IL-5 is involved in a number of immune responses such as helminth infection and allergy. IL-5 also plays roles in innate immunity by maintaining B-1 B cells and mucosal IgA production. However, the identity of IL-5–producing cells has not been unambiguously characterized. In this report, we describe the generation of an IL-5 reporter mouse and identify IL-5–producing non-T lymphoid cells that reside in the intestine, peritoneal cavity, and lungs in naive mice. They share many characteristics with natural helper cells, nuocytes, and Ih2 cells, including surface Ags and responsiveness to cytokines. However, these phenotypes do not completely overlap with any particular one of these cell types. Innate non-T IL-5–producing cells localized most abundantly in the lung and proliferated and upregulated IL-5 production in response to IL-25 and IL-33. IL-33 was more effective than IL-25. These cells contribute to maintaining sufficient numbers of lung eosinophils and are important for eosinophil recruitment mediated by IL-25 and IL-33. Given that eosinophils are shown to possess antitumor activity, we studied lung tumor metastasis and showed that innate IL-5–producing cells were increased in response to tumor invasion, and their regulation of eosinophils is critical to suppress tumor metastasis. Genetic blockade or neutralization of IL-5 impaired eosinophil recruitment into the lung and resulted in increased tumor metastasis. Conversely, exogenous IL-5 treatment resulted in suppressed tumor metastasis and augmented eosinophil infiltration. These newly identified innate IL-5–producing cells thus play a role in tumor surveillance through lung eosinophils and may contribute to development of novel immunotherapies for cancer.
Interleukin-5 is produced by CD4+ Th2 cells, mast cells, eosinophils, and basophils and is involved in many aspects of eosinophil development (1–4). In animal models of allergic diseases or helminth infection, IL-5 induces a massive proliferation of eosinophil progenitors in the bone marrow (3, 5), promotes eosinophil recruitment with eotaxins (6–8), and prolongs eosinophil survival in local tissues (9, 10). IL-5 binds the IL-5R complex, which consists of an IL-5Rα chain specific for IL-5 and a common β-chain that is shared by the receptors for IL-3 and GM-CSF. IL-5 signaling is also necessary for the maintenance of B-1 B cells in the peritoneal and pleural cavities (11, 12) and has been implicated to play a critical role in mucosal immunity, such as IgA production. IL-5 transgenic mice exhibited expansion of B-1 B cells and elevated levels of serum IgA in addition to massive eosinophilia in various tissues (13).
To understand the physiological roles of IL-5 more precisely, it is important to identify and characterize IL-5–producing cells in local tissues and investigate their regulation. Although the majority of research on IL-5–producing cells has focused on Th2 cells, other types of cells have been indicated to produce IL-5 (14–17). Recently, innate production of Th2 cytokines from newly identified cell populations, termed natural helper cells, nuocytes, MPPtype2, and Ih2 cells, has been reported (18–21). Natural helper cells are located in lymphoid clusters in adipose tissues, and nuocytes and MPPtype2 cells exist in mesenteric lymph nodes (MLNs) and the intestine. IL-25 and/or IL-33 are potent cytokines required for their activation and for promoting Th2 immune responses. Although natural helper cells and nuocytes are capable of secreting IL-5, they localize to particular tissues such as MLNs and intestine and IL-5 has been demonstrated to work in various tissues. We sought to investigate IL-5 production systemically by generating an IL-5 reporter mouse in which living IL-5–producing cells are detected by the expression of modified yellow fluorescence protein.
The innate immune system responds rapidly to invading pathogens, recognizing pathogen-associated molecular patterns, and subsequently promotes Ag-specific adaptive immune responses to the pathogens. It provides a variety of defenses against not only exogenous pathogens, but also endogenous, life-threatening tumors. In addition to a number of reports describing the roles of Th1 cytokines, such as IL-12 and IFN-γ, in Ag-specific tumor immunity mediated by CD8+ CTL, there has been accumulating evidence that Th2 cells are also involved in tumor rejection by employing activation of innate immune cells such as macrophages and eosinophils (22–28). One of the intriguing functions of IL-4, a well-characterized Th2 cytokine, is to induce eosinophil influx into tumor-growing sites (25, 26) and prevent angiogenesis (27). Interestingly, exogenous treatment of tumor-bearing mice with IL-5 led to prolonged survival of the mice (29). In this case, IL-5–dependent augmentation of eosinophil infiltration into tumor sites was necessary for tumor surveillance (30).
In this study, we describe the generation of IL-5/Venus knockin (KI) mice and used these mice to identify innate IL-5–producing cells in the small and large intestines, peritoneal cavity, and the lung. The innate IL-5–producing cells possess several similar characteristics with natural helper cells, nuocytes, and Ih2 cells (18, 19, 21). We found unique tissue localization of innate IL-5–producing cells in the lungs of untreated mice. We describe in this study innate IL-5–producing cells that regulate eosinophil infiltration into the lung and appear to prevent tumor metastasis. In addition to their potential roles in regulating Th2 immune responses, innate IL-5–producing cells may contribute to antitumor immunity via lung eosinophils, indicating a role in tumor surveillance.
Materials and Methods
Generation of IL-5/Venus KI mice
Embryonic stem (ES) cells derived from the 129S4/SvJae mouse strain were electroporated with the linearized targeting vector (Fig. 1A), including the HSV thymidine kinase. ES cells transfected with the targeting vector were screened with G418 and gancyclovir, selected, and separately expanded to obtain genomic DNA. Two of the ES cells were chosen to inject into C57BL/6 blastocysts, and the resulting male mice were crossed to C57BL/6 female mice. Offspring derived from ES cells carrying the IL-5/Venus KI allele were identified by their coat color and backcrossed to C57BL/6 mice at least 10 generations. During the period of backcrossing, mice were mated with CAG-cre transgenic mice (31) to delete the neomycin resistance gene from the germline. Those mice were intercrossed to obtain mice expressing IL-5+/+, IL-5+/Venus (IL-5+/V), and IL-5V/V genotypes. The primers used for genotyping are as follows: IL-5 forward, 5′-CTCTTCCTTTGCTGAAGGCCAG-3′; IL-5 reverse, 5′-GAGTAGGGACAGGAAGCCTC ATCG-3′; and Venus reverse, 5′-CTCGATGTTGTGGCGG ATCTT-3′.
32, 33). To neutralize IL-5, a single bolus of 500 μg anti–IL-5 mAb (NC17) (34, 35) or a control rat IgG Ab (Sigma-Aldrich, St. Louis, MO) was i.p. administrated just before the injection of tumor cells. The indicated number of tumor cells was i.v. injected into the tail vein. All experiments were performed according to the guidelines for the care and treatment of experimental animals at the University of Toyama.
Cell culture for Th1 or Th2 differentiation
RPMI 1640 supplemented with 10% FCS, 50 μm 2-ME, 100 U/ml penicillin, and 100 μg/ml streptomycin was used. Purified splenic CD4+
Tumor cells and cell culture
A lung metastatic melanoma cell line, B16F10, was kindly provided by Dr. S. Wakuzawa (Hokuriku University, Kanazawa, Japan). B16F10 cells were maintained in RPMI 1640 supplemented with 10% FCS, 50 μm 2-ME, 100 U/ml penicillin, and 100 μg/ml streptomycin.
Abs and reagents for flow cytometry
Preparation of lung cells
Preparation of lamina propria cells
To obtain lamina propria (LP) cells, the small intestine was harvested, and Peyer’s patches (PPs) were removed. The resulting small intestine was then reversed using a polyethylene tube and cut into five to seven pieces. The large intestine was harvested and cecal patches, and colonic patches were removed. The large intestine was then excised, opened longitudinally, and cut into three pieces. The small and large intestines were washed twice with 40 ml Ca2+- and Mg2+
Immunofluorescence histology and immunohistochemistry
To detect Venus+2O2 methanol solution for 5 min and 5% BSA for 1 min. Subsequently, sections were incubated with M.O.M mouse Ig blocking reagent for 1 h, followed by EG2 Ab (Kabi Pharmacia, Uppsala, Sweden) application and intermittent microwave treatment for 30 min (250 W, 4 s on, 3 s off). After washing, a peroxidase-conjugated envision kit for mouse (DakoCytomation) was applied and subjected to microwave intermittently for 30 min. Following washing, sections were immersed in DAB solution (Sigma-Aldrich) with H2O2. Melanoma cells were detected by Fontana-Masson staining. Images were acquired using a D-33E digital camera (Olympus, Tokyo, Japan).
RNA preparation for real-time quantitative RT-PCR
The p values were calculated using the two-tailed Student t test.
Generation of IL-5/Venus KI mice
To monitor IL-5 production in vivo, the Venus gene (36) was inserted at the translation initiation site in exon 1 of the IL-5 gene locus by homologous recombination (Fig. 1A). Successful homologous recombination was confirmed by Southern blot analysis (Fig. 1B, 1C). CD4+ splenocytes from IL-5+/+, IL-5+/V, and IL-5V/V mice were cultured under Th2-skewing conditions, and the expression of IL-5 along with Venus was measured (Fig. 1D). As expected, no IL-5 but strong Venus expression was detected in IL-5V/V CD4+ T cells. IL-5+/V T cells showed heterogenic expression patterns, namely IL-5+Venus−, IL-5+Venus+, and IL-5−Venus+, indicating that IL-5 loci are biallelically expressed. No Venus signal was observed in CD4+ T cells cultured under Th1-skewing conditions, consistent with a lack of IL-5 production by Th1 T cells (Fig. 1E). B-1 B cells in IL-5V/V mice were smaller in size, as observed in IL-5Rα knockout (KO) mice (11), than those in IL-5+/+ mice, confirming deletion of IL-5 in homozygous IL-5/Venus KI mice (Fig. 1F).
Identification and characterization of Venus+ cells
We next examined the localization of Venus+ cells in immune-related tissues and organs. Venus+ cells were barely detectable in the bone marrow, spleen, lymph nodes, MLNs, and PPs. In contrast, Venus+ lymphoid cells were found in LP from the small and large intestines, peritoneal cavities, and the lungs of unimmunized naive mice (Fig. 2A). Detectable levels of Venus were not observed in monocyte/macrophages, eosinophils, or mast cells (Supplemental Fig. 1). The Venus+ lymphoid cells consisted of at least three different populations: c-Kit+, CD3ε+, and double-negative (c-Kit−CD3ε−) cells. In these tissues, c-Kit+ cells are the majority of Venus+ IL-5 producers. Calculation of the absolute cell numbers revealed that c-Kit+Venus+ cells localized most abundantly in the lung (Fig. 2B). The majority of Venus+ cells in various tissues were lineage negative (B220, CD3ε, CD4, CD8α, CD5, TCRβ, CD11b, CD11c, CD19, DX5, FcεRIα Gr-1, NK1.1, TCRγδ, and TER-119) and expressed Sca-1, Thy1.2, CD25, CD27, CD44, CD69, IL-7Rα, and T1/ST2, a subunit of IL-33R (Fig. 2C). This is reminiscent of natural helper cells as well as of nuocytes (18, 19). Thy1, CD25, CD27, and T1/ST2 were heterogeneously expressed among Venus+ cells found in the indicated organs.
In vitro analysis of innate IL-5–producing cells in the lung
The existence of relatively abundant c-Kit+ IL-5–producing cells in the lung (Fig. 2B) compared with intestine led us to assess expression of Th2 cytokines and Venus in the lung innate lymphoid cells. Cells prepared from lung were cultured with or without PMA plus ionomycin. The majority of the Venus+ cells were c-Kit+CD3ε− cells representing non-T innate cells (Fig. 3A). The expression of IL-5, IL-4, and IL-13 among the Venus+ cells was assessed by intracellular staining (Fig. 3B). There were IL-5+Venus−, IL-5+Venus+, and IL-5−Venus+, indicating that IL-5 loci are biallelically expressed in lung innate IL-5–producing cells (as seen in cultured Th cells). This also indicated that Venus+ lung innate cells are indeed producing IL-5 protein. Concerning other Th2 cytokines, approximately one-third of Venus+ cells also expressed IL-13. Thus, there was a population of cells competent to produce IL-5 (Venus+) but not IL-13. Venus+ cells rarely expressed IL-4.
Because innate IL-5–producing cells expressed receptors for IL-2, IL-7, and IL-33 (Fig. 2C) and IL-25 and IL-33 are well known to induce Th2 responses, we cultured lung cells in the presence of IL-2, IL-7, IL-25, or IL-33 and examined proliferation and IL-5 production. IL-33 treatment increased the number, size, and Venus expression levels of Venus+ cells (Fig. 3C, 3D). The majority of Venus+ cells maintained a c-Kit+CD3ε− phenotype. Essentially the same results were obtained with the stimulation with IL-25, but to a lesser extent compared with IL-33. Reflecting the increased Venus intensity, the amounts of IL-5 secreted in cultured supernatants upon stimulation with IL-33 were significantly augmented (Fig. 3E). The lung cells from IL-5+/+ mice carrying both intact IL-5 alleles produced more IL-5 in response to IL-33 than those from IL-5+/V mice.
Innate IL-5–producing cells in the lung respond to IL-25 and IL-33, leading to inducing eosinophil migration in vivo
As we demonstrated by in vitro analysis that innate IL-5–producing cells responded to IL-25 and IL-33 (Fig. 3C), we examined whether innate IL-5–producing cells in the lung of IL-5+/V mice could be expanded by passive administration of rIL-25 and rIL-33 in vivo. IL-25 treatment resulted in an ∼2-fold increase of c-Kit+ Venus+ cells, whereas IL-33 dramatically increased the number of these cells, ∼10-fold, with larger cell size and higher Venus intensity (Fig. 4A, 4B). Expansion of Venus+ cells was not seen in nonlymphoid cell fractions. Immunofluorescent histological analysis also showed increased Venus+ cells in response to IL-33 and that many of them resided around bronchioles and bronchia (Fig. 4C). Most Venus+ cells did not express CD3ε.
The effect of in vivo administration of IL-25 and IL-33 on IL-5 production in the intestines was also analyzed (Supplemental Fig. 2). In the small intestine, a larger number of Venus+ cells were observed in rIL-25–treated mice than in rIL-33–treated mice, whereas IL-33 treatment was more effective to induce Venus+ cells than IL-25 treatment in the large intestine (Supplemental Fig. 2A). This was further confirmed by flow cytometric analysis of LP cells from the large intestine (Supplemental Fig. 2B). We then evaluated the IL-5 and IL-13 mRNA expression in the small and large intestines in response to IL-25 and IL-33 by quantitative RT-PCR analysis (Supplemental Fig. 2C, 2D). IL-13–producing cells respond almost equally to IL-25 and IL-33 in the intestine or MLNs (19, 21). To compare the expression patterns of IL-5 and IL-13, we analyzed IL-13 mRNA expression (Supplemental Fig. 2D). Intriguingly, IL-13 mRNA expression was enhanced by IL-25 more significantly than by IL-33, both in the small and large intestines. We also examined the effect of IL-25 and IL-33 on IL-5 production in MLNs (Supplemental Fig. 2E) and PPs (Supplemental Fig. 2F). IL-25 and IL-33 also induced Venus+ cells in MLNs and, to a lesser degree, in PPs. In both cases, IL-33 was more effective than IL-25.
It has been shown that in vivo treatment of IL-25 and IL-33 in mice attracts eosinophils into local tissues. When we treated wild-type (WT) with rIL-25 and rIL-33 locally, there was clear eosinophil infiltration in the lung, whereas the effect of IL-25 and IL-33 was not observed in IL-5V/V mice (Fig. 5A). Treatment of WT mice by IL-25 and IL-33 resulted in a significant increase of eosinophils in the lung, ∼3-fold in cell number in comparison with PBS-injected controls, which was not observed in IL-5V/V mice (Fig. 5B). Notably, numbers of eosinophil in the lungs of naive IL-5V/V mice were significantly decreased compared with those from WT mice. These data clearly demonstrate roles of IL-25 and IL-33 in the regulation of innate IL-5–producing cells and subsequent accumulation and migration of eosinophils into the lung.
Mice deficient in IL-5 signaling or treated with IL-5–neutralizing Abs exhibit increased lung tumor metastasis and impaired eosinophil regulation
One of the functions of eosinophils primed by Th2 cytokines is tumor surveillance and clearance (25, 30). Given that deficiency of IL-5 causes impaired eosinophil recruitment to the lung (Fig. 5), we used an experimental model of lung tumor metastasis using a melanoma cell line (B16F10). Melanoma cells were i.v. injected into IL-5+/+, IL-5+/V, or IL-5V/V mice, and tumor nodules were counted (Fig. 6A). The number of nodules in IL-5V/V mice was significantly increased compared with that in IL-5+/+ and IL-5+/V mice (Fig. 6B). Increased tumor metastasis was also observed in IL-5Rα KO mice (Supplemental Fig. 3A–C). In IL-5V/V and IL-5Rα KO mice, IL-5 signaling is constitutively defective. To investigate the effect of secreted IL-5 after tumor injection, IL-5 was neutralized with anti–IL-5 mAb. A bolus injection of anti–IL-5 mAb resulted in increased numbers of tumor nodules (Fig. 6C). Importantly, neutralization of IL-5 completely blocked the recruitment of eosinophils into the lung (Fig. 6D). We also evaluated the migration of eosinophils into the lungs of WT and IL-5Rα KO mice on days 1, 3, 5, and 7 after tumor inoculation (Supplemental Fig. 3D). There was a notable difference in the migration of eosinophils into the lung between WT and IL-5Rα KO mice.
Exogenous IL-5 treatment induces an elevated level of eosinophilia and prolonged eosinophil migration toward tumor cells
To further evaluate the cooperative relationship between IL-5 and eosinophils in antimetastatic activity in the lung, we induced lung eosinophilia by administering rIL-5. This treatment successfully recruited eosinophils into the lung (Fig. 7A, 7B). As IL-5 overexpression resulted in increased tumor-infiltrating eosinophils (30), the location of eosinophils in relation to tumor metastases was analyzed by H&E staining (Supplemental Fig. 3E) and immunohistochemistry (Fig. 7C). Metastasized melanoma cells developed clusters of cells that localized to the capillary vessels of alveolar walls (Supplemental Fig. 3E). Although eosinophils were not reliably detectable in tissues stained with H&E, combination staining with Fontana-Masson followed by immunohistological analysis clearly demonstrated eosinophils around melanoma cell clusters. Eosinophils visible in a high-power field (HPF), centered on melanoma cells, were collected and counted. Quantitation revealed that the numbers of migrating eosinophil toward tumor cells were significantly greater in the rIL-5–treated mice at 48 h post-tumor injection than in IL-5–nontreated controls (Fig. 7C, 7D). We infer from these results that IL-5 is able to prolong eosinophil migration toward tumor cells in the lung tissue.
Invading tumor cells enhance local IL-5 production from innate IL-5–producing cells
As eosinophil infiltration in the lung was increased after tumor inoculation (Fig. 6D), IL-5 production may be upregulated by melanoma cell infiltration into the lung. To more directly measure the role of IL-5 in preventing tumor metastases, we analyzed IL-5 mRNA expression in the sorted lung CD45+Mac-1− fraction from WT mice at 6 and 24 h post-tumor injections (Fig. 8A). Compared to the PBS control group, IL-5 mRNA was significantly upregulated ∼3-fold at 6 h and 15-fold at 24 h after the melanoma injection. This IL-5 upregulation was not observed in other sites (Fig. 8B). These results suggest that upregulation of IL-5 expression in the lung contributes to induce eosinophil infiltration, suppression of metastasis, or progression of tumor cells in the lung. In support of this idea, administration of rIL-5 into the bronchus suppressed progression or metastasis of melanoma cells (Fig. 8C).
The source of IL-5 was primarily c-Kit+Venus+ lymphoid cells (Fig. 8D), and they showed an ∼2.5-fold increase on day 3 after tumor injection (Fig. 8E). This appeared to be well correlated with the numerical increase observed in eosinophils post-tumor inoculation (Fig. 6D, Supplemental Fig. 3D). As innate IL-5–producing cells in the lung responded to IL-25 and IL-33 (Fig. 4), the mRNA expression of IL-25 and IL-33 was analyzed periodically after melanoma inoculation (Fig. 8F). The mRNA expression of IL-2 and IL-7 was also assessed. IL-33 expression was transiently upregulated ∼2-fold at 6 h post-tumor injection and declined to control level at 24 h, whereas IL-25 expression was not detected in the lung. The mRNA expression of IL-2 and IL-7 was not upregulated significantly. These results imply that upregulation of IL-5 mRNA expression in the lung of melanoma-bearing mice may be caused by the transient increase in IL-33 expression.
We have generated IL-5/Venus KI mice that enabled us to identify innate IL-5–producing cells residing in the small and large intestines, peritoneal cavities, and the lung. We have shown that T cells are not the major IL-5 producers in the steady state as well as under the conditions promoted by IL-25 and IL-33. Innate IL-5–producing cells share several characteristics with natural helper cells, nuocytes, MPPtype2, and Ih2 cells, such as tissue localization, surface makers, namely, lineage-negative c-Kit+Sca-1+, and responsiveness to IL-25 and/or IL-33. In contrast, there are several phenotypic and regulatory differences as well. Innate IL-5–producing cells are phenotypically different from MPPtype2 cells that do not express IL-7Rα, T1/ST2, Thy1.2, and CD44. As previously suggested for other innate cells (37), innate IL-5–producing cells are possibly differentiated from MPPtype2 cells. In vitro experiments showed that there were innate lymphoid cells that produced IL-5, IL-13, or both of these cytokines, raising a possibility that IL-5– and IL-13–producing cells might be regulated differently. Although it is difficult to perform a numerical comparison among the innate cells due to different reporter genes used, innate IL-5–producing cells seem to be more abundant than IL-13–producing cells in the lung and peritoneal cavities in unprimed mice (21). In addition, IL-25 is better at inducing IL-13 production in the small and large intestines, compared with IL-33, which showed a larger effect on IL-5 production in the large intestine, lung, MLNs, and PPs. These data support the idea that innate IL-5– and IL-13–producing cells may differ, although a significant fraction of them have the ability to produce both cytokines. Innate IL-5–producing cells are phenotypically and functionally closer to natural helper cells because natural helper cells express the surface Ags examined in this study and produce Th2 cytokines naturally, and both were reported to maintain B-1 B cells in the peritoneal cavity (18). As demonstrated in innate IL-5–producing cells by the current study, natural helper cells possibly express Th2 cytokines heterogeneously. In the current study, we have not examined the function of innate IL-5–producing cells in the same experimental model used to study other newly identified innate cell populations. It is therefore not clear whether all of them including innate IL-5–producing cells are identical in vivo in the case of parasite infection or immune responses such as allergies. Further work is required to more precisely define and categorize these novel innate cells.
To investigate the role(s) of IL-5 in the lung, we used an experimental tumor metastasis model because eosinophils are potent effectors for tumor rejection and accumulating evidence has established that IL-5 is essential for eosinophil development. We have demonstrated that deficiency of IL-5 signaling and neutralization of IL-5 resulted in increased lung tumor metastasis, very likely due to a shortage of lung eosinophils and impaired eosinophil recruitment in response to invading tumors. IL-5/Venus KI mice demonstrated that enhanced IL-5 production mediated by tumor cells was mainly carried out by lung innate IL-5–producing cells, and their use for time-course analysis was helpful for understanding temporal and spatial IL-5 production in vivo. In support of our study, one report described that eosinophil recruitment to tumor cells was an early event and independent of CD4+ T cells (38). Although different ways of inoculating tumor cells were employed, innate IL-5–producing non-T cells might also contribute to such early eosinophil regulation because IL-5 was upregulated immediately after tumor inoculation. The reduction of eosinophils in the IL-5 or IL-5Rα KO mice tends to be more severe in the lung than in the bone marrow and the blood in steady-state, naive animals (11, 12). Together with our finding that innate IL-5–producing cells were most abundant in the lung, it seems likely that lung eosinophil largely depends on IL-5. IL-5 may also participate in prolonged activation of eosinophils in the context of a persistent tumor burden, as indicated previously (30), and in fact, rIL-5 treatment prolonged eosinophil migration toward tumor cells in this study. Natural IL-5 production thus recruits sufficient eosinophils to the lung and generates a microenvironment that prevents establishment of tumor cells and potentially of invasion by foreign pathogens.
In regard to the mechanism of IL-5 production, the transient upregulation of IL-33 observed immediately after tumor injection suggests its role in the initiation of an antitumor response. IL-33 may also act directly on eosinophils together with IL-5 to create an antitumorigenic environment. Recent reports have described that IL-33 induces the degranulation of eosinophils (39) and elicits eosinophilia in the lung (21, 40). However, we should remain alert to the potential of potent Th2-skewing conditions negatively impacting Th1 antitumor adaptive immunity (41, 42), as eosinophils have been suggested to possess both anti- and protumorigenic activity. It is, however, difficult to determine the role of eosinophils in tumorigenesis because it can be affected by tumor type (43–45), and the phase analyzed is often diverse. One hypothesis supposed by the current study and the experiments using IL-5 transgenic mice (30) is that the number of existing eosinophils when tumor cells are inoculated or induced by a carcinogen largely affects the outcome. In early stages of tumor development, eosinophils may serve an antitumorigenic purpose, as a part of immune surveillance. In contrast, in the later stages of tumor development, eosinophils might contribute to angiogenesis to support tumor growth by secreting vascular endothelial growth factor, which may be enhanced by IL-5 (46). Therefore, it would be beneficial to design immunotherapies that enhance innate antitumor activity and adaptive Th1 antitumor immunity early in tumor development while preventing excessive Th2 skewing later in the development.
The role(s) of innate IL-5–producing cells in allergic diseases, such as asthma, was not investigated in this study. Present therapies for asthma are not appropriate for some people. IL-5 is a potent therapeutic target due to its causative relationship to eosinophilia. Therapies for asthma targeting IL-5 are currently being re-evaluated, and promising results have been obtained (47–50). Therefore, innate IL-5–producing cells should be preferable candidates to study allergic diseases.
In this report, we describe the generation of an IL-5/Venus KI mouse that led to the identification of innate IL-5–producing lymphoid cells that play a role in orchestrating eosinophil responses in the lung. This study together with recent reports that have identified novel populations of innate cells will provide new insights into mucosal immunity and Th2 immune responses, and future research in this area will hopefully lead to a fundamental resolution of debilitating immunological disorders such as allergic diseases.
The authors have no financial conflicts of interest.
We thank Dr. T. Hori, Dr. J. Fukuoka, and H. Hatta for technical support and Dr. A. Miyawaki for providing the Venus gene. We also thank Toyama prefecture and the sponsor companies supporting our laboratory.
↵1 M.I. and T.Y. equally contributed to this work.
This work was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology and by the Ministry of Education, Culture, Sports, Science and Technology Regional Innovation Cluster Program, Toyama/Ishikawa Region, Hokuriku Innovation Cluster for Health Science, Japan.
The online version of this article contains supplemental material.
Abbreviations used in this article:
- embryonic stem
- high-power field
- lamina propria
- mesenteric lymph node
- Peyer’s patch
- Received May 2, 2011.
- Accepted November 13, 2011.
- Copyright © 2012 by The American Association of Immunologists, Inc.