Effects of different treatment of faecal microbiota transplantation techniques on ulcerative colitis in rats

Background: Ulcerative colitis (UC) is a chronic non-specic inammatory bowel disease with abdominal pain, mucus, pus, and blood in the stool as the main clinical manifestations. The pathogenesis of UC is still not completely clear, and multiple factors such as genetic susceptibility, immune response, intestinal microecological changes, and environmental factors together lead to the onset of UC. In recent years, the role of intestinal ora disturbance on the pathogenesis of UC has received widespread attention. Therefore, fecal microbiota transplantation(FMT), which changes the intestinal microecological environment of UC patients by transplantation of normal fecal bacteria, has attracted increasing attention from researchers. However, there are no guidelines at home and abroad to recommend fresh FMT or frozen FMT in the treatment of UC, and there are a few studies on this. Therefore the purpose of this experiment was to explore the effects of fresh and frozen fecal microbiota transplantation methods on the treatment of experimental ulcerative colitis models in rats. Results: Compared with the model control group, all faecal microbiota transplantation groups achieved better ecacy, mainly manifested as weight gain by the rats, improvements in faecal characteristics and blood stools, reduced inammatory factors, and normal bacterial ora. The ecacy of the frozen faecal microbiota transplantation group was better than that of the fresh faecal microbiota transplantation group in terms of behaviour and colon length . Conclusions: FMT is a feasible method for treating UC. The mechanism of action may be via competitive inhibition of pathogenic microorganisms, improved immune metabolism, and reduced inammatory response to mitigate the damage to the intestinal barrier and cause UC remission. Compared with fresh FMT, the therapeutic effect of frozen FMT may be greater.


Background
Ulcerative colitis (UC) is a chronic non-speci c in ammatory bowel disease involving abdominal pain, mucus, pus, and blood in the stool as the main clinical manifestations (1,2). The pathogenesis of UC is not completely clear, and multiple factors such as genetic susceptibility, immune response, intestinal microecological changes, and environmental factors together lead to the onset of UC (3). In recent years, the role of intestinal ora disturbance on the pathogenesis of UC has received widespread attention, as disruption of the gut bacteria can destroy the intestinal mucosa; in uence T cell subgroup differentiation; cause imbalances in Thelper (Th)l, Th2, and Th17, and regulatory T cells; and lead to secretion of a large number of in ammatory mediators, such as interleukin (IL)-1β, IL-5, IL-6, IL-17, and tumour necrosis factor alpha (TNF-α) (3,4). Additionally, changes in intestinal immune functions can occur, leading to UC and potentially to further complications. Changes in various intestinal ora genera play a key role in the incidence of UC (5,6). Thus, maintaining the balance of the intestinal ora is essential for alleviating the symptoms of UC. Therefore, faecal microbiota transplantation (FMT), which changes the intestinal microecological environment of patients with UC via transplantation of normal faecal bacteria, has attracted increasing attention (7,8).
At present, FMT is mainly used to treat diseases such as Clostridium di cile infectious diarrhoea (9,10), in ammatory bowel disease (7,11), irritable bowel syndrome (12,13), and non-alcoholic fatty liver (14,15). FMT has a clear effect on recurrent C. di cile infectious diarrhoea and is recommended by the European Society of Microbiology (16). The Joint Guidelines of the British Society of Gastroenterology and the Medical Infectious Society recommend that frozen FMT materials used to treat C. di cile infection should be considered as superior to fresh FMT preparations (level of evidence: high; recommended strength: strong) (17). However, there are no guidelines for recommending fresh FMT or frozen FMT in the treatment of UC, and these treatments have not been widely examined. Therefore, we explored whether fresh and frozen FMT methods were effective for treating experimental UC model rats.

Changes in general conditions of rats in each group
To evaluate the effect of FMT in chronic intestinal in ammation conditions similar to in human patients with in ammatory bowel disease, experimental enteritis in rats was induced by administering 2,4,6trinitrobenzenesulfonic acid (TNBS). After starting TNBS treatment, faeces from normal rats were repeatedly administered to UC model rats by gavage. Figure 1A shows the curve of the average body weight in each group. Before developing the TNBS-induced UC models, there was no signi cant difference between groups (P > 0.05). After model creation for 72 h, the weight of the rats in each model group decreased to a similar extent, with differences observed between the normal group and other groups (P < 0.05). After intervention, there were signi cant differences in body weight between the frozen FMT group, mesalazine group, and UC model group (P < 0.05), whereas there was no signi cant difference between the UC model group and fresh FMT group. There was no difference among the fresh FMT, frozen FMT, and mesalazine groups (P > 0.05).
After model development, each group began showing various degrees of malaise, arched back, yellow coat, diarrhoea, mucus pus, and blood in the stool, accompanied by slow weight gain or even weight loss.
After intervention, the activity status, bloody stool, and diarrhoea of the rats improved. The disease activity index (DAI) score is shown in Figure 1B.
We also measured the colon length. As shown in Figure 1D, the colons of each group showed varying degrees of dilatation, oedema, and even bleeding ulcers after model creation. However, dilation and oedema of the colon of rats in the three different intervention groups were not as obvious as those in the model group, and there was no obvious ulcer bleeding. According to the colon length statistics, as shown in Figure 1C, only the normal group signi cantly differed from the other four groups (P < 0.05). The remaining four groups showed no signi cant differences (P > 0.05).
Pathological changes and scores of colonic tissues of rats in each group The colon tissue of rats in the normal group showed a complete colonic epithelium, regular crypt structure, and small amount of in ammatory cell in ltration. The pathological changes in rats in the UC model group were as follows: obvious erosion and ulceration was observed in the mucosal epithelium; the number and structure of epithelial crypts were changed, and their arrangement was disordered; goblet cells were signi cantly reduced; high in ammatory cell in ltration was observed in the lamina propria, and a large amount of in ammatory cell in ltration and oedema was observed in the submucosa. However, in pathological sections of the fresh and frozen FMT groups, in ammatory in ltration and destruction of the intestinal wall were milder than those in the UC model group.

Levels of in ammatory factors in each group of rats
According to the intervention results of each group, treatment with FMT reduced intestinal in ammation. The expression of TNF-α in the colon of the two FMT groups was signi cantly lower than that in the model group (P < 0.05) and returned to the level in normal rats. There was no signi cant difference between the fresh and frozen FMT groups (P > 0.05).

FMT treatment improves intestinal ora in UC rats
The pair end reads obtained by Miseq sequencing were rst spliced according to the overlap relationship between the sequences, and the sequence quality was controlled and ltered. The samples are distinguished, after which cluster analysis and species taxonomy analysis were performed. Cluster analysis showed that various diversity index analyses could be performed and the depth of sequencing could be detected. Based on taxonomic information, community structure statistics were performed at each classi cation level.
Data corresponding to the richness and diversity of the gut micro ora are shown in Figure 4. The Chao1 index and Shannon index showed that FMT increased alpha diversity (P < 0.05).
The horizontal analysis of the phylum graph showed that Firmicutes, Bacteroidetes, Actinomycetes, and Proteobacteria were predominant ( Figure 5). After TNBS-induced induction of UC, the intestinal ora of rats differed from that of normal rats at the phylum level. The difference was manifested as a decreased abundance of Bacteroidetes, signi cantly increased relative abundance of Firmicutes, and increased abundance of Actinomycetes and Proteobacteria. After different interventions, at the phylum level, there were signi cant differences between the frozen FMT group and UC model group. The speci c performance was as follows: the relative abundance of Firmicutes decreased, whereas the relative abundance of Actinobacteria and Proteobacteria decreased but that of Bacteroides increased. The fresh FMT group showed a similar trend but did not signi cantly differ from the UC model group (P > 0.05).
At the genus level, the abundance comparison of a single species in each sample group is shown in Figure 6, and the Wilcoxon t rank test was used to detect differences between groups. This study showed that after model creation, the abundance of Bacteroides, Christensenellaceae_R_7_group, Fusicatenibacter, and Allobaculum increased, whereas Prevotellaceae_NK3B31_group, Ruminococcaceae_UCG_013, Ruminococcaceae_UCG_014, and Eubacterium_coprostanoligenes_group decreased. After FMT intervention, the abundance of Prevotella_9 increased, and the relative abundance of Coprococcus_2 and Subdoligranulum decreased.
In terms of β diversity, following intervention, the similarity of the composition of the fresh and frozen FMT group sample communities became more consistent with the normal group, as observed by principal coordinate analysis (Figure 7). In the community column chart shown in Figure 9, the composition of the colonic micro ora also showed similar results: the colonic micro ora composition of the frozen and fresh FMT groups gradually trended toward that in the normal group after intervention, which continued after stopping the intervention for one week.
To identify the speci c types of bacteria altered by treatment, we performed linear discriminant analysis effect size (LEfSe) analysis to determine the characteristics of different groups of the gut microbiota. The clade map produced by LEfSe analysis revealed species with signi cant differences in abundance between groups ( Figure 9). The abundance of Coprococcus 2 and Ileibacterium was signi cantly higher than that in the other groups, which were characteristic bacteria in the UC group. Fusobacteriaceae, c_Alphaproteobacteria, Anaerovibrio, Ruminococcaceae_UCG_008, and Prevotellaceae were characteristic bacteria in the frozen FMT group. Prevotella_9, Acidaminococcaceae, Peptostreptococcaceae, Arcobacter, Phascolarctobacterium, Eubacterium_hallii_group, Candidatus Stoque chus, Arcobacteraceae, Faecalibaculum, and Sulfurovaceae were characteristic bacterial genera in the fresh FMT group. The abundances of Lachnospiraceae NK4A136, Ruminococcaceae_UCG_005, and Romboutsia were signi cantly higher than those in the other groups, which are characteristic bacterial genera in the MS group.

Discussion
The pathogenesis of UC is complex, and it is currently thought that the interaction between the host and gut micro ora is a key factor. Under normal conditions, innate and acquired immunity in the host tolerates the normal micro ora while preventing the invasion of harmful bacteria. When the balance of intestinal ora is disrupted, harmful bacteria in the intestinal tract are rapidly increased, and directly invade and destroy intestinal epithelial cells, leading to immune dysfunction. The release of a large amount of enterotoxin increases the permeability of the intestinal mucosa and damages the intestinal mucosal barrier (18,19). Intestinal mucosal barrier function decreases and intestinal microbial ora shifts, further destroying the intestinal mucosal barrier, causing a vicious cycle, and exacerbating intestinal in ammation.
FMT is the process of transferring faecal bacteria from a donor to a recipient (20,21). In recent years, FMT has made great progress in the treatment of UC. It has been reported that patients with UC treated by FMT had a higher disease remission rate than those administered traditional treatment alone (14,22,23).
Moreover, the diversity of intestinal ora in patients treated with FMT was signi cantly increased compared with that before treatment, and the level of intestinal in ammatory factors decreased (24).
Faecal bacterial transplantation treatment often varies in the choice of the donor, different treatment of faecal bacteria, transplantation method, and method of the exact guide. We used fresh and frozen FMT in a rat model of UC to explore the differences between the two methods.
FMT was found to improve diarrhoea, mucous pus, and blood stool, and weight loss in rats. Histologically and pathologically, FMT effectively restored crypt injury, reduced intestinal in ammation and ulcer injury, and restored damaged villi. In addition, FMT effectively downregulated the levels of in ammatory factors. Moreover, the structure of intestinal ora was further analysed, which showed that after inoculation with exogenous faecal micro ora, the colonic micro ora composition of rats gradually became similar to that of the normal group. This is consistent with previous studies showing that after FMT, the intestinal ora composition of recipients and donors was consistent (25,26).
FMT increased the number of bene cial bacteria in the gut and decreased the number of harmful bacteria. These results suggest that the mechanism of action of FMT is competitive inhibition of pathogenic microorganisms, improved immune metabolism, reduced in ammatory response, improved tight junctions between colon cells, and A reduced damage response of the intestinal barrier to alleviate the progression of UC. This is biologically reasonable, as in the early stages of UC, uctuations in the microbiota are easier to recover (22).
However, compared with the fresh FMT group, the frozen FMT group showed a greater relief effect on UC model rats in terms of behaviour and colon length. In terms of histopathology, in ammatory factors in fresh and frozen FMT groups showed improvement in UC model rats, but the difference was not signi cant. A previous study showed that frozen FMT reduced the amount of gram-negative bacteria in faeces, which may explain why frozen faecal transplants are more effective than fresh faecal transplants (27).
Whether the microorganisms remain viable, particularly the bene cial bacteria, over time is unclear. Studies have shown that there is no difference in the ora of frozen and fresh faecal bacteria at 6 or 7 months. In addition, preservation of frozen bacteria requires further analysis. By studying the aerobic bacterial suspension of 10% glycerol stored at -80°C, Costello (28). The difference in the survival rate of different strains in cryopreservation may be clinically important for long-term preservation of FMT, which should be analysed in further studies.
In summary, we showed that frozen FMT can be used to treat UC, and the curative effect was similar to, or even better than that of fresh FMT. Previous studies on C. di cile infection have shown that frozen FMT does not signi cantly affect the implantation and e cacy of microorganisms (30). A random clinical trial showed that the lyophilised product had a slightly lowered e cacy in C. di cile infection patients compared with fresh product, but resembled other treatments in microbial restoration one month after FMT (31). Another clinical randomised controlled trial with a larger sample size showed that among adults with recurrent or refractory CDI, the use of frozen compared with fresh FMT did not negatively affect the clinical resolution of diarrhoea (32). This is important for practical applications, such as supplier relationships, modes of transport, and cost-effectiveness. Compared with fresh FMT, frozen FMT reduces the frequency of donor screening (32). This approach can be applied in a wide range of healthcare settings. At present, FMT banks have been established in foreign countries by using the faecal intelligent separation system (33). China also established the Chinese Fecal Microbiota Transplantation Bank in 2015, aiming to realise the emergency treatment of faecal microbiota in different locations and the standardisation of faecal bacteria distribution, preparation, and preservation. In addition, collecting frozen stool samples through quarantine, and obtaining screening results may ease concerns that fresh FMT could transmit pathogens from donors to recipients.
There were some limitations to our study. First, the experimental period was short. Although the ora after FMT tended to be normal, it was not yet stable. The experimental period should be extended to detect the composition of colonic micro ora. Second, only one model was used. Studies have shown that multiple model creation is more similar to the chronic colitis model, and multiple FMT signi cantly improves the treatment effect. Further studies are required to evaluate this point.

Conclusions
FMT is a feasible method for treating UC. The mechanism of action may be competitive inhibition of pathogenic microorganisms, improving immune metabolism, and reducing the in ammatory response to reduce the damage to the intestinal barrier and development of UC. Compared with fresh FMT, the therapeutic effect of frozen FMT may be greater. According to the random number table method, the rats were randomly divided into 2 groups: a normal (non-UC) group (n = 8) and UC model group (n = 32). After both groups were subjected to 1 week of adaptive feeding, the UC model group was created. The rats were anaesthetised with 3% pentobarbital sodium (0.15 mL/100 g). A rubber infusion tube was then inserted 8 cm into the upper anus of each rat and injected with 5% TNBS at 100 mg/kg (dissolved in 50% ethanol, total volume of 1 mL) (34,35). To prevent uid leakage after colonic infusion and ensure the induction of colitis, the rats were placed in the Trendlenburg position with their head down for 1 min (36). The daily weight and DAI score of both normal and UC model groups of rats were recorded. Fresh faeces were collected from each group before and after UC model creation. After the experiment, the rats were sacri ced by CO 2 inhalation, and colon tissue specimens and faeces were collected and immediately placed at -80°C.

Animals and groups
The UC model group was randomly divided into four groups 72 h later, with eight animals in each group: UC model group (control), mesalazine group, fresh FMT group, and frozen FMT group. The rats in each group were fed separately in cages.

Faecal bacteria preparation
For frozen FMT preparation, seven days before transplantation, the normal group was treated as donors whose faecal samples were collected in a sterile tube, dissolved in 0.9% NaCl at a ratio of 1 g:10 mL and homogenised for 5 min, followed by adding 0.1 mL 10% glycerol, mixing, and freezing at -80℃ (37). The samples were transferred to -80°C refrigerator for 1 week.
For fresh FMT preparation, on transplant day, the normal group was treated as donors whose samples were collected in a sterile tube, dissolved in 0.9% NaCl at a ratio of 1 g:10 mL and homogenised for 5 min, followed by 0.1 mL of 10% glycerol. The time from sampling to successful transplantation did not exceed 3 h (38).

Intervention pattern of each group
To create four different groups within the UC model group, the rats were administered different treatments once per day for ve consecutive days. The fresh FMT group was treated with 1 mL/100 g of fresh faecal homogenate from the normal group (non-UC) on the same day of transplantation, whereas the frozen FMT group was treated with 1 mL/100 g of frozen faecal homogenate collected from the normal group 7 days prior to transplantation. The mesalazine group was administered 0.3 g/kg of mesalazine by gavage as a positive control. The UC model group and normal group were administered 1 mL/100 g of saline by gavage as a blank control and negative control, respectively.

Specimen collection
The faeces of rats were collected at 72 h after model creation, two days after stopping drug intervention, and seven days after stopping drug intervention, and immediately placed in -80°C until analysis. After stopping the drug intervention for seven days, the rats were sacri ced by CO 2 inhalation. The rat colon tissue was separated and the most severely affected part of the rat colon tissue visually and excess fat around the tissue was cut off. The tissues were repeatedly rinsed with pre-cooled normal saline, after which lter paper was used to absorb the excess liquid, and the colon tissue was placed in 4% poly Fixed (lot:69111800, biosharp) and preserved in formaldehyde solution until haematoxylin-eosin staining. A section of severely diseased colon tissue was placed in a sterile cryopreservation tube for subsequent detection of in ammatory factors.

Detection indicator
Behavioural score Weight was measured once every day. After model creation, the weight loss, stool traits, blood, and stool of each group were observed daily to determine the DAI score, as shown in Table 1 (39).

Histological analysis
Colon tissues xed with pom solution were embedded in para n. After haematoxylin and eosin staining, the tissue structure was observed in detail under a light microscope, and the pathological score was determined according to Table 2 (39,40).

ELISA
The collected colon samples were thawed and tested for TNF-α using an ELISA kit according to the kit instructions. The absorbance of samples in each well of the plate was measured with a multifunctional microplate reader at a wavelength of 450 nm.
Bacterial DNA extraction and 16S rRNA gene sequencing Total genomic DNA was extracted from the samples using the CTAB/SDS method. DNA concentration and purity were monitored on 1% agarose gels. According to the concentration, DNA was diluted to 1 ng/ μL using sterile water. The primers 16S V3-V4: 341F CCTAYGGGRBGCASCAG, 806R GGACTACNNGGGTATCTAAT. 16S/18S rRNA genes were tagged with a barcode and used for analysis. All PCRs were carried out in 30 μL volume containing 15 μL Phusion®High-Fidelity PCR Master Mix (New England Biolabs, Ipswich, MA, USA), 0.2 μM of forward and reverse primers, and approximately 10 ng template DNA. Thermal cycling steps consisted of initial denaturation at 98℃ for 1 min, followed by 30 cycles of denaturation at 98℃ for 10 s, annealing at 50℃ for 30 s, and elongation at 72℃ for 60 s. The last step was performed at 72℃ for 5 min. Equal volumes of 1X loading buffer (contained SYB green) and PCR products was mixed and the samples were evaluated by 2% agarose gel electrophoresis.
Samples showing bands at 400-450 base pairs were further analysed. PCR products were mixed in equivalent ratios and puri ed with an AxyPrepDNA Gel Extraction Kit (AXYGEN, Union City, CA, USA). Sequencing libraries were generated using an NEB Next®Ultra™DNA Library Prep Kit for Illumina (New England Biolabs) following the manufacturer's recommendations, and index codes were added. The library quality was assessed on a Qubit@ 2.0 Fluorometer (Thermo Fisher Scienti c, Waltham, MA, USA) and Agilent Bioanalyzer 2100 system (Agilent Technologies, Santa Clara, CA, USA). The library was sequenced on an Illumina Miseq/HiSeq2500 platform (San Diego, CA, USA) and 250-bp/300-bp pairedend reads were generated.

Statistical analysis
The measured data were expressed as the average ± standard deviation (x ± s). SPSS software (version 20.0, SPSS, Inc., Chicago, IL, USA) was used to analyse the data. The independent sample t-test was used to compared data from the two groups, and one-way analysis of variance was performed to compare data between multiple groups. P < 0.05 indicated statistically signi cant results. Flora data were analysed and graphed by non-parametric test Kruskal Wallis sum rank test and Wilcoxon rank sum test with R language ggplot2, and ggpubr, and the MetagenomeSeq toolkit and Ubuntu Linux conda LEfSe software.

Declarations
Ethics approval and consent to participate The study was approved by the Ethics Committee of Zhejiang Chinese Medical University (IACUC-20200506-12).

Consent for publication
Not applicable.

Availability of data and materials
All data generated or analysed during this study are included in this published article.

Competing interests
The authors declare that they have no competing interests.

Funding
This work was supported by Health Commission of Zhejiang Province (2020378459).

Authors' contributions
The study was conceived by KYF and LYC and they secured the funding. ZFY, LYT, KYF, WJQ, WP and MFX did experiments and performed analysis. ZFY performed the data analysis. The rst draft of the manuscript was written by ZFY. All authors read and approved the nal manuscript. Tables   Table 1. Disease activity index score, DAI score as the sum of weight loss, stool consistency, and occult blood or gross blood in the stool.     Comparison of species abundance at the genus level. Principal coordinate analysis of frozen FMT group and fresh FMT group. Frozen FMT group-A = frozen FMT group before treatment; Frozen FMT group-B = Frozen FMT group 1 week after treatment; Frozen FMT group-C = Frozen FMT group 2 weeks after treatment; Fresh FMT group-A = Fresh FMT group before treatment; Fresh FMT group-B = Fresh FMT group 1 week after treatment; Fresh FMT group-C = Fresh FMT group 2 weeks after treatment.

Figure 7
Bacterial community column chart of different groups. BD-A = Frozen FMT group before treatment; BD-B = Frozen FMT group 1 week after treatment; BD-C = Frozen FMT group 2 weeks after treatment; ZC-C = normal group; XX-A = Fresh FMT group before treatment; XX-B = Fresh FMT group 1 week after treatment; XX-C = Fresh FMT group 2 weeks after treatment.