Ginger exosome-like nanoparticle-derived miRNA therapeutics: A strategic inhibitor of intestinal inflammation

Graphical abstract

This study addresses how ginger exosome like nanoparticles (GELNs)-derived miRNAs are used to target specific host mRNA to alleviate inflammation. GELNs-derived osa-miR164d was identified as a regulator of reprogramming macrophage polarization, ultimately alleviating intestinal inflammation. Osa-miR164d directly targets the TAB1 promoting M2 polarization. To scale up preparation and simplify the delivery system, the biomimetic exosomes loaded with osa-miR164d were designed.

Highlights

• •Ginger exosome-like nanoparticles (GELNs) transfer miRNAs to macrophage.
• •Osa-miR164d specifically and directly modulates the TAB1 expression of host cell.
• •GELNs-derived osa-miR164d mitigates inflammation.
• •Osa-miR164d alleviates colitis-related symptoms by rebalancing polarization of macrophages.
• •Bionic exosome loaded Osa-miR164d as biomimetic GELNs exerts delivery and biological functions.

Abstract

Introduction

MicroRNAs (miRNAs) involve in destabilising messenger RNA or repressing translation of target molecules. Ginger-derived exosome-like nanoparticles (GELNs) play a crucial role in modulating intestinal inflammation. Moreover, GELNs contain highly heterogeneous miRNA. However, the role of miRNAs derived from GELNs in immunomodulation remains unclear.

Objectives

This study aimed to elucidate the molecular basis of the unique biological effects mediated by miRNA derived from GELNs on macrophages.

Methods

GELNs were isolated using a combination of commercial exosome isolation kits and the differential centrifugation method, and the lipid composition of GELNs was determined using liquid chromatography-mass spectrometry. Subsequently, PKH26 labelled GELNs were taken up by macrophages. Furthermore, the modulation of inflammatory and immune responses by GELNs or osa-miR164d was assessed through the RNA-seq, RT-qPCR, online databases, and dual luciferase reporter assays to explore the underlying mechanisms of osa-miR164d. Biomimetic exosomes loaded with osa-miR164d were prepared using a microfluidic mixing device and systematically characterized. The therapeutic effects of osa-miR164d on relieving colitis were evaluated.

Results

We report for the first time that GELNs-derived osa-miR164d is a regulatory factor of reprogramming macrophage polarization, thereby inhibiting the intestinal inflammatory response. Mechanistically, osa-miR164d directly targets the 3′-UTRs of TAB1, which regulates macrophage polarization through the downregulation of NF-κB expression. In addition, We have designed a biomimetic exosome mimicking GELNs to deliver osa-miR164d (osa-miR164d-MGELNs). Notably, the osa-miR164d-MGELNs can efficiently reprogram macrophages to alleviate colitis-related symptoms.

Conclusion

Our findings enhance the systematic understanding of how GELNs-derived osa-miR164d mediates cross-kingdom communication and provide an original engineering paradigm for mimicking GELNs to transfer miRNA.

Introduction

MicroRNAs (miRNAs) are non-coding transcripts that are involved in destabilising messenger RNA or repressing translation of target molecules primarily by binding to the 3′Untransated Region (3′UTR) of complementary genes [1]. MiRNAs serve as potent genetic regulators directly modulating cellular processes at the pathway level through interactions with target genes [1], [2]. Within the immune system, specific miRNAs from animal cells have recently emerged as crucial regulators of inflammation response [3], [4]. This indicates that these miRNAs have evolved to be encapsulated within exosomes to perform their biological functions. Growing evidence supports the view that plant miRNA elicits intercellular crosstalk effects on the physiology of recipient cells [5], [6]. Oral administration of soybean-derived gma-miR159a has shown potential anti-colon cancer effects [6]. Various cells can receive miRNAs as a means of communication, which has set the stage for exploring the functional roles of transferred plant miRNAs in the context of intestinal inflammatory responses.

MiRNAs loaded exosome transferred are emerging as novel regulators of cellular function. As a type of promising natural product, plant-derived exosome-like nanoparticles (PELNs) are 30–300 nm membrane nanovesicles, similar to mammalian exosomes [7], [8]. The lipid bilayer membrane of PELNs protects nucleic acids (DNA, mRNA, microRNAs, and lncRNA), and proteins from degradation [7], [8]. Accumulating evidences support the significant role of PELNs in maintaining mammalian host cell homeostasis and reducing susceptibility to infections, chronic inflammatory diseases, and tumors [9], [10], [11]. PELNs enter the host cell and serve as key mediators for intercellular material exchange and cell signal transduction. The plentiful biological effects of the exosomal system have been ascribed to individuals or groups of miRNAs [12]. The field of PELNs biology has rapidly advanced, and therapeutic potential for the use of PELNs-derived miRNAs in diseases may be initiated in the future. Although the physiological roles of PELNs miRNAs remain largely unknown, growing evidence indicates their potential to improve multiple pathological and physiological characteristics in specific animal disease models [5], [6], [9]. Therefore, identifying miRNAs derived from PELNs may help to further elucidate the mechanisms underlying PELNs-derived therapeutics for the improvement of human intestinal health.

Ginger has been used in traditional Chinese medicine to treat gastrointestinal disorders such as stomach aches, abdominal spasms, and vomiting [13], [14]. In the USA, ginger the Food and Drug Administration (FDA) has labeled as “generally recognized as safe”. The beneficial effects of ginger and its ingredients have been demonstrated in various clinical and experimental inflammatory disorders [13], [14]. As aberrant macrophage polarization occurs during the development of intestinal inflammation, the transition from the pro-inflammatory macrophage (M1) phenotype to the anti-inflammatory macrophage (M2) phenotype has been targeted as an attractive therapeutic strategy for intestinal inflammatory disease [15]. Ginger contains tiny vesicles, similar to exosomes [16], [17]. We and others have reported that ginger exosome-like nanoparticles (GELNs) can attenuate intestinal inflammation [9], [18], [19]. Despite our growing understanding of how GELNs affect immunities, the regulation and function of the GELNs-derived miRNAs remain unclear. Therefore, it is vital to elucidate the molecular basis of the unique biological effects mediated by miRNA derived from GELNs on the macrophages and expand our current conceptual framework of GELNs biology.

Based on the aforementioned background and our published miRNA profile of GELNs [19], we hypothesize that GELNs-derived miRNAs may serve as the key component in promoting macrophage reprogramming. Here, we identified GELNs-derived osa-miR164d that played a crucial role in regulating macrophage polarization. Mechanistically, osa-miR164d directly targeted TAB1, thereby promoting M2 polarization and inhibiting the inflammatory response through the TNF signaling pathway. Additionally, we have developed osa-miR164d-loaded bionic exosome mimicking GELNs (osa-miR164d-MGELNs), which can ameliorate colitis by reprogramming macrophage polarization. Together, our findings advance the systematic understanding of how GELNs-derived osa-miR164d mediates cross-kingdom communication and provide an original paradigm for engineering PELNs-derived miRNAs.

Materials and method

Isolation and characterization of GELNs

GELNs were isolated and purified from fresh ginger juice using a combination of commercial exosome isolation kits and the differential centrifugation method as previously published [19]. Using bovine serum albumin as a standard, the protein contents of GELNs were determined using a BCA protein assay kit (Shanghai Yuanye Biotech Co., LTD). The morphology of the purified GELNs was observed by transmission electron microscopy (TEM) (Hitachi HT7700 instruments) operated at 80 kV. With the aid of a nanotracking system (Particle Metrix, Meerbusch, Germany), the particle counts and sizes of GELNs were measured. The zeta potential of GELNs was measured by laser diffraction spectrometry.

Endocytosis experiments

To observe the uptake of PKH26 labelled GELNs (PKH26-GELNs) in RAW264.7 macrophages, the cells were seeded in confocal dishes and incubated with PKH26-GELNs (2 μg/mL) for 6 h and 12 h. The cells were then washed three times with cold Phosphate Buffer Saline (PBS) solution, fixed with 10 % formaldehyde, and stained with 4′,6-diamidino-2-phenylindole (DAPI) and 3,3′-dioctadecyloxacarbocyanine perchlorate (DIO) to identify the nucleus and membrane. PKH26-GELNs uptake was observed by Olympus FV1000 Confocal microscopy. The PKH26 (MCE, USA) was used for labelling of purified GELNs according to a previously described method [19].

Cytotoxicity assay

RAW264.7 macrophages were seeded at a density of 1 × 10⁵ cells/well in 96-well plates for 24 h. Then, the cells were treated with serial concentrations of GELNs (0, 5, 10, 20, 40, and 80 μg/mL) for 24 h. At the indicated time points, cell viability was measured using the MTT ((3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium)) method [19].

Measurement of GELNs effect of M2-like macrophage polarization

After seeding RAW264.7 cells at a density of 1 × 10⁵ cells/well onto 24-well plates, they were incubated for 24 h with 1 × PBS (Control), 40 μg/mL of GELNs (GELNs-1) or 80 μg/mL of GELNs (GELNs-2), respectively. Next, the cells were administered with either lipopolysaccharide (LPS) (5 μg/mL), or 1 × PBS over 24 h. The cell culture supernatant was collected and detected the production level of inflammatory cytokines, using ELISA kits according to the manufacturer’s procedure (Shanghai YouXuan Biotechnology Co., LTD).

Flow cytometric analysis of macrophage surface markers

RAW264.7 cells were harvested, centrifuged, and washed with PBS. The anti-CD16/32 antibody (Biolegend 101319, California, United States) blocked the Fc receptor on the surface of RAW264.7 cells and incubated on ice for 10 min. Following the manufacturer's instructions, RAW264.7 cells were stained with fluorochrome-conjugated monoclonal antibodies against mouse CD206-APC (Biolegend 141707, California, United States) and mouse CD86-FITC (Biolegend 105005, California, United States). Subsequently, stained cells (10⁵ cells/sample) were analyzed by flow cytometry using a FACSCalibur system (BD Biosciences).

Quantitative real-time polymerase chain reaction (RT-qPCR)

Total RNA was isolated from the macrophages with Trizol Reagent (Sangon Biotech, Shanghai) according to the manufacturer's specifications. Next, the extracted RNA was reverse-transcribed and determined according to the manufacturer's instructions of the RT-qPCR kit. GAPDH levels served as a reference point for normalizing the relative expression of mRNAs. The primers were listed in Table S1 (Supporting Information).

Enzyme-linked immunosorbent assay (ELISA)

RAW 264.7 cells supernatant was obtained and centrifuged to dislodge the pellet and polymer. The concentrations of IL-6, TNF-α, IL-10, IL-1β, and TGF-β were determined in cells using ELISA kits.

Quantification and efficacy delivery of exosomal miRNA

GELNs were treated with a Trizol Reagent to extract total RNA. The relative miRNA expression was quantified as previously described [19]. After adding 40 μg/mL or 80 μg/mL GELNs, RAW264.7 cells were incubated for 24 h. The expression levels of several specific miRNAs (Novel-40, Novel-45, osa-miR164d, stu-miR156f-5p, vvi-miR396b, and zma-miR396g-5p) were evaluated by RT-qPCR, and the miRNA sequence information was shown in Table S2. The samples were reverse transcribed using the miRNA first strand cDNA synthesis kit, and 10 pM of synthetic cel-miR-39-3p was added to each sample as a quantitative. RT-qPCR results were analyzed using the 2^−ΔΔCt method. The primer information was shown in Table S3, and the downstream primer sequence was CGCCCGCCCGCTCCCAAGAT.

Prediction of miRNA-mRNA-inflammatory pathway interaction network

The miRanda and RNAhybrid were used to predict the target genes of miRNAs in the mouse genome (Mus musculus). The KOBAS software was utilized to conduct Kyoto encyclopedia of genes and genomes (KEGG) Pathway enrichment analysis for target genes with GELNs-derived special miRNAs [20]. In this enrichment study, the KEGG pathway was chosen as the unit of analysis, and a hypergeometric test was performed to identify the pathways where target genes significantly enrich the background of the entire genome.

MiRNA mimic transfection

The RAW264.7 macrophages were transfected with osa-miR164d (100 pM), vvi-miR396b (100 pM), and zma-miR396g-5p mimics (100 pM) by Invitrogen Lipofectamine^R 2000 Reagent (Thermo Fisher Scientific Inc). RAW264.7 cells were transfected with three miRNA mimics for 24 h when the cells reached ∼80 % confluency, respectively. Then the above-mentioned plates were incubated with LPS (5 μg/mL) for 12 h. After 36 h of transfection, the level of IL-6, IL-10, and TNF-α was determined by RT-qPCR and ELISA kits. The miRNA mimics mentioned above were provided by Sangon Biotech (Shanghai, China), and the control was a scrambled negative control (NC) sequence purchased by Sangon Biotech (Table S2).

Structure analysis of GELNs-derived miRNAs

Using default values for input conditions, an online application (https://www.unafold.org/mfold/applications/rna-folding-form-v2.php), which was utilized to obtain predictions regarding the secondary structure of miRNAs derived from GELNs, was employed. For each unique GELNs-derived miRNA, the simulated structures with the lowest free energy of formation (ΔG, KJ/moL) were chosen as the secondary structure that had the highest probability of occurring.

RNA-seq

RNA sequencing was performed by Shanghai OE Biotech Co., Ltd. Briefly, the libraries were sequenced on an Illumina Novaseq 6000 platform, generating 150 base pairs reads. Raw reads of fastq format were firstly processed using fastp1 to filter out unqualified sequences. Cleaned reads were then mapped to the reference genome using HISAT2². FPKM³ of each gene was calculated, and the read counts of each gene were obtained using HTSeq-count⁴. Differential expression analysis was performed using the DESeq². Q value <0.05 with fold change >2 or fold change <0.5 was considered a significant differential expression gene (DEGs).

Luciferase reporter assay

Potential binding sites of osa-miR164d to TAB1-3′UTR were predicted online using RNAhybrid. The mutant sequences of the target sites were obtained using the point mutation method. Using both sequences as templates, we amplified approximately 150 bp sequences above and below them through polymerase chain reaction (PCR). By using the luciferase reporter assay, it was determined whether osa-miR164d inhibited the expression of TAB1 in RAW264.7 macrophages, respectively. The 3′UTR target sites of osa-miR164d were constructed and inserted into the pmirGlo luciferase vector. The sites were designed with antisense oligonucleotides complementary to the sequences of specific miRNAs. The matching sequences to the binding site of osa-miR164d are 5́-UGGAGAAGCAGGGCACGUGCU-3′. The mimic NC as negative control and osa-miR164d mimics were cotransfected with the constructed pmirGLo luciferase vector into RAW264.7 cells using Lipofectamine 2000. Using a Dual-LumiTM Luciferase Assay Kit, luciferase activity was found 48 h after transfection.

Lipidomic analysis

To extract total lipids, GELNs were immersed in methanol with butylated hydroxytoluene for 15 min. Then, the samples were immersed in glass test tubes with Teflon-lined caps and vortexed at room temperature for 1 h. To the tubes, deionized water was added, and the mixture was vortexed again. The sample was incubated for 10 min to stratify the organic, followed by centrifugation at 1000 g for 10 min at 25 °C. The lower-phased liquid, which contained lipids, was collected via a Pasteur pipette. The upper organic phase was also collected. Subsequently, methyl tertbutyl ether/methanol/water at a 10:3:2.5 ratio (v/v/v) was added to the lower layer (water and methanol) and extracted again to collect the organic phase. The lipid composition of GELNs was determined using liquid chromatography-mass spectrometry.

Preparation and characterization of osa-miR164d-MGELNs

DLin-MC3-DMA, cholesterol, 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC), and DMG-PEG2000 were purchased from GlpBio Inc. Briefly, osa-miR164d was dissolved in pH4.0 sodium citrate buffer solution. Then, ethanol was used to dissolve DLin-MC3-DMA, cholesterol, DSPC, and PEG-2000 at a molar ratio of 50: 10: 38.5: 1.5. The osa-miR164d /total lipid ratio was 0.06 (wt/wt). The lipid mixture and the osa-miR164d were injected into a microfluidic mixing device at a combined flow rate of 0.1 mL min⁻¹. After the generation of osa-miR164d-MGELNs, the nanoparticles were dialyzed against PBS (pH 7.4) overnight with two buffer exchanges to remove ethanol. The solution was sterilized by a 0.22-mm membrane filter. Particle size and zeta potential were quantified by NanoZS Zeta sizer (Malvern). The morphology of osa-miR164d-MGELNs was characterized by a TEM device. The macrophages were transfected with osa-miR164d-MGELNs including 100 pM of miRNA.

Gel electropheresis

The encapsulation efficiency of osa-miR164d-MGELN was analyzed by gel retardation assay. The osa-miR164d-MGELNs were placed on a 4 % denaturing polyacrylamide gel in the Tris/borate/EDTA (TBE) buffer containing 1 × Gel Red nucleic acid stain. The gel was run at 150 V for 30 min, and the miRNA bands were monitored with a gel documentation system using an ultraviolet transilluminator. The intensities of the miRNA bands were analyzed using the ImageJ program.

Animal studies

Six-week-old male C57BL/6 mice (20 ± 2 g) were purchased from the SCBS Biotechnology Co., Ltd. After cohousing for one week, the mice were divided into four groups (n = 5), including the control group, DSS group (2.5 % DSS), DSS- MGELNs group (2.5 % DSS + MGELNs) and osa-miR164d-MGELNs group (2.5 % DSS + osa-miR164d-MGELNs) (Fig. 7A). During the experiment, the body weight of each mouse, stool consistency, and fecal bleeding were recorded daily. On day 8, the mice in four groups were euthanized, and the colon tissues were collected for subsequent analysis. Orally osa-miR164d-MGELNs and MGELNs were administered for 7 days. Disease activity index (DAI) is employed to evaluate the degree of symptoms, which are based on the summation of body weight loss (0–4), stool consistency (0–4), and fecal bleeding (0–4) [21]. The concentrations of inflammatory cytokine and growth factor were determined in colon tissue using ELISA kits.

Fig. 7.

osa-miR164d-MGELNs mitigates inflammation. (A) Pie chart with a summary of the lipid species in GELNs. PA, phosphatidic acids; PE, phosphatidylethanolamines; PG, phosphatidylglycerol; MGDG, mono/di/glycerols; PS, phosphatidylserine; PI, phosphatidylinositol; PC, phosphatidylcholines; LPC, lyso-phosphatidylcholines; LPG, lyso-phosphatidylglycerol. (B) Schematic representation of osa-miR164d-MGELNs. (C) Dynamic Light Scattering characterization of osa-miR164d-MGELNs and MGELNs. (D) Gel Electropheresis of free osa-miR164d and the supernatants of osa-miR164d-MGELNs. (E) Structure of osa-miR164d-MGELNs and MGELNs visualized by TEM. (F) Cell viability assay of RAW264.7 cells treated with either osa-miR164d-MGELNs or MGELNs. (G) Quantitative analysis of IL-1β, IL-6, IL-10, and TGF-β protein expression.

Hematoxylin–Eosin (H&E) and immunohistochemistry

H&E staining was used to detect the pathological changes of the mice colon. The colon was removed and fixed in 4 % paraformaldehyde. The tissue samples were then routinely dehydrated, embedded, and sliced. After that, the slices were stained with H&E, and the colon was observed under the digital microscope. The histopathological score was determined by two independent investigators who adhered to the confidentiality of the experimental protocol. The scoring criteria were listed in Table S4 [21]. For immunohistochemistry, the colon sections were pre-treated, and the tissues were probed with CD11b antibody (Cell Signaling Technology). Next, the sections were incubated with secondary antibodies followed by staining with Hematoxylin. The sections were visualized under a light microscope. The quantitative analysis of CD11b was performed using ImageJ software.

Statistics

The experiment was performed using GraphPad Prism. Unpaired two-tailed Student's t-tests were used to determine the differences between the two groups. Data from multiple comparisons were analyzed using one-way ANOVA. Statistical differences were considered significant at * P < 0.05, ** P < 0.01, or ***P < 0.001, and ns represents no significance.

Results

Identification and analysis of GELNs

The GELNs were purified using a differential centrifugation method combined with an isolation kit reported (Fig. 1A). The morphological assessment of the isolated particles using TEM revealed an expected cup-shaped structure (Fig. 1B), akin to the recognized characteristics of GELNs in our previous report [19]. Fig. 1C shows an average zeta potential of –23.7 mV for GELNs. Additionally, the average protein concentration of GELNs in the isolates was 3010 ng/μL. Nanoparticle tracking analysis revealed particle concentrations of 2.1 × 10¹⁰ particles/mL and a mean size of 163 ± 28 nm (Fig. 1D). Collectively, these data suggested that we obtained a high proportion and purity of GELNs.

Fig. 1.

Identification and characterization of GELNs. (A) Schematic diagram of GELNs isolation. tracking analysis and (B) TEM images of the GELNs, scale bar = 200 nm. (C) Surface charge of the GELNs. (D) Particle size of GELNs.

GELNs modulate M2 activation of macrophages

Macrophages can undergo both morphological and functional changes due to various inflammatory response signals present in the local organism microenvironment [22]. Given that GELNs can alleviate intestinal inflammatory responses, we investigated whether GELNs treatment could anti-inflammatory effect induced expression of M2 macrophage. We tested the cytotoxic effect of the GELNs at a series of concentrations on the viability of macrophages using the MTT assay (Fig. S1). As indicated in Fig. 2A, GELNs were added to the macrophage after LPS exposure. During the 24 h culture period, GELNs were applied to LPS-induced RAW264.7 cells, and the following two observations were made. Based on analyses of macrophage-associated cellular markers, GELNs supplementation led to a significant expression of M1 macrophage markers (CD86, Mincle, iNOS, and IRF-5) with a concurrent increase of classical M2 markers (Ym1, Arg-1, CD206, and IRF-4) (Fig. 2B and C), suggesting that GELNs treatment could convert M1 macrophages to the anti-inflammatory M2 phenotype. Consistent with these findings, flow cytometric analysis showed that macrophages pretreated with GELNs induced M2 polarization, whereas M1 polarization was decreased (Fig. 2D). As expected, GELNs exerted a blunt inhibitory effect on mRNA expression of TNF-α, IL-6, and IL-1β (Fig. 2E). Moreover, IL-10 mRNA levels were found to be significantly increased by GELNs treatment (Fig. 2E). Similarly, GELNs were most effective in reducing the secretion of pro-inflammatory cytokine (TNF-α, IL-6, and IL-1β), and to some extent increasing anti-inflammatory cytokine (IL-10) (Fig. 2F). Taken together, although GELNs treatment elicited pleiotropic responses in LPS-induced macrophages, the pattern of immune responses in the GELNs-treated group exhibited a significant bias towards M2 polarization (Fig. 2G).

Fig. 2.

GELNs modulate M2 activation of macrophages. (A) Experimental design. (B–C) The expression levels of M1 and M2 macrophage markers in GELNs-treated RAW264.7 cells in the presence of LPS. (D) Flow cytometry of the expression levels of the M1 polarization marker and M2 polarization marker in LPS-induced RAW264.7 cells after culturing with GELNs for 24 h. (E) RT-qPCR analysis of the relative mRNA levels of pro-inflammatory cytokines (TNF-α, IL-6, and IL-1β) and anti-inflammatory cytokines (IL-10) in GELN-treated RAW264.7 cells in the presence of LPS. (F) The concentrations of inflammatory cytokines were detected by ELISA. (G) Schematic images of the reprogramming of macrophage polarization, activated by GELNs.

Delivery of GELN-derived miRNAs to macrophages

The transfer of exosomal miRNAs is emerging as a signaling intermediate that can be utilized to regulate the functions of host cells [4], [8]. To investigate whether GELNs-derived miRNAs might contribute to regulating the inflammatory response, bioinformatic analysis were performed separately for 19 miRNAs derived from GELNs, according to our published miRNA profile of GELNs [19]. The interaction of selected miRNAs with multiple genes was examined to form a network of miRNAs target genes involved in the inflammatory response. Fig. 3A shows the miRNA-mRNA-inflammatory pathway interaction network, suggesting specific miRNAs with the ability to regulate inflammatory response. Our focus was on Novel-40, Novel-45, osa-miR164d, stu-miR156f-5p, vvi-miR396b, and zma-miR396g-5p, which are five of the most abundant miRNAs in GELNs, according to our previous study [19]. To assess whether GELNs-derived miRNAs were delivered to macrophages, we incubated macrophages with GELNs within 12 h (Fig. 3B) and made the following two observations. On the one hand, the time-dependent endocytosis of PKH26-labelled GELNs by macrophages was confirmed via confocal microscopy (Fig. 3C). On the other hand, we examined the expression levels of novel-40, novel-45, osa-miR164d, stu-miR156f-5p, vvi-miR396b, and zma-miR396g-5p in RAW264.7 cells. The results revealed a significantly higher abundance of osa-miR164d and zma-miR396g-5p was significantly higher in macrophages exposed to GELNs than in those exposed to PBS (Fig. 3D). This observation was supported by findings showing that grapefruit-derived exosome-like nanoparticles can deliver numerous mir168A-5P to rat enterocytes [23]. As shown in Fig. 3E, the secondary structures of GELNs-derived osa-miR164d and zma-miR396g-5p have stem-loop, and their ΔG values were negative. The predicted secondary structures of osa-miR164d and zma-miR396g-5p were used to help elucidate the relationship between miRNA sequence and stability. Overall, these data suggested that GELNs can deliver miRNAs to macrophages. Additionally, specific miRNAs derived from GELNs may be involved in the regulation of the inflammatory response.

Fig. 3.

Macrophages take up GELNs-derived mRNAs. (A) Construction of the miRNA-target Gene-Inflammatory pathway interaction network. (B) Experimental design. (C) GELNs were taken up by macrophage. Original magnification was 200× (D) Expression levels of GELNs-derived miRNA in RAW264.7 cells. (E) Predicted secondary structures of GELNs-derived miRNAs.

GELNs-derived osa-miR164d attenuates inflammatory responses and favors the reprogramming of macrophage polarization

To determine the function of GELNs-specific miRNAs in macrophages, we transfected specific miRNA mimics into macrophages and evaluated the functional role of GELNs-derived miRNAs in shaping macrophage polarization. According to the results from Fig. 3D, we selected stable miRNAs (osa-miR164d and zma-miR396g-5p) and unstable vvi-miR396b as the candidate miRNAs that potentially affected the M2 macrophage polarization. Macrophages were induced by LPS and transfected with or without GELNs-specific miRNAs, as depicted in Fig. 4A. The osa-miR164d and zma-miR396g-5p effectively suppressed the mRNA expression of M1-specific marker genes (Fig. 4B). The administration of osa-miR164d and zma-miR396g-5p significantly upregulated the expression of M2-specific marker genes (Fig. 4C). In line with the aforementioned findings, flow cytometry results showed that macrophages pretreated with osa-miR164d induced M2 polarization, while M1 polarization was decreased (Fig. 4D). Osa-miR164d and zma-miR396g-5p from GELNs significantly inhibited the mRNA levels of IL-1β, IL-6, and TNF-α in LPS-induced macrophages (Fig. 4E). In addition, ELISA data revealed that treatment of macrophages with osa-miR164d and zma-miR396g-5p increased TGF-β and IL-10 levels (Fig. 4F). Moreover, osa-miR164d reduced secretion of pro-inflammatory cytokines (TNF-α, IL-1β, and IL-6), whereas zma-miR396g-5p resulted in no significant suppression of TNF-α and IL-6 levels (Fig. 4G). Overall, the results suggested that osa-miR164d can reverse the inflammation-related response associated with the M2 phenotype (Fig. 4H).

Fig. 4.

GELNs-derived osa-miR164d mediates macrophage polarization and attenuates inflammatory responses. (A)Schematic images for the treatment of GELNs-derived miRNAs on LPS-induced RAW264.7 cells. (B and C) The relative expression levels of the M1 and M2 macrophage surface markers were detected by RT-qPCR. (D) The expression levels of the polarization markers CD86 and CD206 in the LPS induced-RAW264.7 cells after culturing with mimic miRNAs for 24 h were assessed using flow cytometry. (E) Quantification of mRNA levels of pro-inflammatory in PBS- or miRNAs-treated LPS-induced RAW264.7 cells. (F and G) ELISA analysis showing the levels of TGF-β inflammatory cytokines in RAW264.7 cells transfected with GELNs-derived miRNAs. (H) Schematic diagram of the shift of macrophage polarization between M1 and M2 with the regulation of GELNs-derived osa-miR164d.

Analysis of osa-miR164d/TAB1 mRNA 3′UTR molecular interactions

Transcriptome analysis was conducted to uncover the potential molecular mechanisms of GELNs-derived osa-miR164d in promoting macrophage polarization from M1 to M2. As indicated in Fig. 5A, osa-miR164d was added to the macrophage after LPS exposure. Venn diagram analysis indicated that the LPS + osa-miR164d group exhibited a lower number of unique differentially expressed genes (DEGs) than the LPS group (Fig. 5B). Volcano plots revealed a total of 1642 significant DEGs. Specifically, the osa-miR164d group exhibited an upregulation of 1064 genes and a downregulation of 578 genes (Fig. 5C). Our results revealed that the genes related to inflammatory and immune responses were up-regulated after treatment with osa-miR164d (Fig. 5D). To identify the biological changes induced by osa-miR164d in pro-inflammatory macrophages, we classified DEGs into three categories based on Gene Ontology (GO) analysis: biological process, molecular function, and cellular component. The top 30 terms were displayed in the Fig. 5E. DEGs in the LPS + osa-miR164d group were enriched in inflammatory responses according to GO biological process statistical analysis. Furthermore, osa-miR164d affected TNF signaling pathways, cytokine-cytokine receptor interaction, mitogen-activated protein kinase (MAPK) signaling pathways (Fig. 5F). These pathways were commonly signaling pathways associated with macrophage polarization, according to KEGG statistics. Based on a comprehensive analysis of transcriptomic statistics (Fig. 5) and the interactions between miRNA, target genes, and inflammatory pathways (Fig. 3A), we selected the TNF pathway for further investigation.

Fig. 5.

Mechanisms underlying Osa-miR164d's effects M1 macrophages toward M2. (A) Transcriptomics analysis was conducted in the total mRNA of macrophages, which was extracted from the LPS group and LPS + osa-miR164d group. (B) Venn diagram. (C) Volcano plots showing different expression genes. (D) Heatmaps illustrating the genes involved in inflammation that were either significantly upregulated or downregulated after osa-miR164d treatment. (E) GO enrichment analysis. (F) KEGG enrichment analysis.

MiRNAs directly repress targeted transcripts that possess a specific miRNA binding site in their 3′UTR. To elucidate the downstream mechanisms of osa-miR164d in mediating M2 activation of macrophages, we integrated target prediction (Fig. 3A) and KEGG data (Fig. 5F) for analysis. It is noteworthy that TAB1 was predicted to be a direct target gene of osa-miR164d. Furthermore, we constructed the wild-type TAB1 luciferase reporter construct with the TAB1 3′UTR to verify the relationship between miRNA and the target gene (Fig. 6A). After transfecting osa-miR164d mimics along with TAB1-WT into RAW264.7 cells, it was observed that osa-miR164d mimics suppressed the luciferase activity in TAB1-osa-miR164d-transfected cells compared to the TAB1-NC group (Fig. 6B). Additionally, NF-κB was found to be a downstream effector of osa-miR164d, regulating the TNF signaling pathway to rebalance macrophage polarization (Fig. 6C). Furthermore, osa-miR164d inhibited the activation of IL-1β in LPS-treated macrophages (Fig. 6D). Taken together, these findings were the first to confirm that GELNs-derived osa-miR164d can specifically and stably bind to TAB1 and modulate macrophage polarization by activating the TNF pathway, thereby playing an important role in causing a cascade change and profound downstream effects on the development of the intestinal inflammation (Fig. 6E).

Fig. 6.

osa-miR164d mediates macrophage polarization by the inhibition of TAB1. (A) Schematic diagram of the putative binding sites of osa-miR164d and 3′UTR of the TAB1 gene. The osa-miR164d seed matched in the 3′UTR of the TAB1 gene is mutated at the positions indicated. wild type (WT), mutant (MUT). (B) Relative luciferase activity of 3′UTR-TAB1-luciferase constructs in RAW264.7 cells after transfection of osa-miR164d mimics. The osa-miR164d seed matched the 3′UTR of the TAB1 gene is mutated at the positions indicated. (C–D) The mRNA levels of IL-1β and NF-κB. (E) Schematic illustration of the signaling pathways relevant to the reprogramming macrophage polarization, activated by GELNs-derived osa-miR164d.

Osa-miR164d-loaded bionic exosome mimicking GELNs mitigates inflammation

We wanted to investigate whether the enriched osa-miR164d in GELNs could influence macrophage polarization. For this purpose, we first synthesized osa-miR164d-loaded bionic exosomes mimicking GELNs (osa-miR164d-MGELNs), followed by treatment with LPS-induced macrophage. GELNs were heterogeneous biological particles enclosed by a phospholipid bilayer [24]. To construct osa-miR164d-MGELNs, we examined the total lipid composition of GELNs using lipidomic profiling assays (Fig. 7A). Digalactosyldiacylglycerol (DGDG) and monogalac-tosyldiacyglycerol (MGDG) are important glycolipids. The lipid data revealed that GELNs were enriched with phosphatidic acids (PC) (∼47.98 % of total lipids), phosphatidic acid (PA) (∼37.38 % of total lipids), DGDG (∼7.13 % of total lipids), and MGDG (∼6.89 % of total lipids). Therefore, we selected the PC derivative to construct osa-miR164d-MGELNs for future study.

Next, we synthesized osa-miR164d-MGELNs using a microfluidic mixing method that involved DLin-MC3-DMA, cholesterol, DSPC (i.e. PC derivative), and DMG-PEG-2000 (Fig. 7B). The bionic exosomes mimicking (MGELNs) exhibited particle sizes of ∼150 nm, and we observed an increase in particle size after loading with osa-miR164d (Fig. 7C). Osa-miR164d-MGELNs exhibited a negative zeta potential (Fig. 7C), similar to that of GELNs (Fig. 1C). Representative TEM imaging revealed that the osa-miR164d-MGELNs and MGELNs had a cup-shaped structure with a diameter of ∼200 nm (Fig. 7E). The particle size was consistent with the dynamic light scattering results, which measured a size of about 200 nm. The encapsulation efficiency of osa-miR164d was assessed by agarose gel electrophoresis. A high encapsulation efficiency (94 %) was observed for the osa-miR164d-MGELN (Fig. 7D).

As shown in Fig. 7F, the osa-miR164d-MGELNs did not alter cell viability (>96 %). To confirm the intracellular uptake of the miR164d-MGELNs, we further assessed the intracellular amount of osa-miR164d delivered by MGELNs. Using RT-qPCR, we identified increased levels of intracellular osa-miR164d in macrophages after 24 h incubation. Additionally, we observed a significantly higher abundance of miRNAs in macrophages exposed to miR164d-MGELNs compared with those exposed to osa-miR164d mimics (Fig. S2). We further investigated whether osa-miR164d-MGELNs exhibit a potent anti-inflammatory effect. Compared with the MGELNs group, miR164d-MGELNs treatment further greatly suppressed the secretion of IL-β and IL-6, along with an increase in the secretion of IL-10 (Fig. 7F). These results confirmed that osa-miR164d-MGELNs could efficiently reduce the degree of inflammation.

Osa-miR164d improves colitis and alters the macrophage phenotype

To further investigate whether osa-miR164d promotes M1 reprogramming to ameliorate colitis development, osa-miR164d-MGELNs containing 9 μg of miRNA were administered directly to mice with DSS-induced colitis by gavage for 8 days (Fig. 8A). Interestingly, a decreased weight loss was observed in the osa-miR164d group compared to the DSS–MGELNs and DSS groups (Fig. 8B). The DAI score indicated that the mice in the osa-miR164d group exhibited a significant reduction in disease sever (Fig. 8C). Additionally, osa-miR164d-MGELNs groups exhibited longer colons and less severe colonic damage after DSS treatment (Fig. 8D).

Fig. 8.

Supplementation with Osa-miR164d in DSS-induced colitis mice restores M2 macrophage content and alleviates intestinal inflammation. (A) Schematic diagram for DSS treatment, osa-miR164d-MGELNs, and MGELNs administration. (B)The change in body weights of mice. (C) DIA scores. (D) Optical photographs of the colon and colonic length statistics. (E) H&E staining images of colonic tissue. (F) Representative images of colonic tissue stained with CD11b. (G) The levels of inflammatory cytokines and TGF-β in colonic tissues were tested by ELISA.

H&E image of the osa-miR164d-MGELNs group showed less inflammatory cell infiltration in lamina propria and a more intact epithelial and goblet cell structure compared with the DSS-MGELNs group, similar to that of the healthy mice (Fig. 8E and Fig. S4). Macrophages at the lamina propria of the colon site showed significantly higher levels of CD11b expression in the DSS and MGELNs group than in the control group, which was mitigated in the osa-miR164d-MGELNs group (Fig. 8F and Fig. S5), demonstrating that the effects on the osa-miR164d-MGELNs may be associated with macrophages polarization. In addition, we observed reduced expression levels of pro-inflammatory cytokines, such as IL-6 and IL-1β, in the osa-miR164d groups, accompanied by increased expression levels of IL-10 and TGF-β. No difference was observed in the DSS-MGELNs group compared to the DSS groups (Fig. 8G). Collectively, these data suggested that osa-miR164d has the potential to improve colitis.

Discussion

Exosomes can transport miRNAs cargo between cells and play a role in human health [8], [20]. GELNs have emerged as potentially important contributors for alleviating the intestinal inflammation [9], [17], [19]. Additionally, GELNs have been explored as potential therapeutic agents for inflammatory-related diseases, including COVID-19, with targeted delivery to the lungs [25]. Dietary miRNAs play key roles in metabolism and inflammatory response [6], [26], [27]. Altogether, the above-mentioned evidence highlighted the relevance of administrating GELNs as a novel therapeutic strategy for inflammation. Nevertheless, the underlying mechanisms of miRNA in GELNs were not well understood. We reported that GELNs functioned as efficient nanoplatforms capable of successfully delivering miRNAs to macrophages. This provides new evidence that PELNs are promising vectors to transfer plant-derived active miRNA to mammalian immune cells. Further, GELNs-derived osa-miR164d modulated the M2 activation process with TAB1 as a direct target gene. To the best of our knowledge, we are the first to show that GELNs-derived osa-miR164d attenuates intestinal inflammation through macrophage polarization. These findings indicate the role of miRNAs transported within PELNs in modulating the host's immunomodulation through cross-kingdom regulation. In addition, we have designed biomimetic exosomes that mimic GELNs to deliver osa-miR164d. Importantly, the osa-miR164d-MGELNs skew the M1 towards M2-like phenotype, significantly ameliorating inflammation in vitro and in vivo.

Growing evidence reveals an emerging paradigm with crosstalk between receptor cells and PELNs in the gut microenvironment [9], [28]. We found that administered GELNs facilitated the transition of macrophages from M1 to M2 in vitro, which inhibited NF-kB and induced an array of proinflammatory cytokines (Fig. 2). Recent studies have shown that plant nanovesicles can be absorbed by mammalian cells [23], [29], [30]. GELNs were spontaneously taken up by macrophages in a time-dependent manner (Fig. 3B). Moreover, owing to the presence of GELNs–derived miRNA in macrophages (Fig. 3C), GELNs transported miRNAs to macrophages. Therefore, we speculated that GELNs rebalanced macrophage polarization to attenuate inflammation by delivering functional miRNAs.

Considering the transportability of exosomes, the role of miRNAs in PELNs has received increased attention. Meanwhile, the widespread application of new technologies has led to rapid progress in RNA research, positioning miRNAs at the forefront of disease genomics research [30], [31]. Plant dietary miRNAs can reach organs and modulate mammalian gene expression through cross-kingdom regulation [9], [32]. Another research provided evidence that vascular endothelium derived-exosomal miR-224-5p and miR-361-3 repressed the upregulation of NR4A3 and PCSK9 genes and improved endothelial functions [24]. We have previously characterized the GELNs used in this study in terms of a miRNA Profile by small RNA sequencing [19]. Notably, specific miRNAs derived from GELNs were stabilized in macrophages in our work. We selected stable miRNAs (osa-miR164d and zma-miR396g-5p) and unstable vvi-miR396b as the candidate miRNAs that potentially effected the M2 macrophage polarization. We demonstrated that the biological effects of GELNs–derived osa-miR164d on immune homeostasis could be amplified through the conversion of M1 macrophages into the anti-inflammatory M2 phenotype (Fig. 4).

We reported that GELNs lipids contain typical plant lipids such as PC, PA, DGDG, and MGDG, as shown in Fig. 7A, consistent with the result reported by Sundaram et al. (2019). [24]. In particular, the PE and PC-enriched outer layer was demonstrated to guide PELNs to target M1-type macrophages via the clathrin-dependent pathway [33]. Overall, GELNs potentially targeted M1-type macrophages. In this study, we designed an MGELNs packaged osa-miR164d system to facilitate the delivery of miRNA to M1 macrophages. Meanwhile, it is confirmed that the osa-miR164d can be specifically sent to the macrophages by MGELNs. Additionally, the osa-miR164d-MGELNs skew the M1 towards M2-like phenotype, significantly ameliorating inflammation in vitro. Another important finding is that the oral administration of osa-miR164d-MGELNs effectively provoked macrophage polarization to M2 phenotype, and downregulated inflammatory responses.

Even though osa-miR164d exerted the attenuation of intestinal inflammation and reprogramming macrophage polarization, the exact mechanism required further elucidation. According to the KEGG pathway mapping and miRNA-target gene-inflammatory pathway interaction network analysis, we selected TNF pathways as potentially associated with intestinal inflammation. Further investigation through bioinformatic bioinformatics analysis and dual luciferase assays revealed TAB1 as the target gene of osa-miR164d derived from GELNs. Osa-miR164d-targeted TAB1 exhibits a tight relation with TNF signaling. TAB1 plays an important role in the innate immune response [33], and the TAK1-TAB1 complex can stimulate NF-κB signaling pathways [34]. The downregulation of TAB1 favored the promotion of M1 toward M2 [35]. In addition, we have developed an osa-miR164d-loaded bionic exosome mimicking GELNs, which can ameliorate colitis by reprogramming macrophage polarization. One study reported that ginger-derived exosomal miRNAs contributed to improving mouse colitis via shaping gut microbiota [9]. Our data suggested that osa-miR164d regulated the expression of target gene TAB1, followed by regulating TNF signaling in macrophage polarization. These findings suggested that dietary GELNs–derived osa-miR164d is a novel functional component of food and Chinese medicine. Additionally, our findings on the mechanism of action of GELNs–specific miRNAs have implications for other known edible PELNs. An understanding of the interplay between plant diet and immune cells may thus provide new therapeutic strategies for the treatment and/or prevention of diseases due to dysregulation of gut immune homeostasis.

Unavoidably, certain limitations exist in our work. The primary constraint on this research is the possibility that osa-miR164d targets several genes involved in modulating macrophage polarization. Specifically, this study only investigated the effect of TAB1, excluding the other genes.

Conclusion

Our findings revealed, for the first time, that GELNs-derived osa-miR164d is a regulator of reprogramming macrophage polarization, ultimately alleviating intestinal inflammatory response. Mechanistically, GELNs-derived osa-miR164d can inhibit inflammatory response by targeting TAB1. In addition, the biomimetic GELNs loaded with osa-miR164d were developed and can be utilized as a safe therapy to alleviate intestinal inflammation by regulating the balance between M1 and M2 macrophages. Our results provide better evidence of GELNs-derived osa-miR164d as a regulatory molecule for cross-kingdom communications and open avenues for the engineering of GELNs by allowing the loading of specific miRNAs.

Compliance with ethics requirements

The experiments involving animals were approved by the Institutional Animal Care and Use Committee of Hefei University of Technology (no. HFUT22166).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary material

The following are the Supplementary data to this article: