CXCL12-CXCR4-Mediated Chemotaxis Supports Accumulation of Mucosal-Associated Invariant T Cells Into the Liver of Patients With PBC

Objectives: To explore the potential role of CD3+CD8+CD161high TCRVα7.2+ mucosal-associated invariant T (MAIT) cells in the pathogenesis of primary biliary cholangitis (PBC). Methods: We enrolled 55 patients with PBC, 69 healthy controls (HCs), and 8 patients with hepatic hemangioma. Circulating MAIT cells and their chemokine receptor profiles and cytokine production were quantified using flow cytometry. Liver-resident MAIT cells were examined by immunofluorescence staining. CXCL12-mediated chemotaxis of MAIT cells was measured using a transwell migration assay. Plasma interleukin (IL)-18 was measured using ELISA, and cytokine production in IL-18-stimulated MAIT cells was detected using flow cytometry. Result: Peripheral MAIT cells were found to be significantly lower in patients with PBC (3.0 ± 3.2% vs. 9.4 ± 8.0%, p < 0.01) and negatively correlated with alkaline phosphatase (ALP) levels (r = −0.3209, p < 0.05). Liver immunofluorescence staining suggested that MAIT cells might accumulate in PBC liver. MAIT cells from patients with PBC expressed higher levels of CXCR4 (84.8 ± 18.0% vs. 58.7 ± 11.4%, p < 0.01), and the expression of CXCL12 was higher in PBC liver. CXCL12 promoted MAIT cell chemotaxis (70.4 ± 6.8% vs. 52.2 ± 3.5%, p < 0.01), which was attenuated by CXCR4 antagonist. MAIT cells from PBC produced significantly more interferon-γ (IFN-γ) (88.3 ± 4.2% vs. 64.2 ± 10.1%, p < 0.01), tumor necrosis factor-α (TNF-α) (93.0 ± 1.1% vs. 80.1 ± 5.3%, p < 0.01), Granzyme B (89.3 ± 3.3% vs. 72.1 ± 7.0%, p < 0.01), and perforin (46.8 ± 6.6% vs. 34.8 ± 7.7%, p < 0.05). MAIT cells from PBC expressed higher levels of IL18-Rα (83.8 ± 10.2% vs. 58.3 ± 8.7%, p < 0.01). Plasma IL-18 was more abundant in patients with PBC (286.8 ± 75.7 pg/ml vs. 132.9 ± 78.1 pg/ml, p < 0.01). IL-18 promoted IFN-γ production in MAIT cells (74.9 ± 6.6% vs. 54.7 ± 6.7%, p < 0.01), which was partially attenuated by blocking IL-18R (68.6 ± 8.3% vs. 43.5 ± 4.2%, p < 0.01). Conclusion: Mucosal-associated invariant T cells from patients with PBC accumulated in the liver via CXCL12-CXCR4-mediated chemotaxis, produced pro-inflammatory cytokines, and contributed to portal inflammation, which was potentially mediated by elevated IL-18. Targeting MAIT cells might be a therapeutic approach for PBC.

Mucosal-associated invariant T cells have been studied in PBC. Setsu et al. (15) report that MAIT cells are persistently activated and exhausted in patients with PBC and are lower in the peripheral blood and the liver. Jiang et al. (16) found reduced circulating MAIT cells but abundant liver-resident MAIT cells in patients with PBC, which is activated by IL-7 produced by cholic acid-stimulated hepatocytes. These studies have suggested that MAIT cells are potentially implicated in PBC. However, whether, and how, MAIT cells aggregate in the liver of patients with PBC remain unknown.
To address this issue, we analyzed the population and immunophenotype of peripheral MAIT cells from patients with PBC. We further analyzed the liverresident MAIT cells in patients with PBC. Finally, we explored the underlying mechanism of liver accumulation and hyperactivation of MAIT cells from patients with PBC.

Patients and Healthy Controls
We enrolled 55 patients with PBC who fulfilled the consensus criteria of PBC and were admitted to the Peking Union Medical College Hospital (PUMCH) from July 2017 to September 2019 ( Table 1) (17). Substantial higher alkaline phosphatase (ALP), gamma-glutamyl transferase, and total bile acid were found in patients with PBC. We also enrolled 69 age-and sex-matched healthy volunteers as healthy controls (HCs). Additionally, eight patients with hepatic hemangioma who underwent surgical resection were enrolled as controls. This study was approved by the institutional review board of PUMCH. Written informed consent was obtained from all participants.

Immunohistochemistry and Immunofluorescence
Liver tissues fixed with paraformaldehyde (PFA) were stained with anti-CXCL12 (R&D Systems, MN, USA) antibody. In brief, the tissue sections were deparaffinized, rehydrated, and treated with an EDTA buffer antigen retrieval solution for 26 min.
After blocking endogenous peroxidase with 3% H 2 O 2 and 3% BSA at room temperature for 30 min, slides were incubated with a mouse anti-CXCL12 antibody (R&D Systems, MN, USA) overnight at 4 • C. The slides were washed and covered with HRPconjugated anti-mouse antibodies (Servicebio, Wuhan, China) at room temperature for 50 min. The slides were developed with diaminiobenzidine (DAB), counterstained in the nucleus with hematoxylin, washed, dehydrated in series-gradient ethanol, washed, and mounted on a microscope. The density of CXCL12 + cells (cells/mm 2 ) was quantified using ImageJ [National Institutes of Health (NIH)]. Immunofluorescence staining was performed on paraffinembedded liver tissues using the following antibodies: rabbit antihuman CD161 (ab197979, Abcam, MA, USA), CY3 goat antirabbit (Servicebio, Wuhan, China), and FITC mouse anti-human TCRVα7.2 antibodies. After staining with DAPI, images were acquired using fluorescent microscopy (NIKON Eclipse Ti). The density of CD161 + TCRVα 7.2 + cells (cells/mm 2 ) was quantified using Image J (NIH).

CHEMOTAXIS ASSAY
CD8 + T cells from patients with PBC were starved in RPMI-1640 medium containing 1% FBS overnight. The cells were seeded in the upper transwell chamber (Corning, NY, USA), and media containing 0, 10, 30, or 50 ng/ml CXCL12 (Peprotech, NJ, USA) were added to the lower chamber for 1.5 h. Cells in both chambers were harvested and mixed with 123 count eBeads (Thermo Fisher Scientific, MA, USA); stained with fluorochrome-conjugated anti-CD3, anti-CD8a, anti-TCRVα7.2, and anti-CD161 antibodies; and processed with a flow cytometer. Chemotaxis rate was calculated as the percentage of MAIT cells in the lower chamber of the total MAIT cells. For blocking CXCR4, the cells were pre-treated with AMD3100 (1 µg/ml) (TOCRIS, Abingdon, UK) for 30 min before loading onto the upper chamber.

Statistical Analysis
The data were summarized as the mean ± SD, median (interquartile range), or number (percentage). For normal distribution data, the Student's t-test was used to compare the differences between the two groups. A paired t-test was used to compare the differences between the paired samples at different conditions. Non-normal distribution data were analyzed using the Kolmogorov-Smirnov test, ANOVA, and the Mann-Whitney U test. Correlations were calculated using Spearman's correlation analysis. A two-sided p < 0.05 was considered statistically significant. The data were analyzed using the SPSS V.17.0 software (IBM, NY, USA).

CXCL12-CXCR4 PATHWAY POTENTIALLY MEDIATES ACCUMULATION OF MAIT CELLS IN THE LIVER
Given that the liver is the target organ for PBC and that it is enriched with MAIT cells, we analyzed the liver-resident MAIT cells using immunofluorescence staining. We found a higher number of CD161 + TCRVα7.2 + MAIT cells in the liver (MAIT cells per mm 2 , 899 ± 431 vs. 189 ± 76, p < 0.01, Figure 3A), including the portal area (Supplementary Figure 3) from patients with PBC than those from controls. Furthermore, the apoptosis of resting (1.5 ± 1.0% vs. 1.3 ± 1.2%, p > 0.05) MAIT cells and activated (1.3 ± 0.5% vs. 1.5 ± 0.6%, p > 0.05) MAIT cells and the proliferation potential (53.1 ± 7.9% vs. 54.5 ± 6.4%, p > 0.05) of MAIT cells from patients with PBC were comparable to those from HCs (Supplementary Figure 4), suggesting that decreased MAIT cells might result from their accumulation in the liver.

DISCUSSION
In this study, we confirmed that circulating MAIT cells were significantly decreased in patients with PBC. MAIT cells from patients with PBC potentially accumulated in the liver, which was promoted by CXCL12-mediated chemotaxis of CXCR4-expressing MAIT cells. Furthermore, MAIT cells from PBC patients were activated and produced pro-inflammatory  control. The data were expressed as mean ± SD. NS, not statistically significant; *p < 0.05, ***p < 0.001, and ****p < 0.0001 by Student's t-test.
cytokines,which was mediated by elevated circulating IL-18 and IL-18Rα on MAIT cells. The present study confirm that peripheral MAIT cells have reportedly decreased in patients with PBC (15,16). Furthermore, peripheral MAIT cells in patients with PBC were inversely correlated with serum ALP, which is a disease activity index of PBC. Although half of the patients with PBC have hyperlipidemia, no association between circulating MAIT cells and lipid profiles was observed. However, the underlying mechanism for the decreased circulation of MAIT cells remains controversial. Jiang et al. (16) have proposed an overactivationexhaustion model. We found that MAIT cells were indeed potentially activated. Exhaustion markers, such as PD-1, CD39, and Tim-3, were expressed by liver-resident MAIT cells, which might be induced by chronic stimuli exposure. Our data showed that circulating MAIT cells expressed lower CD39 in patients with PBC than in HCs (data not shown). However, a higher apoptotic potential in MAIT cells from patients with PBC was not observed, which did not support the hypothesis that decreased peripheral MAIT cells were induced by excessive postactivation of cell death. We noted that MAIT cells in patients with PBC mainly resided in the liver, which was consistent with the findings of Jiang et al. (16). However, Setsu et al. (15) noted that MAIT cells in patients with PBC were scarce in the liver. This discrepancy may be due to different PBC stages among studies. Similarly, excessive liver infiltration of MAIT cells is reported in autoimmune hepatitis and non-alcoholic fatty liver (14,20). Furthermore, T cells and aberrant chemotaxis of MAIT cells are implicated in rheumatoid arthritis and multiple sclerosis, respectively (21,22). Therefore, the data from this study as well as from other studies supported the hypothesis that decreased circulating MAIT cells from patients with PBC were induced by excessive infiltration into the liver through aberrant chemotaxis.
Chemokines and chemokine receptors orchestrated chemotaxis. We screened T cell-expressing receptors, including CXCR4, CCR10, CCR6, CXCR6, CX3CR1, CXCR1, CXCR3, CXCR5, and CCR5, on MAIT cells (23). We found that CXCR4 was abnormally elevated in MAIT cells from patients with PBC. Furthermore, serum CXCL12, the CXCR4 ligand, is higher in patients with PBC (24). CXCL12 is secreted by the bile duct epithelium, especially in inflammatory conditions (25,26). We demonstrated that CXCL12 chemoattracted MAIT cells in vitro. Taken together, the inflammatory milieu of the portal area in patients with PBC might promote the bile duct epithelium to produce CXCL12, which subsequently recruits CXCR4 + MAIT cells. However, CXCR4-expressing CD3 + CD8 + T cells from patients with PBC did not significantly decrease in the peripheral blood, which might attribute to lower levels in the expression of CXCR4 compared to that by MAIT cells. Thus, our study suggested that targeting the CXCL12-CXCR4 pathway might be a potential therapeutic approach to alleviate MAIT cell infiltration into the liver of patients with PBC.
Mucosal-associated invariant T cells from patients with PBC produced higher pro-inflammatory cytokines, Granzyme B, and perforin, indicating that MAIT cells were pathogenic to bile ducts. PBC is characterized by small bile duct inflammation, and TNF-α and IFN-γ produced by MAIT cells might contribute to the apoptosis of the intrahepatic bile duct (23). MAIT cells are activated by TLR ligands, such as viral DNA and lipopolysaccharide, and cytokines (27), such as IL-7, IL-12, and IL-18. Jiang et al. (16) reported that IL-7 promotes the activation of MAIT cells in patients with PBC. Serum IL-18 levels are higher in patients with PBC and are positively correlated with cirrhosis (28). Therefore, we observed that IL-18Rα on MAIT cells from patients with PBC was upregulated, and IL-18 indeed promoted the activation of MAIT cells. Taken together, we identified a new mechanism of IL-18mediated activation of MAIT cells, which added a new layer of hyperactivation of MAIT cells in patients with PBC and suggested that IL-18 and IL-18R might be a therapeutic target for PBC.
In conclusion, MAIT cells from patients with PBC accumulated in liver via CXCL12-CXCR4-mediated chemotaxis, produced pro-inflammatory cytokines, and contributed to portal inflammation, which was potentially mediated by elevated IL-18. Targeting MAIT cells may be a therapeutic approach for PBC.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by Institutional Review Board of PUMCH. The patients/participants provided their written informed consent to participate in this study.

AUTHOR CONTRIBUTIONS
FZ and HC designed and supervized the study. SL, LW, and CH collected samples. ZC and JS collected data. ZC performed the experiments and statistical analysis. ZC and HC drafted and revised the manuscript. FZ obtained funds. All authors contributed to the article and approved the submitted version.