Effective antigen presentation to helper T cells by human eosinophils (original) (raw)

• Journal List
• Immunology
• v.149(4); 2016 Dec
• PMC5095493

Immunology. 2016 Dec; 149(4): 413–422.

Ruhaifah K. Farhan

1Immunity, Infection and Inflammation Programme, Institute of Medical Sciences, School of Medicine, Medical Sciences and Nutrition, University of Aberdeen, Aberdeen, UK

Mark A. Vickers

1Immunity, Infection and Inflammation Programme, Institute of Medical Sciences, School of Medicine, Medical Sciences and Nutrition, University of Aberdeen, Aberdeen, UK

Amir M. Ghaemmaghami

2Cellular Immunology & Allergy Group, Division of Immunology, School of Life Sciences, Faculty of Medicine and Health Sciences, Queen's Medical Centre, Nottingham, UK

Andrew M. Hall

1Immunity, Infection and Inflammation Programme, Institute of Medical Sciences, School of Medicine, Medical Sciences and Nutrition, University of Aberdeen, Aberdeen, UK

Robert N. Barker

1Immunity, Infection and Inflammation Programme, Institute of Medical Sciences, School of Medicine, Medical Sciences and Nutrition, University of Aberdeen, Aberdeen, UK

Garry M. Walsh

1Immunity, Infection and Inflammation Programme, Institute of Medical Sciences, School of Medicine, Medical Sciences and Nutrition, University of Aberdeen, Aberdeen, UK

1Immunity, Infection and Inflammation Programme, Institute of Medical Sciences, School of Medicine, Medical Sciences and Nutrition, University of Aberdeen, Aberdeen, UK

2Cellular Immunology & Allergy Group, Division of Immunology, School of Life Sciences, Faculty of Medicine and Health Sciences, Queen's Medical Centre, Nottingham, UK

Corresponding author.

*Correspondence: Dr Andrew M. Hall, Immunity, Infection and Inflammation Programme, School of Medicine & Dentistry, Institute of Medical Sciences, University of Aberdeen, Foresterhill, Aberdeen AB25 2ZD, UK. Email: ku.ca.ndba@llah.m.a,
Senior authors: Andrew M. Hall, Robert N. Barker, Garry M. Walsh
Emails: ku.ca.ndba@rekrab.n.r (RNB), ku.ca.ndba@hslaw.m.g (GMW)

Received 2016 Apr 18; Revised 2016 Jul 29; Accepted 2016 Aug 1.

Supplementary Materials

Figure S1. Purity of eosinophil isolation.

Figure S2. Control positive staining of B cells for MHC class II and co‐stimulatory molecules.

Figure S3. Blockade of HLA‐DR/DP/DQ inhibits house dust mite‐induced T helper cell proliferation.

GUID: 879C00C8-38BD-43FC-BF99-3256503D28E7

Summary

Although eosinophils are inflammatory cells, there is increasing attention on their immunomodulatory roles. For example, murine eosinophils can present antigen to CD4+ T helper (Th) cells, but it remains unclear whether human eosinophils also have this ability. This study determined whether human eosinophils present a range of antigens, including allergens, to activate Th cells, and characterized their expression of MHC class II and co‐stimulatory molecules required for effective presentation. Human peripheral blood eosinophils purified from non‐allergic donors were pulsed with the antigens house dust mite extract (HDM), Timothy Grass extract (TG) or Mycobacterium tuberculosis purified protein derivative (PPD), before co‐culture with autologous CD4+ Th cells. Proliferative and cytokine responses were measured, with eosinophil expression of HLA‐DR/DP/DQ and the co‐stimulatory molecules CD40, CD80 and CD86 determined by flow cytometry. Eosinophils pulsed with HDM, TG or PPD drove Th cell proliferation, with the response strength dependent on antigen concentration. The cytokine responses varied with donor and antigen, and were not biased towards any particular Th subset, often including combinations of pro‐ and anti‐inflammatory cytokines. Eosinophils up‐regulated surface expression of HLA‐DR/DP/DQ, CD80, CD86 and CD40 in culture, increases that were sustained over 5 days when incubated with antigens, including HDM, or the major allergens it contains, Der p I or Der p II. Human eosinophils can, therefore, act as effective antigen‐presenting cells to stimulate varied Th cell responses against a panel of antigens including HDM, TG or PPD, an ability that may help to determine the development of allergic disease.

Keywords: allergen, antigen presentation, co‐stimulatory molecule, eosinophil, major histocompatibility complex class II

Abbreviations

APC

antigen‐presenting cell

GM‐CSF

granulocyte–macrophage colony‐stimulating factor

HDM

house dust mite extract

IL‐3

interleukin‐3

IFN‐γ

interferon‐γ

PPD

purified protein derivative

rh

recombinant human

TG

Timothy grass extract

Th

helper T

TNF‐α

tumour necrosis factor‐α

Introduction

Eosinophil involvement in inflammatory conditions affecting the skin, gastrointestinal tract and upper and lower airways is well‐documented.1, 2 In addition to their role as degranulating effector cells, more recent findings emphasize the immunomodulatory properties of eosinophils,3 and other important effector functions such as a potential role in maintaining host survival in life‐threatening respiratory viral infections.4 One question that has attracted interest is whether eosinophils can modulate immune responses by acting as antigen‐presenting cells (APC) to stimulate CD4+ helper T (Th) cell responses. It has been known for many years that in vitro culture of eosinophils with granulocyte–macrophage colony‐stimulating factor (GM‐CSF), typically added to prevent their apoptosis, can also induce expression of MHC Class II,5 which could equip them for antigen presentation.

Consistent with APC function, murine eosinophils home to lymphoid tissue and provide a second signal for T‐cell activation through the expression of key co‐stimulatory molecules such as CD80 and CD86.6, 7 Although it has been demonstrated that human eosinophils can express CD86 when taken from hyper‐eosinophilic patients,8 or stimulated with interleukin‐3 (IL‐3),9 it is unclear how commonly, or in what circumstances, they display such co‐stimulatory molecules. There are also reports that human eosinophils can process and present antigen to activate specific T cells,10 but, again, it remains to be established how widespread is such an ability in different individuals and for different antigens. Despite the evidence that eosinophils have the potential to act as APC to drive Th cell responses and thereby propagate inflammation,11, 12, 13 other findings have suggested that this is limited to super‐antigens and peptides rather than proteins that require processing.6, 14

The effects of Th activation are critically dependent on the subset(s) that respond, and the associated cytokines they produce. In type I hypersensitivity responses, Th2 cells produce cytokines, such as IL‐5 and IL‐13,15 that promote eosinophil activation and proliferation,16 but in healthy donors the response to common allergens such as Timothy grass17 or house dust mite18 is associated with Th cells producing both Th1 and Th2 cytokines. However it is not known whether eosinophil antigen presentation preferentially supports responses by any particular Th type.

To address the unanswered questions about their roles as APC, we performed a comprehensive examination of the ability of purified peripheral blood human eosinophils to present a variety of protein antigens to autologous CD4+ Th cells. The focus here was on donors with no history of allergy, as we wished to establish the ability of eosinophils to contribute to helper activation in the absence of any pre‐existing strong pathogenic Th2 response. This work is possible because healthy donors harbour allergen‐specific Th cells.19, 20 Helper responsiveness was tested for eosinophils pulsed with the allergens house dust mite extract (HDM), Timothy Grass extract (TG), Der p I, Der p II, or the microbial recall antigen Mycobacterium tuberculosis purified protein derivative (PPD), with eosinophil expression of MHC Class II molecules and the co‐stimulatory molecules CD40, CD80 and CD86 characterized. It was also determined whether any Th cytokines elicited by eosinophil antigen presentation exhibited a bias towards particular effector or regulatory subsets.

Materials and methods

Materials

CD‐16 immunomagnetic beads, the Human CD4+ T Cell Isolation Kit II and magnetically activated separation columns were from Miltenyi Biotec (Surrey, UK). HDM (Dermatophagoides pteronyssinus) and TG extracts, (both certified lipopolysaccharide‐free and obtained from NIBSC, UK) were dialysed using slide‐A‐Lyzer dialysis cassettes (Thermo Scientific, Paisley, UK) for 24 hr and used at a final concentrations of 500—2500 IU/ml. PPD (Statens Serum Institute, Copenhagen, Denmark) was added to cell cultures at a final concentration of 5 µg/ml. Dermatophagoides pteronyssinus allergens, Der p I or Der p II (Indoor Biotechnologies Ltd, Cardiff, UK) were used at final concentrations 10 µg/ml. All cells were cultured in RPMI‐1640 (Labtech International Ltd, Uckfield, UK) supplemented with HEPES and 100 IU/ml penicillin, 100 µg/ml streptomycin, l‐ glutamine 5% (v/v) (Gibco, Life Technologies, Paisley, UK) and 5% (v/v) heat‐inactivated autologous serum, with 10 nm recombinant human (rh) GM‐CSF (R&D Systems, Abingdon, UK). The latter was essential to prevent eosinophil apoptosis during co‐culture. The following monoclonal antibodies were used in these studies: HLA‐DR/DP/DQ (clone TL2.1), CD1a (clone H1149), CD40 (clone5 C3), CD80 (L307·4) and CD86 (clone 2331) were all from BD Pharmingen (Oxford, UK). Siglec‐8 monoclonal antibody (clone 7CP) was from Biolegend (London, UK). Cytokine levels were measured by Multiplex array for interferon‐γ (IFN‐γ), IL‐3, IL‐6, IL‐10, IL‐17A and tumour necrosis factor‐α (TNF‐α; Luminex, Millipore, Watford, UK) and by ELISA for IL‐9 and IL‐13 (Biolegend).

Eosinophil and Th cell Isolation

Human eosinophils and CD4+ Th cells were purified from individuals with no clinical history of allergy or eosinophilia (≤ 0·5 × 106 eosinophils/ml), and who were not taking any medication for allergic disease. The inclusion of patients with haemochromatosis, who were being routinely bled to treat the disease, allowed relatively large numbers of eosinophils for some experiments to be collected from whole units of blood in the face of normal eosinophil counts. All subjects gave informed consent and the study was approved by the North of Scotland Research Ethics Service (ref 09/S0801/16). Eosinophils were purified from samples of peripheral blood using our standard technique13 using dextran sedimentation and centrifugation on Percoll gradients followed by CD16‐dependent negative immunomagnetic selection. To obtain CD4+ Th cells, peripheral blood mononuclear cells were isolated by density gradient centrifugation21 and non‐target cells were depleted by negative selection following the manufacturer's instructions.

Using these methods, eosinophils and Th cells with respective purities of at least 98% were obtained, as expected from previous reports.22 In particular, the eosinophil preparations were free from any detectable cells expressing the CD1a, CD14 or CD19 markers for professional APC types (see Supplementary material, Fig. S1). All cell preparations exhibited > 98% viability as assessed by trypan blue exclusion.

Cell cultures

Eosinophils (5 × 105/ml) were incubated in culture medium for up to 5 days in the presence of rhGM‐CSF (10 nm) to inhibit their spontaneous apoptosis as previously described,23 with or without the addition of antigens. When co‐cultured with autologous CD4+ Th cells, eosinophils were first pulsed with antigens by overnight incubation, then washed and added at 5 × 105/ml in medium containing GM‐CSF to the Th cells (1 × 106/ml) for 5 days; these conditions were found to give optimal Th responses in pilot experiments (data not shown). Cultures of autologous peripheral blood mononuclear cells (1·25 × 106/ml), with or without antigen, provided controls for comparison of Th responsiveness.

Th cell responses

Proliferative Th cell responses were determined by incorporation of [3H]thymidine in triplicate 100‐μl volumes withdrawn from cultures 5 days after stimulation as previously described, with results presented as counts per minute.24 Cytokine levels in cultures were measured by bead array for IFN‐γ, IL‐3, IL‐6, IL‐10, IL‐17A and TNF‐α and by ELISA for IL‐9 and IL‐13, according to the manufacturers' instructions. Cytokine responses > 2× background in unstimulated wells were considered significant.21 To test the dependency of responses on MHC class II,25 antigen‐pulsed eosinophils were incubated with blocking antibody before co‐culture with autologous CD4+ Th cells.

Immunostaining and flow cytometry

Cell surface expression markers were examined using flow cytometry for HLA‐DR,‐DP,‐DQ (FITC), and co‐stimulatory molecules, CD40 (BD Horizon V450), CD80 (Alexa Fluor 700), and CD86 (allophycocyanin) using established protocols.13 Briefly, eosinophils and CD4+ Th cells were removed from co‐culture, washed and saturating quantities of primary antibodies or specific isotype controls were added to the cells and incubated for 40 min at 4° in the dark, washed and fixed. Human dendritic cells were identified by staining cells with CD1a (allophycocyanin) while eosinophils were identified by staining with a specific marker, siglec‐8 (phycoerythrin).26 An EBV‐transformed B cell line was used as a source of cells that stained positively for HLA‐DR/DP/DQ, CD40, CD80 or CD86 (see Supplementary material, Fig. S2). Multiple panels of conjugated antibodies were used to identify the subpopulation of immune cells and the corresponding specific cell surface markers. Ten thousand events were collected on a flow cytometer, LSRII (BD Biosciences, Oxford, UK) using FACS diva software (BD Biosciences, Oxford, UK). Analysis was performed using flowjo software (Treestar Inc., Ashland, OR). In all plots, dead cells, cellular debris and cell aggregates were excluded during the gating process.

Statistics

Data were analysed using graphpad prism 6 (GraphPad Software, La Jolla, CA), and because D'Agostino–Person Omnibus and Shapiro–Wilk normality tests were failed, non‐parametric methods were used (Wilcoxon‐signed‐rank‐test or Mann–Whitney U test as appropriate). Bonferroni corrections were applied to multiple comparisons. Data were considered to be statically significant if P < 0·05, and are expressed as either median and interquartile range (IQR), or individual data points if n < 4.

Results

Eosinophils pulsed with antigen induce proliferative Th responses

We first examined the ability of human eosinophils pulsed with a range of concentrations of the allergens HDM or TG, or with our standard concentration of the control microbial recall antigen PPD, to induce proliferation by autologous CD4+ Th cells after 5 days of co‐culture. In a series of nine experiments, significant (P < 0·001) increases in proliferation were observed when eosinophils had been pulsed by pre‐incubation with the allergens HDM (Fig. ​1a) or TG (Fig. ​1b) at final concentrations of 1000, 1500 or 2500 IU/ml, compared with medium alone. The strongest responses were induced by HDM or TG at 2500 IU, and this concentration of the allergens was therefore used in all subsequent experiments. Eosinophils pulsed with the control antigen PPD also elicited significant proliferation at the standard concentration of 5 μg/ml (Fig. ​1c). To confirm that responses to each stimulus required the presence of both eosinophils and Th cells, proliferation was compared in cultures containing each cell type alone or together, with or without antigen pulsing (Fig. ​2). Proliferative responses were significant only when both antigen‐pulsed eosinophils and Th cells were added, and, strikingly, these responses were similar in size to those seen when unfractionated peripheral blood mononuclear cells were stimulated with the respective antigen as a positive control. It should be noted that the ability of the eosinophil preparations to drive such strong Th cell responses cannot be explained by the presence of contaminating APC, as none could be detected by flow cytometry (see Supplementary material, Fig. S1). Control analyses also confirmed that proliferative responses in antigen‐stimulated co‐cultures were mediated by Th cells, but not eosinophils (median CD4+ Th cell count increased by 30% in HDM‐stimulated co‐cultures after 5 days, P = 0·016, Wilcoxon signed rank test, n = 8, versus no increase in median eosinophil numbers). The ability of Th cells to respond to antigen‐pulsed eosinophils was MHC class II dependent, as incubating HDM‐pulsed eosinophils with blocking mAb specific for HLA‐DR/DP/DQ25 significantly (P < 0·001) reduced T‐cell proliferation by 77% (see Supplementary material, Fig. S3).

Eosinophils pulsed with allergen stimulate T helper (Th) cell proliferative responses. Panels show proliferation in co‐cultures of peripheral blood CD4+ Th cells and autologous eosinophils that have been pulsed with the antigens house dust mite (HDM) (a), Timothy grass extract (TG) (b), or purified protein derivative (PPD (c). Results are expressed as median counts per minute (CPM) and interquartile range (IQR) from nine independent experiments (*P < 0·01, Mann–Whitney _U_‐test with Bonferroni correction).

Eosinophils pulsed with antigen stimulate T helper (Th) proliferation comparable to peripheral blood mononuclear cells (PBMC) responses. Panels show proliferation in cultures containing peripheral blood CD4+ Th cells and autologous eosinophils, either alone or together, with or without pulsing of the eosinophils with the antigens house dust mite (HDM) (a), Timothy grass extract (TG) (b) or purified protein derivative (PPD) (c). Proliferation in peripheral blood mononuclear cells (PBMC) cultures, either untreated or antigen stimulated, is included for comparison. Eo = purified eosinophils, T = purified CD4+ T cells (*P < 0·05, Mann–Whitney _U_‐test with Bonferroni correction).

Eosinophil expression of MHC class II and co‐stimulatory molecules

Effective APC require expression of both MHC class II and co‐stimulatory molecules, so we next tested whether cultured eosinophils display HLA‐DR/DP/DQ and CD40, CD80 and CD86. Examples of flow cytometric analyses, and graphical summaries of data from nine independent experiments, demonstrate eosinophil expression of HLA‐DR/DP/DQ (Fig. ​3) and CD40, CD80 and CD86 (Fig. ​4) during 5‐day cultures, with or without HDM addition. It can be seen that, within 24 hr, eosinophil expression of MHC class II, and the co‐stimulatory molecules CD40, CD80 and CD86, was significantly increased in all cultures, an effect that may be at least partly due to the presence of GM‐CSF, which was added to the medium to prevent eosinophil apoptosis.5 However, there was a further effect of HDM. Addition of the antigen sustained the elevated levels of MHC class II, CD40, CD80 and CD86 over the course of the incubation, because, without HDM, expression of all the markers fell back by day 5 to levels not significantly above those seen at the beginning of the culture.

Cultured eosinophils express MHC class II. Representative flow cytometric histograms (n = 9) demonstate HLA DR/DP/DQ expression by purified eosinophils incubated with house dust mite (HDM) at day 0 (a) and day 5 (b) of culture (solid line = stained cells, dotted line = unstained cells, dashed line = isotype control). The gate indicates the percentage of eosinophils staining positively for HLA‐DR/DP/DQ. HLA‐DR/DP/DQ expression on unstimulated (c) or HDM‐stimulated (d) eosinophils over 5 days of culture is summarized in bar charts, with results expressed as median and interquartile range (IQR) of % eosinophils positive for HLA DR\DP\DQ staining from nine independent experiments (*P < 0·01, Mann–Whitney _U_‐test with Bonferroni correction).

Cultured eosinophils express co‐stimulatory molecules. Representative flow cytometric histograms (n = 9) demonstate CD40 (a), CD80 (b) and CD86 (c) expression by purified eosinophils incubated with house dust mite (HDM) at day 0 (left panels) and day 5 (right panels) of culture (solid line = stained cells, dotted line = unstained cells, dashed line = isotype control). The gate indicates the percentage of eosinophils staining positively for each co‐stimulatory molecule. CD40 (d), CD80 (e) and CD86 (f) expression by unstimulated (left panels) or HDM‐stimulated (right panels) eosinophils over 5 days of culture is summarized in the bar charts, with results expressed as median and interquartile range (IQR) of % eosinophils positive for each marker from nine independent experiments (*P < 0·01, Mann–Whitney _U_‐test with Bonferroni correction).

The HDM preparation is a simple extract, so we next tested whether the major allergens it contains, Der p I and Der p II, recapitulate its ability to sustain enhanced expression of MHC class II and co‐stimulatory molecules by eosinophils. Eosinophils incubated with purified Der p I or Der p II exhibited similar sustained increases in expression of HLA‐DR\DP\DQ, CD40, CD80 and CD86 compared with those elicited by the crude HDM preparation (Fig. ​5).

Der p I and Der p II antigens share the ability of house dust mite (HDM) to sustain eosinophil surface expression of MHC class II and co‐stimulatory molecules. Comparison of the effects of incubation with the purified allergens Der p I or Der p II, or the allergen extract HDM, on numbers of eosinophils that express HLA‐DR\DP\DQ (a), CD40 (b), CD80 (c) and CD86 (d) after 5 days of culture. Results are expressed as median and interquartile range (IQR) of % eosinophils positive for each marker from six independent experiments (*P < 0·01, Mann–Whitney _U_‐test with Bonferroni correction).

Cytokine responses induced by eosinophils pulsed with antigen

Having demonstrated the ability of eosinophils to present antigen to drive Th proliferative responses, the question arises as to whether pro‐inflammatory or anti‐inflammatory cytokines, or cytokines associated with any particular CD4+ subset are produced. Signature cytokines for the major subsets Th1 (IFN‐γ), Th2 (IL‐13), Th9 (IL‐9), regulatory T cell (IL‐10), Th17 (IL‐17A) and the inflammatory cytokines TNF‐α, IL‐3 and IL‐6, were measured in co‐cultures of CD4+ Th cells and eosinophils with, or without, pulsing with the antigens HDM, TG or PPD. Different patterns of cytokine response to the antigens were seen in cultures from each of the donors tested (n = 6), with examples illustrated in Fig. ​6, and all results summarized in Table 1. It can be seen that antigen‐pulsed eosinophils are capable of eliciting a wide range of cytokines tested. Although IL‐6 production was the most frequently seen response, cultures of antigen‐pulsed eosinophils could also contain complex mixtures of both pro‐inflammatory and anti‐inflammatory cytokines, and, overall, there was no clear or consistent bias towards any particular cytokine type.

Production of multiple cytokines in co‐cultures of CD4+ T helper (Th) cells and antigen‐pulsed eosinophils. Examples are shown (donors 1 and 5, n = 6) of different patterns of cytokine secretion by co‐cultures of CD4+ Th cells and eosinophils pulsed with the antigens house dust mite (HDM), Timothy grass extract (TG) or purified protein derivative (PPD).

Table 1

Cytokine production by co‐cultures of T helper (Th) cells and eosinophils presenting antigen

Discussion

The present study has established the ability of human eosinophils to present a wide variety of protein antigens, including allergens, to stimulate proliferative and cytokine responses by CD4+ Th cells. In line with their ability to elicit responses to antigen, eosinophils exhibited up‐regulated MHC class II and co‐stimulatory molecules in culture. Since the donors tested here had no clinical history of allergy, the results raise the possibility that eosinophils acting as APC can help to determine whether Th responses to allergen are elicited or become pathogenic.

There is now a substantial body of evidence demonstrating that murine eosinophils presenting antigen have the ability to stimulate Th cells, including the induction of primary responses, and that they act as true professional APC in homing to lymph nodes, where they can interact with Th cells.12 We show here, for the first time, that human eosinophils are also effective APC for a variety of protein antigens in vitro, able to evoke significant Th proliferative responses comparable in magnitude to those seen when paired peripheral blood mononuclear cells samples are challenged with the same antigen. Hence, human eosinophil presentation to Th cells is limited neither to peptide antigens nor to super‐antigens that do not require processing, as has been suggested,6, 14 but may instead contribute to responses against many different proteins. The question arises as to why previous studies have not clearly demonstrated such APC activity by human eosinophils, particularly when this ability would be predicted from murine studies.6, 7 Although it is not possible to pinpoint with any confidence which factor, or combination of factors, accounts for the great sensitivity of our study versus those carried out several years ago in other laboratories, we can offer a number of suggestions. Our cultures are optimized to support proliferative responses by antigen‐specific T cells present at low precursor frequency, with relatively long time courses allowed for responses to develop, and supplementation with autologous serum to enhance sensitivity, rather than the fetal calf serum used in other studies. We also exploited polyclonal T cells as detectors of antigen presentation by eosinophils, rather than rely on T‐cell clones raised against other APC types. Whatever the explanation, whether the ability we describe of human eosinophils to act as APC in vitro is replicated in vivo, or in patients with hypersensitivity, now needs to be established. It is noteworthy that up‐regulation of eosinophil MHC class II expression observed in asthma,27 chronic eosinophilic pneumonia,28 and eosinophilic oesophagitis29 is consistent with such an immune modulatory role.10, 12

Function of APC is dependent on display of MHC class II and co‐stimulatory molecules.30 We not only confirm that cultured human eosinophils express HLA‐DR/DP/DQ, but also show expression of all three major co‐stimulatory molecules, CD40, CD80 and CD86 by these cells. Eosinophils in culture require the addition of GM‐CSF to prevent apoptosis, and the presence of this cytokine is likely to have contributed to the APC phenotype seen here, as eosinophils purified from the spleens of IL‐5 transgenic mice were also observed to express MHC Class II, CD40, CD80 and CD86 when stimulated with GM‐CSF.10 Previous studies of human eosinophils have described up‐regulation of MHC class II and CD86 in response to cytokine or super‐antigen exposure,9, 28, 31 but, to our knowledge, we are the first to report expression of such a complete APC surface phenotype in cultured human eosinophils. Although it could be argued that the precise conditions in vitro do not reflect those in vivo, the results nevertheless establish that eosinophils have the potential to present antigen very effectively, and the induction of APC function by stimuli such as GM‐CSF in vivo may well represent an important mechanism by which eosinophils influence immune responses to allergens. In addition, GM‐CSF was not the only factor up‐regulating APC surface markers, as we demonstrated that the increases in the expression of HLA‐DR/DP/DQ, CD40, CD80 and CD86 were sustained for up to 5 days of co‐culture by eosinophils stimulated with whole HDM extract or the major HDM allergens Der p I and Der p II. A number of studies have identified similar effects of HDM extract, Der p I or Der p II on other cell types. For example, Der p I stimulation of monocyte‐derived dendritic cells isolated from donors allergic to HDM increased CD86 expression, whereas control non‐allergic subjects had significant increases in CD80 expression,32 and another study showed that Der p I stimulated human peripheral blood dendritic cells to increase expression of HLA‐DR, CD80 and CD86.33 The underlying mechanisms remain to be established, but may include proteolytic activity of the allergen,34 or its interaction with pattern recognition receptors. Der p II in particular has auto‐adjuvant properties due to structural and functional homology with MD‐2, the lipopolysaccharide‐binding component of the Toll‐like receptor 4 signalling complex.35

The cytokines elicited during responses to antigen play a key role in determining both protection from infection and immune pathology. Here, eosinophils acting as APC supported production of a wide range of cytokines that differed between individuals and antigens, but with no clear preference for any Th response type. However, our study was of cells from donors with no clinical allergic disease, and so the possibility remains open that eosinophils may skew helper responses towards the pathogenic Th2 subset in patients with overt allergy, or a susceptibility to atopic disease. The notion that eosinophils can drive a variety of Th subsets is supported by a previous study, which also tested co‐cultures of human peripheral blood eosinophils and autologous CD4+ Th cells, and demonstrated both Th1 and Th2 cytokine responses to the super‐antigen staphylococcal enterotoxin B. These workers also demonstrated HLA‐DR expression by peripheral blood eosinophils isolated from 50% of the subjects, attributed to GM‐CSF added to cultures, but stimulation with staphylococcal enterotoxin B did not induce eosinophil expression of CD80 or CD86.36 The reasons for the differences between this result and the present study may well reflect the use of antigens versus polyclonal activator for stimulation. Others have described the effect of stimulation with HDM antigen on eosinophil function. For example, HDM stimulation of eosinophils in vitro led to the production of IL‐9 that may promote a Th2 immune response,37 but, although we detect this cytokine in some co‐cultures of Th cells and antigen‐pulsed eosinophils, HDM did not elicit the response more frequently than other antigens, and IL‐9 was not associated with any clear Th subset bias.

Taken together, our data demonstrate that human eosinophils can act as effective APC to stimulate Th responses against a variety of antigens, including the allergens HDM or TG: a property that may contribute to the regulation of responses in vivo and to induction or control of pathology in allergic disease, depending on the cytokines elicited. The findings add to the accumulating evidence that eosinophils possess more complex immunomodulatory roles in allergic disease than previously suspected. Furthermore, any ability of HDM, Der p I or Der p II to act not only as antigens, but also to increase eosinophil co‐stimulatory molecule expression, may enhance their immunogenicity. Having demonstrated the ability of human eosinophils to present antigen, this study opens up new questions as to how important they are in initiating, skewing, amplifying or regulating allergic responses in patients.

Disclosures

The authors declare no conflict of interest.

Supporting information

Figure S1. Purity of eosinophil isolation.

Figure S2. Control positive staining of B cells for MHC class II and co‐stimulatory molecules.

Figure S3. Blockade of HLA‐DR/DP/DQ inhibits house dust mite‐induced T helper cell proliferation.

Acknowledgements

This study was funded by grants from The Embassy of the State of Kuwait. The authors wish to thank Alison Scaife for invaluble technical assistance and we are also grateful for the willingness of those who volunteered to give blood for this study.

Notes

Joint authorship: AMH, RNB and GMW contributed equally to this article.

References

2. Lacy P, Rosenberg HF, Walsh GM. Eosinophil overview: structure, biological properties, and key functions. Methods Mol Biol2014; 1178:1–12. [PubMed] [Google Scholar]

3. Rosenberg HF, Dyer KD, Foster PS. Eosinophils: changing perspectives in health and disease. Nat Rev Immunol2013; 13:9–22. [PMC free article] [PubMed] [Google Scholar]

4. Percopo CM, Dyer KD, Ochkur SI, Luo JL, Fischer ER, Lee JJ_et al_ Activated mouse eosinophils protect against lethal respiratory virus infection. Blood2014; 123:743–52. [PMC free article] [PubMed] [Google Scholar]

5. Lucey DR, Nicholson‐Weller A, Weller PF. Mature human eosinophils have the capacity to express HLA‐DR. Proc Natl Acad Sci U S A1989; 86:1348–51. [PMC free article] [PubMed] [Google Scholar]

6. van Rijt LS, Vos N, Hijdra D, de Vries VC, Hoogsteden HC, Lambrecht BN. Airway eosinophils accumulate in the mediastinal lymph nodes but lack antigen‐presenting potential for naive T cells. J Immunol2003; 171:3372–8. [PubMed] [Google Scholar]

7. Shi HZ, Humbles A, Gerard C, Jin Z, Weller PF. Lymph node trafficking and antigen presentation by endobronchial eosinophils. J Clin Invest2000; 105:945–53. [PMC free article] [PubMed] [Google Scholar]

8. Woerly G, Roger N, Loiseau S, Dombrowicz D, Capron A, Capron M. Expression of CD28 and CD86 by human eosinophils and role in the secretion of type 1 cytokines (interleukin 2 and interferon γ): inhibition by immunoglobulin a complexes. J Exp Med1999; 190:487–95. [PMC free article] [PubMed] [Google Scholar]

9. Celestin J, Rotschke O, Falk K, Ramesh N, Jabara H, Strominger J_et al_ IL‐3 induces B7.2 (CD86) expression and costimulatory activity in human eosinophils. J Immunol2001; 167:6097–104. [PubMed] [Google Scholar]

10. Wang HB, Ghiran I, Matthaei K, Weller PF. Airway eosinophils: allergic inflammation recruited professional antigen‐presenting cells. J Immunol2007; 179:7585–92. [PMC free article] [PubMed] [Google Scholar]

11. Hansel TT, De Vries IJ, Carballido JM, Braun RK, Carballido‐Perrig N, Rihs S_et al_ Induction and function of eosinophil intercellular adhesion molecule‐1 and HLA‐DR. J Immunol1992; 149:2130–6. [PubMed] [Google Scholar]

12. Akuthota P, Wang H, Weller PF. Eosinophils as antigen‐presenting cells in allergic upper airway disease. Curr Opin Allergy Clin Immunol2010; 10:14–9. [PMC free article] [PubMed] [Google Scholar]

13. Robinson AJ, Kashanin D, O'Dowd F, Fitzgerald K, Williams V, Walsh GM. Fluvastatin and lovastatin inhibit granulocyte macrophage‐colony stimulating factor‐stimulated human eosinophil adhesion to inter‐cellular adhesion molecule‐1 under flow conditions. Clin Exp Allergy2009; 39:1866–74. [PubMed] [Google Scholar]

14. Mawhorter SD, Pearlman E, Kazura JW, Boom WH. Class II major histocompatibility complex molecule expression on murine eosinophils activated in vivo by Brugia malayi . Infect Immun1993; 61:5410–2. [PMC free article] [PubMed] [Google Scholar]

15. Wynn TA. Type 2 cytokines: mechanisms and therapeutic strategies. Nat Rev Immunol2015; 15:271–82. [PubMed] [Google Scholar]

16. Mazzarella G, Bianco A, Catena E, De Palma R, Abbate GF. Th1/Th2 lymphocyte polarization in asthma. Allergy2000; 55(Suppl. 61):6–9. [PubMed] [Google Scholar]

17. Hinz D, Seumois G, Gholami AM, Greenbaum JA, Lane J, White B_et al_ Lack of allergy to timothy grass pollen is not a passive phenomenon but associated with allergen‐specific modulation of immune reactivity. Clin Exp Allergy2015; 46:705–19. [PMC free article] [PubMed] [Google Scholar]

18. Akdis M, Verhagen J, Taylor A, Karamloo F, Karagiannidis C, Crameri R_et al_ Immune responses in healthy and allergic individuals are characterized by a fine balance between allergen‐specific T regulatory 1 and T helper 2 cells. J Exp Med2004; 199:1567–75. [PMC free article] [PubMed] [Google Scholar]

19. Wierenga EA, Snoek M, Bos JD, Jansen HM, Kapsenberg ML. Comparison of diversity and function of house dust mite‐specific T lymphocyte clones from atopic and non‐atopic donors. Eur J Immunol1990; 20:1519–26. [PubMed] [Google Scholar]

20. Till S, Dickason R, Huston D, O'Brien F, Lamb J, Corrigan C_et al_ IL‐5 secretion by allergen‐stimulated CD4+ T cells in primary culture: relationship to expression of allergic disease. J Allergy Clin Immunol1997; 99:563–9. [PubMed] [Google Scholar]

21. Devereux G, Hall AM, Barker RN. Measurement of T‐helper cytokines secreted by cord blood mononuclear cells in response to allergens. J Immunol Methods2000; 234:13–22. [PubMed] [Google Scholar]

22. Akuthota P, Shamri R, Weller PF. Isolation of human eosinophils. Curr Protoc Immunol2012; 7:31. [PMC free article] [PubMed] [Google Scholar]

23. Blaylock MG, Sexton DW, Walsh GM. Ligation of CD45 and the isoforms CD45RA and CD45RB accelerates the rate of constitutive apoptosis in human eosinophils. J Allergy Clin Immunol1999; 104:1244–50. [PubMed] [Google Scholar]

24. Cairns LS, Phelps RG, Bowie L, Hall AM, Saweirs WW, Rees AJ_et al_ The fine specificity and cytokine profile of T‐helper cells responsive to the _α_3 chain of type IV collagen in Goodpasture's disease. J Am Soc Nephrol2003; 14:2801–12. [PubMed] [Google Scholar]

25. Sukati H, Watson HG, Urbaniak SJ, Barker RN. Mapping helper T‐cell epitopes on platelet membrane glycoprotein IIIa in chronic autoimmune thrombocytopenic purpura. Blood2007; 109:4528–38. [PubMed] [Google Scholar]

26. Floyd H, Ni J, Cornish AL, Zeng Z, Liu D, Carter KC_et al_ Siglec‐8. A novel eosinophil‐specific member of the immunoglobulin superfamily. J Biol Chem2000; 275:861–6. [PubMed] [Google Scholar]

27. Hansel TT, Braunstein JB, Walker C, Blaser K, Bruijnzeel PL, Virchow JC Jr_et al_ Sputum eosinophils from asthmatics express ICAM‐1 and HLA‐DR. Clin Exp Immunol1991; 86:271–7. [PMC free article] [PubMed] [Google Scholar]

28. Beninati W, Derdak S, Dixon PF, Grider DJ, Strollo DC, Hensley RE_et al_ Pulmonary eosinophils express HLA‐DR in chronic eosinophilic pneumonia. J Allergy Clin Immunol1993; 92:442–9. [PubMed] [Google Scholar]

29. Patel AJ, Fuentebella J, Gernez Y, Nguyen T, Bass D, Berquist W_et al_ Increased HLA‐DR expression on tissue eosinophils in eosinophilic esophagitis. J Pediatr Gastroenterol Nutr2010; 51:290–4. [PubMed] [Google Scholar]

30. Unanue ER. Perspective on antigen processing and presentation. Immunol Rev2002; 185:86–102. [PubMed] [Google Scholar]

31. Mawhorter SD, Kazura JW, Boom WH. Human eosinophils as antigen‐presenting cells: relative efficiency for superantigen‐ and antigen‐induced CD4+ T‐cell proliferation. Immunology1994; 81:584–91. [PMC free article] [PubMed] [Google Scholar]

32. Hammad H, Charbonnier AS, Duez C, Jacquet A, Stewart GA, Tonnel AB_et al_ Th2 polarization by Der p I‐pulsed monocyte‐derived dendritic cells is due to the allergic status of the donors. Blood2001; 98:1135–41. [PubMed] [Google Scholar]

33. Charbonnier AS, Hammad H, Gosset P, Stewart GA, Alkan S, Tonnel AB_et al_ Der p I‐pulsed myeloid and plasmacytoid dendritic cells from house dust mite‐sensitized allergic patients dysregulate the T cell response. J Leukoc Biol2003; 73:91–9. [PubMed] [Google Scholar]

34. Shakib F, Schulz O, Sewell H. A mite subversive: cleavage of CD23 and CD25 by Der p I enhances allergenicity. Immunol Today1998; 19:313–6. [PubMed] [Google Scholar]

35. Trompette A, Divanovic S, Visintin A, Blanchard C, Hedge RS, Madan R_et al_ Allergenicity resulting from functional mimicry of a toll‐like receptor complex protein. Nature2009; 457:585–8. [PMC free article] [PubMed] [Google Scholar]

36. Liu LY, Mathur SK, Sedgwick JB, Jarjour NN, Busse WW, Kelly EA. Human airway and peripheral blood eosinophils enhance Th1 and Th2 cytokine secretion. Allergy2006; 61:589–97. [PubMed] [Google Scholar]

37. Fujisawa T, Katsumata H, Kato Y. House dust mite extract induces interleukin‐9 expression in human eosinophils. Allergol Int2008; 57:141–6. [PubMed] [Google Scholar]

Articles from Immunology are provided here courtesy of British Society for Immunology