Recruitment of Neutrophils Mediated by Vγ2 γδ T Cells Deteriorates Liver Fibrosis Induced by Schistosoma japonicum Infection in C57BL/6 Mice

ABSTRACT Conventional adaptive T cell responses contribute to the pathogenesis of Schistosoma japonicum infection, leading to liver fibrosis. However, the role of gamma-delta (γδ) T cells in this disease is less clear. γδ T cells are known to secrete interleukin-17 (IL-17) in response to infection, exerting either protective or pathogenic functions. In the present study, mice infected with S. japonicum are used to characterize the role of γδ T cells. Combined with the infection of S. japonicum, an extremely significant increase in the percentage of neutrophils in the CD45+ cells was detected (from approximately 2.45% to 46.10% in blood and from 0.18% to 7.34% in spleen). Further analysis identified two different γδ T cell subsets that have different functions in the formation of granulomas in S. japonicum-infected mice. The Vγ1 T cells secrete gamma interferon (IFN-γ) only, while the Vγ2 T cells secrete both IL-17A and IFN-γ. Both subtypes lose the ability to secrete cytokine during the late stage of infection (12 weeks postinfection). When we depleted the Vγ2 T cells in infected mice, the percentage of neutrophils in blood and spleen decreased significantly, the liver fibrosis in the granulomas was reduced, and the level of IL-17A in the serum decreased (P < 0.05). These results suggest that during S. japonicum infection, Vγ2 T cells can recruit neutrophils and aggravate liver fibrosis by secreting IL-17A. This is the first report that a subset of γδ T cells plays a partial role in the pathological process of schistosome infection.

S chistosomiasis is endemic to many tropical and subtropical regions globally, including China and the Philippines, where Schistosoma japonicum is endemic (1). In total, schistosomiasis affects more than 200 million people (2). Progression of schistosomiasis from the time of egg deposition through the development of mature granulomas in the liver and hepatic fibrosis has been associated with distinct temporal gene expression patterns (3). Based on these patterns, neutrophils may play a significant role in determining the outcome of S. japonicum infection. The recruitment of neutrophils to the liver has been associated with the development of fibrosis in other chronic liver diseases (4)(5)(6), suggesting they could contribute to fibrosis in schistosomiasis.
Interleukin-17 (IL-17) has been linked to neutrophil infiltration in the liver during schistosomiasis caused by S. japonicum (7,8) and is related to the development of liver fibrosis (1). Of the cells known to secrete IL-17, gamma-delta (␥␦) T cells play a crucial role in the immune system. These cells represent a small population of the overall T lymphocytes (0.5% to 5%) and are known to be the first line of host defense against pathogens, including those causing malaria and tuberculosis (9). ␥␦ T cells have been shown to secrete Th1 (gamma interferon [IFN-␥] and tumor necrosis factor alpha [TNF-␣]), Th2 (IL-4 and IL-10), and antigen-presenting cells like cytokines (IL-12) under different circumstances. They have also been shown to produce IFN-␥, IL-17, IL-4, IL-5, IL-10, IL-13, transforming growth factor beta (TGF-␤), and granulocyte-macrophage colony-stimulating factor (GM-CSF). There are several distinct subsets of ␥␦ T cells that have different functions in inflammation and autoimmunity (10,11). Ex vivo, ␥␦ T cells in the intestinal intraepithelial lymphocyte (IEL) population can be activated with anti-T cell receptor (TCR) antibody, and the V␥2 T cells in particular produce IL-17 (12).
It is unclear whether ␥␦ T cells contribute to liver fibrosis during S. japonicum infection (13,14). In one study in which ␥␦ Ϫ\Ϫ mice were infected with S. japonicum, there was no evidence that ␥␦ T cells were required for granuloma formation in the liver (15). Moreover, when RAG Ϫ/Ϫ mice were infected, granulomas still developed (16). However, another study found that IL-17 secretion by liver lymphocytes was significantly enhanced by S. japonicum infection (1). ␥␦ T cells have been shown to recruit neutrophils by secreting IL-17A in models such as breast cancer and Listeria monocytogenes infection (17). Sporadic reports address the behavior of ␥␦ T cells in S. japonicum-infected mice, but they do not address interaction with neutrophils (18)(19)(20). Therefore, to better understand the role of ␥␦ T cells in S. japonicum infection, we used the C57BL/6 mouse infection model to characterize the cytokine profile and effects of ␥␦ T cell function on neutrophils.

RESULTS
The percentages of neutrophils in the blood and spleen were increased following S. japonicum infection. To begin, we first characterized the expansion of neutrophils in the peripheral blood and spleen of S. japonicum-infected C57BL/6 mice. Following infection, the percentage of neutrophils (CD45 ϩ CD11b ϩ Ly6g ϩ Ly6c ϩ F4/80 Ϫ ) in the white blood cell (CD45 ϩ ) population remained low in the peripheral blood from 0 (2.45%) to 4 (1.57%) weeks postinfection, and then it increased beginning 6 weeks (8.18%) and peaked 8 weeks (46.1%) postinfection, still remaining high at the 12th week (45.5%). The difference between the groups was significant (P Ͻ 0.001). The differences between the uninfected and 8-week groups (P Ͻ 0.01), uninfected and 12-week groups (P Ͻ 0.001), 4-and 8-week groups (P Ͻ 0.01), 4-and 12-week groups (P Ͻ 0.001), 6-and 8-week groups (P Ͻ 0.01), and 6-and 12-week groups (P Ͻ 0.01) were significant. Similar trends were observed in the spleen, which were 0.18%, 0.06%, 1.12%, 2.56%, and 7.34%, respectively. The differences between the groups was significant (P value of Ͻ0.001 by one-way analysis of variance [ANOVA]). The difference between uninfected and 12-week groups (P Ͻ 0.001), 4-and 12-week groups (P Ͻ 0.001), 6-and 12-week groups (P Ͻ 0.01), and 8-and 12-week groups (P Ͻ 0.01) were significant ( Fig. 1A and B). In addition, a population of CD11b Ϫ Ly6g low cells/lymphocytes (CD45 ϩ ) was approximately 0.38% to 6.76% in the blood, 0.17% to 0.64% in the spleen, and 2.50% to 40.10% in the liver of both the uninfected and infected mice. On the other hand, during the later stage of infection (after 8 weeks), the CD11b ϩ Ly6G low population increased markedly and may represent either eosinophils or macrophages, as most were F4/80 ϩ (data not shown) (21).
Given that the neutrophils were CD11b ϩ Ly6G ϩ , similar to myeloid-derived suppressor cells (MDSCs), we cocultured CD3 ϩ T cells labeled with carboxyfluorescein succinimidyl ester (CFSE) isolated from spleens of uninfected mice with neutrophils isolated from the spleens of mice which were 8 weeks postinfection with S. japonicum at a 1:1 ratio to determine whether the neutrophils could inhibit T cell proliferation (22). The neutrophils significantly reduced the number of proliferated CD4 ϩ T cells (69.4% T cells alone versus 12.3% in the coculture; P Ͻ 0.001) and CD8 ϩ T cells (68.6% in T cells alone versus 16.1% in coculture; P Ͻ 0.001) ( Fig. 1C and D). Therefore, it is likely that the neutrophils expanding during S. japonicum infection in this model play an inhibitory role and block the function of T cells (23).
Pathological changes in the liver caused by neutrophils. The formation of egg granulomas in the liver is the primary pathology associated with S. japonicum infection in C57BL/6 mice (13). Therefore, we used immunofluorescence staining to describe the relationship between neutrophils and egg granulomas in the liver.
Seven weeks postinfection, neutrophils tended to cluster around the edge of the granuloma ( Fig. 2A). By 9 weeks postinfection, the granulomas had enlarged and Ly6g high neutrophils began to appear around the new eggs ( Fig. 2A). Finally, during the late stage of the disease (11 weeks postinfection), the edges of the granuloma had blurred and the neutrophils were distributed throughout the liver ( Fig. 2A). Interestingly, 8 weeks postinfection, the CD11b ϩ Ly6g ϩ neutrophils were observed situated around the edge of the granulomas in the liver (Fig. 2B). As in the flow cytometry analysis of the situation in the blood and spleen, a population of CD11b Ϫ Ly6G ϩ cells was present in the liver regardless of infection status (Fig. 2C).
Changes in the cytokine profile following S. japonicum infection. Previous studies have shown that cytokines, particularly IL-17, may be associated with neutrophil infiltration into the liver (24). In addition, several studies have suggested roles for IFN-␥, IL-1␤, IL-10, IL-4, and G-CSF in the pathogenesis of S. japonicum infection (13). Therefore, we used a cytokine bead array assay to measure the levels of IL-17, IFN-␥, IL-1␤, IL-10, IL-4, and G-CSF in the serum over time (Fig. 3).
The first cytokine that increased was IFN-␥, which was elevated from 4 weeks postinfection, decreased after the fifth week, and was restored to normal levels after the ninth week. The difference between the groups was significant (P Ͻ 0.001). The peak point was approximately 28,988 fg/ml at 5 weeks postinfection and was significantly different from those of the other groups (P Ͻ 0.001).
IL-1␤ increased from 5 weeks postinfection, decreased after the sixth week, and returned to uninfected levels after the ninth week. The difference between the groups was significant (P Ͻ 0.001). The peak point was approximately 9,290 fg/ml at the sixth week after infection and was significantly different from those of other groups (P Ͻ 0.001).
IL-10 was elevated 6 weeks after infection, decreased after the seventh week, and returned to normal levels after the tenth week. The difference between the groups was  significant (P Ͻ 0.01). The 7-week group (30,435 fg/ml) was significantly different from the other groups (P Ͻ 0.05), except for the 6-week (6,576 fg/ml) and 9-week (4,203 fg/ml) groups (P Ͼ 0.05).
IL-17 increased from 6 weeks postinfection, declined after the seventh week, and returned to uninfected levels after the ninth week. The difference between the groups was significant (P Ͻ 0.05). The difference between the 7 (1,606 fg/ml)-and 10 (28.8 fg/ml)-week groups was significant (P Ͻ 0.05).
IL-4 slowly increased from 5 weeks postinfection, maintaining a high level from the 7th week to the end of the observation period. The difference between the groups was significant (P Ͻ 0.001). The 7-week group (429 pg/ml) was significantly different from the uninfected (0 pg/ml), 4-week (2.8 pg/ml), and 5-week (12.8 pg/ml) groups (P value of Ͻ0.05 for each). The 9-week group was significantly different from the uninfected (P Ͻ 0.01), 4-week (P Ͻ 0.01), and 5-week (P Ͻ 0.05) groups, respectively. G-CSF began increasing 4 weeks postinfection, severely declined at week 7, and then increased again during the ninth week. The difference between the groups was significant (P Ͻ 0.01). The difference between the uninfected (32.6 pg/ml) and 9-week (663 pg/ml) groups was significant (P Ͻ 0.05).

The function of different subsets of ␥␦ T cells declined postinfection. ␥␦ T cells
have the capacity to produce a wide array of cytokines and function as an innate and adaptive immune cell, primarily in the mucosa and, to a lesser extent, in the blood and secondary lymphatic organs (25). Given that ␥␦ T cells can produce all of the cytokines that were altered during S. japonicum infection, we next assessed whether the quantity of ␥␦ T cells changed during S. japonicum infection in this model. However, there was no significant change in the proportion of the total ␥␦ T cells/CD3 ϩ T cells (data not shown).
Therefore, we looked more closely at the function of the V␥1 and V␥2 T cell subsets to determine if functional changes were occurring. Both subsets produced IFN-␥; however, only the V␥2 T cells produced IL-17 (Fig. 4A). Cytokine production from both the V␥1 and V␥2 T cell subsets decreased during the course of the infection (Fig. 4A and  B). On the other hand, during observation of liver slices, the V␥2 T cells were located beside the eggs or lying inside the granuloma (Fig. 4C). The percentage of IL-17A ϩ /V␥2 T cells changed following infection (P Ͻ 0.001). The 8-week (2.6%) group was significantly different from the uninfected (10.6%), 4-week (9.2%), and 6-week (7.8%) groups (P Ͻ 0.05). The 12-week (0.7%) group was significantly different from the uninfected (P Ͻ 0.001), 4-week (P Ͻ 0.01), and 6-week (P Ͻ 0.01) groups.

The function of V␥2 T cells in vivo.
Previous studies have shown that ␥␦ T cells are the major producers of IL-17A, rather than Th17 ␣␤ T cells, and that IL-17A induces an inflammatory response in neutrophils (8,(26)(27)(28). To determine the relationship between ␥␦ T cells and neutrophils during infection, we depleted ␥␦ T cells from mice infected with S. japonicum from the fifth week after the parasites had laid eggs in the liver to the seventh week, when the level of IL-17A had increased. While there were no significant differences between the depleted and nondepleted groups (P Ͼ 0.05), there was a reduction in the average percentage of neutrophils/CD45 ϩ cells (28.6% in the blood, 2.1% in the spleen) and the serum level of IL-17A (875 fg/ml) ( Fig. 5A and B). When the V␥2 T cells were specifically depleted, the percentage of neutrophils in the blood and spleen declined significantly, from 43.30% to 24.36% in the blood (P Ͻ 0.01) and 3.13% to 1.78% in the spleen (P Ͻ 0.01) (Fig. 5A). The percentages of neutrophils/ CD45 ϩ cells were significantly different when IL-17A was depleted, with an average level of 18.4% in the blood (P Ͻ 0.01) and 0.5% in the spleen (P Ͻ 0.001) (Fig. 5A). The absolute level of IL-17A in the serum was concurrently reduced from 1,797 fg/ml to 243 fg/ml (P Ͻ 0.05) (Fig. 5B). On the other hand, there was no significant difference after blocking the V␥1 T cells regarding both the percentage of neutrophils/CD45 ϩ cells and the level of IL-17A (P Ͼ 0.05).
The levels of hyaluronic acid (HA) and collagen type III (PC-III) in the serum were tested to analyze the degree of liver fibrosis. When the V␥2 T cells were blocked, HA declined from 1,639 to 1,018 ng/ml (P Ͻ 0.05) and the PC-III level dropped from 112 to 81 ng/ml (P Ͻ 0.01). When IL-17A was depleted, HA declined to 880 ng/ml (P Ͻ 0.05) and the PC-III level dropped to 78 ng/ml (P Ͻ 0.01). There was no significant difference when the ␥␦ T and V␥2 T cells were depleted (P Ͼ 0.05) (Fig. 5C). The percentage of collagen fiber/granuloma in the liver was significantly reduced when the V␥2 T cells (from 18.8% to 13.6%; P Ͻ 0.01) or IL-17A (from 18% to 13.4%; P Ͻ 0.01) was blocked ( Fig. 5D and E). The difference in the anti-␥␦ T (19.5%) or anti-V␥1 T (21.1%) cell group was not significant (P Ͼ 0.05).
Adoptive transfer of ␥␦ T cells to RAG ؊/؊ mice prior to infection with S. japonicum. To confirm the function of ␥␦ T cells in mice with schistosomiasis, the adoptive transfer of the cells from the spleen of wild-type (WT) mice into RAG Ϫ/Ϫ mice, which lacked mature T cells and B cells, was implemented. First, the RAG Ϫ/Ϫ mice and the WT mice were infected with S. japonicum at the same time. As a result, there were lots of neutrophils around the eggs of S. japonicum in the center of the egg granuloma in the infected RAG Ϫ/Ϫ mice (Fig. 6A). It is suggested that neutrophils are recruited by the eggs with or without the help of T/B cells in the RAG Ϫ/Ϫ mice. Compared to the egg granuloma in liver of WT mice owning plenty of eosinophils (CD170 ϩ cells), there were few eosinophils in the granuloma in liver of RAG Ϫ/Ϫ mice during the early stage of infection (5 to 6 weeks).
Six weeks postinfection with S. japonicum, the ␥␦ T cells and CD4 ϩ T cells from spleen of infected WT mice were isolated and then transferred into different groups of infected RAG Ϫ/Ϫ mice. One week after the adoptive transfer, none of the neutrophil proportions in the blood, spleen, or liver of RAG Ϫ/Ϫ mice were increased compared to those of the nontransferred group (P Ͼ 0.05) (Fig. 6B).
On the other hand, in the group to which CD4 ϩ T cells were transferred, the eosinophils appeared around the edge of egg granuloma in the liver of RAG Ϫ/Ϫ mice. As shown in Fig. 6A, the eosinophils were newly recruited because they were not scattered in the egg granuloma, as is the case for WT mice. It is suggested that the recruitment of eosinophils is involved in the help of CD4 ϩ T cells, while the ␥␦ T cells could not have the same effect. The RAG Ϫ/Ϫ mice began dying beginning at the eighth week postinfection with S. japonicum, and 2 out of 20 mice were still alive at the tenth week.

DISCUSSION
The main focus of our study was to understand the factors driving neutrophil accumulation in the liver during S. japonicum infection, particularly in the context of ␥␦ T cells. In our model, the number of neutrophils in the liver, blood, and spleen began to increase 6 weeks after infection and remained elevated through 12 weeks postinfection. Concurrent with the increase in neutrophils, IL-17 levels were increased. While the number of IL-17-producing V␥2 T cells did not markedly increase following infection, we showed by depleting ␥␦ T cells that they were helping to recruit neutrophils to granulomas caused by S. japonicum eggs and aggravating liver fibrosis. The serum levels of IL-17A and IFN-␥ were also reduced in the ␥␦ T cell-depleted mice. Interestingly, the ability of ␥␦ T cells to produce cytokines (IL-17A and IFN-␥) decreased as the infection progressed. The ␥␦ T cells may also be recruiting neutrophils by secreting CCL22, CCL1, IL-23␣, IL-3, or IL-4.
While it is likely that ␥␦ T cells contribute to the pathogenesis of schistosomiasis, it is also important to note that they were not necessary for granuloma formation, as mice depleted of ␥␦ T cells formed granulomas. In addition, neutrophils were found around the liver granulomas when we depleted the V␥2 TCR. There is also evidence that blocking IL-17A reduces the number of neutrophils present in the liver but does not eliminate them (1). On the other hand, the granuloma formation in the livers of RAG Ϫ/Ϫ mice also suggests that the formation is dominated by innate immunity. Thus, it is likely that the presence of S. japonicum eggs recruits neutrophils to the liver with or without the help of T or B cells (29,30).
Granulomas progress through several key stages as they mature. During the early stage of formation, the presence of parasite eggs rapidly recruits neutrophils that are similar to MDSCs in terms of phenotype and function (31,32). During granuloma enlargement, neutrophils with a CD11b ϩ Ly6G ϩ phenotype were present at the border of the granulomas. Our results suggest that these neutrophils are distinct (high  's t test). fluorescence) from the neutrophils in the middle of the granuloma (low fluorescence), suggesting they have a different function. In addition, eosinophils tended to be grouped toward the middle of the granuloma while the neutrophils were on the periphery, suggesting they have antagonistic functions. During late infection (Ͼ8 weeks postinfection), the CD4 ϩ T cells, ␥␦ T cells, and CD8 ϩ T cells all tend to have reduced cytokine-producing abilities (33)(34)(35)(36)(37). Notably in our study, ␥␦ T cells were no longer capable of producing IL-17A. These observations suggest a complex interaction whereby ␥␦ T cells enhance granulation by producing IL-17A and recruiting neutrophils, which then inhibit T cell function (e.g., proliferation) (22, 38) and cause fibrosis, preventing granulation from continuing with unlimited expansion (39). A subset of innate IL-17-producing ␥␦ T cells is produced during fetal development (IL-1R ϩ IL-23R ϩ CCR6 ϩ ), termed T␥␦17, which have an innate ability to make IL-17 (40,41). In adults, T␥␦17 cells in the thymus preferentially express V␥4 (42). Their ability to produce IL-17 is not related to TCR triggering but to the transcription factors SOX4, SOX13, TCF1, and LEF1. SOX4 and SOX13 directly regulate the two requisite T␥␦17 cell-specific genes Rorc and Blk. The TCF1 and LEF1 transcription factors counter the SOX proteins and induce expression of alternative effector genes. In addition, the development of natural T␥␦17 cells is also regulated by the transcription factor ETV5 (43,44). Mature ␥␦ T cells no longer require TCR stimulation to produce IL-17 and can produce the cytokine in response to IL-1 and IL-23 alone (17,30). However, in this study, when we blocked the ␥␦ TCR, the IL-17A level in the serum declined.
The V␥1 and V␥2 subsets of ␥␦ T cells seem to have distinct roles in S. japonicum infection. The V␥1 cells produce IFN-␥, which is an inflammatory cytokine that helps to recruit macrophages and eosinophils. In contrast, V␥2 T cells produce both IL-17A and IFN-␥. IL-17A recruits neutrophils with an MDSC-like phenotype that inhibit inflammation. In previous studies, cells expressing V␥1 have been shown to reduce bacterial clearance in a Listeria model, although ␥␦ T cells as a whole promote clearance (3,13). In this study, both subsets lost the ability to produce cytokine as the disease progressed. It is possible that IL-4 or other immunosuppressive cytokines have contributed to the loss of function by creating immunosuppressive microenvironments. This is the first finding that subsets of ␥␦ T cells play a role, albeit partial, in the pathological process of schistosome infection. The neutrophils comprised a large part of the cells forming granulomas in this model. However, our study is limited in the following ways: (i) we did not transfer the V␥2 T cells to the RAG Ϫ/Ϫ mice to demonstrate their function by acquisition of a trait; (ii) we did not measure the IL-17A in the serum after the cells were adoptively transferred to a RAG Ϫ/Ϫ host; and (iii) fully defining the interaction between CD4 ϩ T cells and CD170 ϩ eosinophils requires further study. Animals. C57BL/6 (WT) mice, 6 to 8 weeks of age, were purchased from Shanghai Laboratory Animal Center, Chinese Academy of Sciences. RAG1 Ϫ/Ϫ mice were obtained from the Model Animal Research Center of Nanjing University (China). All mice were maintained under specific-pathogen-free (SPF) conditions at the National Institute of Parasitic Diseases.

MATERIALS AND METHODS
Cercariae were provided by the Key Laboratory of Parasite and Vector Biology, Ministry of Health (Shanghai, China). Each mouse was infected through the skin with 20 cercariae.
For the adoptive transfer assay, the CD3 ϩ cells from spleen of WT mouse infected for 4 weeks with S. japonicum were purified by a magnetic column (Miltenyi Biotec, Bergisch Gladbach, Germany). The CD4 ϩ T cells and ␥␦ T cells were isolated by BD FACSAria III flow cytometry and subjected to CD4 ϩ or ␥␦ TCR ϩ labeling. The CD4 ϩ T cells and ␥␦ T cells (2 ϫ 10 5 ) were injected through the tail vein into each RAG1 Ϫ/Ϫ mouse.
Flow cytometry. Blood samples were collected in tubes containing heparin. Spleen samples were mashed through a 70-mm cell strainer. Liver samples were incubated in collagenase IV for 30 min, mashed through a 70-mm cell strainer, and then separated from hepatocytes by centrifugal mixing with 35% isotonic Percoll. All single-cell suspensions were treated with NH 4 Cl erythrocyte lysis buffer. Cells were stained with directly conjugated antibodies (listed below) for 30 min at 4°C in the dark in phosphate-buffered saline-1% bovine serum albumin (PBS-1% BSA). Fixable viability dye eFluor 780 (1:1,000; eBioscience, San Diego, CA, USA) was added to exclude dead cells. For intracellular staining, single-cell suspensions were stimulated in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum (FBS), 100 IU/ml penicillin, 100 mg/ml streptomycin, 0.5% ␤-mercaptoethanol, 81 nM phorbol myristate acetate, 1.34 M ionomycin, and 10.6 M brefeldin A (eBioscience) for 4 h at 37°C. Surface antigens were stained first, followed by fixation and permeabilization using the Foxp3 staining buffer set (eBioscience) and then staining of intracellular proteins. All experiments were performed using a Cyan ADP flow cytometer with Submit software. Data analyses used FlowJo software, version 9.7.1.
All antibodies were purchased from eBioscience, except that Ly6G and IFN-␥ were from BD Biosciences (San Diego CA, USA), V␥1 was from BioLegend (San Diego CA, USA), CCR2 was from R&D Systems Immunofluorescence. Tissues were embedded in optimal cutting temperature compound (OCT), frozen in liquid nitrogen, and then cut into 6-m pieces. Tissue on the plate was washed three times with PBS, fixed in acetone for 15 min, and then washed three times with PBS. The tissue was then incubated in blocking buffer (10% goat serum, 5% BSA) for 1 h at 25°C prior to addition of mixed monoclonal antibody culture supernatant at a 1:50 dilution at 4°C overnight. The antibodies then were removed by repeated washing in PBS, and the coverslips were mounted on 10 l of mounting liquid (including 4=,6-diamidino-2-phenylindole [DAPI]). Ly6G-PE-CF594 and CD11b-FITC were used for neutrophils. Images were taken on a confocal microscope (A1; Nikon) using the provided software (NIS-Elements).
Masson's trichrome staining. Fresh liver tissues (round pieces about 7 mm in diameter cut from the edge of the liver) were fixed in 4% formaldehyde overnight and routinely paraffin embedded. Paraffin sections (5 m) were prepared from each liver tissue sample. The liver tissue sections were stained by Masson's trichrome staining to evaluate collagen content and distribution. The collagen fibers were stained blue, cell nuclei were stained black, and the background was stained red. Each stained section was examined by optical microscopy with ϫ200 magnification and identical settings. Thirty pictures were taken from three sections from each tissue, which included the egg granulomas in the center. The percentage of granuloma areas with collagen-positive color (blue), i.e., the positive blue color area/ granuloma area (measured as a percentage), was analyzed using Image-Pro Plus 6.0 software. Every picture was evaluated in double-blind fashion by two independent investigators.
Cytokine analysis. Cytometric bead array was performed according to the manufacturer's instructions to detect the serum levels of IFN, IL-10, IL-1␤, IL-17A, IL-4, and G-CSF at different infection stages.
HA was measured using an enzyme-linked immunosorbent assay (ELISA) kit purchased from R&D Systems (Minneapolis, MN, USA). PC-III was detected by an ELISA kit obtained from Jiancheng Bioengineering Institute (Nanjing, Jiangsu, China).
Real-time PCR. Total RNA was extracted from fresh liver tissue homogenized in TRIzol reagent (Invitrogen) according to the manufacturer's protocol. RNA purity and concentration were assessed by spectrophotometry. Reverse transcriptase (RT) reactions for cDNA synthesis were performed using PrimeScript RT master mix (TaKaRa Bio, Tokyo, Japan). The relative mRNA expression level was determined by real-time quantitative PCR (qPCR) with a SYBR green I PCR master (TaKaRa) kit on an ABI ViiATM7 machine according to the manufacturer's protocol.
For the qPCR array assay, ␥␦ T cells were isolated using a BD FACSAria III flow cytometer with CD3 ϩ ␥␦ TCR ϩ labeling. The purity of isolated ␥␦ T cells was validated by flow cytometry. Since the ␥␦ T cells are only a small part of whole white blood, cells from spleens of 10 normal mice and 4 infected mice were used for each sample.
Statistics. All of the data were analyzed using SPSS 13.0 and GraphPad Prism, version 5, using one-way analysis of variance, except for the blocking tests, which used a t test, with a P value of Ͻ0.05 indicating significance.