mRNA delivery (for example, P = 0.9980 and P = 9998, respectively, at 24 h).

Fig. 4. Oral delivery of IL-10-mRNA-RNACaps to the rat intestines enables in vivo mRNA transfection.

(A) Experimental setup for oral administration of IL-10 (or EGFP)-mRNA-RNACaps to rats for blood sample collection. (B) ELISA analysis of IL-10 l concentration in serum from untreated rats, as well as rats treated with empty-RNACaps or IL-10 (or EGFP)-mRNA-RNACaps for 24 h, 48 h and 72 h. IL-10 (or EGFP)-mRNA: 50 μg in 6 RNACaps per rat. (C) Experimental setup for intravenous (I.V.) injection of IL-10 -mRNA NPs to rats for blood sample collection. (D) ELISA analysis of IL-10 concentration in serum after I.V. administration. IL-10-mRNA: 50 μg in 0.5 mL PBS per rat. (E) Structure of engineered non-snap-cap rigid capsules (Rigid-Cap) and the experimental setup for oral administration of IL-10-mRNA loaded in rigid capsules (IL-10-mRNA-Rigid-Cap) to rats for blood sample collection. (F) ELISA analysis of IL-10 concentration in serum after treatment with IL-10-mRNA-Rigid-Caps. IL-10-mRNA: 50 μg in 6 Rigid-Caps per rat. (G) Experimental setup for oral administration of IL-10-mRNA-RNACaps (mRNA in COVID LNPs) to rats for blood sample collection. Here, mRNA was first formulated with widely used COVID LNPs instead of the lead NPs before being loaded into the RNACaps. (H) ELISA analysis of IL-10 concentration in serum after treatment with IL-10-mRNA-RNACaps (COVID LNPs). IL-10-mRNA: 50 μg in 6 RNACaps per rat. (I) Real-time reverse transcription polymerase chain reaction (RT-PCR) analysis of IL-10 mRNA concentration in different parts of the intestines of untreated rats or rats treated with IL-10-mRNA-RNACaps after 24 h. (J) ELISA analysis of IL-10 concentration in different parts of intestinal tissues from untreated rats or rats treated with IL-10-mRNA-RNACaps after 24 h. (K) Immunofluorescence images of intestinal tissue sections from rats treated with IL-10-mRNA-RNACaps for 24 h. IL-10 was stained with an anti-IL-10 antibody (green); Villi were stained with an anti-villi antibody (orange); nuclei, 4′, 6-diamidino-2-phenylindole (DAPI) (blue). Scale bars, 1000 μm, 2000 μm, 100 μm and 200 μm. (L and M) Quantification of fluorescence intensity ratio of IL-10 to DAPI in the small intestine (L) and colon (M). (N) Western blot analysis of IL-10 expression in different parts of intestinal tissues from untreated rats or rats treated with IL-10-mRNA-RNACaps for 24 h. (O) Hematoxylin and eosin (H&E) staining images of intestinal tissue sections from rats treated with IL-10-mRNA-RNACaps for 24 h. Scale bars, 1000 μm and 100 μm. Data are presented as mean ± S.D. Dots represent individual sample replicates. Statistical significance was evaluated by one-way ANOVA with Tukey’s post hoc analysis in (B), (D), (F) and (H) and by unpaired two-tailed Student’s t-test in (I), (J), (L) and (M). *P < 0.05, **P < 0.01, ****P < 0.0001. Data in (B, D, F, H, I, J, K, L, M, N, O) are representative of n = 3 animals per group. All the schematic illustrations were created using Adobe Illustrator.

We next compared our oral mRNA delivery method with the systemic mRNA delivery method. To this end, the same amount of IL-10-mRNA NPs used for oral delivery was injected into rats through the tail vein (Fig. 4C). Rats receiving intravenous (I.V.) injections of IL-10-mRNA NPs displayed lower blood IL-10 concentration than those treated with IL-10-mRNA-RNACap at 24 h and 48 h post-oral administration (Fig. 4D). This result is not surprising, as the NPs used here are optimized for rapid oral (local) delivery not systemic delivery.

Our design leverages intestinal peristalsis to trigger the rapid release of cargos from RNACap. To test whether peristalsis is indeed critical for the delivery mechanism, we fabricated a non-snap-cap rigid capsule (Rigid-Cap) that cannot be deformed by peristalsis for oral mRNA delivery (Fig. 4E and fig. S22, A and B). RNACaps soften in the neutral pH of the intestine, allowing intestinal contractions to squeeze and release their cargo. In contrast, non-snap-cap Rigid-Cap remain rigid even after the removal of enteric coating due to the presence of an additional ductile coating membrane, making it resistant to contraction force; thus, no contraction-driven release occurs (fig. S22B). We next compared the in vitro release of RB from RNACap and Rigid-Cap at a neutral intestinal pH of 7.4 (fig. S23A). The squeezing force, mimicking the intestinal contraction, applied to the capsules at time points of 5 min and 15 min led to deformation of RNACap immediately, but not to Rigid-Cap (fig. S23, B and C). No obvious RB release from Rigid-Cap was observed over a period of 2 h, whereas, each squeezing event triggered rapid RB release from RNACap, reaching 50% RB release at 13 min. Rats treated with IL-10-mRNA-Rigid-Caps displayed similar blood IL-10 concentration to the untreated rats at all time points and significantly lower blood IL-10 concentration than those treated with IL-10-mRNA-RNACaps at 24 h and 48 h post-oral administration (P = 0.0431 and P = 0.0058 at 24 h and 48 h, respectively) (Fig. 4, E and F). These findings indicate that the intestinal peristalsis-induced cargo snap release mechanism of RNACap is critical for efficient oral mRNA delivery.

We compared our lead NPs with the gold-standard lipid nanoparticles (LNP) used in the mRNA-1273 COVID-19 vaccine (fig. S24). The COVID-19 LNP formulation includes SM-102 as ionizable lipid, DSPC as helper lipid, cholesterol, and DMG-PEG2000 in a molar ratio of 50:10:38.5:1.5, respectively (42, 43). Upon oral administration of RNACaps containing IL-10-mRNA in COVID LNPs, rats displayed lower blood IL-10 concentration compared with those treated with IL-10-mRNA-RNACap at 48 h post-oral administration (P = 0.0114) (Fig. 4, G and H).

Subsequently, the IL-10-mRNA concentration in the intestines of rats was determined by real-time reverse transcription polymerase chain reaction (RT-PCR) at 24 h post-oral administration. IL-10-mRNA-RNACap-treated rats showed higher IL-10 mRNA concentration in the duodenum, jejunum, and colon compared with untreated rats (Fig. 4I). Elevated IL-10 expression was also observed in different parts of the intestines of the IL-10-mRNA-RNACap-treated rats, as indicated by ELISA analysis (Fig. 4J). IL-10 expression in small intestines and colons of IL-10-mRNA-RNACap-treated rats was visualized by immunofluorescence staining. Strong green fluorescence representing IL-10 expression was observed in IL-10-mRNA-RNACap-treated rats, whereas minimal green fluorescence was observed in untreated rats (Fig. 4, K to M). Different parts of the intestines of untreated or IL-10-mRNA-RNACap-treated rats were further analyzed by western blot, indicating that oral administration of IL-10-mRNA-RNACaps induced significant IL-10 expression in the jejunum and colon (for example, P = 0.0128) (Fig. 4N and fig. S25). We observed no evident toxicity or damage to intestines was observed with the oral administration of IL-10-mRNA-RNACaps, as demonstrated by the H&E staining (Fig. 4O). Unlike other non-degradable capsule devices that must be expelled from the body after releasing their cargos, the RNACaps developed here can be squeezed and dissolves quickly in the intestine after releasing the mRNA cargos, leaving only the soft coating and sealing membranes. The total transit time for the tiny soft membranes from the RNACaps is estimated to be 18-24 h, based on evaluation of the rat GI tract at different time points after administration.

IL-10-mRNA-RNACaps mitigate colitis in rats

Having demonstrated the in vivo intestinal expression of IL-10 facilitated by RNACaps, we proceeded to assess the therapeutic effect of orally administered IL-10-mRNA-RNACaps in a rat colitis model induced by DSS. DSS-induced colitis results in noticeable symptoms such as body weight loss, shortened colon length, and increased disease activity index (DAI) (44), which is determined based on parameters like weight loss, diarrhea, and rectal bleeding (45, 46). Due to the potent anti-inflammatory properties of IL-10, we expected that intestinal IL-10 expression would alleviate inflammation. To model acute colitis, rats were administered DSS by allowing free access to drinking water supplemented with 6.0% (w/w) DSS for the duration of the 8 days. After the initiation of colitis with 6% DSS, IL-10-mRNA-RNACaps were orally administered to rats on day 2, 5 and 8 (Fig. 5A). When compared with the DSS group, the DSS plus IL-10-mRNA-RNACap group exhibited less body weight loss, lower DAI increase, and longer colon length (Fig. 5, B to D and fig. S26). We then examined the inflammation-related cytokine and protein concentration in colon tissue and blood, including IL-10, TNF-α, IL-1β, IL-6, IL-17A and monocyte chemoattractant protein-1 (MCP-1). In colon tissue, ELISA analysis revealed that the DSS group had lower protein concentration of anti-inflammatory IL-10 and higher proinflammatory IL-1β, IL-6, IL-17A and MCP-1 than the untreated water control group, indicating the severe inflammation induced by DSS (Fig. 5, E to J). Compared with the DSS group, the DSS plus IL-10-mRNA-RNACap group showed less DSS-induced inflammation, with higher IL-10 and lower IL-1β, IL-6, IL-17A and MCP-1 proteins. In blood, the IL-10-mRNA-RNACap treated DSS group had a higher IL-10 and lower TNF-α, IL-1β, IL-17A, IL-6 and MCP-1 circulating protein than the DSS group (Fig. 5, K to P). Finally, the severity of inflammation in all groups was assessed by H&E staining. The DSS plus IL-10-mRNA-RNACap group exhibited less inflammation, including reduced epithelial injury, inflammatory cell infiltration, and submucosal edema, compared with the DSS group (Fig. 5Q).

Fig. 5. Oral administration of IL-10-mRNA-RNACaps ameliorates colitis in rats.

(A) Experimental timeline for oral administration of IL-10-mRNA-RNACaps to acute colitis rat models. IL-10-mRNA: 25 μg in 3 RNACaps per rat. Acute colitis was induced in rats by providing free access to drinking water supplemented with 6.0% (w/w) dextran sulfate sodium (DSS) for 8 days. Rats were treated with RNACaps on day 2, 5 and 8. (B and C) Relative body weight (B) and disease activity index (DAI) (C) of healthy (plain water-treated), DSS-treated or DSS plus IL-10-mRNA-RNACap-treated rats were monitored daily. (D) Quantification of colon length on day 8. (E to J) Colon tissue protein expression of IL-10 (E), tumor necrosis factor-alpha (TNF-α) (F), interleukin-1β (IL-1β) (G), IL-6 (H), IL-17A (I) and monocyte chemoattractant protein-1 (MCP-1) (J) by ELISA. (K to P) Quantification of protein expression in blood of IL-10 (K), TNF-α (L), IL-1β (M), IL-6 (N), IL-17A (O) and MCP-1 (P) by ELISA. (Q) H&E staining images of the colon tissue sections. Dashed box indicates inset. Scale bars, 500 μm and 400 μm. Data are presented as mean ± S.D. Dots represent individual sample replicates. Statistical significance was evaluated by one-way ANOVA with Tukey’s post hoc analysis in (B to P). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. n = 5 animals per group for all panels. All the schematic illustrations were created using Adobe Illustrator.

IL-10-mRNA-RNACaps mitigate colitis in rats in a delayed therapeutic setting

To further assess the therapeutic potential of IL-10-mRNA-RNACaps in a more clinically relevant setting, we explored whether oral administration of IL-10-mRNA-RNACaps could alleviate colitis in rats in a delayed therapeutic scenario. Rats were given free access to drinking water supplemented with 8.0% (w/w) DSS for 10 days to establish a colitis model. The development of colitis is highly dependent on the concentration of DSS and experimental setup, therefore, to establish a colitis model with meaningful disease parameters, we optimized the DSS concentration for the delayed therapeutic setting. We observed that colitis development was slower in the experimental setup of the delayed therapeutic setting compared with the acute therapeutic setting. Therefore, we increased the DSS concentration from 6.0 % to 8.0% to establish the colitis model within a similar time frame. After DSS administration for 10 days, plain water was provided, and rats were orally administered RNACaps on days 11, 14 and 17 or sulfasalazine (SSZ, standard therapy, 100 mg/kg/day by oral administration) daily (Fig. 6A). Compared with the DSS group, both the DSS plus IL-10-mRNA-RNACap and DSS plus SSZ groups exhibited reduced body weight loss, lower DAI increase, and longer colon length (Fig. 6, B to D and fig. S27). We then examined the inflammation-related cytokine and protein concentration in colon tissue and blood, including IL-10, TNF-α, IL-1β, IL-6, IL-17A and MCP-1. ELISA analysis of colon tissue showed that the DSS group had lower anti-inflammatory cytokine IL-10 and higher proinflammatory markers such as TNF-α, IL-1β, IL-6, IL-17A and MCP-1 compared with the water group, indicating severe inflammation induced by DSS (Fig. 6, E to J). In contrast, both the DSS plus IL-10-mRNA-RNACap and DSS plus SSZ groups showed less DSS-induced inflammation, with higher IL-10 and lower TNF-α, IL-1β, IL-6, IL-17A and MCP-1. Similarly, in blood samples, both the DSS plus IL-10-mRNA-RNACap and DSS plus SSZ groups had higher IL-10 and lower TNF-α, IL-1β, IL-6 and MCP-1 compared with the DSS group (Fig. 6, K to P). H&E staining further revealed that both the DSS plus IL-10-mRNA-RNACap and DSS plus SSZ groups exhibited reduced inflammation, including less crypt loss and architectural distortion, epithelial damage, and inflammatory cell infiltration, compared with the DSS group (Fig. 6Q). These findings underscore the promising therapeutic potential of RNACap-mediated oral mRNA delivery for the treatment of IBD and potentially other diseases.

Fig. 6. Oral administration of IL-10-mRNA-RNACaps ameliorates colitis in rats in a delayed therapeutic setting.

(A) Experimental timeline for the oral administration of IL-10-mRNA-RNACaps in acute colitis rat models. Rats were given free access to drinking water supplemented with 8.0% (w/w) DSS for 10 days to induce colitis. Afterward, plain water was provided, and rats were treated with RNACaps (IL-10-mRNA: 25 μg in 3 RNACaps per rat.) on days 11, 14 and 17 or sulfasalazine (SSZ, standard therapy, 100 mg/kg/day) daily. (B and C) Relative body weight (B) and DAI (C) of healthy (plain water-treated), DSS-treated, DSS plus IL-10-mRNA-RNACap-treated or DSS plus SSZ-treated rats were monitored daily. (D) Quantification of colon length on day 17. (E to J) Quantification tissue protein expression of IL-10 (E), TNF-α (F), IL-1β (G), IL-6 (H), IL-17A (I) and MCP-1 (J) by ELISA. (K to P) Quantification of protein expression in blood of IL-10 (K), TNF-α (L), IL-1β (M), IL-6 (N), IL-17A (O) and MCP-1 (P) by ELISA. (Q) H&E staining images of the corresponding colon tissue sections. Dashed box indicates inset. Scale bars, 100 μm and 400 μm. Data are presented as mean ± S.D. Dots represent individual sample replicates. Statistical significance was evaluated by one-way ANOVA with Tukey’s post hoc analysis in (B to P). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. n = 5 animals per group for all panels. All the schematic illustrations were created using Adobe Illustrator.

RNACaps demonstrate in vivo safety in rats

To evaluate the in vivo toxicity of RNACaps, healthy rats were orally administered IL-10-mRNA-RNACaps on days 0, 3 and 6, with untreated rats serving as the control group (Fig. 7A). The rats were euthanized after the final administration to harvest intestines, other major organs, and blood samples for analysis. The IL-10-mRNA-RNACap group exhibited no significant difference in plasma alanine transaminase (ALT), aspartate aminotransferase (AST) or blood urea nitrogen (BUN) concentrations compared with the untreated group, suggesting that the oral administration of RNACap did not cause liver or renal toxicity (P > 0.05) (Fig. 7B). Furthermore, the administration of RNACap did not lead to significant alterations in a series of hematological parameters, when compared with the untreated group (P > 0.05) (Fig. 7, B and C). In addition, histopathological examination by H&E staining revealed that oral RNACap administration did not cause any noticeable damage to intestines or other major organs (Fig. 7D; figs. S28 and S29). These findings collectively demonstrate that RNACap can serve as a non-toxic platform for the oral delivery of liquid mRNA therapeutics.

Fig. 7. RNACaps demonstrate in vivo safety in rats.

(A) Experimental timeline for oral administration of IL-10-mRNA-RNACaps to rats for safety assessment. IL-10-mRNA: 25 μg in 3 RNACaps per rat. (B and C) Analysis of blood chemistry, including aminotransferase (ALT), aspartate aminotransferase (AST), and blood urea nitrogen (BUN) and complete blood count analysis, including white blood cell count (WBC), neutrophil (NE) (B), lymphocyte (LY), red blood cell count (RBC), hemoglobin (Hb), hematocrit (HCT), mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), and platelets (PLT) (C). (D) The rats were euthanized at the end of the study, and the indicated organs were sectioned and stained with H&E. Scale bars, 100 μm and 400 μm. (E and F) Cytokine concentration in the serum 6 h following oral administration of Fluc-mRNA-RNACap (Fluc-mRNA: 25 μg in 3 RNACaps per rat) to healthy rats were measured using ELISA and Luminex. Cytokines measured include IL-1ra, IL-1α, IL-1β, IL-2, IL-4, IL-5, IL-6, and IL-10 (E) or IL-12p70, IL-13, IL-17A, IL-18, IFN-γ, TNF-α, GM-CSF, and VEGF (F). Data are presented as mean ± S.D. Dots represent individual sample replicates. Statistical significance was evaluated by unpaired two-tailed Student’s t-test in (B, C, E, F). *P < 0.05, ***P < 0.001. n = 3 animals per group for all panels. All the schematic illustrations were created using Adobe Illustrator.

Oral administration of RNACaps does not cause obvious inflammatory toxicity

The pathways involved in immune reactivity to mRNAs have been well described (47), particularly regarding cytokine expression changes in serum following I.V. administration of mRNA formulations. Understanding these cytokine responses to oral mRNA exposure is therefore valuable for our study. To this end, rats were orally administered with Fluc-mRNA-RNACaps, and cytokines in the serum of both treated and untreated rats were measured 6 h post-administration using ELISA and Luminex. Although elevated serum cytokine concentrations of IL-1Ra, IL-5, and IL-6 were observed in Fluc-mRNA-RNACap-treated rats compared with untreated rats, the majority of serum cytokines evaluated were similar between both groups (P > 0.05), suggesting that no obvious inflammatory toxicities were observed after the oral administration of mRNA (Fig. 7, E and F and fig. S30, A to C). These findings are consistent with previous studies showing that murine leukocytes respond to RNA vaccines by upregulating anti-inflammatory IL-1Ra and protecting mice from cytokine-mediated toxicities (47).

RNACap mediates intestinal mRNA expression in a large swine model

Having demonstrated the ability of RNACap to deliver therapeutic IL-10-mRNA to the intestines for disease treatment in several rat models, we further investigated its feasibility for delivering multiple mRNAs to the intestines of a large swine model. First, we administered Cy5-loaded RNACaps (Free Cy5 dye solution was used here) to the stomach and intestines of each swine through a surgical procedure under anesthesia. (Fig. 8A). No Cy5 release from the Cy5-RNACaps was detected in the stomach fluid over a 2h period, indicating the stability of RNACap in the acidic stomach environment (Fig. 8B). Based on our design, RNACap sheds its enteric coating, becoming softened and uncapped in the intestine due to the neutral pH and the resulting pressure release. Although the contractions are known to be weaker under anesthesia, they should still be strong enough to compress the RNACap. Indeed, squeezed Cy5-RNACaps were recovered from the dissected intestines at end of the study (Fig. 8C and fig. S31). We observed the soft membrane remnants of RNACap 8.5 h post-administration (Fig. 8D), allowing us to conclude that the total transit time for the soft membrane remnants of RNACap in swine exceeds 8.5 h. The excised intestines (small intestine) of Cy5-RNACap-treated swine exhibited strong Cy5 fluorescence, whereas the intestines of untreated swine showed minimal fluorescence (Fig. 8, D to F). All three treated swine displayed higher fluorescence intensity compared with the untreated ones (Fig. 8G).

Fig. 8. RNACap mediates intestinal mRNA expression in a large swine model.

(A) Experimental timeline for dosing Cy5-loaded RNACaps (Cy5-RNACap, Cy5: 50 μg per RNACap, 3 RANCaps per swine) to a swine model. (B) Stomach fluids were collected at predetermined time points to determine the release of Cy5 from RNACaps. (C) The intestines of Cy5-RNACap-treated swine were dissected, and squeezed Cy5-RNACaps were recovered. (D to F) The small intestines of Cy5-RNACap-treated swine (swine 1 (D), 2 (E), 3 (F)) were isolated after 8.5 h, and the Cy5 fluorescence of the intestines was detected using an in vivo imaging system (IVIS). (G) The fluorescence of the intestines was quantified. (H to J) FLuc-mRNA-NP-loaded RNACaps (FLuc-mRNA-RNACap, FLuc-mRNA: 300 μg per RNACap, 3 RANCaps per swine) were dosed to a swine model (swine 1 (H), 2 (I), 3 (J)). The intestines of FLuc-mRNA-RNACap-treated swine were isolated after 8.5 h, and the Fluc bioluminescence of the intestines was detected using IVIS. (K) The bioluminescence of the intestines was quantified. (L) Experimental timeline for dosing IL-10-mRNA-NP-loaded RNACaps (IL-10-mRNA-RNACap, IL-10-mRNA: 300 μg per RNACap, 3 RANCaps per swine) to a swine model. Blood samples were collected every 30 min, and the intestines were isolated after 8.5 h. (M and N) The IL-10 concentration in the intestines were determined by ELISA (M) and immunofluorescence staining (N). IL-10 was stained with anti-IL-10 antibody (orange); cell nuclei were stained with DAPI (blue). Scale bar, 200 μm. (O) The fluorescence intensity ratios of IL-10 to DAPI in intestines were determined. (P) The IL-10 concentration in serum were determined by ELISA. (Q) Representative H&E staining of intestines and stomach collected 8.5 h post-treatment. Scale bars, 1000 μm and 200 μm. Data are presented as mean ± S.D. Dots represent individual sample replicates. Statistical significance was evaluated by an unpaired two-tailed Student’s t-test in (G, K, M, O) and by one-way ANOVA with Tukey’s post hoc analysis in (K, M, P). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. P-values of < 0.05 were considered statistically significant. n = 3 animals per group for all panels. All the schematic illustrations were created using Adobe Illustrator.

Next, FLuc-mRNA-NP-loaded RNACaps (3 RNACaps per swine) were dosed to each swine. FLuc bioluminescence was detected in the intestines of all three treated swine within just 8.5 h post-administration (Fig. 8, H to J). Among the treated swine, swine 3 exhibited higher bioluminescence intensity compared with swine 1 and 2 and all treated animals had higher intensity than untreated control (Fig. 8K). Similarly, EGFP-mRNA-RNACaps were dosed to swine (fig. S32, A to C). The successful expression of EGFP in the intestines was confirmed by the strong green fluorescence observed in the immunofluorescence images (fig. S32B), while EGFP fluorescence was barely detectable in untreated intestines (fig. S32C). These results demonstrate that RNACap not only releases its liquid mRNA cargo into the intestines of swine but also enables rapid mRNA expression within a few hours.

Having established the capability of RNACap to deliver reporter mRNAs to swine intestines, we next assessed whether the RNACap could deliver therapeutic IL-10-mRNA to the intestines of swine. IL-10-mRNA-RNACaps were dosed to each swine after 12 h of fasting (Fig. 8L). Blood samples were collected every 30 min, and the intestines were isolated after 8.5 h. As expected, squeezed IL-10-mRNA-RNACaps were recovered from the dissected intestines at the end of the study (fig. S33). ELISA analysis revealed that swine treated with IL-10-RNACap had higher IL-10 protein concentration in their intestines compared with untreated swine. (Fig. 8M). High IL-10 expression in the intestines of the IL-10-RNACap-teated swine was further confirmed by the immunofluorescence staining, which showed strong orange fluorescence labeling IL-10 (Fig. 8, N and O). In all three treated animals, we observed a statistically significant elevation of IL-10 concentration in the blood was observed between 7 and 8 h post-administration (P = 0.0010, P = and P < 0.0118, respectively) (Fig. 8P). H&E staining revealed that the dosing of RNACap did not cause any noticeable damage to the intestines, stomach (Fig. 8Q) or other major organs (figs. S34 to 36). These findings in a large swine model further highlight the clinical translation potential of our RNACap platform, due to the similarities in the GI system between human and swine.

DISCUSSION

mRNA therapeutics have substantially transformed the medical landscape, particularly by offering highly effective vaccines against SARS-CoV-2 at an unprecedented speed (1-3). This success has accelerated the development of many other highly effective mRNA vaccines and therapies for a wide range of diseases (4-7). Although current mRNA therapeutics are primarily administered through needle injections, the potential of non-invasive and self-administered mRNA treatments, such as oral mRNA therapeutics, holds great promise. Here, we introduce RNACap, a capsule-based device designed for specific oral delivery of liquid mRNA therapeutics to the intestines. Several key design features make the RNACap a promising platform for oral mRNA delivery. First, capacity for liquid mRNA formulations: RNACap can accommodate liquid mRNA formulations, avoiding the costly and time-consuming lyophilization process required for dry mRNA formulations (31). In addition, the liquid formulation is readily absorbable; Second, protection in the stomach environment: RNACap protects mRNA formulations from the harsh acidic and degradative conditions of the stomach. The optimized pH-sensitive polymer coating ensures the RNACap integrity in the stomach’s acidic pH while allowing its rapid dissolution in the neutral pH of intestines; Third, pressure-driven cargo release: The sealing mechanism of RNACap relies on the pressure exerted by the capsule cap, which rapidly dissolves in the intestines, releasing the pressure on the sealing membrane. This process triggers the fast departure of the sealing membrane, allowing the release of the mRNA cargo; Fourth, peristalsis (contraction)-driven cargo release: Unlike a 3D-printed rigid plastic device, dissolution of RNACap’s rigid layer in the intestines results in a softer structure. This enables the RNACap to be easily squeezed by the force generated by intestinal peristalsis, leading to the rapid release of cargo; Fifth, efficient intestinal transfection: The optimized NP formulation within RNACap facilitates the transport of mRNA from the intestinal lumen to the intestinal cells, where mRNA expression occurs. This allows for efficient intestinal transfection with various mRNA types.

One advantage of IL-10-mRNA-based therapy is its ability to enable efficient uptake of IL-10-mRNA by intestinal cells. This not only increases the concentration of anti-inflammatory cytokine IL-10 in the disease-affected tissue but also elevates IL-10 in the bloodstream due to this protein’s secreted nature. The IL-10 present in the blood can circulate to the disease-affected tissue, further enhancing therapeutic outcomes. In fact, oral administration of IL-10-mRNA-RNACap resulted in increased IL-10 protein in the blood.

Although the impairment of peristalsis during IBD has been well-documented, it is unlikely to affect the peristalsis-induced release of RNACap in this study. Based on our design, the dissolution of the enteric coating of RNACap facilitates the softening and dissolution of the gelatin capsules, leaving only the very soft, ductile coating membranes to accommodate the cargos. As a result, even minimal contractile force is sufficient to drive the release of cargo from the RNACap. A study on colonic motility in both healthy individuals and patients with ulcerative colitis demonstrated that although contractility was reduced in patients with ulcerative colitis compared with healthy individuals, it did not cease entirely (48). Meanwhile, the low-amplitude propagating contractions were observed to increase in patients with ulcerative colitis (48). Therefore, the residual peristalsis in IBD remains adequate to trigger the release of RNACap.

Single or multiple mRNA doses ranging from 0.25 mg/kg to 10 mg/kg have been demonstrated to be safe both in preclinical and clinical trials (49, 50). The total mRNA dose used in the present efficacy study was about 0.2 mg/kg, which is relatively low compared with the doses reported in these studies, suggesting the mRNA dose used here is likely safe. Compared with the small-molecule drug SSZ, mRNA offers comparable therapeutic efficacy with a substantially lower drug dose and reduced dosing frequency. This reduction in dose and frequency may also lower the risk of potential toxicity or adverse effects associated with high drug doses and frequent administration and warrants further investigation.

Our study is not without limitations. The present study focuses on the design and fabrication of oral RNACap, and future efforts can concentrate on extending its shelf life. A longer shelf life would enable the distribution of mRNA therapeutics to remote and resource-limited regions, which is especially crucial for vaccine distribution. This can be achieved through mRNA engineering, such as utilizing circular mRNAs, known for their enhanced stability (51-53). Alternatively, new mRNA carriers (for example, new NP formulations) could be developed to improve stability (54). Lastly, although the RNACap can deliver mRNAs to the intestines for robust transfection, the expression of mRNA in other major organs is not observed. Future efforts could be directed towards the development of the next generation of RNACap, which would enable mRNA therapeutics to reach systemic circulation after oral administration, allowing the treatment of both local and systemic diseases.

MATERIALS AND METHODS

Study design

This study was designed to investigate the potential for oral delivery of liquid mRNA therapeutics using an engineered capsule-based device. Rat colitis models were used to validate the therapeutic effect of the orally administered IL-10-mRNA-RNACaps. Body weight and inflammation-associated parameters were monitored daily, and cytokine concentrations in intestinal tissue and blood were determined using ELISA. The ability of RNACap to deliver multiple mRNAs in a large animal model (swine) was also examined. Following the administration of EGFP, Fluc or IL-10-mRNA-RNACaps, the corresponding protein expression was assessed using IVIS, ELISA, immunofluorescence staining, among other methods. All experiments presented in the main text were performed at least three times, with no outliers or data points excluded. All animal procedures and experiments were conducted with approval from the Institutional Animal Care and Use Committee of Brigham and Women’s Hospital or the Institutional Animal Ethics Committee of Zhejiang University, in accordance with National Institutes of Health animal care guidelines. Adult rats or swine were randomly assigned to different groups, and treatments were administered in a blinded manner. The treatments were only revealed during data analysis. Details regarding sample sizes, biological replicates, and statistical methodologies are provided in the corresponding figure legends. Details for all methods are available in the Supplementary Materials.

Animals

The Institutional Animal Care and Use Committee of Harvard Medical School/Brigham and Women’s Hospital, and the Institutional Animal Ethics Committee of Zhejiang University approved all in vivo experimental protocols for this study. The ethical approval numbers for all rat and pig studies are 2021N000236 and 2022N000096, respectively. Male Sprague Dawley (SD) rats with an average weight of 500 g were used. Swine (Yorkshire) with a weight of 30-50 kg were used in this study. Swine were randomly selected regardless of sex; both males and females were included. All in vivo studies were performed in accordance with the National Institutes of Health animal care guidelines, and in strict pathogen-free conditions in the animal facility of Brigham and Women’s Hospital and Zhejiang University.

Fabrication of RNACap for oral mRNA delivery to rats

An FDA-approved gelatin capsule (size 9h, Torpac) that can rapidly dissolve in an aqueous solution was used. The inner surface of the capsule’s body was first coated with a ductile membrane A. After loading 10 μL mRNA NP solution, an inductile membrane B was employed to cover the opening of the capsule’s body. The capsule was sealed by attaching the capsule cap. The final RNACap was obtained by coating with a pH-sensitive polymer (Eudragit L100-55 or Eudragit L100, 20% in methanol, w/w, Evonik Industries AG) and dried under air.

Fabrication of non-snap-cap Rigid-Cap for oral mRNA delivery to rats

The non-snap-cap, rigid capsule (Rigid-Cap) was fabricated by introducing an additional ductile coating membrane to the outer surface of the capsule’s body and cap. These additional membranes protect the capsule from softening and dissolving in the neutral pH environment of the intestine. Briefly, the inner and outer surfaces of the gelatin capsule’s body (size 9h, Torpac) were first coated with ductile membrane A. After loading 10 μL mRNA NP solution, an inductile membrane B was employed to cover the opening of the capsule’s body. The capsule was then sealed by attaching the capsule cap, which had been pre-coated with ductile membrane A on its outer surface. The final Rigid-Cap was obtained by coating with a pH-sensitive polymer by dipping it into a polymer solution (Eudragit L100-55, 20% in methanol, w/w, Evonik Industries AG) and allowing it to air-dry.

In vivo distribution of Cy5-mRNA-RNACap

Sprague Dawley (SD) rats (n = 3 rats per group) with an average weight of 500 g were fasted for 12-16 h before administration of Cy5-mRNA-RNACaps (L100-55) containing 50 μg Cy5-mRNA (6 RNACaps) via oral gavage. Notably, the RNACap developed in this study is specifically designed for intestinal delivery rather than stomach delivery. To this end, the RNACap must pass through the stomach and enter the intestine. However, the significant presence of food in the stomach may obstruct the pylorus, the opening between the stomach and the intestine, thereby reducing the chance of the RNACap reaching the intestine. This could affect the consistency and effectiveness of the treatment. In addition, unlike humans, who eat individual meals, rats eat continuously, keeping their stomachs filled with food at all times. As a result, brief fasting better simulates the human condition. Thus, fasted rats were used in this study, a practice commonly reported in many other oral therapy studies (55). At predetermined time points (15 min, 1 h, 2 h, 4 h, and 6 h), the rats were euthanized, and the stomach, intestines and other major organs, including heart, lungs, spleen, kidneys, were harvested for ex vivo imaging using an IVIS Lumina S5 (PerkinElmer) imaging system. Similarly, L-100-coated Cy5-mRNA-RNACaps were also orally administered to rats, which were euthanized and imaged at 6 h post-administration.

RNACap-mediated in vivo absorption of mRNA

SD rats (n = 3 rats per group) with an average weight of 500 g were fasted for 12-16 h before administration of Cy5-mRNA-RNACaps containing 50 μg Cy5-mRNA (six RNACaps) via oral gavage. The rats were euthanized at 4 h post-oral administration and intestinal tissues were collected and immediately frozen in Tissue-Tek O.C.T. using liquid nitrogen. Cross-sections of the intestinal tissues with a thickness of 12 μm were obtained using a LeicaCM1900 cryostat. The obtained sections were stained with Hoechst 33342 for nuclei staining, followed by washes with PBS. The samples were observed by a confocal laser scanning microscope (Olympus FV1000).

Transit time of RNACap remnants

SD rats (n = 3 rats per group) with an average weight of 500 g were fasted for 12-16 h before administration of RNACaps containing 50 μg IL-10-mRNA (6 RNACaps) by oral gavage. At 16 or 24 h post-oral administration, rats were euthanized. The GI tract of the rats was carefully collected and dissected, and all the solid food residues and feces in the GI tract were retrieved. The food residues and feces were then washed with water, portion by portion, to identify the tiny soft membranes from the RNACap (remnants).

Oral delivery of reporter mRNA by RNACap

SD rats (n = 3 rats per group) with an average weight of 500 g were fasted for 12-16 h before administration of EGFP-mRNA-RNACaps containing 50 μg EGFP-mRNA (6 RNACaps) by oral gavage. At 24 h post-oral administration, rats were euthanized, and the intestines were collected for further analysis. The immunofluorescence and H&E staining of the intestines were performed according to section “General procedure for immunofluorescence and H&E staining of tissues”. Antibodies dilutions: GFP (D5.1) Rabbit mAb (1:200, 2956, Cell Signaling Technology), villin polyclonal antibody (1:200, 16488-1-AP, Proteintech), SOX9 rabbit mAb (1:200, ab185230, Abcam)

Immune reactivity to mRNAs

SD rats (n = 3 rats per group) with an average weight of 500 g were fasted for 12-16 h before administration of Fluc-mRNA-RNACaps containing 25 μg Fluc-mRNA (3 RNACaps) by oral gavage. Six h post-administration, serum was collected for further analysis. Rat IL-1ra concentrations were measured using an ELISA kit (ThermoFisher Scientific). Serum cytokines/chemokine concentrations were determined using MILLIPLEX Rat Cytokine/Chemokine Magnetic Bead Panel (RECYTMAG-65K) according to the manufacturer’s instructions. The analytes included EGF, Eotaxin/CCL11, Fractalkine, G-CSF, GM-CSF, GRO/KC, IFN-γ, IL-1α, IL-1β, IL-2, IL-4, IL-5, IL-6, IL-10, IL-12 (p70), IL-13, IL-17A, IL-18, IP-10, Leptin, LIX, MCP-1, MIP-1α, MIP-2, RANTES, TNF-α, VEGF.

Oral delivery of therapeutic mRNA by RNACap

For blood: To determine the blood IL-10 concentrations in rats following oral administration of RNACaps, SD rats (n = 3 rats per group) with an average weight of 500 g were fasted for 12-16 h before receiving empty RNACaps, EGFP-mRNA-RNACaps or IL-10-mRNA-RNACaps containing 50 μg of IL-10 (or EGFP)-mRNA (6 RNACaps) via oral gavage; To determine the blood IL-10 concentrations in rats following intravenous (I.V.) administration, SD rats (n = 3 rats per group) with an average weight of 500 g were fasted for 12-16 h before receiving IL-10-mRNA NPs containing 50 μg of IL-10-mRNA (in 0.5 mL PBS) via I.V. injection through the tail vein; To determine the blood IL-10 concentrations in rats following oral administration of Rigid-Caps, SD rats (n = 3 rats per group) with an average weight of 500 g were fasted for 12-16 h before receiving IL-10-mRNA-Rigid-Caps containing 50 μg of IL-10-mRNA (6 Rigid-Caps) via oral gavage; To determine the blood IL-10 concentrations in rats following oral administration of RNACaps with mRNA in COVID LNPs, SD rats (n = 3 rats per group) with an average weight of 500 g were fasted for 12-16 h before receiving IL-10-mRNA-RNACaps (COVID LNPs) containing 50 μg of IL-10-mRNA (6 RNACaps) via oral gavage. Blood samples were collected at 24 h, 48 h and 72 h post-oral administration, followed by the IL-10 measurement by ELISA.

For tissue: SD rats (n = 3 rats per group) with an average weight of 500 g were fasted for 12-16 h before administration of IL-10-mRNA-RNACaps containing 50 μg IL-10-mRNA (6 RNACaps) via oral gavage. At 24 h post-oral administration, rats were euthanized, and the intestines were collected for further analysis.

The RT-PCR of the intestines was performed according to the section “Reverse transcriptase polymerase chain reaction (RT-PCR) assay of tissue.” The duodenum, jejunum, ileum and colon were analyzed. Primer sequences for RT-PCR (Integrated DNA Technologies): IL-10, forward GCTCTTACTGACTGGCATGAG, reverse CGCAGCTCTAGGAGCATGTG; GAPDH, forward CAAGTTCAACGGCACAGTCA, reverse CCATTTGATGTTAGCGGGAT.

The intestines were divided into 9 parts and cut into small pieces. Then, 200 mg of each part were added to 4 mL of tissue protein extraction reagent (78510, Thermo Fisher Scientific Inc.) containing protease inhibitors (87785, Thermo Fisher Scientific Inc.). The tissues were homogenized and incubated in ice for 30 min, followed by centrifugation (10, 000 g, 4 °C, 10 min). The supernatant (protein) was collected, and the IL-10 concentrations were analyzed by ELISA.

The immunofluorescence and H&E staining of the intestines were performed according to the section “General procedure for immunofluorescence and H&E staining of tissues.” Antibodies dilutions: Mouse/Rat IL-10 Antibody (1:40, AF519, R&D Systems), Villin Polyclonal antibody (1:200, 16488-1-AP, Cell Signaling Technology).

For protein western blotting, the supernatant (protein) was denatured by adding SDS buffer (SDS: 2-mercaptoethanol = 9:1, v/v) and subsequently heated at 100 °C for 10 min. The denatured protein samples were immediately used for western blot analysis or stored at −20 °C. Antibody dilutions: anti-IL-10 antibody mouse mAb (1:250, sc-365858, Santa Cruz Biotechnology), β-Actin rabbit Ab (1:1000, 4967, Cell Signaling Technology), HRP-linked anti-rabbit IgG (1:1000, 7074, Cell Signaling Technology) and anti-mouse IgG (1:1000, 7076, Cell Signaling Technology).

Therapeutic efficacy evaluation of IL-10-mRNA-RNACap in rats

The therapeutic efficacy of IL-10-mRNA-RNACap was evaluated in a DSS-induced rat colitis model. Acute colitis was induced in rats (n = 5 rats per group) by providing free access to drinking water supplemented with 6.0% (w/w) DSS (molecular weight: 36k-50k Da, 23250, Cayman Chemicals) for 8 days. After the initiation of colitis induction with 6% DSS, IL-10-mRNA-RNACaps containing 25 μg IL-10-mRNA (3 RNACaps) were orally administered to the rats on days 2, 5 and 8. Rats treated with water or only DSS were used as control groups. The body weight, stool consistency, and bleeding of the rats were monitored daily. The disease activity index (DAI) of the rats was calculated based on the inflammation-associated parameters such as body weight loss (the percentage of weight loss relative to the initial body weight, where 0, no loss; 1, 1-5%; 2, 6%-10%; 3, 11%-20%; 4, >20%), stool consistency (0, normal; 2, loose stool; 4, diarrhea) and bleeding (0, no blood; 1, hemoccult positive; 2, hemoccult positive and visual pellet bleeding; 4, gross bleeding, blood around anus). At the end of the treatment, blood samples were collected, and the rats were euthanized. The length of the colon was determined, and the colons were cut into small pieces. Subsequently, 200 mg of colon tissue was added to 4 mL of tissue protein extraction reagent (78510, Thermo Fisher Scientific Inc.) containing protease inhibitors (87785, Thermo Fisher Scientific Inc.). The tissues were homogenized and incubated in ice for 30 min, followed by centrifugation (10, 000 g, 4 °C, 10 min). The supernatant (protein) was collected. Then, the IL-10, TNF-α, IL-1β, IL-6, IL-17A and MCP-1 concentrations in colon tissues and serum were analyzed by ELISA. Rat TNF-α ELISA (438204, Biolegend), Rat IL-1 beta/IL-1F2 ELISA (DY501-05, R&D Systems), Rat IL-6 ELISA (DY506-05, R&D Systems), Rat IL-17A ELISA (437904, Biolegend), Rat JE/MCP-1/CCL2 ELISA (DY3144-05, R&D Systems). For H&E staining: The H&E staining of the colons was performed according to the section “General procedure for immunofluorescence and H&E staining of tissues.”

Therapeutic efficacy evaluation of IL-10-mRNA-RNACap in rats in a delayed therapeutic setting

The therapeutic efficacy of IL-10-mRNA-RNACap was evaluated in a DSS-induced rat colitis model. Rats (n = 5 rats per group) were given free access to drinking water supplemented with 8.0% (w/w) DSS for 10 days to establish a colitis model. Afterward, plain water was provided, and rats were orally administered IL-10-mRNA-RNACaps containing 25 μg of IL-10-mRNA (3 RNACaps) on days 11, 14 and 17 or sulfasalazine (SSZ, standard therapy, 100mg/kg/day) daily. SSZ was suspended in distilled water containing 0.5% sodium carboxymethyl cellulose. Rats treated with water or only DSS were used as control groups. The rats and samples were monitored and analyzed as described in section “Therapeutic efficacy evaluation of IL-10-mRNA-RNACap in rats”.

In vivo safety evaluation of IL-10-mRNA-RNACap

The safety profile of IL-10-mRNA-RNACap was assessed in healthy rats (n = 3 rats per group). IL-10-mRNA-RNACaps containing 25 μg IL-10-mRNA (3 RNACaps) were orally administered to the rats on days 2, 5 and 8. Untreated rats were used as control groups. At the end of the treatment, blood samples were collected. The protein concentrations in plasma and hematological parameters of blood were analyzed, including alanine transaminase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (BUN) concentrations, white blood cell count (WBC), neutrophil (NE), lymphocyte (LY), red blood cell count (RBC), platelets (PLT), hemoglobin (Hb), hematocrit (HCT), mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH) and mean corpuscular hemoglobin concentration (MCHC). Then the rats were euthanized, and the intestines and other major organs were collected for further analysis. For H&E staining: The H&E staining of the colons and other major organs was performed according to the section “General procedure for immunofluorescence and H&E staining of tissues.”

Fabrication of RNACap for delivery of mRNAs to a large swine model

An FDA-approved gelatin capsule (size 000), which can rapidly dissolve in an aqueous solution, was used. RNACap was engineered similarly to the previously described size 9h or size 5 capsule-based RNACap. However, a much larger volume of 1mL of Cy5 solution (50 μg per RNACap) or FLuc-, EGFP or IL-10-mRNA NP solutions (300 μg per RNACap) were loaded into the RNACap.

In vivo delivery of multiple mRNAs to the intestines of swine by RNACap

Swine (Yorkshire) (n = 3 rats per group)with a weight of 30-50 kg were used in this study. During the experiment, the swine were under anesthesia. One or three Cy5-loaded RNACaps (Cy5-RNACap, Cy5: 50 μg per RNACap) were dosed to the stomach or intestines of each swine via surgery. The stomach fluid was collected from the stomach at 0, 15, 30, 45, 60, 90 and 120 min post-administration to determine the release of Cy5 from the Cy5-RNACaps. Then, the intestines of Cy5-RNACap-treated swine were isolated after 8.5 h, and the Cy5 fluorescence of the intestines was detected using an IVIS. The fluorescence of the intestines was also quantified. Similarly, three FLuc-mRNA-NP-loaded RNACaps (FLuc-mRNA-RNACap, FLuc-mRNA: 300 μg per RNACap), EGFP-mRNA-NP-loaded RNACaps (EGFP-mRNA-RNACap, EGFP-mRNA: 300 μg per RNACap) or IL-10-mRNA-NP-loaded RNACaps (IL-10-mRNA-RNACap, IL-10-mRNA: 300 μg per RNACap) were dosed to each swine. The Fluc bioluminescence of the intestines was detected using IVIS. The EGFP concentrations in the intestine were detected by immunofluorescence staining. The IL-10 concentration in the intestinal tissues and blood was measured by ELISA, and the IL-10 concentration in the intestinal tissues was further detected by immunofluorescence staining. The intestines, stomachs and other major organs were collected 8.5 h post-treatment, followed by sectioning and staining with H&E.

Statistical analysis

All experiments were performed in triplicate unless otherwise specified, and all results are presented as the means ± SD. A two-tailed unpaired Student’s t-test was used for comparisons between two groups, while a one-way analysis of variance (ANOVA) with Tukey’s correction for multiple comparisons was applied when comparing more than two groups. Normality was tested using the Shapiro-Wilk test. All statistical analyses were conducted using GraphPad Prism 7 software. P values of <0.05 were considered statistically significant. Statistical significance in all figures is represented as follows: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Sample sizes (n) are provided in the figure legends. All individual-level data are available in data file S1.

Supplementary Material

Materials and Methods

Figs. S1 to S36

Data file S1

MDAR Reproducibility Checklist

References (41-43, 56-57)