IL-5 Production by NK Cells Contributes to Infiltration in a Mouse Model of Allergic

This information is current as Christoph Walker, James Checkel, Salvatore Cammisuli, Paul of October 2, 2021. J. Leibson and Gerald J. Gleich J Immunol 1998; 161:1962-1969; ; http://www.jimmunol.org/content/161/4/1962 Downloaded from References This article cites 51 articles, 17 of which you can access for free at: http://www.jimmunol.org/content/161/4/1962.full#ref-list-1

Why The JI? Submit online. http://www.jimmunol.org/

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on October 2, 2021 Subscription Information about subscribing to The Journal of is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 1998 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. IL-5 Production by NK Cells Contributes to Eosinophil Infiltration in a Mouse Model of Allergic Inflammation1

Christoph Walker,2*† James Checkel,* Salvatore Cammisuli,† Paul J. Leibson,* and Gerald J. Gleich*

IL-5 production in vivo plays a unique role in the production, activation, and localization of in a variety of allergic conditions. The current paradigm suggests that -specific Th2 cells are the main source for the IL-5 production. The experiments outlined in this work, however, suggest that in vivo production of IL-5 by NK cells can separately influence eosinophil- associated inflammatory responses. Specifically, a mouse model of allergic inflammation was used in which C57BL/6 mice were immunized and challenged with a short ragweed Ag extract, known to induce a selective within the peritoneal cavity. Peritoneal lavage fluids from these mice also contained increased numbers of T cells and NK cells, as well as significantly elevated

levels of IL-4, IL-5, and IFN-␥. Flow-cytometric analysis of -producing cells in peritoneal lavage fluid revealed increased Downloaded from numbers of IL-5-producing cells in both and NK cell populations following allergen exposure. Depletion of NK cells by treatment with NK1.1 Abs selectively reduced the number of infiltrating eosinophils by more than 50%. Moreover, the inhibition of the infiltration of eosinophils was accompanied by a complete loss of IL-5-producing NK cells and significantly reduced levels of peritoneal lavage fluid IL-5, whereas the number of IL-5-producing T cells was not affected. Thus, the results presented in this study provide clear evidence for a novel immunoregulatory function of NK cells in vivo, promoting allergen-induced eosinophilic inflammatory responses by the production of IL-5. The Journal of Immunology, 1998, 161: 1962–1969. http://www.jimmunol.org/

lood and tissue eosinophilia is a characteristic abnormal- lasting and selective eosinophilia, whereas IL-5-deficient mice ity in and , and eosinophil-derived are unable to produce increased numbers of eosinophils in re- contribute to specific pathologic features such as epithe- sponse to specific Ags (5, 6, 18, 19). B ϩ lial and bronchial hyperresponsiveness (1–4). Increas- Allergen-specific CD4 Th2 cells producing IL-4, IL-5, and ing evidence indicates a unique role for IL-5 in the regulation of IL-10, but no IL-2 or IFN-␥, are currently regarded as the key this selective eosinophilia. IL-5 not only regulates the terminal regulatory cells controlling allergic eosinophilic inflammatory re- differentiation of committed eosinophil precursors, but also acti- sponses (11–13, 20–22). For example, in atopic individuals, aller- vates mature eosinophils, prolongs their survival, and enhances gen-specific T cell clones produce a Th2-like pattern of , by guest on October 2, 2021 (5–8). The central role of IL-5 in eosinophilic in- whereas other Ag-specific T cell clones from the same patients flammations has been demonstrated in various in vivo conditions. have a Th1-like pattern of cytokine production, secreting IL-2 and In humans, IL-5 is expressed during allergen-induced cutaneous IFN-␥, but no IL-4 or IL-5 (20). Furthermore, the infiltrating cells late phase reactions in atopic subjects and is detectable in bron- in allergen-induced late phase skin reactions, as well as after seg- chial mucosal biopsies of patients with asthma (9, 10). Further- mental allergen challenge in the of asthmatic and more, increased IL-5 levels have been demonstrated in bronchoal- patients express mRNA and produce proteins for IL-3, IL-4, IL-5, veolar lavage fluid from patients with mild to moderate allergic and GM-CSF, but not for IL-2 or IFN-␥ (9, 12, 13). Moreover, it and nonallergic asthma, as well as following segmental allergen has been demonstrated that the development of Ag-induced pul- challenge, closely related to the number of eosinophils and monary eosinophilia and airway hyperresponsiveness in murine severity (11–13). In experimental animal models, inhibition of models of allergic inflammation is closely associated and depen- IL-5 by neutralizing mAbs prevents the terminal differentiation of dent on CD4ϩ T cells producing a Th2 cell type pattern of cyto- eosinophils, suppresses the infiltration of mature cells into in- kines. Administration of anti-CD4 Abs before Ag challenge com- flamed tissues, and prevents the induction of bronchial hyperre- pletely prevented these responses (23, 24). sponsiveness (14–17). The unique role of IL-5 in the production, On the other hand, recent studies have demonstrated that human activation, and localization of eosinophils is further supported NK cells can be induced to produce IL-5 in vitro, suggesting that by the findings that mice overexpressing IL-5 develop a long these cells may contribute to the development of eosinophilic in- flammation (25, 26). Cytokines produced by NK cells have indeed been shown to play important immunoregulatory functions in the *Mayo Clinic, Department of Immunology, Rochester, MN 55902; and † Horsham Research Centre, Horsham, United Kingdom early responses to viral, bacterial, and parasitic , as well Received for publication June 18, 1997. Accepted for publication April 14, 1998. as in the development of T cell responses to these infectious agents The costs of publication of this article were defrayed in part by the payment of page (27, 28). For example, it is well established that NK cells are major charges. This article must therefore be hereby marked advertisement in accordance producers of IFN-␥ in vivo, thereby directing the differentiation of with 18 U.S.C. Section 1734 solely to indicate this fact. Th cells into IL-2- and IFN-␥-producing Th1-type cells (29–34). 1 This work was supported in part by National Institutes of Health Grant AI 345677. However, several in vitro studies have demonstrated that polariz- 2 Address correspondence and reprint requests to Dr. Christoph Walker, Novartis ing stimuli such as IL-4 and IL-12 profoundly affect the cytokine Horsham Research Centre, Department of Respiratory , Wimblehurst ␥ Road, Horsham, West Sussex RH12 4AB U.K. E-mail address: christoph. pattern produced by NK cells (25, 26). The production of IFN- in [email protected] cultures of purified peripheral blood NK cells was inhibited by

Copyright © 1998 by The American Association of Immunologists 0022-1767/98/$02.00 The Journal of Immunology 1963

IL-4, but significantly enhanced by IL-12. In contrast, IL-4 aug- Table I. Distribution of leukocytes in naive, ragweed sensitized and mented, whereas IL-12 inhibited the production of IL-5. Thus, challenged C57BL/6 mice similar to the generation of Th1 or Th2 cells, cytokines present in the local microenvironment may differentially affect the develop- Sensitized plus a b c ment of distinct cytokine-producing NK cell subsets. Conse- Naive Sensitized Challenged quently, one set of cytokines released by NK cells (IFN-␥) may Total cells 3.31 Ϯ 0.41d 3.98 Ϯ 0.54 11.13 Ϯ 1.53* favor the development of a characteristic Th1-type immune re- Eosinophils 0.04 Ϯ 0.01 0.26 Ϯ 0.06* 2.61 Ϯ 0.42* Ϯ Ϯ Ϯ sponse, whereas other cytokines such as IL-5 may contribute to 0.72 0.08 0.92 0.13 2.57 0.37* 0.01 Ϯ 0.01 0.02 Ϯ 0.01 0.05 Ϯ 0.02* eosinophilic inflammation. However, whether NK cells produce 2.45 Ϯ 0.32 2.71 Ϯ 0.39 5.82 Ϯ 0.86* IL-5 in vivo and thereby contribute to the development of eosin- Mast cells 0.09 Ϯ 0.02 0.05 Ϯ 0.01 0.06 Ϯ 0.01 ophilic inflammatory responses is not yet known. To address this a Naive mice: injected with saline during the sensitization and challenged period. question, we analyzed the distribution and cytokine production of b Sensitized mice: immunized with ragweed and challenged with saline. NK cells in a well-established murine model of allergic inflam- c Sensitized and challenged mice: immunized and challenged with ragweed. d Cells ϫ 106; mean values Ϯ SD from 12 to 18 mice per group. mation, known to be associated with a selective tissue accumula- * Significantly different from the corresponding control group (p Ͻ 0.05). tion of eosinophils (35–37). The data presented in this study clearly demonstrate that NK cells indeed produce IL-5 in vivo, and thereby exert an important regulatory function in allergen-induced of CyChrome-conjugated anti-CD3 mAb and FITC-conjugated anti-pan eosinophilic inflammation. NK or NK1.1. mAbs (PharMingen) in the dark on ice for 30 min. Cells were then washed twice with staining buffer and fixed with a 1% para- Materials and Methods formaldehyde solution (pH 7.4 in PBS). Cytofluorometric analysis was Downloaded from Animals, sensitization, and challenge procedure performed using laser excitation at 488 nm, and the number of immun- ofluorescence-positive cells was determined per 10,000 analyzed cells. C57BL/6 mice (females, 18 to 25 g, 6 to 8 wk of age) were obtained from Specific binding of mAbs was controlled by subtraction of - The Jackson Laboratory (Bar Harbor, ME). Mice were immunized s.c. with matched control Abs. short ragweed Ag extract (1/10,000 dilution; Greer Laboratories, Lenoir, NC) in 0.2 ml of saline containing penicillin/streptomycin (50 U/ml and 5 Determination of cytokine-producing subpopulation

␮g/ml, respectively; Sigma, St. Louis, MO) on days 1 and 8. Sham-im- http://www.jimmunol.org/ Cytokine-producing peritoneal lavage cells were determined as recently munized mice received two injections of saline alone. Seven days after the described (38). Briefly, peritoneal lavage cells were incubated for4hat last immunization (day 15), animals were challenged by i.p. injection of 0.2 37°C in RPMI containing 10% FCS and brefeldin A (10 ␮g/ml; Sigma) to ml ragweed Ag extract. Saline- or ragweed-immunized control groups re- disaggregate the Golgi complex, enabling newly synthesized proteins to ceived an i.p. injection of 0.2 ml of saline. Forty eight hours after the Ag accumulate intracellularly (39). Cells were then washed with PBS and in- provocation, animals were killed and peritoneal lavages were performed ϩ ϩ cubated for 30 min at 4°C with optimal concentration of CyChrome-con- with 3 ml of HBSS (without Ca2 and Mg2 ; Celox, Hopkins, MN) con- jugated anti-CD3 mAb and FITC-conjugated pan NK or NK1.1. mAbs taining 0.1% BSA (Sigma). Total cell numbers and leukocyte differentials (PharMingen). In some experiments, the anti-CD3 Ab was replaced by a were performed as described below. Lavages were centrifuged and super- mixture of anti-␣/␤ and ␥/␦ TCR mAbs (Fig. 7; PharMingen). Cells were natants were frozen at Ϫ20°C until use for cytokine measurements. washed again and fixed in 100 ␮l solution A (Fix & Perm cell permeabi-

Depletion of NK cells lization kit; Caltag Laboratories, San Francisco, CA) for 15 min at room by guest on October 2, 2021 temperature, washed in PBS, and resuspended in 100 ␮l of permeabiliza- Mice were depleted of NK cells by i.v. administration of anti-mouse NK1.1 tion solution B (Fix & Perm kit; Caltag Laboratories) containing phyco- mAbs (30 ␮g/mouse/day; PharMingen, San Diego, CA) either before al- erythrin-labeled anti-IL-5 (TRFK5), anti-IFN-␥ (4S.B3), or isotype- lergen challenge (days 14 and 15) or both during immunization and before matched control Abs (all purchased from PharMingen). Cells were challenge (days Ϫ1, 0, ϩ1, 14, and 15). Control groups of mice received incubated for another 15 min at room temperature, washed in PBS, and an isotype-matched control mAb (mouse IgG2a, 30 ␮g/mouse/day; immediately analyzed using a FACScan (Becton Dickinson, San Jose, CA). PharMingen) at the same time intervals. Specificity of the staining was controlled by isotype-matched control mAbs and by preincubation of the permeabilized cells with nonconjugated anti- Determination of total cells and leukocyte differentials IL-5 or anti-IFN-␥ mAbs before adding the phycoerythrin-conjugated anti- The total nucleated cell count was determined microscopically following cytokine mAb, reducing the specific immunofluorescence signal to back- staining of peritoneal lavage cells by Randolph’s stain, and calculated as ground levels. total cells per recovered volume. Cytologic examinations of peritoneal Statistical analysis lavage cells were done after cytocentrifugation and staining with May- Gruenwald-Giemsa. The relative proportions of the various leukocyte sub- Statistical anlysis was performed using the two-tailed Mann-Whitney U populations were determined by a cell differential count of 1000 cells. test. Differences associated with probability values of p Ͻ 0.05 were con- sidered significant. Quantitation of cytokines in peritoneal lavage fluid IL-5, IL-4, and IFN-␥ were measured by sandwich ELISA using two mAbs Results recognizing different of the specific cytokine. Abs used for mea- Leukocytes and cytokines in peritoneal lavage fluid from suring IL-5 (TRFK5 and biotinylated TRFK4), IL-4 (BVD4-1D11 and bi- otinylated BVD6-24G2), or IFN-␥ (R4-6A2 and biotinylated XMG1.2) ragweed-immunized and -challenged C57BL/6 mice were all purchased from PharMingen. In all cases, binding of the second Intraperitoneal injection of Ag in short ragweed Ag-immunized Ab was analyzed by stepwise incubation with streptavidin-alkaline phos- mice induced an infiltration of various leukocytes into the perito- phatase conjugate (Mabtech, Stockholm, Sweden) and 4-nitrophenylphos- phate disodium salts (Sigma). OD was measured at 405 nM, and cytokine neal cavity (Table I). Significantly increased total numbers of eo- concentration was calculated based on the results from serial dilutions of sinophils, lymphocytes, neutrophils, and macrophages were found standard recombinant mouse IL-5, IL-4, and IFN-␥, respectively. The sen- in allergen-challenged mice compared with the nonimmunized or sitivity of the cytokine ELISAs was about 10 pg/ml. sham-challenged group with the most pronouced change in the Determination of lymphocyte subpopulation by number of eosinophils. Both the absolute as well as the relative immunofluorescence numbers of eosinophils were increased significantly in allergen- challenged mice (1.1% Ϯ 0.1 in naive, 5.9% Ϯ 0.7 in nonchal- Specific binding of mAbs was analyzed by direct immunofluorescence us- Ϯ ing a FACScan flow cytometer (Becton Dickinson, Mountain View, CA). lenged, and 23.7% 1.4 in challenged mice), whereas the relative Briefly, 1 ϫ 105 cells in staining buffer (PBS containing 2% FCS and 0.1% numbers of lymphocytes, neutrophils, or macrophages were not sodium azide) were incubated in the presence of saturating concentrations changed or decreased. 1964 REGULATION OF EOSINOPHIL INFILTRATION BY NK CELLS

Table II. Production of Th1 and Th2 cell cytokines following ragweed challenge

Sensitized plus Naivea Sensitizedb Challengedc

IL-4 (pg/ml) 11.0 Ϯ 7.6d 21.9 Ϯ 9.1 61.4 Ϯ 6.7* IL-5 (pg/ml) 7.4 Ϯ 1.1 15.2 Ϯ 4.0 51.6 Ϯ 11.2* IFN-␥ (pg/ml) 3.6 Ϯ 2.3 11.7 Ϯ 7.5 28.6 Ϯ 8.8*

a Naive mice: injected with saline during the sensitization and challenge period. b Sensitized mice: immunized with ragweed and challenged with saline. c Sensitized and challenged mice: immunized and challenged with ragweed. d Mean values Ϯ SD from six to eight mice per group. * Significantly different from the corresponding control group (p Ͻ 0.05). FIGURE 1. Distribution of lymphocytes, T cells, and NK cells in peri- toneal lavage fluid from ragweed-sensitized and -challenged mice. Total number of lymphocytes, CD3-positive T cells (CD3), NK1.1.-expressing Parallel to the selective eosinophil infiltration, peritoneal la- CD3-positive T cells (CD3ϩNKϩ), and CD3-negative, NK1.1.-positive ϩ vages obtained from allergen-exposed, ragweed-sensitized mice NK cells (NK ) in peritoneal lavage fluids from ragweed-sensitized, sa- contained significantly increased levels of the predominantly Th2 line-challenged (shaded bars), and ragweed-sensitized and -challenged cell type-derived cytokines IL-4 and IL-5 (Table II). The concen- (solid bars) mice. Flow-cytometric analysis was performed on cells stained with CyChrome-conjugated anti-CD3 and FITC-conjugated NK1.1. mAbs. tration of IFN-␥, mainly produced by Th1 and NK cells, was also Results are expressed as mean values Ϯ SEM from 16 to 18 mice per increased in the peritoneal cavity from these animals. However, group. *Denotes significantly different values compared with the corre- Downloaded from ␥ the absolute IFN- concentration after allergen challenge was con- sponding control groups (p Ͻ 0.05). siderably lower compared with the levels of IL-4 or IL-5. Taken together, i.p. provocation of actively immunized C57BL/6 mice with ragweed Ags induced a selective accumulation of eosinophils number of NK cells. These results suggest that all three cell types in the peritoneal cavity and the production of a predominantly Th2 may participate in the regulation of an eosinophilic inflammation. cell type cytokine pattern, both characteristic features of an allergic This is further supported by the fact that T cells, NK1.1.-bearing T http://www.jimmunol.org/ inflammatory response. cells, as well as NK cells were all shown to be capable of IL-5 production in vitro (20–22, 26, 26, 40, 41). Increased numbers of IL-5-producing T cells and NK cells in Indeed, by using a flow-cytometric technique that allows the peritoneal lavages following ragweed Ag challenge measurement of intracellular accumulated, newly synthesized cy- To identify the cellular source for the increased levels of Th1 and tokines on a single cell level (38, 42, 43), we could demonstrate Th2 cell type cytokines present in peritoneal lavages after allergen that increased numbers of T cells as well as NK cells produced provocation, the total number and cytokine-producing T cells and IL-5 following ragweed provocation (Figs. 2 and 3). Figure 2 NK cells were determined using immunofluorescence-staining shows a representative example of a three-color immunofluores- techniques and flow cytometry. First, the number of T cells and cence used to determine the number of IL-5-producing T and NK by guest on October 2, 2021 NK cells present in peritoneal lavage fluids from the various cells. Based on their expression of CD3 or NK1.1 Ags, cells were groups of ragweed-sensitized and -challenged mice were deter- electronically separated into CD3-expressing T cells and CD3- mined by staining peritoneal lavage cells with mAbs directed negative NK cells (Fig. 2A), followed by analysis of their fluores- against CD3 or against the NK cell marker NK1.1. Double fluo- cence profile after staining with either anti-IL-5 or isotype- rescence analysis of lymphocytes from these animals revealed the matched control Abs. The immunofluorescence signal obtained presence of three clearly separable populations of Ab-stained lym- from anti-IL-5-stained T cells (Fig. 2C) and NK cells (Fig. 2E) was phocytes, a CD3-positive, NK1.1-negative T cell population, a clearly separable from the measurements of Ig isotype-matched small population of double-positive T cells, as well as a population control samples (Fig. 2, B and C) and completely inhibitable by of CD3-negative NK1.1.-positive cells representing NK cells. As preincubation of the permeabilized cells with unconjugated anti- shown in Figure 1, all three subpopulations of lymphocytes were cytokine Abs (data not shown), clearly demonstrating the presence increased significantly in peritoneal lavage fluids from ragweed- of IL-5-producing T cells and NK cells in peritoneal lavages from challenged mice, with the most pronounced relative increase in the ragweed-challenged mice. As shown in Figure 3, peritoneal lavage

FIGURE 2. Identification of IL-5-producing T cells and NK cells in peritoneal lavage fluid from ragweed-sensitized and -challenged mice. Dot-plot analysis of peritoneal lavage cells from ragweed-challenged mice stained with Cy- Chrome-conjugated anti-CD3 (FL3), FITC-con- jugated NK1.1. (FL1), and phycoerythrin-labeled anti-IL-5 or isotype-matched control Abs (FL2). A, Representative double immunofluorescence of CD3- and NK1.1.-stained cells with gates set around the CD3-positive (CD3) and the CD3- negative NK1.1. (NK)-positive population. Dis- tribution of control (B) and anti-IL-5 (C)-stained and -gated CD3-positive T cells or control (D)or anti-IL-5 (E)-stained and -gated NK cells. The Journal of Immunology 1965

ing T cells (Fig. 5A), whereas the number of NK cell Ags express- ing CD3-positive T cells (Fig. 5B) was significantly, but not com- pletely, reduced. Similar results were obtained by staining the cells with either fluorescent labeled NK1.1. Abs or a pan NK cell marker that recognizes a different molecule on NK cells with a very similar cell distribution compared with NK1.1. (data not shown). Next, the effect of NK cell depletion on the allergen-induced eosinophil accumulation into the peritoneal cavity was analyzed. As shown in Figure 6, treatment with NK1.1, but not isotype- matched control Abs resulted in a significant and selective reduc- tion of eosinophils present in peritoneal lavages obtained from ragweed-challenged and -immunized mice. The total cell numbers FIGURE 3. Increased numbers of IL-5-producing T cells and NK cells as well as the number of lymphocytes were slightly decreased, in peritoneal lavages from ragweed-sensitized and -challenged mice. Num- whereas a slight increase of macrophages was observed in these bers of IL-5-producing NK cells (A) and T cells (B) from allergen-sensi- NK cell-depleted mice. Furthermore, no differences were found by tized, nonchallenged (control), or ragweed-sensitized and -challenged (rag- comparing the effect of NK1.1. treatment before challenge or dur- weed) mice, as measured by triple immunofluorescence (for details, see Materials and Methods and Fig. 2). Results are expressed as mean val- ing immunization and challenge, suggesting that the main eosin- Ϯ ophil active effect of NK cells takes place directly in response to ues SEM from 10 to 12 mice per group. *Denotes significantly different Downloaded from values compared with the corresponding control groups (p Ͻ 0.05). the allergen challenge. Depletion of NK cells reduces the level of IL-5 in peritoneal lavage fluids, but did not change the number of IL-5-producing fluids obtained from these allergen-exposed mice contained sig- T cells nificantly increased numbers of IL-5-producing T cells and NK cells as compared with ragweed-sensitized, saline-challenged an- To further analyze whether the depletion of NK cells was also http://www.jimmunol.org/ imals. Thus, both T cells and NK cells produce IL-5 in response to accompanied by a reduced number of IL-5-producing cells and allergen exposure in vivo, and may thereby contribute to the Ag- diminished levels of IL-5, peritoneal lavages obtained from NK induced, IL-5-mediated tissue accumulation of eosinophils. Simi- cell-depleted and allergen-sensitized and -challenged mice were lar results were obtained by analyzing the number of IL-5-produc- analyzed for their content of cytokines as well as IL-5- and IFN- ing, NK1.1.-expressing CD3 cells, which were elevated ␥-producing T cells and NK cells. As shown in Table III, depletion significantly following allergen exposure (0.19 ϫ 103 in nonchal- of NK cells significantly reduced the concentration of IL-5, IL-4, lenged mice vs 1.01 ϫ 103 in ragweed-challenged mice). How- and IFN-␥ in peritoneal lavage fluid from allergen-challenged ever, due to the low number of IL-5-expressing cells within this mice. The inhibition of IL-5 was much more pronounced com- ␥ lymphocyte subpopulation, this type of analysis was not feasible in pared with the changes of IL-4 and IFN- levels. Again, no effect by guest on October 2, 2021 all analyzed mice. was found in control Ab-treated animals, and no significant dif- ferences were observed by comparing the effect of NK1.1. treat- Depletion of NK cells results in a selective reduction of the ment before challenge or before immunization and challenge. Ag-induced eosinophil infiltration Moreover, analysis of cytokine-producing T cells and NK cells To further analyze whether IL-5 derived from NK cells contributes clearly demonstrated that treatment with NK1.1 Abs completely ␥ to the ragweed Ag-induced eosinophil infiltration, C57BL/6 mice eliminated the total number as well as the IL-5- and IFN- -pro- were treated with anti-mouse NK1.1 mAbs, a well-established ducing NK cells in the peritoneal cavity (Fig. 7, B, D, and F). In method to deplete NK cells in mice (44–46). Abs against NK1.1 contrast, the total number as well as the proportion of cytokine- or isotype-matched control Abs were administered i.v. either producing T cells was not altered significantly by the Ab treatment shortly before the ragweed challenge or before both the immuni- (Fig. 7, A, C, and E), suggesting that NK cells directly affect the zation and challenge procedures. Figure 4 shows a representative allergen-induced tissue accumulation of eosinophils by the pro- flow-cytometric measurement from cells stained with anti-CD3 duction of IL-5 without inhibiting cytokine release from T cells. and an anti-NK cell Ab obtained from ragweed-challenged, NK1.1.-treated and -challenged, or control Ab-treated and -chal- Discussion lenged mice. Treatment with NK1.1. Abs completely depleted the It is well established that NK cells exert potent immunoregulatory CD3-negative NK cell population, whereas a significant proportion activities in vivo and in vitro (27, 28). In particular, the secretion of the double-positive CD3/NK1.1-expressing T cells remained of high levels of IFN-␥ is thought to play an important role in present. This is also shown in Figure 5, demonstrating that NK cell immune responses against various by determining the depletion by NK1.1. Abs did not alter the number of CD3-express- differentiation of Ag-specific Th cells producing a Th1 cytokine

FIGURE 4. Flow-cytometric analysis of the distribution of T cells and NK cells in NK1.1.- treated mice. Representative dot-plot analysis of peritoneal lavage cells obtained from ragweed- sensitized and -challenged mice stained with Cy- Chrome-conjugated anti-CD3 (FL3) and FITC- conjugated NK1.1. (FL1) mAbs. Distribution analysis from nontreated (A), NK1.1.-treated (B), and control Ab-treated mice (C). 1966 REGULATION OF EOSINOPHIL INFILTRATION BY NK CELLS

FIGURE 5. Distribution of CD3ϩ, CD3ϩNK1.1.ϩ, and NK cells in peritoneal la- vage fluid of NK1.1.-treated mice. Numbers of CD3ϩNK1.1.Ϫ (A), CD3ϩNK1.1.ϩ (B), and CD3ϪNK1.1.ϩ (C) peritoneal lavage lympho- cytes from ragweed-sensitized, nonchallenged (c), ragweed-challenged (Rg), control Ab- treated and -challenged (cmAb), and NK1.1.- treated and -challenged (NK1.1.) mice. Results are expressed as mean values ϩ SEM from 10 to 12 mice per group. *Denotes significantly different values compared with the correspond- ing control group.

pattern with IL-2 and IFN-␥, but no IL-4 or IL-5 (27–31). How- Many studies have focused on identifying the cellular source of ever, the recent description of IL-5 production by human NK cells IL-5 in allergic inflammation. CD4ϩ and CD8ϩ T cells, mast cells, in vitro raised the question as to whether these cells produce a and eosinophils have all been shown to be capable of IL-5 pro- distinct, polarized cytokine profile similar to Th1 and Th2 cell duction in vivo and in vitro (47–50). With regard to T cell subsets, Downloaded from types, and thereby contribute to allergen-induced eosinophilic in- CD4ϩ cells appear to be a much more important source for IL-5 flammatory responses (25, 26). To answer this question, we ana- than CD8ϩ T cells. This could be demonstrated in experiments in lyzed the distribution and cytokine production of NK cells in a which mice were depleted of CD4- or CD8-positive T cells by ragweed-induced murine allergic model. The results treatment with specific Abs (23, 24, 37). Under these conditions, presented in this study provide clear evidence for the presence of Ag-induced IL-5 release and eosinophil infiltration into the mouse IL-5-producing NK cells in vivo. Moreover, depletion of NK cells bronchial tissue of peritoneal cavity were suppressed in CD4, but not only reduced the levels of IL-5 in peritoneal lavage fluid, but not CD8, T cell-depleted animals, supporting the concept of an http://www.jimmunol.org/ also selectively inhibited the infiltration of eosinophils into the IL-5 and CD4 T cell dependency of eosinophilic inflammations in peritoneal cavity following allergen challenge. Thus, these results mice. Similar conclusions were reached from clinical studies with clearly indicate that NK cells have indeed the capacity to regulate patients with asthma, in which IL-5 produced in the bronchial mu- tissue accumulations of eosinophils by overproduction of IL-5. cosa appeared to be mainly associated with CD4ϩ T cells, and the The results obtained with the ragweed allergen-induced murine CD4 T cell activation profile correlated with the concentration of allergic peritonitis model confirm earlier studies using other mouse IL-5 in serum and bronchoalveolar lavage fluids (10, 11, 51). More strains and/or other (35–37). In those studies, i.p. aller- controversial are studies on the relative importance of mast cells as

gen challenge of sensitized mice also resulted in a massive accu- cytokine-producing cells with eosinophilia-regulating capabilities. by guest on October 2, 2021 mulation of eosinophils within the peritoneal cavity, which was Although mast cells have been shown to produce and contain pre- associated with increased levels of IL-5 and IFN-␥. In addition, formed stores of IL-4 and IL-5, and IgE-dependent activation re- treatment with neutralizing anti-IL-5 mAb completely inhibited sults in the secretion of these eosinophil-regulating factors, mast the tissue accumulation of eosinophils (35, 37), demonstrating the cell-deficient mice do not show any defect in allergen-induced eo- IL-5 dependency of this allergic eosinophilia. sinophil recruitment (52, 53). These data indicate that mast cells

FIGURE 6. Effect of depletion of NK cells on Ag-induced eosinophil infiltration in immunized C57BL/6 mice. Ragweed- immunized mice were treated with NK1.1 mAb (NK1.1) or isotype-matched control mAb (CmAb) either before the ragweed challenge (challenge) or during the immuni- zation and before challenge (sensitization ϩ challenge). Results are expressed as percent- age (mean values Ϯ SEM from 12 to 18 mice) of the baseline values (see Table I) obtained from ragweed-immunized and -challenged mice without Ab treatment (rag- weed). (A, total cells; B, eosinophils; C, lymphocytes; and D, macrophages.) *De- notes values significantly different from the other group, at least p Ͻ 0.05. The Journal of Immunology 1967

Table III. Effect of NK cell depletion on peritoneal lavage cytokine produce IL-5 and IFN-␥ following stimulation with IL-2 in pri- levels mary and secondary cultures. The production of IL-5 was IL-2 dependent and required the presence of accessory cells or NK- b NK1.1a cmAb NK1.1 cmAb specific target cells. Moreover, the secretion of IL-5 and IFN-␥ ϩ c ϩ (s c) (s c) (c) (c) was affected profoundly by the addition of IL-4, IL-12, IL-10, and IL-4 63 Ϯ 13d* 92 Ϯ 871Ϯ 11* 93 Ϯ 5 IL-15 and unrelated to the cytolytic function of these cells. We IL-5 29 Ϯ 8* 94 Ϯ 16 41 Ϯ 9* 95 Ϯ 8 have confirmed and extended these observations and demonstrated ␥ Ϯ Ϯ Ϯ Ϯ IFN- 55 12* 92 10 70 12* 103 9 the existence of distinct IL-5 high and low producing NK cell a Mice treated with NK1.1 Abs during the sensitization and/or challenged period. clones producing a differential cytokine pattern (unpublished ob- b Mice treated with isotype-matched control Abs (cmAb). servation). Moreover, the polarized cytokine profile of both IL-5 c Abs administered either during the sensitization and challenge period (s ϩ c) or shortly before the allergen challenge (c). high and low producing NK cell subsets was found to be stable d Results are expressed as percent of the baseline values obtained from ragweed over time and most likely determined by exposure to specific cy- sensitized and challenged mice without Ab treatment (see Table II) (mean values Ϯ SD from six to eight mice per group). tokines during initial activation events. This was demonstrated in * Significantly different from the control group (p Ͻ 0.05). experiments in which NK cells cultured in the presence of IL-12 lost their ability to produce IL-5 in response to IL-2, despite pro- ducing similar levels of IFN-␥ in parallel cultures without IL-12. are of limited importance as a cellular source of IL-5 in the reg- Thus, the cytokine environment present at the beginning of an ulation of allergen-induced eosinophil accumulation in inflamed inflammatory response may induce the production of a polarized tissues. Besides T cells and mast cells, eosinophils themselves cytokine pattern in NK cells, which, similar to the well-known Th1 Downloaded from have been shown to produce IL-5. However, the current view and Th2 cell types, may influence the outcome of a specific im- about eosinophil-derived IL-5 is that this cytokine acts predomi- mune response. The present study further extends these observa- nantly as an autocrine survival factor within the tissue. In fact, we tions by demonstrating first an increased infiltration of NK cells in were not able to detect IL-5-producing eosinophils in our analysis response to allergen exposure of sensitized mice, and second, IL-5 of cytokine-producing cells obtained from peritoneal lavage fluids production by these NK cells in vivo. The allergen-induced in- of allergen-sensitized and -challenged mice. This might be due to crease in the total number as well as number of IL-5-producing NK the sensitivity of the method used to detect intracellular cytokines, cells was comparable with the results obtained with T cells, al- http://www.jimmunol.org/ but this result suggests that the amount of IL-5 produced by eo- though the absolute values were about 10 times lower for NK cells. sinophils appears to be considerably lower compared with T cells Nevertheless, these results suggest that significant amounts of the and NK cells. IL-5 detectable at the site of inflammation are derived from NK More recently, H. Warren et al. have demonstrated IL-5 pro- cells. However, conclusions about the relative contribution of T duction by human NK cells in vitro (25, 26). In their studies, pu- cells and NK cells to the overall IL-5 production are difficult, be- rified peripheral blood NK cells were analyzed for their capacity to cytokine-producing cells were only analyzed at a single by guest on October 2, 2021

FIGURE 7. Effect of depletion of NK cells on IL-5- and IFN-␥-producing T cells and NK cells. Ragweed-immunized mice were treated with NK1.1 mAb (solid bars) or isotype-matched control mAb (shaded bars) either before the ragweed challenge (challenge) or during the immunization and before challenge (sensitization ϩ challenge). Results are expressed as percentage (mean values Ϯ SEM from 10 to 12 mice) of the baseline values obtained from ragweed-immunized and -challenged mice without Ab treatment (Control, striped bars). (A, Total CD3 or ␣/␤ and ␥/␦ TCR-expressing lymphocytes; B, total NK1.1-expressing, CD3- or TCR-negative lymphocytes; C, total IL-5-producing T cells; D, total IL-5-producing NK cells; E, total IFN-␥-producing T cells; F, total IFN-␥-producing NK cells.) *Denotes values significantly different from the other group, at least p Ͻ 0.05. 1968 REGULATION OF EOSINOPHIL INFILTRATION BY NK CELLS point in time. Therefore, we cannot exclude the possibility that 3. Calhoun, W. J., J. Sedwick, and W. W. Busse. 1991. The role of eosinophils in kinetic differences in the cytokine production profile between T the pathophysiology of asthma. Ann. N.Y. Acad. Sci. 629:62. 4. Wardlaw, A. J., R. Moqbel, and A. B. Kay. 1995. Eosinophils: biology and role cells and NK cells following Ag provocation may differentially in disease. Adv. Immunol. 60:151. affect the levels of IL-5 as well as the number of infiltrating eo- 5. Dent, L. A., M. Strath, A. L. Mellor, and C. J. Sanderson. 1990. Eosinophilia in transgenic mice expressing IL-5. J. Exp. Med. 172:1425. sinophils. On the other hand, the results obtained with NK cell- 6. Sanderson, C. J. 1992. IL-5, eosinophils and disease. Blood 79:3101. depleted mice, demonstrating a reduction of the IL-5 levels of 7. Takatsu, K., S. Takaki, and Y. Hitoshi. 1994. IL-5 and its receptor system: im- more than 50%, rather suggest a higher contribution of NK cells to plication in the and inflammation. Adv. Immunol. 57:145. 8. Walker, C., R. K. Braun, C. Boer, C. Kroegel, J. C. Virchow, and T. T. Hansel. the total amount of IL-5 produced than would be expected from the 1994. Cytokine control of eosinophils in pulmonary diseases. J. Allergy Clin. total number of cytokine-producing cells. Immunol. 94:1262. We cannot exclude the possibility that depletion of NK cells by 9. Kay, A. B., S. Ying, V. Varney, M. Gaga, S. R. Surham, R. Moqbel, S. J. Wardlaw, and Q. Hamid. 1991. mRNA expression of the cytokine gene NK1.1 Abs has additional effects, such as changes in the overall cluster IL-3, IL-4, IL-5 and GM-CSF in allergen induced late-phase cutaneous cell-cell communication network as well as depletion of additional, reactions in atopic subjects. J. Exp. Med. 173:774. IL-5-producing cell types. Indeed, it recently has been demon- 10. Bentley, A. M., Q. Meng, D. S. Robinson, Q. Hamid, A. B. Kay, and S. R. Durham. 1993. Increases in activated T cells, eosinophils and cytokine strated that a small subpopulation of T cells bearing the NK1.1 Ag mRNA expression for IL-5 and GM-CSF in bronchial biopsies after allergen rapidly produces IL-4, IL-5, and IFN-␥ after in vivo activation via inhalation challenge in atopic asthmatics. Am. J. Respir. Cell Mol. Biol. 8:35. 11. Walker, C., E. Bode, L. Boer, T. T. Hansel, K. Blaser, and J. C. Virchow. 1992. the CD3/TCR complex (41, 42). These observations suggest that Allergic and non-allergic asthmatics have distinct patterns of T cell activation and these cells may provide the initial source of IL-4 required for the cytokine production in peripheral blood and BAL. Am. Rev. Respir. Dis. 146:109. priming of CD4ϩ T cells to develop into a Th2 cell phenotype, as 12. Robinson, S., Q. Hamid, S. Bentley, S. Ying, A. B. Kay, and S. R. Durham. 1993. Activation of CD4ϩ T cells, increased TH2-type cytokine mRNA expression and well as for the initiation of IgE production by B cells (47, 48). eosinophil recruitment in bronchoalveolar lavage after allergen inhalation chal- Downloaded from Consequently, one might expect that treatment of mice with anti- lenge in patients with atopic asthma. J. Allergy Clin. Immunol. 152:351. NK1.1 Abs, which not only deplete NK cells, but also those spe- 13. Virchow, J. C., C. Walker, D. Hafner, C. Kortsik, P. Werner, H. Matthys, and C. Kroegel. 1995. T cells and cytokines in bronchoalveolar lavage fluid after cific NK1.1-bearing T cell subsets, would prevent the development segmental allergen provocation in atopic asthma. Am. J. Respir. Crit. Care Med. of Th2 cells and reduce the production of the Th2 cell-derived 151:960. cytokines IL-4 and IL-5, as well as the synthesis of allergen-spe- 14. Coffman, R. L., B. W. P. Seymour, S. Huddak, J. Jackson, and D. Rennick. 1989. to IL-5 inhibits helminth induced eosinophilia in mice. Science 245: cific IgE. Our results, however, clearly demonstrate that treatment 308. of mice with anti-NK1.1. Abs almost completely eliminated NK 15. Mauser, P. J., A. Pitman, X. Fernandez, J. A. Zurcher, T. Kung, A. S. Watnick, http://www.jimmunol.org/ R. W. Egan, W. Kreutner, and G. K. Adams III. 1993. Inhibitory effect of the cells within the pertioneal cavity, whereas the numbers of NK1.1.- TRFK-5 anti-IL-5 antibody in a guinea pig model of asthma. Am. Rev. Respir. expressing CD3-positive T cells were reduced by only 50%. Sim- Dis. 148:1623. ilarly, the number of IL-5-producing T cells bearing the NK1.1. Ag 16. Mauser, P. J., A. M. Pitman, X. Fernandez, S. K. Foran, G. K. Admas, W. Kreutner, R. W. Egan, and R. W. Chapman. 1995. Effects of an antibody to was 5- to 10-fold lower compared with the number of IL-5-pro- IL-5 in a monkey model of asthma. Am. J. Respir. Crit. Care Med. 152:467. ducing NK cells. Moreover, treatment of mice with anti-NK1.1 17. Van Oosterhout, A. J. M., A. R. C. Ladenius, H. F. J. Savelkoul, I. Van Ark, Abs either before the allergen challenge or during the immuniza- K. C. Delsman, and F. P. Nijkamp. 1993. Effect of anti-IL-5 and IL-5 on airway hyperreactivity and eosinophilia in guinea pigs. Am. Rev. Respir. Dis. 147:548. tion and challenge procedure resulted in comparable suppression 18. Foster, P. S., S. P. Hogan, A. J. Ramsay, K. I. Matthaei, and I. G. Young. 1996. 5 deficiency abolishes eosinophilia, airway hyperreactivity and lung of eosinophil infiltration and cytokine production, with no overall by guest on October 2, 2021 change of the Th2 cell cytokine pattern, suggesting that treatment damage in a mouse asthma model. J. Exp. Med. 183:195. 19. Kopf, M., F. Brombacher, P. D. Hodgkin, A. J. Ramsay, E. A. Milboune, with NK1.1 Abs during the immunization did not alter the devel- W. J. Dai, K. S. Ovington, C. A. Behm, G. Koehler, I. G. Young, and opment of Ag-specific Th2 cells nor the production of IgE. Fur- K. I. Matthaei. 1996. IL-5 deficient mice have a developmental defect in CD5ϩ B1 cells and lack eosinophilia but have normal antibody and thermore, no significant changes were found in the number of in- responses. 4:15. filtrating and cytokine-producing T cells, suggesting no T cell 20. Wierenga, E. A., M. Snoek, C. De Groot, I. Cretien, J. D. de Bos, H. M. Jansen, inhibitory effect through the administration of NK1.1. Abs. Taken and M. L. Kapsenberg. 1990. Evidence for compartmentalization of functional subsets of CD4ϩ T lymphocytes in atopic patients. J. Immunol. 144:4651. together, these data clearly indicate that, at least in our model, 21. Robinson, D. R., Q. Hamid, S. Ying, A. Tsicopoulos, J. Barkans, A. M. Bentley, NK1.1.-bearing T cells are of limited importance in the regulation C. Corrigan, S. R. Durham, and A. B. Kay. 1992. Predominant Th2-like bron- of eosinophilia, eosinophil infiltration, and IL-5 production. choalveolar lavage T-lymphocyte population in atopic asthma. N. Engl. J. Med. 326:298. In conclusion, the data presented in this study clearly demon- 22. Ying, S., S. R. Durham, C. J. Corrigan, Q. Hamid, and A. B. Kay. 1995. Phe- strate that NK cells are capable of IL-5 production in vivo, and notype of cells expressing mRNA for TH2-type (IL-4 and IL-5) and TH1 type ␥ thereby contribute to the development of an eosinophilic inflam- (IL-2 and IFN- ) cytokines in bronchoalveolar lavage and bronchial biopsies from atopic asthmatic and normal control subjects. Am. J. Respir. Cell Mol. Biol. matory response. These results, together with our recent data with 12:477. 23. Nakajima, H., I. Iwamoto, S. Tomoe, R. Matsumara, H. Tomioka, K. Takatsu, NK cell clones demonstrating clear differences in signal require- ϩ ␥ and S. Yoshida. 1992. CD4 lymphocytes and IL-5 mediate induced ment for the induction of IL-5 and IFN- production and the ex- eosinophil infiltration into the mouse . Am. Rev. Respir. Dis. 144:374. istence of distinct IL-5 high and low producing NK cell clones, 24. Coyle, A. J., G. LeGros, C. Bertrand, S. Tsuyuki, C. H. Heusser, M. Kopf, and suggest that cytokines available in the local environment pro- G. P. Anderson. 1995. is required for the induction of lung TH2 mucosal immunity. Am. J. Respir. Cell Mol. Biol. 13:54. foundly affect the range of cytokines produced by activated NK 25. Warren, H. S., B. F. Kinnear, J. H. Phillips, and L. L. Lanier. 1995. Production cells. As a consequence, one set of cytokines released by NK cells of IL-5 by human NK cells and regulation of IL-5 secretion by IL-4, IL-10 and may favor the development of a characteristic Th1-type immune IL-12. J. Immunol. 154:5144. 26. Warren, H. S., B. F. Kinnear, R. L. Kastelein, and L. L. Lanier. 1996. Analysis response, whereas other cytokines such as IL-5 may contribute to of the costimulatory role of IL-2 and IL-15 in initiating proliferation of resting eosinophilic inflammations, as found in patients with asthma or (CD56 dim) human NK cells. J. Immunol. 156:3234. parasitic infections. 27. Naume, B., and T. Espevik. 1994. Immunoregulatory effects of cytokines on NK cells. Scand. J. Immunol. 40:128. 28. Trinchieri, G. 1995. Natural killer cells wear different hats: effector cells of innate resistance and regulatory cells of adaptive immunity and of hematopoiesis. Se- References min. Immunol. 7:83. 29. Brunda, M. J. 1994. Interleukin 12. J. Leukocyte Biol. 55:280. 1. Gleich, G. J. 1990. The eosinophil and bronchial asthma: current understanding. 30. Trinchieri, G., and F. Gerosa. 1996. Immunoregulation by IL-12. J. Leukocyte J. Allergy Clin. Immunol. 85:422. Biol. 59:505. 2. Bousquet, J., P. Chanez, J. Y. Lacoste, G. Barneon, N. Ghevanian, I. Enander, 31. Scharton, T. M., and P. Schott. 1993. Natural killer cells are a source of IFN ␥ P. Venge, S. Ahlstedt, J. Simony-Lafontaine, P. Godard, and F. Michel. 1991. that drives differentiation of CD4ϩ T cell subsets and induces early resistance to Eosinophilic inflammation in asthma. N. Engl. J. Med. 324:1033. major in mice. J. Exp. Med. 178:567. The Journal of Immunology 1969

32. Wherry, J. C., R. D. Schreiber, and E. R. Unanue. 1991. Regulation of IFN-␥ 43. Picker, L. J., M. K. Singh, Z. Zvadreski, J. R. Treer, S. L. Waldrop, production by NK cells in scid mice: roles of TNF and bacterial stimuli. Infect. P. R. Bergstresser, and V. C. Maino. 1995. Direct demonstration of cytokine Immun. 59:1709. synthesis heterogeneity among human memory/effector T cells by flow cytom- 33. Hunter, C. A., R. Chizzonite, and J. S. Remington. 1995. IL-1␤ is required for etry. Blood 86:1408. IL-12 to induce production of IFN-␥ by NK cells: a role for IL-1␤ in the T cell 44. Koo, G., F. Dumont, M. Tutt, J. Hackett, and V. Kumar. 1986. The NK1.1Ϫ independent mechanism of resistance against intracellular pathogens. J. Immunol. mouse: a model to study differentiation of murine NK cells. J. Immunol. 137: 155:4347. 3742. 34. Heremans, H., C. Dillen, J. vanDamme, and A. Billiau. 1994. Essential role for 45. Karlhofer, F. M., and W. M. Yokoyama. 1991. Stimulation of murine NK cells NK cells in the lethal LPS induced Schwarzman-like reaction in mice. Eur. J. Im- by a mAb specific for the NK1.1 antigen. J. Immunol. 146:3662. munol. 24:1155. 35. Kaneko, M., Y. Hitoshi, K. Takatsu, and S. Matsumoto. 1991. Role of interleukin 46. Wilder, J. A., C. Y. Koh, and D. Yuan. 1996. The role of NK cells during in vivo 5 in local accumulation of eosinophils in mouse allergic peritonitis. Int. Arch. antigen-specific antibody responses. J. Immunol. 156:146. Allergy Appl. Immunol. 96:41. 47. Anderson, G. P., and A. J. Coyle. 1994. TH2 and TH2-like cells in allergy and 36. Zuany-Amorim, C., D. Leduc, B. B. Vargaftig, and M. Pretolani. 1993. Charac- asthma: pharmacological perspectives. Trends Pharmacol. Sci. 15:324. terization and pharmacological modulation of antigen induced peritonitis in ac- 48. Mosman, T. R., and S. Sas. 1996. The expanding universe of T cell subsets: Th1, tively sensitized mice. Br. J. Pharmacol. 110:917. Th2 and more. Immunol. Today 17:138. 37. Zuany-Amorim, C., C. Creminon, M. C. Nevers, M. A. Nahori, B. B. Vargaftig, 49. Bradding, P., J. A. Roberts, K. M. Britten, S. Montefort, R. Djukanovic, and M. Pretolani. 1996. Modulation by IL-10 of antigen induced IL-5 generation ␣ ϩ P. H. Howarth, and S. T. Holgate. 1994. Interleukin-4, -5, -6 and TNF in normal and CD4 T lymphocyte and eosinophil infiltration into the mouse peritoneal and asthmatic airways: evidence for the human as an important source cavity. J. Immunol. 157:377. of these cytokines. Am. J. Respir. Cell Mol. Biol. 10:471. 38. Ferrick, D. A., M. D. Schrenzel, T. Mulvania, B. Hsieh, W. G. Ferlin, and 50. Brodie, D. H., M. M. Paine, and G. S. Firestein. 1992. Eosinophils express IL-5 H. Lepper. 1995. Differential production of IFN-␥ and interleukin-4 in response ␥␦ and GM-CSF mRNA at sites of allergic inflammation in asthmatics. J. Clin. to Th1 and Th2 stimulating pathogens by T cells in vivo. 373:255. Invest. 90:1414. 39. Klausner, R. D., J. G. Donaldson, and J. Lippincott-Schwartz. 1992. Brefeldin A: insights into the control of membrane traffic and organelle structure. J. Cell Biol. 51. Corrigan, C. J., A. Haczku, V. Gemou-Engesaeth, S. Doi, Y. Kikuchi, K. Takatsu, 116:1071. S. R. Durham, and A. B. Kay. 1993. CD4 T-lymphoycte activation in asthma is Downloaded from 40. Yoshimoto, T., and W. E. Paul. 1994. CD4 positive NK1.1 positive T cells accompanied by increased serum concentrations of IL-5: effect of promptly produce interleukin 4 in response to in vivo challenge with anti-CD3. therapy. Am. Rev. Respir. Dis. 147:540. J. Exp. Med. 179:1285. 52. Plaut, M., J. H. Pierce, C. J. Watson, J. Hanley-Hyden, R. P. Nordan, and 41. Vicai, A. P., and A. Zlotnik. 1996. Mouse NK1.1.ϩ T cells: a new family of T W. E. Paul. 1989. Mast cell lines produce lymphokines in response to cross- cells. Immunol. Today 17:71. linkage of Fc⑀RI or to calcium ionophore. Nature 339:64. 42. Openshaw, P., E. E. Murphy, N. A. Hosken, V. Maino, K. Davis, K. Murphy, and 53. Brusselle, G. G., J. C. Kips, J. H. Tavernier, J. G. Van der Heyden, A. O’Garra. 1995. Heterogeneity of intracellular cytokine synthesis at the single C. A. Cuvelier, R. A. Pauwels, and H. Bluetham. 1994. Attenuation of allergic cell level in polarized T helper 1 and T helper 2 populations. J. Exp. Med. 182:1. inflammation in IL-4 deficient mice. Clin. Exp. Allergy 24:73. http://www.jimmunol.org/ by guest on October 2, 2021