| Part Ⅰ. Endomicroscopy will track injected mesenchymal stem cellsBackground & aims:Mesenchymal stem cells (MSC) is pluripotent stem cells which can exert in vivo tissue repairing and immunomodulatory effects. MSC have demonstrated significant potentials for the treatment of inflammatory bowel disease (IBD). The in vivo fate of transplanted MSC can be highly heterogeneous and depends on the source of MSC, transplantation route, receptor’s immune status, and previous transplantation history. However, clinical feasible methods to individually document the MSC recruitment to intestinal mucosa is lacking. Conventional radiology methods cannot efficiently discriminate between the debris of dead cells and live MSC. And they cannot evaluate the MSC homing efficiency to the intestinal mucosa. Confocal laser endomicroscope is a newly invited endoscope that can non-invasively provide the in vivo histology with*1000 magnification and cellular resolution. The illuminating and receiving wavelength of endomicroscope was 488nm and 505～585nm respectively. Thus the endomicroscope was compatible with the fluorescent feature of enhanced green fluorescent protein. Here we proposed that endomicroscopy could non-invasively track MSC in vivo at cellular resolution.Method:Isolated Sprague-Dawley rat MSC (rMSC) was characterized and fluorescently labelled by Carboxyfluorescein succinimidyl ester (CFSE). Then the labelled rMSC was transplanted to DSS induced colitis rat model and then imaged ex vivo using endomicroscope. The eGFP-MSC was transplanted to DSS induced colitis rat models through peritoneal or venous injection. Then eGFP labelled MSC was tracked in vivo, and acquired images were compared to immunofluorescence, immunohistology and fluorescent in situ hybridization results. The endomicroscopic features of eGFP-MSC on the receptors’mucosa was summarized. Then we further compared the MSC s homing efficiency to the receptor’s inflamed colonic mucosa under different conditions. Four transplantation model were made, 1) rat-rat allograft model,2) human-rat xenograft model,3) hypoxia treated rateGFP-MSC allograft model,4) immunosuppressed receptor by busulphan and allograft model. All four groups were evaluated by endomicroscope, and the eGFP-MSC was tracked in vivo. An automatic image processing program was developed to count the MSC automatically. And the frequency of MSC on endomicroscopic images was calculated.Results:Rat bone marrow MSC was successfully isolated from the bone marrow of donor rats. The MSC can differentiate to bone tissue, adipose tissue and cartilage tissue under proper conditions. Flow-cytometry confirmed that the MSC express CD29 (+), CD 44 (+), CD73 (+), and CD90 (+), while they do not express CD34 (-), CD45 (-)and CDllb/c (-). This was consistent with the definition of rat MSC. Endomicroscopy visualized clearly the eGFP labelled or CFSE stained MSC ex vivo. And the endomicroscopic features matched that on fluorescent microscope. Endomicroscopy using the FIVE1 system could track eGFP-MSC with distinct in vivo features. The in vivo eGFP-MSC was represented by bright, round and heterogenous nucleus. Immunofluorescence, immunohistochemistry and fluorescent in situ hybridization confirmed the presence of eGFP positive cells. In vivo endomicroscopy could quantify the transplanted MSC that homed to colonic mucosa of the recipient rat in multiplemodels, including the rat-to-rat allograft, human-to-rat xenograft, hypoxia induced MSC, and busulphan immunosuppressed recipient rat models. The stem cell quantification can be assisted by the automatic image processing programme. After hypoxia induction, there was a trend of enhanced rMSC homing to the inflamed mucosa as visualized by endomicroscopy (114.1 in hypoxia group vs 34.3 in other three group combined, t =2.14, P=0.0644).Conclusion:Endomicroscopy is a novel and promising tool to track transplanted MSC to the colonic mucosa. This clinical available non-invasive cellular tracking method may provide new insight to individualize each recipient’s regimen in the future. Significance:This study was the first to use endomicroscopic equipment to track stem cells in vivo. And this method provided cellular level resolution. In comparison to previous radiological tracking methods, CLE will not be interfered by the debris of dead cells and only the live MSC can be imaged and quantified. Our method has obvious superiority in stem cell tracking at the gastrointestinal mucosa.Part Ⅱ. Human colorectal mucosal microbiota correlates with its host niche physiology revealed by endomicroscopyBackground & aims:A huge amount of microbes inhabits the intestinal of human. In the recent years, the development of second generation sequencing promoted people’s understanding of gut microbial composition. The human gut microbiota plays a pivotal role in the maintenance of health, but how the microbiota interacts with the host at the colorectal mucosa is poorly understood. Previous study had reported that the colorectal mucosal microbiota was affect by the physiological status of host mucosa. For example, the oxygen pressure could significantly affect the energy metabolism of intestinal microbiota. The intestinal microbiota is likely to be affected by multiple host physiological factors simultaneously. However, no one has correlated the mucosal microbiota with multiple physiological factors. This was probably limited by the clinical feasibility, that multiple physiological indexes would require the corresponding number of biopsy samples from the same patient. Confocal laser endomicroscope was a novel equipment that entered clinical practice in recent years. It can non-invasively image in the receptor’s gastrointestinal tract. The real-time image provided by endomicroscope can reflect multiple physical factors of the colorectal mucosa. In this study, we proposed that confocal laser endomicroscopy (CLE) might help to untangle this relationship by providing in vivo physiological information of the mucosa. This help to understand the rules of host-microbiota interaction.Method:During November 2013 to April 2014, recipients were recruited to participate this study. The enrolled recipients were examed by CLE. For each patient, the whole colon was first examed by white light endoscopy, then the diseased lesion and a nearby normal control lesion were examed by probe based CLE. We used CLE to evaluate the in vivo physiology of human colorectal mucosa. Nine features of the colorectal mucosal morphology and physiology were scored according to the CLE images. They are 1) density of goblet cells,2) crypt enlargement,3) asterisk crypt,4) tubular crypt,5) epithelial leakage to fluorescein,6) thick epithelium,7) fluorescein leakage into crypt,8) fluorescein leakagethrough the vessel, and 9) infiltrated cell mass. The scored sites were then target biopsied, and the mucosal microbiota was quantified using 16s rDNA pyrosequencing. The raw sequencing data was processed, and each speices’abundance in each sample was calculated. The relationship between microbial composition, biopsy locus, and disease statues was explored in diversity analysis, cluster analysis, primary component analysis, non-redundancy analysis, and LEfSe analysis. The correlation network between colorectal mucosal microbiota and morphological or physiological features was calculated and visualized. The mucosal microbial metagenome was further predicted using PICRUSt. The metagenome’s correlation between the mucosal physiology was analyzed in network analysis.Results:A total of 267 recipients were screened, and 39 of them were taken into the final data analysis. These samples included 25 colorectal adenoma,5 ulcerative colitis, and 39 normal sites under white light endoscope. A total of 554,070 reads were finally analyzed after sequence de-noising, trimming and chimera picking. We found no significant differences in microbial diversity among disease types. The intra-adenoma distance and the inter-class distance were significantly higher than the intra-normaldistances (p<0.05 for both). The human mucosal microbiota agglomerated to three major clusters dominated by Prevotella, Bacteroides and Lactococcus. The mucosal microbiota clusters did not significantly correlate with the disease status or biopsy sites but closely correlated with the mucosal niche physiology, which was non-invasively revealed by CLE. Clear visualization showed that the mucosalmicrobiota formed two major mutualistic subnetworks. The correlationswere all positive within each subnetwork, and the correlations between different subnetworks were allnegative. Inflammation tilted two subnetworks within the mucosal microbiota. Fusobacterium was most closely related to the mucosal physiological factors but showed little relationship to the rest of the microbial community. Infiltration of inflammatory cells significantly correlated with multiple components in the predicted metagenome, such as the VirD2 component of the type Ⅳ secretory pathway.Conclusion:Our data suggest that a close correlation exists between the mucosal microbiota and the colorectal mucosal physiology, and CLE is a clinically available tool that can be used to facilitate the study of the in vivo correlation between colorectal mucosal physiology and the mucosal microbiota.Significance:This study is the first to use endoscopic equipment in gut microbiota study. Our method provided a new and clinical feasible method to study the host-microbiota interaction. The correlation between the mucosal microbial metagenome and the mucosal physiology identified new targets for future mechanistic research. |
PDF Full Text Request