Adipose tissue‐derived mesenchymal stem cells' acellular product extracellular vesicles as a potential therapy for Crohn's disease

Abstract The breakdown of gastrointestinal tract immune homeostasis leads to Crohn's disease (CD). Mesenchymal stem cells (MSCs) have demonstrated clinical efficacy in treating CD in clinical trials, but there is little known about the mechanism of healing. Considering the critical roles of macrophage polarization in CD and immunomodulatory properties of MSCs, we sought to decipher the interaction between adipose‐derived MSCs and macrophages, including their cytokine production, regulation of differentiation, and pro‐/anti‐inflammatory function. RNA extraction and next generation sequencing was performed in adipose tissue from healthy control patients' mesentery (n = 3) and CD mesentery (n = 3). Infiltrated macrophage activation in the CD mesentery was tested, MSCs and extracellular vesicles (EVs) were isolated to compare the regulation of macrophage differentiation, cytokines production, and self‐renewal capacities in vitro. CD patients' mesentery has increased M1 macrophage polarization and elevated activation. MSCs and their derived EVs, isolated from inflamed Crohn's mesentery, leads to a rapid differentiation of monocytes to a M1‐like polarized phenotype. Conversely, MSCs and their derived EVs from healthy, non‐Crohn's patients results in monocyte polarization into a M2 phenotype; this is seen regardless of the adipose source of MSCs (subcutaneous fat, omentum, normal mesentery). EVs derived from MSCs have the ability to regulate macrophage differentiation. Healthy MSCs and their associated EVs have the ability to drive monocytes to a M2 subset, effectively reversing an inflammatory phenotype. This mechanism supports why MSCs may be an effective therapeutic in CD and highlights EVs as a novel therapeutic for further exploration.


| INTRODUCTION
Tissue damage provides a need for the host to neutralize invasive microorganisms, clear the damaged cell and debris sites, and then restore and rebuild the tissue (Faz-Lopez et al., 2016). Therefore, the initiation of an inflammatory defensive reaction has developed to facilitate leukocyte and macrophage migration from the systemic circulation to local damaged tissue, removal of invasive pathogenic agents, and initiate tissue repair and functional recovery (Laskin et al., 2011;Wynn & Vannella, 2016). Ideally, the inflammatory response is an independent mechanism leading to complete resolution of inflammation and a rapid return to tissue homeostasis (Schett & Neurath, 2018;Sugimoto et al., 2016). In recent decades, inflammation resolution has been considered a passive operation; unique molecular and cellular mechanisms involved in inflammation resolution have been identified in which macrophages play an important role in preventing excessive immune responses Serhan et al., 2007;Sugimoto et al., 2016). Indeed, after damage to tissue, phagocytosis of apoptotic cells (efferocytosis) causes a functional transition to an anti-inflammatory transcript program, which is close to the alternative macrophage activation pattern, with cytokine production, growth factors, and lipid mediators necessary for the reintroduction of homeostasis (Arienti et al., 2019;Fadeel et al., 2010).
In the healthy intestinal lamina propria, macrophages are the most abundant mononuclear phagocyte to facilitate essential immune functions . Following the microenvironment changes, these macrophages are polarized into two classic macrophage phenotypes, a proinflammatory (M1) and anti-inflammatory macrophage (M2) populations (Atri et al., 2018). During an inflammatory response, both phenotypes may turn into one another.
The M1 and M2 cells are believed to be able to conduct doublehanded activities either by inducing inflammatory effects to enhance inflammation or by anti-inflammatory functions to facilitate healing (Alvarez & Liu, 2016). Therefore, the immune system responds according to the macrophage phenotype, with or without inflammatory or noninflammatory cytokines, which is highly necessary for inflammation progression and forecast (Labonte et al., 2014). Under normal conditions, inflammatory pathways inhibit immunopathogenesis. Thus, imbalances between M1 and M2 switching are key problems that cause a number of disorders like Crohn's disease (CD) (Lee et al., 2018). In addition to preventing unnecessary inflammation of harmless commensal microorganisms, intestinal macrophages foster tolerance by favoring the expansion of mesenchymal stem cells (MSCs), particularly by developing their anti-inflammatory and regenerative function. MSC-based treatment is intended to resolve CD by adjusting the immune response to restore equilibrium and repair intestinal tissue damage including immune cell balance and microbiome diversity. Based on this data, a new perspective on the pathogenesis of CD and treatment methods could be given by understanding the function of macrophages and MSCs in the intestinal inflammatory response in CD specifically.

| Bioinformatics
Bioinformatic analysis was performed by the Cleveland Institute for Computational Biology. Sequencing reads generated from the Illumina platform were assessed for quality and trimmed for adapter sequences using TrimGalore! v0.4.2 (Babraham Bioinformatics), a wrapper script for FastQC and cutadapt. Reads that passed quality control were then aligned to the human reference genome (GRCh37) using the STAR aligner v2.5.1. The alignment for the sequences were guided using the GENCODE annotation for hg19. The aligned reads were analyzed for differential expression using Cufflinks v2.2.1, a RNASeq analysis package which reports the fragments per kilobase of exon per million fragments mapped (FPKM) for each gene.
Differential analysis report was generated using Cuffdiff. Differential genes were identified using a significance cutoff of q < 0.05.

| Pathway analysis
Pathway analysis of differentially expressed genes was performed with Qiagen's Ingenuity Pathway Analysis (IPA) software. All genes related to macrophages as identified by IPA's "Diseases and Functions" were sorted out of the main data set and a new Core gene expression analysis was run for all six groups. Focus was further placed on all the differentially expressed genes in "Activation of Macrophages" and "Migration of Macrophages" Functions. These genes were plotted as a heat map of the normalized FPKM z-score for all groups. A table of the gene's log fold change and predicted effect on the function is listed in Supporting Information: Tables 1 and 2. The predicted effect on the functions were calculated in IPA using a z-score statistical method. |z| > 2 were considered significant.

| MSC isolation and culture
To isolate MSCs, fresh tissues were minced and digested with 1 mg/ ml Collagenase from Clostridium histolyticum (Sigma-Aldrich) at 37°C for 30 min. The digest was passed through a 100 µm cell strainer and centrifuged for 5 min at 1200 rpm. The cell pellets were resuspended in red blood cell lysis buffer and incubated for 1-2 min at 37°C. THP-1 cells were cultured in RPMI plus 10% FBS and 1% penicillin/streptomycin (Cleveland Clinic Media Production Core).
Cells were cocultured with MSCs or EVs for indicated time points before harvest. Media were collected and centrifuged at 12,000 rpm for 5 min. Cells were scrapped in phosphate-buffered saline (PBS) and pelleted at 5000 rpm for 5 min.

| Extracellular vesicle isolation
MSC cultured media were collected from confluent cells (between passages 1-3) after 48-72 h. Media were centrifuged at 3000 rmp for 5 min to remove large cell debris and EVs were extracted and concentrated. Briefly, the cultured media were filtered through a 0.22 μm filter to remove cell debris and large vesicles, followed by ultracentrifugation at 30,000g for 20 min to pellet larger microvesicles. The supernatants were then subjected to ultracentrifugation at 120,000g for 3 h to sediment the MSC-derived EVs. Supernatant was discarded and the EV pellet was resuspended in 200 µl 0.1 µM filtered PBS. For the TEM sample, 50 µl of EV suspension was mixed 1:1 with 4% 0.1 µM filtered PFA in PBS and stored at 4°C until processing. Remaining 50 µl aliquots were stored at −80°C. Transmission electron microscopy (TEM) sample processing and imaging was performed by the Cleveland Clinic Imaging Core on 2% PFAfixed EVs. Zetaview analysis was performed by the Cleveland Clinic Flow Cytometry Core. Densitometry was performed and results were normalized to β-actin.

| Western blot
Normalized phosphorylated protein values were further normalized to total protein values.

| Enzyme-linked immunosorbent assay (ELISA)
ELISA plates were coated with 2 µg/ml capture (purified) antibodies overnight at 4°C. Wells were washed three times with TBST and blocked for 1 h with 1% bovine serum albumin. Block was removed and recombinant protein standards and samples were incubated for 2 h. Wells were washed three times and 2 µg/ml of biotynilated antibody was incubated for 1 h. Wells were washed three times and a 1:1000 dilution of HRP-Avidin antibody was incubated for 40 min.
Wells were washed three times and 75 µl TMB was added.
The reaction was stopped with 75 µl 10% sulfuric acid and plates

| Statistical analysis
Statistical analyses were carried out using GraphPad Prism v.7 or JMP v.12 (SAS Institute). The data were checked to confirm normality and that groups had equal variance. One-way analysis of variance (ANOVA) with Tukey's multiple comparison tests was employed to determine significant differences between sample groups. Results from these tests were reported as significant if p < 0.05, with results from these tests shown as mean ± SEM.  (Figure 2b,c). Under inflammatory conditions, the total protein content of MEK1/2 and ERK was increased.

| Inflamed MSC derived extracellular vesicles induce high protein expression of M1 polarization
The balance of tolerogenic or differentiation function from proinflammatory microenvironments is characterized by the balance of macrophage stability, especially given that these populations have proven to be in either M1 or M2 cohorts (Belkaid & Hand, 2014) | 3007 variation in the ability to regulate proinflammatory cells and their subsequential cytokine production may result in disease susceptibility (Balding et al., 2004;Cantor et al., 2005). Pathognomonic to CD is inflamed and hypervascularized mesentery which wraps around the bowel wall, aka creeping fat, and has recently been seen as an active player in the pathophysiology of CD. However, the reason why diseased mesentery plays an active role in disease progression remains unknown (Coffey et al., 2018;Ha et al., 2020). We herein found that both MSCs and their associated EVs derived from diseased mesentery characteristic of CD may contribute to an ongoing inflammation and disease progression due to a M1 proinflammatory phenotype.
Total RNA sequencing was used to assess the mRNA expression signature of macrophage related genes in adipose tissue from healthy patients, adipose tissue from CD subcutaneous layer, and adipose tissue from diseased Crohn's mesentery. Pathway analysis indicated an increase in the activation and migration of macrophages with distinct expression profiles in between healthy versus Crohn's subcutaneous tissue versus CD mesentery tissue. Results from other papers and our own previous investigation have reported that the elevated activity of the mitogen-activated protein kinase (MAPK) signaling cascade is found in the majority of mouse colon tissue and is known to regulate proliferation, survival and invasion of macrophage populations (Guereño et al., 2020;Hernandez-Padilla et al., 2020;Hernandez-Silva et al., 2020;Yi et al., 2020;Zhao et al., 2020). This process involves inflammatory factors, such as MEK and ERK.
Consistent with these findings, we found that MEK and ERK act as potential upstream regulators of several differentially expressed genes related to the activation and migration of macrophages and were able to confirm elevated protein levels of MEK and ERK in inflamed compared to noninflamed mesentery of CD patients.
While MSCs are known to be a safe and effective therapeutic for the treatment of perianal CD (Cho et al., 2015;de la Portilla et al., 2013;Garcia-Olmo et al., 2009;Pan et al., 2014;Panés et al., 2016), likely due to their anti-inflammatory and immunomodulatory properties which promote tissue repair (Kinchen et al., 2018;Mao et al., 2017;Soontararak et al., 2018;Turse et al., 2018;Zhang et al., 2017), the source of MSCs is critical to the phenotype and function of the cell.
When MSCs are extracted from diseased mesentery they appear to be proinflammatory, largely due to their effect on macrophage polarization. There are limitations to our study worth mentioning. First, we did not look to see if healthy MSCs and healthy MSC-derived extracellular vesicles reversed macrophage polarization in the setting of inflamed F I G U R E 5 Extracellular vesicles favor macrophage reprogramming during the resolution of inflammation. THP-1 monocytes were differentiated under the PMA conditions with extracellular vesicles treatment over 72 h. Culture supernatants were then collected and proinflammatory cytokine secretion were determined by selective ELISA. Levels of IL-6 (a), IL-12p40 (b), IL-12 (c), IL33 (d), IFN-a2 (e) are significantly increased in the diseased mesentery extracellular vesicles group compared to the other groups while TNF-a (f), TGF-b (g), and IL-10 (h) are significantly decreased in the diseased mesentery group. Results are mean ± SEM from four independent experiments. *p< 0.05. ELISA, enzyme-linked immunosorbent assay; IL-6, interleukin 6; PMA, phorbol-12-myristate-13-acetate; TNF-a, tumor necrosis factor-a.
states within an in vivo model. Second, we did not investigate whether the concentration of MSCs or extracellular vesicles affected macrophage polarization to understand if this finding was dose-dependent and thus suggesting a critical level of inflammation would be required to see the same results. Third, we did not investigate the systemic effects of the MSC and MSC-derived extracellular vesicles which would be important in future investigations.
In conclusion, diseased Crohn's mesentery exhibits a proinflammatory state through the polarization of macrophages to a M1 phenotype. This may be driven by MSCs in the resident tissue and their associated extracellular vesicles resulting in a proinflammatory phenotype.
Conversely, healthy donor-derived MSCs and their associated extracellular vesicles have the ability to drive monocytes to a M2 subset, effectively reversing an inflammatory phenotype. This mechanism supports why MSCs may be an effective therapeutic in CD and highlights extracellular vesicles as a novel therapeutic for further exploration.

AUTHOR CONTRIBUTIONS
We confirm that the manuscript has been read and approved by all named authors and that there are no other persons who satisfied the criteria for authorship but are not listed. We further confirm that the order of authors listed in the manuscript has been approved by all of us.
F I G U R E 6 Enhanced cleaved caspase 3, MEK, ERK, STAT3, and suppressed Arg-1 signaling in inflamed extracellular vesicles incubated macrophages. THP-1 cells were exposed to inflamed, subcutaneous or noninflamed MSCs or extracellular vesicles for 72 h in macrophage differentiation conditions before protein harvest and western blot analysis. (a) The levels of cleaved caspase 3, pMEK, MEK, pERK, ERK, STAT3, and ARG-1 after MSCs treatment were determined by immunoblotting. (b) The levels of cleaved caspase 3, pMEK, MEK, pERK, ERK, STAT3, and ARG-1 after extracellular vesicles treatment were determined by immunoblotting. Relative band intensities were quantified from five individual experiments and measured by densitometry analysis. Beta-actin was used as a loading control. Phosphorylated proteins were further normalized by total protein values. Mes, diseased mesentery; MSC, mesenchymal stem cells; Subq, subcutaneous; WO, without treatment. n = 3 patients. *p < 0.05.