Treatment with Recombinant Trichinella spiralis Cathepsin B–like Protein Ameliorates Intestinal Ischemia/Reperfusion Injury in Mice by Promoting a Switch from M1 to M2 Macrophages

Intestinal ischemia/reperfusion (I/R) injury is a grave condition resulting from acute mesenteric ischemia; hemorrhagic, traumatic, or septic shock; severe burns; and some surgical procedures, including small bowel transplantation and abdominal aortic surgery, and plays an important role in the pathogenesis of systemic inflammation and multiple organ dysfunction (1), which contributes to a high morbidity and mortality (2, 3). The precise mechanisms involved have not been fully elucidated despite intensive investigations, as such effective preventive strategies are still lacking.

It is believed that both the innate and adaptive immune systems have an important role in I/R injury (4–6). Previous studies focused on the role of innate immune response such as reactive oxygen species, cytokines and chemokines, the complement system, and neutrophils (4, 7). In contrast, the adaptive immune system has not yet been appointed a determinant role in the production of tissue injury in classical models of I/R injury. Macrophages are an essential element in the orchestration and expression of innate immunity and adaptive immune responses. Increasing evidence has proved that macrophages participate actively in I/R injury, and macrophage depletion has been shown to diminish organ damage in models of intestinal I/R injury (8–13). Moreover, macrophages seem to modulate the recruitment of neutrophils that occurs hours after intestinal I/R injury (12). However, the mechanism by which macrophages contribute to intestinal I/R injury is not completely understood.

Macrophages exhibit a variety of phenotypes broadly classified into classically (M1) and alternatively (M2) activated macrophages, a phenomenon that has been described as macrophage polarization or heterogeneity (14, 15). Different phenotypes play various roles in damage and maintenance of tissues. For example, M1 macrophages are induced by exposure to LPS or Th 1 cytokines IFN-γ and TNF-α, and they are characterized by the expression of high levels of inducible NO synthase 2 (NOS2), as well as many proinflammatory cytokines such as IL-1β, TNF-α, and IL-6, enhancing immune responses and tissue injury (16). In contrast, exposure of macrophages to Th2 cytokines IL-4 or IL-13, inhibits expression of these proinflammatory markers and instead activates expression of high levels of arginase-1 (Arg-1) and mannose receptor, releasing IL-10 and TGF-β1 to modulate the inflammatory response and to promote tissue repair (17). Therefore, the switch from a proinflammatory (M1) to an anti-inflammatory phenotype (M2) supports the transition from tissue injury to tissue repair in I/R injury. However, whether the macrophage phenotypes switch is involved in intestinal injury induced by intestinal I/R is not clear.

In recent years, experimental and clinical evidence has accumulatively demonstrated that helminth infections or their excretory/secretory (ES) products could protect hosts from immune-mediated diseases, such as inflammatory bowel diseases, multiple sclerosis and allergic disease (18–22). Moreover, the proteins released from helminthes preferentially activate Th2 response and suppress Th1-prone gastrointestinal inflammation, and they promote the switch from M1 to M2 macrophages through STAT6 signaling (19, 23). Treatment with recombinant Trichinella spiralis–specific 53-kDa glycoprotein (rTsP53) showed a preventive effect on trinitrobenzene sulfonic acid–induced colitis by evoking an M2-bias innate immunity and stimulating Th2 responses in our previous study (24). Furthermore, in an in vitro study, rTsP53 showed anti-inflammatory function by inducing M2 macrophages via acting dependently on STAT6 (25). Recently, vaccination with recombinant Trichinella spiralis cathepsin B–like protein (rTsCPB) cloned/expressed by our team also showed an IgG1 (Th2)-prone response in mice (26). However, whether rTsCPB can regulate intestinal immune response by promoting STAT6 dependent M1 to M2 transition and show intestinal protective benefits after intestinal I/R remains unclear.

Based on the above findings, the current study was to 1) explore whether macrophage phenotypes switch is involved in intestinal injury induced by intestinal I/R and 2) investigate the effect of rTsCPB on intestinal I/R-induced intestinal injury and the potential mechanisms related to macrophage phenotypes switch involving STAT6 signaling.

The cDNAs encoding full-length TsCPB were subcloned into the pET-28a (+) vector. The recombinant proteins were expressed in Escherichia coli BL21 (DE3) and purified with Ni-affinity chromatography as described previously by our team (26). Endotoxin was removed from rTsCPB as in our previous study (26).

The study was approved by the Animal Care Committee of Sun Yat-sen University, China, and the handling of experimental animals was performed in compliance with National Institutes of Health guidelines. Male BALB/c mice aged 4–5 wk were purchased from Guangdong Medical Laboratory Animal Center, China. Mice were housed under controlled temperature and humidity conditions with a 12-h light-dark cycle, and were fed commercial mouse chow with water ad libitum for 1 wk before protocol entry.

Three vaccine groups of mice were administered rTsCPB (groups T and AS, please see Experimental Protocols) formulated with complete or incomplete Freund’s adjuvants (Sigma), or Freund’s adjuvants (group F) only, respectively. Each mice was vaccinated s.c. with 50 μg rTsCPB formulated with different adjuvants in a total volume of 100 μl as described previously by our team (26). Mice were boosted twice with the same dose and formulation of rTsCPB at 2-wk intervals. The injury group (group I) was given PBS only. Mice serum samples were collected after immunization for measurement of Ab response before the surgical procedure.

All animals aged 11–12 wk, weighing 26–32 g, were fasted overnight before the experiment, but they had free access to water ad libitum. Animals were anesthetized with sevoflurane and were placed in a supine position with their paws taped to the operating table. The model of intestinal I/R was established as in our previous study (27). Briefly, the small intestine was exteriorized with a 1-cm midline laparotomy, and the superior mesenteric artery (SMA) was identified. The SMA was occluded with a noncrushing microvascular clip for 60 min as described previously (28, 29). Ischemia was recognized by the existence of pulseless or pale color of the small intestine. The laparotomy incision was temporarily closed during SMA occlusion and than reopened to remove the clip. The return of the pulses and the reestablishment of the pink color were assumed to be the valid reperfusion of the intestine. During the study period, body temperature was maintained at 37°C with the help of a heating pad. The animals were resuscitated with a 0.5-ml s.c. injection of saline solution just after reperfusion and provided free access to chow and water.

Sham mice underwent laparotomy with identification of the SMA and normal saline resuscitation, but the SMA was not occluded. Daily food intakes from days 1 through 7 and body weight changes on days 1, 2, 4, and 7 were measured.

As Fig. 1 shows, the mice were allocated randomly to five groups: 1) sham operation group (group S); 2) injury group (group I), in which SMA was occluded for 60 min followed by 2 h or 7 d reperfusion; 3) rTsCPB vaccinated group (group T); 4) AS1517499 (Gene Operation, Ann Arbor, MI) treatment group (group AS), in which AS1517499, a selective STAT6 inhibitor, was administered i.p. 1 h before intestinal ischemia and from day 1 to day 7 after reperfusion at a dose of 10 mg/kg/d in a total volume of 100 μl as described previously (30); the rest of the procedures were performed as in the group T; and 5) Freund’s adjuvant–vaccinated group (group F).

In a separate group of animals receiving the same protocols, survival was monitored for 7 d. Animals were monitored every 2 h for the initial 8 h after SMA occlusion, every 8 h for the following 16 h, and every 24 h until the end of the 7 d study period.

Survived mice were sacrificed with an overdose of sevoflurane and laparotomy was performed at 2 h or on day 7 after reperfusion. A segment of 4.0 to 5.0 cm intestine was cut 5 cm away from the terminal ileum (ileocecal valve) and was divided into two segments. The first segment was paraffin embedded for morphologic analysis. The second was washed with PBS and preserved at −80°C for detection of cytokine levels and Western blot analysis (see the following).

Serum samples were used in an ELISA to measure the levels of IgG1 and IgG2a against rTsCPB as described previously (26). Briefly, flat-bottom 96-well microtiter plates (Greiner, Germany) were coated for 2 h at 37°C with 100 μl of rTsCPB at a concentration of 5 μg/ml in 0.05 M bicarbonate buffer (pH 9.6) per well. Serum samples diluted 1:100 (IgG1; Bethyl Laboratories) and 1:100 (IgG2a; Bethyl Laboratories) with 0.05% Tween-20 buffer in PBS (100 μl/well) were then added in duplicate.

The segment of small intestine was stained with H&E, blind-coded by a different person, and evaluated in blinded manner by two independent pathologists. The degree of injury was assessed 2 h after reperfusion using a modified Chiu method (31) according to changes of the villus and glands of the intestinal mucosa. Modified Chiu score was graded as: 0, normal villus and gland; 1, changes at the top of villus and initial formation of subepidermal Gruenhagen antrum; 2, formation of subepidermal Gruenhagen antrum and slightly injured gland; 3, enlargement of subepidermal gap and engorgement of capillary vessel; 4, epidermis moderately isolated with lamina propria and injured gland; 5, top villus shedding; 6, obvious villus shedding and capillary vessel dilating; 7, lamina propria villus shedding, and distinct injured gland; 8, initially decomposed lamina propria; and 9, hemorrhage and ulceration.

Villus height and crypt depth on day 7 after reperfusion were measured in 8–10 well-oriented villus/crypt units for each mouse per group (n = 7–10). All measurements were taken using magnification ×200.

Total RNA was extracted using Trizol reagent according to the manufacturer’s protocol (Invitrogen). The cDNA of each sample was synthesized from 1 μg total RNA using SuperScript II according to the manufacturer’s instruction (Invitrogen). Th1 cytokine (IL-6 and IFN-γ), Th2 cytokine (IL-4 and IL-10), M1 markers (NOS2 and CCR7), and M2 markers (Arg-1, and found in inflammatory zone protein [Fizz1]), were all assessed with quantitative real-time PCR, which was done with a real-time PCR kit in a Bio-Rad iQ5 real-time PCR machine using SYBR Green detection protocol. House-keeping gene β-actin was used as the endogenous control. Gene-specific primers were used to amplify genes, including: IL-6: sense primer, 5′-GACTGATGCTGGTGACAACC-3′; antisense primer, 5′-AGACAGGTCTGTTGGGAGTG-3′; IFN-γ: sense primer, 5′-AACTCAAGTGGCATAGATGTGGA-3′; antisense primer, 5′-CAGGTGTGATTCAATGACGCT-3′; IL-4: sense primer, 5′-AGTTGTCATCCTGCTCTTCTTTCT-3′; antisense primer, 5′-TGTGGTGTTCTTCGTTGCTGT-3′; IL-10: sense primer, 5′-CAGCCAGGTGAAGACTTTCT-3′; antisense primer, 5′-CATTTCCGATAAGGCTTGGC-3′; Arg-1: sense primer, 5′-ACCTGGCCTTTGTTGATGTC-3′; antisense primer, 5′-CAGCACCACACTGACTCTTC-3′; Fizz1: sense primer, 5′- CCCTTCTCATCTGCATCTCC -3′; antisense primer, 5′- CAGTAGCAGTCATCCCAGCA -3′; NOS2: sense primer, 5′-TTTGACGCTCGGAACTGTAG-3′; antisense primer, 5′-GAAGTCATGTTTGCCGTCAC-3′; CCR7: sense primer, 5′-GGTGGCTCTCCTTGTCATTTTC-3′; antisense primer, 5′-AGGTTGAGCAGGTAGGTATCCG-3′; β-actin: sense primer, 5′-GTGACGTTGACATCCGTAAAGA-3′; antisense primer, 5′-GTAACAGTCCGCCTAGAAGCAC-3′.

Dissociation of the PCR products by a melting curve analysis protocol showed consistently specific single melting peaks for all used primer pairs. Relative quantification was performed using the comparative cycle threshold (C_T) method as described previously (32). Relative mRNA expression of target genes was obtained by normalizing to the level of β-actin in sham group with the comparative ΔCt method using the equation 2^−ΔΔCt.

The concentrations of IFN-γ, IL-4, IL-6, and IL-10 in intestine tissue were measured with an ELISA kit according to the manufacturer’s protocol (Nanjing Jiancheng Biochemicals, Nanjing, China). Briefly, 100 mg intestinal tissue was boiled in 1 ml of a mixture of 1 M acetate and 20 mM hydrochloride for 10 min at 100°C, and then centrifuged at 10,000 × g for 10 min at 4°C. This extracted peptide solution was applied to the ELISA plate. All measurements were performed in triplicate, and the intra-assay and interassay variability were <10%. Results were calculated as picograms per milliliter.

To investigate the role of the switch between M1 and M2 macrophages in I/R-induced intestinal injury and the effects of rTsCPB on intestinal injury, protein expression of M1 markers (NOS2 and CCR7), M2 markers (Arg-1 and Fizz1), and phosphorylation of STAT6 involving the switch from M1 to M2 in the intestine were detected with Western blot analysis. Total proteins were prepared as described previously (33). Protein (50 μg/lane) was electrophoresed on 12% SDS-polyacrylamide gels and then electrotransferred onto polyvinylidene fluoride membrane. Abs against NOS2, CCR7, Arg-1, Fizz1, phosphorylated STAT6 (p-STAT6), and STAT6 (Santa Cruz Biotechnology) were used, and the intensities of bands were quantified with Image J software. β-Actin was used as an internal housekeeping control to calculate relative ratio of OD with values compared with those of sham controls.

The small intestines were perfusion fixed with 4% paraformaldehyde and processed for immunostaining (4-μm cryosections). The sections were blocked at room temperature using saline containing 0.1% BSA and 10% goat serum. The following primary Abs were used: anti-CD163 (Santa Cruz Biotechnology), anti-NOS2 (Santa Cruz Biotechnology), anti-myeloperoxidase (MPO; Santa Cruz Biotechnology), and anti-Ki-67 (Dako Cytomation, Carpinteria, CA). For negative controls, tissue sections were treated with mouse or rabbit nonimmune serum instead of the primary Abs. Quantification of cells expressing the specified marker was performed in a blinded fashion by counting positive cells in five randomly chosen 200× fields of each sample (five mice per group). Images were taken at 200× using an Olympus microscopy system.

Data are expressed as mean ± SEM. Biochemical assays were performed in duplicate or triplicate for each specific sample. Therefore, all the data points are means of numbers that themselves are means of duplicate or triplicate measurements for these parameters. Significance was evaluated using one-way ANOVA (SPSS 13.0; SPSS, Chicago, IL) followed by Tukey post hoc multiple comparisons or paired two-tailed Student t test. Survival statistics were compared with a Kaplan–Meier curve and log rank test. Correlation between different variables was assessed by Spearman coefficient; p < 0.05 was considered statistically significant.

The gene encoding rTsCPB was 990 bp in length and encoded a protein of 329 aa with a m.w. of 38 kDa (26). The highly expressed fusion protein was purified by Ni-NTA affinity column. The purified thrombin-cleaved products were analyzed with SDS-PAGE, with a single band in size as expected (Fig. 2A).

As shown in Fig. 3A, the intestine of the sham mice exhibited normal mucosal architecture, which had no obvious intestinal damage, with the mucosal surface intact and the villus height maintained. Mice subjected to I/R showed damage to the intestinal wall with reduction in villus height and marked destruction of epithelial cells within the villi. Mice in groups I, AS, and F (Fig. 1) had comparable levels of intestinal injury; the modified Chiu scores 2 h after reperfusion were 5.56 ± 0.15, 5.48 ± 0.20, and 5.44 ± 0.25, respectively. Mucosal injury scores (4.67 ± 0.23) were significantly reduced in group T compared with those in groups I, AS, and F (p < 0.05; F = 141.361; Fig. 3B).

Measurement of villi from histologic sections demonstrated a statistically significant increase in villus height on day 7 after reperfusion in group T (38.14 ± 0.76 μm) compared with group I (34.89 ± 0.87 μm), group AS (35.15 ± 0.29 μm), and group F (35.27 ± 0.30 μm; p < 0.01; F = 6.3), with no changes in crypt depth (Fig. 3C, 3D). There were no significant differences in villus height between group S (37.03 ± 0.30 μm) and group T (Fig. 3C).

On day 1, mice in groups I, T, AS, and F consumed only a small volume of chow, but food intake in group T quickly increased and approached that of sham animals. Chow intakes from days 2 through 7 were significantly higher in group T than those in groups I, AS, and F (Fig. 4A).

The I/R animals lost more body weight than the sham mice on days 1, 2, 4, and 7 (Fig. 4B). However, mice in group T gained significantly more weight on days 4 and 7 than those in groups I, AS, and F, which quickly increased and approached that of sham animals.

All sham mice survived the entire 7-d observation period. Compared with group I (survival rate 34%), group AS (survival rate 40%), and group F (37%), mice in group T had significantly improved survival rate (66%) on day 7 (p < 0.05; χ² = 6.914, 4.644, and 5.719, respectively; Fig. 4C).

To assess the potential of rTsCPB in eliciting host-protective Abs and immunogenicity, the titers of Ab isotypes IgG1 and IgG2a against rTsCPB, were measured in the sera of mice by indirect ELISA and compared in terms of OD value (450 nm). The rTsCPB treated mice produced significantly higher levels of rTsCPB-specific IgG1 compared with other groups (p < 0.001; Fig. 2B). The rTsCPB-specific IgG2a in the rTsCPB-treated mice was also significantly higher than in other groups (p < 0.001), but significantly lower than in in the rTsCPB treated mice (p < 0.001; Fig. 2B). The results suggest that rTsCPB vaccination seems to induce an IgG1 (Th2)-dominate Ab response.

To elucidate the effect of this Th2-prone immune status on local mucosal immunity in I/R mouse, we investigated the mRNA and protein levels of Th1 cytokines (IL-6 and IFN-γ) and Th2 cytokines (IL-4 and IL-10) in the intestine. Expression of IL-6 and IFN-γ significantly increased, whereas expression of IL-4 and IL-10 significantly decreased on day 7 after reperfusion when compared with those at 2 h in groups I and F (Fig. 5). At 2 h and on day 7 after reperfusion in group T, expression of Th 1 cytokines (IL-6 and IFN-γ) and Th2 cytokines (IL-4 and IL-10) was significantly upregulated. However, fold changes of Th2 cytokines on day 7 to 2 h after reperfusion in group T were significantly higher than those of Th1 cytokines (Fig. 5). Thus, rTsCPB slightly upregulated Th1 cytokines and significantly upregulated Th2 cytokines. Consistent with the serum changes of IgG1 and IgG2a, these results confirm that rTsCPB induced a mixed Th1/Th2 but Th2-predominant response. There were no significant differences in the expression of IL-6, IFN-γ, and IL-4 between groups T and AS. However, in mice treated with rTsCPB, the administration of the STAT6 inhibitor AS1517499 significantly downregulated IL-10 expression (Fig. 5D), which indicates the suppression of M2 polarization.

Macrophages are polarized in different microenvironment. Studies showed that M1 macrophages were induced in Th1 microenvironment and M2 macrophages were induced in Th2 microenvironment (34). Therefore, to elucidate the effect of intestinal I/R and rTsCPB on macrophage switch, we determined the mRNA and protein levels of M1 markers (NOS2 and CCR7) and M2 markers (Arg-1 and Fizz1) in the intestine. At 2 h after reperfusion, only weak macrophages polarization was induced in all groups, as there were weak mRNA and protein levels of NOS2, CCR7, Arg-1, and Fizz1 (Fig. 6). On day 7 after reperfusion in group I, macrophage polarization shifted to M1 as NOS2 and CCR7 expression increased and Arg-1 and Fizz1 expression decreased, whereas, in group T, macrophage polarization shifted to M2 as they expressed high levels of Arg-1 and Fizz1 and very low levels of NOS2 and CCR7 (Fig. 6). Mice in group AS had simultaneous changes in mRNA and protein levels of NOS2 and CCR7 compared with those in group T. However, the administration of AS1517499 significantly downregulated Arg-1 and Fizz1 expression (Fig. 6C, 6D), which further indicates the suppression of M2 polarization.

The shift between M1 and M2 macrophages was further confirmed by immunostaining with NOS2, a marker for M1 macrophages, and CD163, a marker for M2 macrophages (Fig. 7). Compared with group S, the number of NOS2-positive cells (M1 macrophages) in groups I, T, AS, and F was significantly higher at 2 h and on day 7 after reperfusion (p < 0.001; Fig. 7D; F = 28.213 at 2 h and 60.236 on day 7, respectively). There were no significant differences among groups I, T, AS, and F in the number of M1 macrophages at different reperfusions (Fig. 7D). The number of CD163-positive cells (M2 macrophages) at 2 h and on day 7 after reperfusion in group T significantly increased than those of groups S, I, AS, and F (p < 0.001; Fig. 7C; F = 98.376 at 2 h and 174.932 on day 7, respectively). On day 7, the number of M2 macrophages in groups S, I, and F significantly decreased (p < 0.01), whereas the number of M2 macrophages significantly increased in group T (p < 0.01), when compared with those at 2 h after reperfusion. Meanwhile, in mice treated with rTsCPB, the increase in M2 macrophages was significantly reduced under the treatment of STAT6 inhibitor, AS1517499 (Fig. 7C).

Because previous studies showed that the switch of M1 to M2 macrophages was dependent on the activation of STAT6 (25), we used the selective inhibitor of STAT6, AS1517499, to identify whether the intestinal protective effect of rTsCPB is related to the switch of M1 to M2 macrophages. As shown in Fig. 6E, expression of STAT6 and pSTAT6 was significantly upregulated in mice treated with rTsCPB, whereas phosphorylation of STAT6 in group AS was inhibited by AS1517499, with no changes in total STAT6 expression. Thus, AS1517499 can inhibit rTsCPB-induced STAT6 phosphorylation, which could account for the suppression of rTsCPB on IL-10, Arg-1, and Fizz-1 expression and M2 polarization (Figs. 5D, 6C, 6D, 7C). These data suggest that one of the possible mechanisms for the shift from M1 to M2 macrophages is STAT6 dependent.

To further demonstrate the switch between M1 and M2 macrophages, we compared the ratio of M2/M1 macrophages. There were no significant differences among groups in the ratio of M2/M1 macrophages at 2 h after reperfusion (Fig. 7E). On day 7, the ratio of M2/M1 macrophages in groups I, AS, and F significantly decreased (p < 0.001), whereas the ratio significantly increased in group T (p < 0.01), when compared with those at 2 h after reperfusion (Fig. 7E).

Together, these results indicate that intestinal I/R led to a switch from M2 to M1 macrophages and that rTsCPB promoted a switch from M1 to M2 macrophages, which was suppressed by the administration of a selective STAT6 inhibitor.

To confirm further the relationship between macrophage polarization and intestinal I/R injury and restitution, the correlation analysis for NOS2, Arg-1 expression, M1 or M2 macrophages, and modified Chiu scores (reflecting intestinal injury) or villus height (reflecting intestinal restitution) was performed. In all groups, significant positive correlations between modified Chiu scores and NOS2 expression or M1 macrophages (r = 0.621 and 0.723; p < 0.01; Fig. 8A, 8B), between villus height and Arg-1 expression or M2 macrophages (r = 0.543 and 0.616; p < 0.01; Fig. 8C, 8D), were identified.

Immunostaining with anti–Ki-67, a marker for cell proliferation, showed that the number of Ki-67–positive cells at 2 h after reperfusion in group T was significantly higher than those in groups I, AS, and F (p < 0.05; F = 6.985). On day 7, the number of Ki-67–positive cells in group T was significantly higher than those in groups S, I, AS, and F (p < 0.05; Fig. 9; F = 13.75).

As shown in Fig. 10, MPO-positive cells, indicating polymorphonuclear neutrophil infiltration, were stained dark brown in the cytoplasm. Intestinal I/R induced an increase of MPO-positive cells in groups I, AS, and F, whereas few MPO-positive cells were found in groups S and T at 2 h after reperfusion (p < 0.001; F = 39.802). On day 7, the number of MPO-positive cells in groups I, AS, and F was still significantly higher than those in groups S and T (p < 0.01 and p < 0.05, respectively; F = 9.094).

In the current study, we found that intestinal I/R resulted in significant intestinal injury accompanied by a switch from M2 to M1 macrophages evidenced by a decrease in mRNA and protein levels of M2 markers (Arg-1 and Fizz1) and Th2 cytokines (IL-4 and IL-10), an increase in levels of M1 markers (NOS2 and CCR7) and Th1 cytokines (IL-6 and IFN-γ), a decrease in M2 cells, and an increase in M1 cells from 2 h to 7 d after reperfusion. More interestingly, the present findings also showed that the rTsCPB, a Trichinella secretory product, could attenuate intestinal injury induced by intestinal I/R as evidenced by a decrease in modified Chiu score and an increase in villus height with no changes in crypt depth in the intestinal mucosa. Moreover, rTsCPB improved intestinal function and the outcome of injured animals, as evidenced by increased body weight and food intakes and improved survival. Of note, we found that intestinal protection of rTsCPB was related to promoting a switch from M1 to M2 macrophages, as evidenced by a slight increase in mRNA and protein levels of M1 markers and Th1 cytokines, a marked increase in levels of M2 markers and Th2 cytokines, and a significant increase in the ratio of M2/M1 macrophages, from 2 h to 7 d after reperfusion. We also found that rTsCPB-induced switch from M1 to M2 macrophages is STAT6 dependent.

For several years, it has been known that different phenotypes of macrophages coexist in vivo and exert different functions depending on their microenvironmental stimulus. The M1 phenotype has an important role in mediating tissue damage and initiating inflammatory responses by expressing high levels of proinflammatory cytokines and reactive nitrogen and oxygen intermediates (35, 36). In contrast, M2 macrophages are considered to be involved in promoting tissue remodeling and to have immunoregulatory function (17). The dualistic roles of distinctly polarized macrophage populations have been reported in several I/R models, including cerebral (37), myocardial (38), and renal (17) I/R injury. However, publications related to intestinal I/R injury and macrophages subsets are limited. To determine whether the macrophages present in the early phase of injury differ phenotypically or functionally from those present during recovery of injury, mRNA and protein levels of M1 and M2 markers were detected in the intestinal I/R mice. At 2 h after reperfusion in group I, expression of M1 and M2 markers increased compared with the sham mice, whereas expression of M2 markers decreased and expression of M1 markers increased significantly in the intestine on day 7. These results suggest that intestinal I/R-induced macrophage activation in the early phase of intestine injury and switched M2 to M1 macrophages during the recovery of injury. The switch was further confirmed with immunostaining, which showed a decrease in the ratio of M2/M1 macrophages on day 7. Although macrophages are critical for the induction of an effective immune response to pathogens, inappropriate and sustained M1 macrophages polarization leads to tissue damage and immune dysfunction. It has been suggested that NO may be an important protective molecule at the onset of intestinal I/R (39). However, excessive NO production through NOS2 has been proved to be detrimental in the pathogenesis of intestinal I/R (40). Suppression of excessive NO production using various NO inhibitors has beneficial effects on ameliorating intestinal I/R injury (41, 42). In the current study, intestinal I/R resulted in severe intestine damage and delayed intestine repair, as evidenced by the increase in modified Chiu scores and decrease in villus height. Moreover, significant positive correlation between modified Chiu scores and NOS2 expression was identified. The switch from M2 to M1 macrophages could be one of the mechanisms for I/R-induced intestinal injury.

Recent studies suggested that the ES Ags of T. spiralis are likely involved in immunomodulation (26, 43). The consideration of using rTsCPB as an immunomodulatory agent has been derived from its striking antigenicity of stimulating strong IgG1, with no specific IgG2a production in mice immunized with rTsCPB in our previous study (26). Because it stimulates strong IgG1 production and significantly higher levels of Th2 cytokines (IL-4 and IL-10), rTsCPB can be considered as a Th2-prone immunomodulator. In this study, our data showed that levels of M1 and M2 markers increased significantly compared with the sham mice from 2 h to 7 d after reperfusion in group T. Contrary to the changes in group I, macrophage polarization shifted to M2 as they expressed high levels of M2 markers and very low levels of M1 markers on day 7 after reperfusion. Furthermore, immunostaining with CD163, an M2 marker, and NOS2 showed that the number of M1 and M2 macrophages at different reperfusions and the ratio of M2/M1 macrophages on day 7 significantly increased in group T, when compared with that at 2 h after reperfusion. These results suggest that rTsCPB activated macrophages, reversed I/R-induced switch from M2 to M1 macrophages, and promoted the switch from M1 to M2 macrophages during the process of I/R-induced intestine injury to intestine repair. The level of Arg-1 was significantly higher in the rTsCPB group on day 7 after reperfusion. Arg-1 is the counterregulatory enzyme to NOS2 and can thus act to suppress NO production. Arg-1 also has well-documented roles in tissue repair (44). Similarly, significant positive correlation between villus height and Arg-1 expression was identified in the current study. Therefore, we conjectured that the attenuation of intestinal I/R injury by rTsCPB in the current study might be linked to the switch from M1 to M2 macrophages in intestinal mucosa. Our results are consistent with a previous report of switch from M1 to M2 promoting tubular repair after I/R injury (17). Moreover, rTsCPB significantly stimulated cell proliferation, increased daily food intakes from days 2 through 7 and body weight on days 4 and 7, and significantly enhanced survival, also suggesting that an M2-bias innate immunity might promote recovery of gut function and improve the outcome of mice. Thus, the switch from M1 to M2 macrophages could be one of the mechanisms for restitution from I/R-induced intestinal injury.

Both M1 markers and Th1 cytokines were increased in rTsCPB-treated mice relative to groups I and S in the current study. However, the increased extents of M1 markers and Th1 cytokines were significantly lower than those of M2 markers and Th2 cytokines, which is consistent with the serum changes of IgG1 and IgG2a. These results confirm that rTsCPB induced a mixed Th1/Th2 but Th2-predominant response, which is consistent with previous work by Little et al. (45) showing a mixed Th1/Th2 response to infection with the gastrointestinal nematode parasite Trichuris muris in C57BL/6 mice. However, the adaptive immune response to helminth infections varies considerably between different strains of mouse, and even different levels of infection (45, 46). For example, BALB/c mice mounted a strong Th2 response (and also a weak and delayed Th17 response) to T. muris, whereas C57BL/6 mice mounted strong Th1, Th2, and Th17 responses (1). What’s more, high-level infection resulted in a mixed Th1/Th2 response to T. muris, and a low-level infection resulted in a Th1 response in C57BL/6 mice. Therefore, the effects of rTsCPB-induced adaptive immune response between different strains of mouse and the relative mechanisms should be investigated further in future studies.

STAT6 is a Th2-associated transcription factor that is activated through phosphorylation and that translocates into the nucleus during the activation of M2 macrophages (47). Previous studies have demonstrated that helminth infections or the treatment with their ES products led to significant M1-to-M2 macrophage transition, which is dependent on activation of the STAT6 signaling (25, 48). Studies using STAT6 knockout mice also demonstrated a defect in M2-associated responses in helminth infections (23). STAT6 is one of the transfactors that bind to the promoter of Arg-1 and Fizz1 (49). As rTsCPB activates Arg-1 and Fizz1 expression, we investigated whether these effects were mediated by the STAT6 signaling. AS1517499, a potent selective STAT6 inhibitor, has been shown to inhibit the phosphorylation of STAT6 induced by IL-4/IL-13, and to inhibit the Th2 cell differentiation without influencing the Th1 cell differentiation (30, 50). In the current study, we found that the addition of AS1517499 to rTsCPB-treated mice significantly downregulated rTsCPB-induced Arg-1 and Fizz1 expression and inhibited M1 to M2 transition mediated by STAT6 activation, accompanied by severer intestinal I/R injury and worse survival rate. Together, these results show that the mechanism of rTsCPB-induced switch from M1 to M2 macrophages is STAT6 dependent, establishing a causal relationship between the increased M2 macrophages and protection from I/R injury.

Although vaccination with rTsCPB showed a Th2-prone response and evoked an M2-bias innate immunity in I/R mice, it surprisingly upregulated expression of IL-6 in the intestinal tissue, as shown in our results that mRNA level was significantly higher in the rTsCPB group than in other groups. IL-6 is known to play an important role in the development of the inflammatory response to ischemic insult (51). However, IL-6 is a multifunctional cytokine with a variety of effects on cells and tissues. In intestine, IL-6 is induced in several injury models, including intestinal I/R injury and after surgical intestinal manipulation, and it promotes the inflammatory response (52, 53). On the contrary, in vitro study has shown that IL-6 prevents death of rat intestinal epithelial cells subjected to hypoxia (54). IL-6 has also been found to preserve intestinal graft function by maintaining graft blood flow and reducing proinflammatory cytokine upregulation and neutrophil infiltration (33). Another study demonstrated that exogenous administration of IL-6 resulted in an increase in small bowel mass and in intestinal villus height, and increased enterocyte resistance to extrinsic apoptosis, intrinsic apoptosis, and oxidative injuries, leading to less intestinal injury and improved barrier function following ischemia reperfusion of the small bowel (55). Similarly, our data also showed a significant increase in intestinal villus height and less neutrophil infiltration in the intestine of rTsCPB-treated I/R mice, suggesting a potential role for IL-6 in intestinal homeostasis. The exact mechanism of overexpression of IL-6 in rTsCPB-treated I/R mice remains to be clarified by additional studies.

In this study, we chose 60 min of intestinal ischemia as described previously (28, 29), a rather severe ischemic insult causing mortality from 52.9 to 100% (56, 57), for our murine model. The lethal intestinal damage may minimize the difference in mortality; therefore, more animals may be needed in each group to detect the difference. Stringa et al. (58) obtained a 100% mortality rate 6 h after reperfusion in BALB/c mice that underwent 40 or 45 min of intestinal ischemia, in which the animals were anesthetized with ketamine and diazepam. However, we obtained a 64% mortality rate on day 7 after reperfusion in group I with sevoflurane anesthesia. The variation in anesthetics, surgical conditions, operators, and even animals may attribute to the different mortality rates. For example, Souza et al. (57) obtained a 100% mortality rate 1 h after reperfusion in C57BL6 mice that had been anesthetized with urethane and underwent 60 min of intestinal ischemia. However, Watson et al. (59) obtained a 70% mortality rate in C57BL6 mice on day 7 after reperfusion with isoflurane anesthesia.

There could be some limitations in this study. First, rTsCPB dosage was determined based on our previous study (26), without evaluating possible effects of rTsCPB at various dosage and inoculation times on intestinal I/R injury. In future studies, optimal dosage and inoculation time of rTsCPB should be examined in intestinal I/R models. Second, although the switch from M1 to M2 macrophages could be one of the mechanisms for restitution from I/R-induced intestinal injury, whether rTsCPB acts on macrophages directly or indirectly remains unknown. Studies have shown that ES-62 from filarial nematodes modulates cytokine production in dendritic cells and macrophages via a TLR4-mediated pathway (60, 61). Therefore, further investigation is needed to investigate whether the M2 macrophages in the intestinal mucosa induced by s.c. inoculated rTsCPB is executed through direct interaction between rTsCPB and macrophages or through other immune cells, such as dendritic cells. Lastly, the macrophage phenotypes switch occurs over time and parallels the course of inflammation or infection. Macrophage infiltration peaks at 24–48 h and remains elevated for at least 6 d following reperfusion injury (62). Therefore, an extended observation period (e.g., 14 d after reperfusion) is necessary in future studies.

In summary, the current study demonstrates that intestinal I/R injury is accompanied by a switch from M2 to M1 macrophages and that rTsCPB is effective in ameliorating intestinal injury and improving intestinal function following intestinal I/R. rTsCPB mediates the efficacy by reversing intestinal I/R-induced switch from M2 to M1 macrophages and promoting the switch from M1 to M2 macrophages, which is STAT6 dependent. rTsCPB may provide a potential therapeutic application in intestinal I/R-induced intestinal injury.

We thank Dr. Guan-Rong Zhang for professional help and advice about the statistical analysis, Dr. Ri-Dong Wu for assistance with the real-time PCR, and Dr. Yi-Hong Ling and Dr. Yuan-Zhong Yang for contributing to the histologic analysis.

The authors have no financial conflicts of interest.