Male Lewis rats (9–12 weeks old) weighing 270–320 g and male Dark Agouti (DA) rats (12–16 weeks old) weighing 260–290 g were prepared. The whole liver graft was transplanted after 1 h of cold storage in phosphate-buffered saline. The median weight of grafts from Lewis and Dark Agouti rats was 10.465 g (range, 9.250–11.600) and 8.001 g (range, 7.510–8.888), respectively. Rats were divided into two groups after orthotopic liver transplantation (OLT): 1) the syngeneic group (n = 6), in which both the donors and recipients were Lewis rats; and 2) the allogeneic group (n = 6), in which the donors were DA rats and the recipients were Lewis rats. Fecal, blood, and histological (liver and small intestine) samples were obtained at four fixed time points (days 1, 3, 7, and 10) after OLT. In total, 48 OLTs were performed for 24 individuals in each group (6 individuals × 4 time points). Six healthy Lewis rats were used as the controls. As the present basic research is an exploratory study, a power calculation was not performed. We selected this relatively small sample size empirically because the GM and fecal OAs were measured simultaneously for rats for the first time, and therefore, the initial intention was to gather basic evidence regarding the transitions of these variables that could be utilized in future human studies. All OLT procedures were performed under inhalation anesthesia using 1.5% isoflurane with endotracheal intubation and artificial respiration, according to our previous techniques, with hepatic artery reconstruction (16) and without fasting, intravenous drip, antibiotic administration, or immunosuppression. Animals were housed under specific pathogen-free conditions in a temperature- and humidity-controlled environment under a 12-h light/dark cycle. The rats were fed a standard diet (F-2; Oriental Bio Service, Kyoto, Japan) and tap water ad libitum.

All experiments were conducted in accordance with the Animal Research: Reporting of In Vivo Experiments (ARRIVE) Guidelines. The institutional ethics committee of Kyoto University approved the experimental protocol (MedKyo18537).

Under inhalation anesthesia, portal venous pressure (PVP) was measured and monitored via a pressure transducer using the following procedure: a segment of the mesenteric branch vein was cannulated with a 24-g cannula needle, and the tip of the cannula was advanced into the trunk of the superior mesenteric vein. Blood samples were collected from the inferior vena cava and feces from the rectum. Each individual was euthanized at each time point after sample collection.

The fecal samples were placed directly into two tubes (∼1.0 g/tube); one tube contained 2 ml of RNAlater® (Ambion, Austin, TX, United States), and the other was empty. The samples with RNAlater® were held at room temperature for 10 min before storage at 4°C (for the analysis of GM), and the others were placed in a freezer at −80°C (for the analysis of fecal OA concentrations) within 30 min of excretion. Samples were sent to the Yakult Central Institute at −20°C for analysis.

GM composition was analyzed by the 16S and 23S rRNA-targeted RT–qPCR system using Yakult Intestinal Flora-SCAN (YIF-SCAN®). The mechanisms and advantages of YIF-SCAN® for measuring bacterial counts in fecal and blood samples have been previously described elsewhere (17,18,19). Briefly, three serial dilutions of the extracted RNA sample were used for bacterial rRNA-targeted RT-qPCR, and threshold cycle values in the linear range of the assay were applied to the standard curve to obtain the corresponding bacterial cell count in each nucleic acid sample. These data were then used to calculate bacterial counts per sample. The specificity of the RT-qPCR assay using group-, genus-, or species-specific primers was determined as previously described (Supplementary Table S1).

The bacteria examined included obligate anaerobes (Clostridium coccoides group, C. leptum subgroup, Bacteroides fragilis group, genus Bifidobacterium, Atopobium cluster, genus Prevotella, Clostridioides difficile, and C. perfringens), facultative anaerobes (family Enterobacteriaceae, genus Enterococcus, genus Streptococcus, and genus Staphylococcus), and aerobes (genus Pseudomonas).

A portion of the feces was homogenized in four volumes of 0.15 mol/L perchloric acid and stored at 4°C for 12 h. The homogenate was centrifuged at 4°C at 20,400 ×g for 10 min, and the resulting supernatant was passed through a membrane filter with a pore size of 0.45-μm (Millipore Japan Ltd., Tokyo, Japan). The sample was analyzed by high-performance liquid chromatography using a Waters system with Waters 432 Conductivity Detector (Waters Co., Milford, MA) equipped with two columns (Shodex RS pack KC-811; Showa Denko Co. Ltd., Tokyo, Japan).

In this study, the SCFAs included acetic acid, butyric acid, and propionic acid.

Blood tests, including complete blood count, peripheral neutrophil/lymphocyte ratio (NLR), aspartate aminotransferase, alanine aminotransferase (ALT), total bilirubin (T-Bil), serum albumin, and serum creatinine, were performed in a professional clinical laboratory (Japan Clinical Laboratories, Kyoto, Japan).

To evaluate the immune function, the CD4/CD8 T-cell ratio was analyzed. The conjugated mouse anti-rat monoclonal antibodies used for flow cytometry, APC-conjugated CD3, FITC-conjugated CD4, and PE-conjugated CD8a, were commercially available (BD Biosciences, San Josè, CA, United States). Samples were acquired using a BD Accuri C6 (BD Biosciences).

The fecal IgA content was determined to evaluate intestinal barrier function by enzyme-linked immunosorbent assay (ELISA) using a rat IgA ELISA kit (Bethyl Laboratories, Inc., Montgomery, TX, United States). Serum lipopolysaccharide (LPS) levels were evaluated using a rat LPS ELISA kit (CUSABIO, Wuhan, China).

Formalin-fixed, paraffin-embedded sections (4-μm thickness) of rat liver grafts and small intestines were stained with hematoxylin and eosin (H&E). For electron microscopy, rat small intestines were perfused through the aorta with a mixture of 2% glutaraldehyde and 4% paraformaldehyde and then extracted. The intestines were cut into small pieces and stored overnight at 4°C. The sections were stained with saturated uranyl acetate and lead citrate and observed using a Hitachi H-7650 electron microscope (Hitachi, Tokyo, Japan) for transmission electron microscopy (TEM). Two independent investigators examined all the tissue sections in a blinded manner. The severity of TCMR was evaluated in accordance with the Banff classification (20).

The results of the fecal GM and OA analyses are expressed as the mean ± standard error. For statistical calculation, a value of half of the detection limit was assigned when the count or concentration was below the detection limit. Longitudinal data of these variables were analyzed using a linear mixed-effects model, which included the study group, time after LT, and interaction of the study group with the time after LT. Other continuous variables were presented as the median and range or interquartile range (IQR), as appropriate. Categorical variables were presented as numbers and percentages. Correlations between two variables were determined using Spearman’s rank correlation coefficient. Statistical significance was set at a p value < 0.05. JMP 14.0 (SAS Institute, Cary, NC, United States) was used for all statistical analyses.

Representative biochemical analyses (Figure 1) and histopathological findings (Figures 2, 3) are presented. Briefly, after OLT, the hepatic graft suffered from ischemia/reperfusion injury in both syngeneic and allogeneic groups on day 1, with elevated liver enzyme levels (Figure 1) and histological periportal edema (Figures 2A,E). However, the graft recovered to nearly normal levels both functionally and histologically in the syngeneic group (Figures 2B–D), whereas progressive TCMR led to irreversible graft failure after day 7 and by day 10 in the allogeneic group (Figures 2G,H). Figures 2F–H represent “mild,” “moderate,” and “severe” by Banff classification, respectively. As liver enzyme levels increased and cholestasis progressed, synthetic ability decreased, and portal venous pressure increased (Figure 1). Small intestine histology in the allogeneic group showed worsening submucosal edema and disarrangement of the epithelium as TCMR progressed from day 7 to day 10 (Figures 2M–P). Transmission electron microscopy images of the small intestine demonstrated epithelial cell structural destruction as TCMR progressed from day 7, indicating disruption of barrier function (Figure 3).

The median survival period of allogeneic liver grafts was 11 days (range, 10–13), whereas all syngeneic liver grafts survived.

Dynamic comparisons of the representative fecal microbiota are shown in Figure 4A. Dysbiosis progressed as the liver function declined in the allogeneic group, mainly on days 7 and 10, whereas it recovered as the liver function improved in the syngeneic group. The number of predominant obligate anaerobes (POAs), such as the C. coccoides group, B. fragilis group, and Bifidobacterium, decreased as liver function declined in the allogeneic group. These changes were more remarkable in obligate and facultative anaerobes, some of which are responsible for opportunistic infections. Total lactobacilli and its subgroup showed a significant decrease, and Clostridioides difficile, Enterobacteriaceae, Enterococcus, Streptococcus, and Staphylococcus showed a significant increase in the allogeneic group on days 7 and 10, as liver function declined. Meanwhile, GM seemed to be restored to normal by day 10 in the syngeneic group as liver function improved. C. perfringens, Lacticaseibacillus, and Pseudomonas were below the detection limits.

Dynamic comparisons of representative OAs are shown in Figure 4B. Overall, the allogeneic group showed significantly lower OA concentrations as the liver function declined. More specifically, in the allogeneic group, SCFA concentrations were depleted by day 10, with a slight recovery trend from days 1–3. In contrast, the concentration of SCFAs in the syngeneic group recovered to nearly normal levels after depletion on day 1. The remaining values are listed in Supplementary Table S2. The transitions in these values were similar between the groups.

In short, although the disturbance was slight, transient, and reversible in the syngeneic model, the allogeneic TCMR model demonstrated distinct dysbiosis and depletion of fecal OAs.

In the syngeneic group, all measured values gradually returned to normal by day 10 after LT (Figure 4C). In contrast, immune function, represented by the NLR and CD4/CD8 ratio, and intestinal barrier function, represented by the fecal IgA level, decreased as liver function declined in the allogeneic group. On days 7 and 10, the allogeneic group showed a significantly higher NLR, a lower CD4/CD8 ratio, and decreased fecal IgA levels than the syngeneic group. Consequently, extremely high LPS levels were observed from days 7–10, implying the occurrence of BT.

Spearman’s rank correlation coefficient was calculated using the data of all 48 OLTs because no single individual was sampled multiple times. ALT was significantly negatively correlated with C. coccoides group and Prevotella and positively correlated with Enterobacteriaceae and Enterococcus (Figure 5A). With regard to OA, there were significant negative correlations between ALT and fecal concentrations of total OA, butyric acid, and propionic acid. T-Bil was significantly negatively correlated with total lactobacilli and Lactobacillus and positively correlated with Enterobacteriaceae and Enterococcus (Figure 5B). With regard to OA, there were significant negative correlations between T-Bil and fecal concentrations of total OA, acetic acid, butyric acid, and propionic acid.

In summary, liver function affected the counts of Enterobacteriaceae and Enterococcus, and the fecal concentration of SCFAs.