Multiplex immuno uorescence of spatial distribution of in ltrating T cells in different regions of hepatic lobules during liver transplantation rejection

Shipeng Li Capital Medical University A liated Beijing Friendship Hospital Guangpeng Zhou Capital Medical University A liated Beijing Friendship Hospital Jie Sun Capital Medical University A liated Beijing Friendship Hospital Bin Cui Capital Medical University A liated Beijing Friendship Hospital Haiming Zhang Capital Medical University A liated Beijing Friendship Hospital Lin Wei Capital Medical University A liated Beijing Friendship Hospital Liying Sun Capital Medical University A liated Beijing Friendship Hospital Zhijun Zhu (  zhu-zhijun@outlook.com ) Capital Medical University A liated Beijing Friendship Hospital https://orcid.org/0000-0001-7031-2083


Introduction
Rejection is a major obstacle to the long-term survival of liver grafts and the prognosis of patients after liver transplantation 1,2 . Rejection includes cellmediated cell rejection and humoral antibody mediated immune response 3 . In recent years, it has been proved that the changes of T cell subsets are different in transplant rejection. CD4 + T cells can stimulate T cell clonal expansion and differentiation by recognizing allogeneic antigen and secreting interleukin (IL)-2, and induction of acute rejection by activating CD8 + cytotoxic T cells 4 . T helper (Th)17 cells are immune cells that mainly secrete IL-17A, IL-17F and other cytokines, and mediate the in ammatory response 5 . It has been showed that T regulatory cells (Tregs) play an important role in inhibiting transplantation immune rejection and inducing immune tolerance 6 . The proportion of CD4 + T cells, CD8 + T cells and Th17 in peripheral blood of patients with acute rejection is increased, but the proportion of Tregs is increased in peripheral blood of patients without acute rejection 7 . At the same time, CD4 + T cells, CD8 + T cells, Th17 and Tregs in liver tissue also change accordingly.
To date, the exact phenotype of these immune cells has not been speci ed. However, different phenotypic and molecular studies have found that the composition of the in ltrating lymphocytes may be important. The liver consists of about 1 million hepatic lobules, and blood ows from the peripheral portal area to the central vein 8 . There are gradient changes in nutrition, oxygen concentration, hormones and other aspects in the hepatocyte plate of the hepatic lobule, thus forming different metabolic functional areas 9 . The hepatic lobules are divided into three regions according to their different metabolic functions: perivenous, transitional and periportal zones 8, 9 . Halpern divided the porto-central lobule axis into nine layers and developed a probabilistic inference algorithm to calculate the likelihood of each cell belonging to any of these layers based on the expression of our panel of landmark genes 10 . The functional status of the corresponding non-hepatic parenchymal cells (endothelial cells, immune cells, etc.) may be different due to the different partition of the hepatic lobules.
Previous studies have found that Kupffer cells are enriched near periportal regions, with those closest to portal triads having distinct phenotypic properties 11 . However, the distribution of diverse resident immune cell types, the cellular and molecular mechanisms underlying the spatial organization of the liver immune system, and the functional consequences of asymmetric immune cell localization remain to be clari ed. Kupffer cells are enriched in periportal regions and there are quantitative data detailing this patterning, with MHCII hi or MHCII int Kupffer cells showing a similar distance to the central vein 12 . These data reveal that the hepatic lobules contain a spatially polarized immune system 'immune zonation'. However, there are no reports about the spatial distribution of in ltrating T cells and the changes in immune microenvironment in different areas of hepatic lobules during liver transplantation rejection. Due to the complexity of the structure of the hepatic lobules, the gene expression pattern of the spatial axis from the central vein to the portal zone is different, and its function also change gradually. Especially in the traditional hepatic lobule transitional zone, many important genes are expressed. Curiously, There is a question as to whether there are differences in the types and functional status of immune cells in different regions of hepatic lobules. In response to this question, we studied the in ltration and distribution of T cells in different regions of hepatic lobules, and the spatial distribution of T cells in different regions of hepatic lobules under conditions of acute and non-acute rejection after liver transplantation.
Previous studies have used immunohistochemical or immuno uorescence techniques to detect changes in CD4, CD8 and other indicators in grafts 13,14 , but these techniques can only show a certain type of T cell distribution by one or two indicators. Here, we present a direct method that generated readouts for a comprehensive panel of biomarkers from liver biopsy from liver transplantation patients. Multiplex immuno uorescence staining can simultaneously label multi-cellular markers to observe various cell types and their functional status and interactions. It reveals details of the spatial localization of immune cells in different tissues, and gives a clear understanding of the immune microenvironment of organs and tissues 15 . Our research indicated that the in ltrated immune cells had spatial speci city and some exciting phenomena were found after liver transplantation. The distribution of CD8 + T cells in different regions of hepatic lobules is closely related to rejection level after liver transplantation. Acute rejection of liver transplantation may occur when the percentage of CD8 + T cells in perivenous zone increases. Although the percentage of regional CD4 + T cells could not re ect the level of rejection, the number of CD4 + T cells and CD8 + T cells in different regions was closely related to the rejection level.

Patients and clinical data collection
Patients who underwent liver transplantation at Liver Transplantation Center, Beijing Friendship Hospital, Capital Medical University between January 2017 and July 2020 were included. Recipients' demographic, medical, transplantation, and follow-up data were collected (Table 1). Medical data comprised etiology of liver disease and laboratory data. Data related to transplantation were ABO blood group, donor sex and age, graft-type, operation time, and warm ischemic time. Laboratory data included trough concentration of tacrolimus and liver function tests such as albumin, aspartate aminotransferase (AST), alanine aminotransferase (ALT), and total bilirubin (TB). All liver grafts were voluntarily donated after cardiac death or by living donors, and all donations were approved by the Ethics Committee of Beijing Friendship Hospital, Capital Medical University. Group management of liver transplantation patients All liver transplantation patients were divided into the acute rejection and non-acute rejection groups according to the Banff schema for grading liver allograft acute rejection 16 , and a normal group was set up. The total score of rejection activity index (RAI) was 9. The patients with a score less than 3 were classi ed as non-acute rejection, and 3-9 as acute rejection.

Liver histology
Liver biopsies were performed using the Menghini method under deep sedation with propofol 17 . Biopsies were performed by two experienced examiners with a 17-gauge needle. The specimen length was ≥ 2 cm. Normal liver tissues were obtained from living donors. The liver tissues were xed in 4% mediosilicic isotonic formaldehyde for 24 h, dehydrated and embedded in para n. Four-micrometer-thick sections were cut from each para n embedded tissue and stained with hematoxylin and eosin (HE) to evaluate the pathological changes in the liver.

Multiplex immuno uorescence assays 18
Tissue multiplex immuno uorescent staining was performed using the Opal Polaris 5 color IHC staining kit (Akoya Biosciences). Brie y, tissue sections of 4µm thick formalin-xed, para n-embedded (FFPE) were baked for 2 h at 60°C before staining. Slides were rehydrated with a series of graded ethanol to deionized water. Antigen retrieval was performed at pH 6 for 20 min at 95°C. Slides were serially stained with the following antibodies: anti-CD4, -CD8, -IL-17 and -FOXP3. Anti-mouse / rabbit horseradish peroxidase (Cell Signaling Technology, Danvers, MA, USA) was used as the primary antibody. TSA-conjugated uorophores (PerkinElmer) were used to visualize each biomarker: Opal 690 (IL-17), Opal 620 (FoxP3), Opal 570 (CD4) and Opal 520 (CD8), incubation time per primary antibody was 1 h. Subsequently, anti-rabbit/mouse Polymeric Horseradish Peroxidase (Opal IHC Detection Kit, Akoya Biosciences) was applied as a secondary label with an incubation time of 10 min. Antibody signal was visualized using the corresponding Opal Fluorophore (Akoya Biosciences) by incubating the slides for 10 min. Slides were mounted with anti-fade mounting medium (P36965, Life Technologies) and stored at 4°C before imaging. Image acquisition was performed using the Vectra Polaris multispectral imaging platform (Akoya Biosciences), whole slide image was scanned and 3-5 representative regions of interest were chosen by the pathologist at 200× resolution as multispectral images.

Hepatic lobule zone
The hepatic lobules are divided into three regions: perivenous, transitional and periportal zones, as previously described 8 .

Statistical analysis
The data were expressed as mean ± standard deviation. Differences among groups were analyzed by one-way analysis of variance and Newman-Keuls test.
SPSS version 22.0 was used for these analyses. A P value < 0.05 was considered as statistically signi cant.

Pathological changes in liver
The histopathological features in acute rejection consisted of three major components of the lesion: in ammation of the portal tracts, bile duct damage, and endotheliitis. The portal tract showed a pleomorphic in ltrate of lymphocytes, neutrophils, and plasma cells. Considerable balloon like degeneration, cholestasis, necrosis, phlebitis and brosis were observed in the acute rejection group as compared with the normal group. Compared with acute rejection group, he staining showed that the in ltration of immune cells in portal area and hepatic lobule was signi cantly reduced in non-rejection group, bile duct injury and endothelial injury were rare, and hepatocyte edema and necrosis were reduced (Fig. 1A).
Expression and distribution of CD4, CD8, IL-17 and FOXP3 in liver tissues Expression and distribution of CD4, CD8, IL-17 and FOXP3 were detected by immunohistochemistry (Fig. 1B). CD4 was mainly expressed in the membrane and cytoplasm of sinusoidal endothelial cells and lymphocytes. In normal liver tissue, CD4 was mainly expressed in liver sinusoidal endothelial cells, while CD4 + T cells were few in number. In the acute rejection group, there were more in ltrating immune cells and CD4 + T cells, but fewer CD4 + liver sinusoidal endothelial cells. The in ltrating immune cells and CD4 + T cells in the non-acute rejection group were signi cantly reduced, while the number of CD4 + liver sinusoidal endothelial cells was increased. The expression and distribution of CD8 T cells were similar to those of CD4 + T cells. In normal liver tissue, there were few CD8 + T cells, mainly concentrated in the portal area, and fewer in ltrated the hepatic lobules. CD8 + T cells increased signi cantly in the acute rejection group. Compared with the acute rejection group, CD8 + T cells decreased signi cantly in the non-acute rejection group. IL-17 was mainly expressed in hepatocytes, endothelial cells and immune cells cytoplasm. Compared with the normal and non-acute rejection groups, the expression of IL-17 in lymphocytes was higher in the acute rejection group. FOXP3 was mainly expressed in nucleus and immune cells. FOXP3 + T cells were almost not found in normal liver tissue. There were signi cantly more FOXP3 + T cells in the acute rejection than non-acute rejection group.

Multiplex immuno uorescence to measure dynamic changes in in ltrating T cells in liver acute rejection
To visualize changes in the composition of the immune in ltrate in liver acute rejection, we performed multiplexed immuno uorescence staining for CD4, CD8, IL-17 and FOXP3 ( Fig. 2A). We calculated the proportion of various types of T cells in each slice (N% = number of certain T cells / total number of cells in each slice×100%), and compared the differences in T cells in each group. This revealed an increase of in ltrating CD4 + T cells in the acute rejection and non-acute rejection groups as compared with the normal group, The proportion of CD4 + T cells was highest in the acute rejection group, and the difference was statistically signi cant (Fig. 2B, P < 0.001). The number of in ltrating CD8 + T cells was highly variable among patients in the normal group (Fig. 2C, P < 0.001). We also detected numbers of in ltrating FOXP3 + regulatory T cells (CD4 + FOXP3 + T and CD8 + FOXP3 + T cells) in the liver. It was di cult to observe CD4 + FOXP3 + T (Tregs) and CD8 + FOXP3 + T cells in the normal and non-acute rejection groups, but the proportion of Tregs and CD8 + FOXP3 + T cells in the acute rejection group was the highest, and the difference was signi cant (Fig. 2D and 2E, P < 0.001). We also counted the number of CD4 + IL-17 + T (Th17) and CD8 + IL-17 + T cells, and the proportion of Th17 cells in the acute rejection group was the highest, and the difference was signi cant (Fig. 2F, P < 0.001). The number of CD8 + IL-17 + T cells in liver tissue was low, although the proportion was still the highest in the acute rejection group (Fig. 2G, P < 0.001). The above results showed that the proportion of CD4 + T cells, CD8 + T cells, Tregs, and Th17 increased in acute rejection after liver transplantation, and these cells were involved in the formation of acute rejection or immune tolerance after liver transplantation. When the rejection level decreased or there was no rejection, these types of T cells decreased or even disappeared.

Distribution of in ltrating T cells in hepatic lobule and portal area
Due to the difference in the structure of hepatic lobule and portal area (Fig. 3A and 3B), the distribution of in ammatory cell in ltration is also different. We observed the difference in T cell distribution in hepatic lobules and portal area by multiple staining technique, and summarized the characteristics of T cell spatial distribution in the liver under different immune states of liver transplantation (Fig. 3C). In the normal group, there were few in ltrating T cells in the hepatic lobules, and a small number of CD8 + T cells, CD4 + T cells, CD4 + FOXP3 + T cells (Tregs), CD4 + IL-17 + T cells (Th17). In the acute rejection group, the number of in ltrating T cells was signi cantly increased, including CD8 + T cells, while the number of CD8 + T cells was higher than that of CD4 + T cells. The number of in ltrating T lymphocyte in the non-acute rejection group was lower than that in the acute rejection group; CD8 + T cells and CD4 + T cells were seen, but Tregs and Th17 could rarely be observed. In the normal group, in ltrating T cells could be seen in the portal area, mainly CD8 + T cells and CD4 + T cells. A small number of Th17 were also seen, but there were no Tregs. In the acute rejection group, in ltrating T cells were seen in the portal area, with a large number of CD8 + T cells and CD4 + T cells, and Tregs and Th17 were also observed. The number of in ltrating T cells in non-acute rejection group was lower than that in acute rejection group, CD8 + T cells and CD4 + T cells were seen, but there were fewer Tregs and Th17 (Fig. 3C). The above results showed that the distribution of in ltrating T cells was different in the hepatic lobules and portal area. The in ltrating T cells mainly concentrated in the portal area, and a few in ltrated the hepatic lobules.

Distribution of T cell in ltration in different areas of hepatic lobules
The hepatic lobule is divided into the perivenous, transitional and the periportal zones ( Fig. 4A and 4B). The percentage of CD4 + T cells, CD8 + T cells and Tregs in each region was statistically analyzed (n% = the number of T cells / total number of cells in the each zone×100%), and then the differences were compared in the different rejection states of liver transplantation. As showed in Fig. 4D, we compared the spatial distribution of periportal zone CD4 + T cells, CD8 + T cells and Tregs in the normal, acute rejection and non-acute rejection groups. Multiplex immuno uorescence staining showed that the percentage of Tregs in the acute rejection group was signi cantly higher than that of other two groups (P < 0.001). The percentage of CD4 + T cells and CD8 + T cells was higher than in the normal group (P < 0.01). The percentage of CD4 + T cells and CD8 + T cells in the non-rejection group was higher than that of the normal group (P < 0.05). The percentage of CD4 + T cells in non-rejection group was signi cantly lower than that in acute rejection group, but signi cantly higher than that in normal group (P < 0.05). In the transitional zone, the percentage of CD4 + T cells and Tregs in the acute rejection group was higher than that of the other two groups (P < 0.01), the percentage of CD8 + T cells was higher than that of the normal group (P < 0.05); the percentage of CD4 + T cells in the nonrejection group was signi cantly higher than in the non-rejection group (P < 0.05). The percentage of CD4 + T cells, CD8 + T cells and Tregs was lower than in the acute rejection group, but higher than in the normal group (P < 0.05). At the same time, the proportion of CD4 + T cells, CD8 + T cells and Tregs in perivenous zone in the acute rejection group was higher than that of the other two groups (P < 0.05). The percentage of CD4 + T cells, CD8 + T cells in the nonrejection group was lower in the acute rejection group, but higher than in the normal group (P < 0.05).
We observed the difference in T cell spatial distribution in different regions of liver tissue in the same group (Fig. 4E). In the normal group, the percentage of CD4 + T cells and Tregs in different regions was different, but not statistically (P > 0.05). The percentage of transitional zone CD4 + T cells was signi cantly higher than in the other two regions in the normal group (P < 0.05). Similarly, the percentage of CD4 + T cells from the periportal to perivenous zones was different in the acute rejection group, but not statistically (P > 0.05). Interestingly, the percentage of CD8 + T cells from the periportal to perivenous zones increased gradually, the percentage of CD8 + T cells in perivenous zone was higher than in other two zones (P < 0.05). The percentage of transitional zone Tregs was higher than that of perivenus zone (P < 0.05). The percentage of CD4 + T cells and CD8 + T cells from the periportal to perivenous zones was different in the non-acute rejection group, but not signi cantly (P > 0.05). The percentage of Tregs in the transitional zone in the non-rejection group was signi cantly higher than in the other two regions (P < 0.05).

Discussion
T cells play an important role in the rejection of allogeneic liver transplantation 19 , and participate in the recognition of transplantation antigens, immune regulation, cell dissolving effect and other immune responses 20,21 . In order to study the phenotype and quantity of T cells in different rejection levels after liver transplantation, the changes of CD4 + T cells, CD8 + T cells, Tregs and Th17 in peripheral blood were detected by ow cytometry or immunohistochemistry. After liver transplantation, the levels of blood CD4 + T cells peaked in the acute rejection group and was signi cantly higher than that of the non-acute rejection group 22 . Tregs and Th17 are known to be involved in the alloreactive responses in organ transplantation. Wang et al 23 investigated whether the circulating Tregs/Th17 ratio is associated with acute allograft rejection in liver transplantation. The frequency of circulating Tregs was signi cantly decreased, whereas the frequency of circulating Th17 cells was signi cantly increased in acute rejection. Tregs/Th17 ratio had a negative correlation with liver damage indices and the score of rejection activity index after liver transplantation.
When immune rejection occurs, a large number of T cells in ltrate the graft, such as CD4 and CD8 subsets, which increases with the aggravation of rejection 24 . The most important histological changes of acute rejection are mixed in ammatory cell in ltration of the con uence area 25 . Krukemeyer et al 26 found that more B lymphocytes and plasma cells migrated to the liver in patients with rejection than in non-rejection patients through liver tissue studies. But these experimental techniques have some limitations. The exact composition of leukocyte in ltration during liver allograft rejection is di cult to comprehend and visualize on the same biopsy slide. With the development of science and technology, multiplex immuno uorescence assays has been favored by researchers.
Multiplexed imaging platforms to simultaneously detect multiple epitopes in the same tissue section have emerged in recent years as powerful tools to study the immune context of organs and tissues 27 . Multiplex immuno uorescence assays can improve the understanding of the tissue microenvironment, new therapeutic targets, prognostic and predictive biomarkers, and translational studies 28 . Until now, this new technology has mainly been used in oncology to better characterize immune cell in ltration in tumors in situ [29][30][31] . Previous studies have used immunohistochemical or immuno uorescence techniques to detect the changes in CD4 + T cells, CD8 + T cells, Th17, Tregs and other indicators in grafts. However, using serial tissue sections makes the cellular interplay and the precise location of the cells di cult to analyze. We present multiplex immuno uorescence assays that generated readouts for a comprehensive panel of biomarkers from liver biopsies from liver transplantation patients. Our ndings revealed increased in ltration of CD4 + T and CD8 + T cells in the acute rejection and non-acute rejection groups as compared with the normal group. The proportion of CD4 + and CD8 + T cells in the acute rejection group was the highest. There were also some interesting note was that it was di cult to observe Tregs in the normal and non-acute rejection groups, but the proportion of Tregs and Th17 was the highest in the acute rejection group. The above results showed that the proportion of CD4 + T cells, CD8 + T cells, Tregs and Th17 increased during acute rejection after liver transplantation, and these cells were involved in the development of acute rejection or immune tolerance after liver transplantation. Calvani et al 32 detected simultaneously NK cells, macrophages, and T cells to determine their intra-or extravascular localization using multiplex immuno uorescence assays, demonstrating the feasibility and utility of multiplex immuno uorescence imaging to study and better understand the kidney allograft rejection process. The above results show that different types of lymphocytes in ltrate different immune states after organ transplantation. The mechanism through which different immune cells participate in rejection or immune tolerance is different.
When acute rejection occurs, interlobular bile duct and vascular endothelial cells are the primary targets of immune attack 33 . Many immune cells in ltrate the mixed portal area and hepatic lobules, and cause venous endothelial in ammation and bile duct in ammation 34 . Similarly, our immuno uorescence staining results showed that in ltrating T cells could be seen in the portal area, including a large number of CD8 + T cells, CD4 + T cells, Tregs and Th17 in the acute rejection group. The number of in ltrating T lymphocytes in the non-acute rejection group was less than that in the acute rejection group. We found a novel phenomenon: more lymphocytes in ltrated hepatic lobules under higher levels of the rejection, but the number of CD4 + sinusoidal endothelial cells decreased signi cantly. Previous studies have showed that tissue-resident immune cells are important for organ homeostasis and defense, and the epithelium may contribute to these functions directly or by crosstalk with immune cells 35 . The mechanism of interaction between T cells and sinusoidal endothelial cells in acute rejection of liver transplantation needs further study.
The above results showed that the distribution of in ltrating T cells was different in the hepatic lobules and portal area. Due to the complexity of hepatic lobule structure, it can be divided into three regions according to different metabolic functions. The gene expression patterns of the spatial axis from the central vein to portal area are different, at the same time, its function is gradual 10 . Because liver lobules contain a spatially polarized immune system "immune zonation", the types and numbers of immune cells in ltrating hepatic lobules are different in different disease states 12 . Kupffer cells can localize in ltrating neutrophils to the periportal regions of the lobule, limiting damage to cells that reside around the central vein 36 . It remains unclear whether there are differences in the types and functional status of immune cells in different regions of hepatic lobules under different rejection conditions after liver transplantation. We used multiple immuno uorescence staining to study the in ltration and distribution of T cells in different areas of hepatic lobules, and the spatial distribution of various types of T cells in different areas of hepatic lobules under acute rejection and non-acute rejection after liver transplantation. The results showed that the in ltrated immune cells had some spatial speci city and some exciting phenomena were found after liver transplantation. In the same area of the hepatic lobule, the percentages of CD4 + T cells, CD8 + T cells and Tregs in the rejection group was higher than that in the other two groups, which indicated that the percentage of these T cells increased in each area of the hepatic lobule when rejection occurred. Within three group, percentage of CD4 + T cells, CD8 + T cells and Tregs from periportal zone to perivenous zone were initially increased and then decreased. Interestingly, the percentage of CD8 + T cells from the periportal to perivenous zones increased gradually, the percentage of CD8 + T cells in the perivenous zone was higher than that in the other two zones in the rejection group. Renal allograft biopsies showing c-aABMR show a predominance of in ltrating CD8 + T cells, and increased numbers of interstitial FOXP3 + T cells are associated with inferior allograft survival 37 . Similarly, the distribution of CD8 + T cells in different regions of hepatic lobules may be closely related to rejection level after liver transplantation.
There were some limitations in our study, such as the small number of clinical samples, the di culty in obtaining liver biopsy specimens, and the incomplete structure of hepatic lobules in liver biopsy tissue. These de ciencies may have led to instability of the results. Therefore, this study needs to be further improved to nd other novel results. In conclusion, we believe that the percentage of CD8 + T cells in different regions of hepatic lobules is closely related to rejection level after liver transplantation. We observed an important heterogeneity in the global composition of the in ammatory burden during liver allograft rejection. Acute rejection of liver transplantation may occur when the percentage of CD8 + T cells in perivenous zone increase, and Tregs increases reactively.
The number and proportion of CD8 + T cells increased gradually from the periportal to perivenous zones under different rejection levels. Although the percentage of regional CD4 + T did not re ect the rejection level, the number of CD4 + and CD8 + T cells in different regions was closely related to the rejection level.

Declarations
Ethics approval and consent to participate The study protocol was approved by the ethics committee of Beijing Friendship Hospital, Capital Medical University, Beijing, China Consent for publication All authors have reviewed the manuscript and have given consent for publication.