Human group 2 innate lymphoid cells do not express the IL-5 receptor.

To the Editor:

Eosinophils, cardinal effector cells of type 2 inflammation, contribute to the clinical and immunopathologic manifestations of asthma and chronic obstructive pulmonary disease inflammatory endotypes. Eosinophil biology is governed by IL-5, a cytokine that binds with high affinity to a specific IL-5 receptor α subunit (IL-5Rα) before forming a heterodimeric receptor complex with the β subunit. IL-5 signaling promotes differentiation, maturation, and survival of eosinophil-committed progenitors while acting on mature eosinophils to enhance their migratory potential and effector responses. IL-5 signaling also promotes alternative splicing of the IL-5Rα gene to generate transmembrane forms of IL-5Rα. Because the IL-5-IL-5R axis appears to be restricted to eosinophils and basophils (and their progenitors), therapeutic regulation of these cells through the neutralization of circulating IL-5 (eg, mepolizumab and reslizumab) or IL-5Rα ligation (eg, benralizumab),1, 2 have emerged as effective strategies to deplete blood, tissue, and airway eosinophils and consequently reduce exacerbation rates and improve lung function.

Basophils and group 2 innate lymphoid cells (ILC2s) are important innate sources of type 2 cytokines, including IL-5, in response to epithelial-derived cytokines such as IL-33.7, 8 ILC2s, however, have emerged as central and critical innate coordinators of steady-state eosinophilopoiesis and epithelial cell–driven, type 2 immunopathology in asthma. As an important upstream regulator of eosinophil function we asked whether ILC2s, like their basophil counterparts, express the IL-5Rα subunit because this may have important implications for our understanding of the role of the IL-5-IL-5R axis in disease and therapeutic targeting of these rare but important innate immune cells. Detailed methods are provided in this article's Online Repository at www.jacionline.org.

ILC2s are rare innate lymphocytes that lack the T-cell receptor complex and all known lineage markers but, similar to other type 2 cytokine-producing cells such as TH2 cells, eosinophils, and basophils, express the type 2 prostaglandin D2 receptor, DP2/CRTH2 (CD294). We defined blood ILC2s as cells with singlet, lymphocyte light scatter properties (Fig 1, A and B), lineage (CD2, 3, 14, 16, 19, 56, and 235a)− but CD294+ (Fig 1, C). A large proportion of cells within this gate were basophils, (CD123+ cells in Fig 1, D); however, CD294+, CD123− ILC2s were present (Fig 1, D, and light scatter in E). Basophils displayed a higher level of CD125-phycoerythrin (PE) staining than did the ILC2s (Fig 1, D) that could be blocked in the presence of rhIL-5 (Fig 1, F), confirming the specificity of the antibody and basophil IL-5Rα expression. In contrast to basophils, there was no change in CD125-PE signal intensity when ILC2s were incubated with rhIL-5, indicating that there was no IL-5Rα expression on these cells (Fig 1, F). To confirm the reproducibility of these findings, we recruited 6 control and 13 volunteers with asthma (see clinical details in Table E1 in this article's Online Repository at www.jacionline.org) and measured IL-5Rα expression. The total mean (SD) % of ILC2s within the lymphocyte gate for both control and asthma groups were 0.1% ± 0.1% and 0.03% ± 0.03%, respectively (P = .06, Mann-Whitney). In total, combining asthma and control data sets, the basophil CD125 geometric mean fluorescence intensity (GMFI) (mean ± SD) in the absence or presence of rhIL-5 was 1285 ± 614 and 401 ± 315 respectively, representing a significant reduction in GMFI (Fig 1, G and I). In contrast, the ILC2 CD125 GMFI (mean ± SD) in the absence or presence of rhIL-5 was 45 ± 45 and 42 ± 33, respectively (Fig 1, H and I). In a subset of samples (n = 3), where basophils and ILC2s could be detected in whole blood, data were qualitatively similar. There were no significant differences between the control and asthma subjects for any of the flow cytometric measurements obtained.

Fig 1

CD125 protein is absent on blood ILC2s. A, Total PBMCs, singlets within boxed region. B, Singlet light scatter with lymphocyte gate overlaid. C, Lineage−, CD294+ cells were identified (boxed region) encompassing (D) CD123+ basophils and CD123− ILC2s, in the upper and lower left quadrant, respectively, with CD125 GMFI values. E, ILC2 light scatter properties. F, Overlay of basophil and ILC2 CD125-stained cells in the absence (red) or presence (blue) of rhIL-5. (G) Basophil and (H) ILC2 CD125 GMFI paired data (mean ± SD) from 6 healthy volunteers (open) and 13 (filled) volunteers with asthma, also shown in (I) unpaired on the same axes. Statistical comparisons were made using a Wilcoxon matched-pairs signed rank test. APC, Allophycocyanin; FITC, fluorescein isothiocyanate; FSC-A, forward scatter-area; FSC-H, forward scatter-height; SSC-A, side scatter-area.

Table E1

Clinical characteristics of the study participants recruited for blood collection

Characteristic	Control	Asthma	P value∗
n	6	13
Age (y)†	50 ± 15	59 ± 13	.18
Sex: M:F	4:2	6:7	.63
Smoking history (never, ex, current)‡	5, 1, 0	9, 4, 0	1.0
BMI (kg/m2)†	27 ± 4	29 ± 6	.47
Post-BD FEV1 (L)†	3.6 ± 0.8	2.5 ± 0.9	.02
Post-BD FEV1 (% predicted)†	109 ± 16	92 ± 24	.17
FEV1/FVC (ratio)†	0.82 ± 0.06	0.71 ± 0.09	.01
WBC (×109/L)†	6.4 ± 0.9§	7.1 ± 2.8‖	.63
Eosinophils (×109/L)†	0.16 ± 0.04§	0.28 ± 0.16‖	.12
Eosinophil (%)	2.5 ± 0.6§	4.3 ± 2.4‖	.13
Basophils (×109/L)†	0.04 ± 0.02§	0.04 ± 0.02‖	.97
GINA 1, 2, 3, 4, 5	NA	1, 1, 0, 7, 4	NA
Beclomethasone dipropionate equivalents (μg)¶	NA	1000 (650-1600)	NA

BD, Bronchodilator; BMI, body mass index; F, female; FVC, forced vital capacity; GINA, Global Initiative for Asthma; M, male; NA, not applicable; WBC, white blood cell.

Unpaired t test.

Mean ± SD.

All ex-smokers have quit smoking for 3+ years.

n = 5.

n = 12.

Median (interquartile range).

We next sought to confirm whether the IL-5Rα data were reflected at the RNA level using real-time PCR probes spanning the exon boundaries present within IL-5Rα subunit variants (see Fig E1, A, in this article's Online Repository at www.jacionline.org). For this, ILC2s were isolated from 3 additional donors and cultured in the absence or presence of cytokines known to enhance survival (IL-2 and IL-7) and activate ILC2s (IL-25 and IL-33). RNA was isolated at 6 time points (day 0, 1, 2, 4, 7, and 14) and converted into cDNA. Eosinophil cDNA was included as a positive control. The IL-5Rα transcripts were not detected in ILC2s at any time point; however, an increase in IL-5 mRNA was observed, indicating that the culture conditions were sufficient to activate the cells (Fig E1, B and C).

Fig E1

ILC2s do not express IL-5Rα transcripts. A, Schematic representation of IL-5Rα mRNA showing the splice variants and probe sets Hs00602482_m1 (detecting soluble and transmembrane variants) and Hs00236871_m1 (transmembrane only) used in this study. ΔCT values were calculated relative to 18S mRNA for (B) all IL-5Rα transcript variants using the Hs00602482_m1 probe and (C) IL-5 transcripts using Hs00174200_m1. In Fig E1, B and C, data are derived from isolated ILC2s (n = 3) that were grown for up to 14 days in the presence of rhIL-2, rhIL-7, rhIL-25, and rhIL-33. In Fig E1, B, cDNA generated from freshly isolated eosinophils was also included as a positive control in each IL-5Rα assay. In Fig E1, C, IL-5 transcript data are shown to confirm activation in the presence of cytokines; low ΔCT values reflect high abundance of the IL-5 transcript. n.d., Not detected.

Finally, using a similar approach to the blood analyses (Fig 1) and using surgically removed lung tissue from 7 subjects (see Table E2 in this article's Online Repository at www.jacionline.org), we asked whether tissue-derived ILC2s and basophils expressed IL-5Rα. Tissue-derived basophils were defined as viable, CD45+Lin−CD294+CD123+CD127− cells and ILC2s as viable, CD45+Lin−CD294+CD123−CD127+ cells present within the singlet lymphocyte gate (Fig 2, A-E). Basophil CD125-PE GMFI (mean ± SD, 612 ± 327) could be reversed in the presence of rhIL-5 to a mean of 317 ± 91 (example in Fig 2, F, and cumulative data in Fig 2, H). Using the super enhanced Dmax (SED) algorithm to compare basophil populations stained in the absence or presence of IL-5 for each of the 7 donors revealed that 71.1% ± 4.2% basophils were CD125 positive (Fig 2, I). In contrast, ILC2 CD125 GMFI (mean ± SD) was significantly lower than that of basophils and showed little change in the presence of rhIL-5 (190 ± 42 and 145 ± 49, respectively) (example in Fig 2, G, and cumulative data in Fig 2, H). SED analysis of ILC2s stained in the absence or presence of rhIL-5 for each of the 7 donors (Fig 2, I) did not reveal a CD125-positive subset (mean ± SD, 3.3% ± 0.89%). Data from these 7 subjects consistently show that tissue-derived basophils express the IL-5Rα whereas tissue-derived ILC2s do not (Fig 2, H).

Table E2

Clinical characteristics of the study participants recruited for lung tissue collection

Characteristic	Control
n	7
Age (y)∗	66 ± 9
Sex: M:F	4:3
Smoking history (never, ex, current)†	0, 4, 3
WBC (×109/L)∗	9.5 ± 3.7
Eosinophils (×109/L)∗	0.24 ± 0.18
Eosinophil (%)∗	2.4 ± 1.1
Basophils (×109/L)∗	0.04 ± 0.03
Diagnosis	Adenocarcinoma (n = 4)Squamous cell carcinoma (n = 2)Lung volume reduction surgery (n = 1)

F, Female; M, male; WBC, white blood cell.

Mean ± SD.

All ex-smokers have quit smoking for 2+ years.

Fig 2

CD125 protein is absent on lung-derived ILC2s. A, Total lung tissue cells, singlets within boxed region. B, Singlet light scatter properties with lymphocyte gate overlaid. C, Viable CD45+ cells (boxed region). D, Lineage− CD294+ cells (boxed region). E, ILC2s (CD123−CD127+) and basophils (CD123+CD127−) were identified. (F) Basophil CD125 fluorescence and (G) ILC2 CD125 fluorescence in the absence (red) or presence of rhIL-5 (blue). H, Collated basophil and ILC2 CD125 GMFI (mean ± SD) data (n = 7). I, Proportion of CD125+ basophils and ILC2s (calculated using SED); exemplar values given in Fig 2, F and G. Statistical comparisons were made using a Wilcoxon matched-pairs signed rank test. FITC, Fluorescein isothiocyanate; FSC-A, forward scatter-area; FSC-H, forward scatter-height; SSC-A, side scatter-area.

The strength(s) of this study lies in the fact that we have measured protein IL-5Rα subunit expression in a rare critical regulator of type 2 inflammation, ILC2s, and in the context of rhIL-5–mediated receptor blockade/downregulation, rather than an isotype control, which is less robust. The robustness of our data using human peripheral blood (from donors with asthma), ex vivo activated cells, and lung tissue cells diminishes the likelihood that IL-5Rα–expressing ILC2s are present in asthmatic tissue. Although it would be desirable to extend our blood and tissue observations to investigate whether ILC2s in bronchial biopsies from patients with asthma express IL-5Rα, the paucity of tissue ILCs and the need for multiple immunological markers to positively identify them severely limits this approach. These results extend the list of cells that are known not to express the IL-5Rα subunit, specific for the biological activities of IL-5. Moreover, we show that rare tissue-derived basophils, like their blood counterparts, express the IL-5Rα. Our data suggest that the success of therapeutic interventions targeting IL-5/R is unlikely to be mediated directly on ILC2s but may function via both eosinophils and basophils.