Sequential engagement of adhesion molecules and cytokine receptors impacts both piecemeal and anaphylactic degranulation of human basophils

Basophils are rare granulocytes in circulation which home to tissues in a process depending on rolling, adhesion and cytokine exposure. However, it is still unclear how these steps affect basophil degranulation. Our aim was to imitate these processes associated with homing by sequential crosslinking of adhesion molecules and cytokine exposure and evaluate the effect on basophil piecemeal (PMD) and anaphylactic degranulation (AND). Blood donors with or without allergic asthma were recruited from an ongoing cohort study. Basophils were subjected to CD62L‐, CD49d‐ or CD11b crosslinking and IL‐3 or IL‐33 stimulation in different orders followed by anti‐IgE and fMLP stimulation. Basophil CD203c and CD63 expression were analysed by flow cytometry to determine PMD and AND, respectively. IL‐3 induced PMD in basophils and combined with CD62L‐ or CD11b crosslinking, IL‐3 potentiated the degranulation regardless of sequential order. IL‐3 priming followed by adhesion molecule crosslinking induced AND and potentiated the effect of anti‐IgE. CD62L‐ and CD11b crosslinking did not further potentiate this effect. CD49d crosslinking followed by IL‐3 increased CD63 expression following anti‐IgE. IL‐3 potentiated the effect of fMLP on AND while adhesion molecule crosslinking did not. IL‐33 had impact on PMD only when followed by adhesion molecule crosslinking but did not potentiate neither IgE‐dependent nor IgE‐independent degranulation. Our data indicate that sequential interactions between basophils, cytokines and adhesion molecule ligands have a decisive effect on basophil degranulation and that these interactions are operational for fine‐tuning the activity of tissue dwelling basophils. These data should be considered when the effect of different pharmaceutical on basophil function is studied.

IL-3 induced PMD in basophils and combined with CD62L-or CD11b crosslinking, IL-3 potentiated the degranulation regardless of sequential order.IL-3 priming followed by adhesion molecule crosslinking induced AND and potentiated the effect of anti-IgE.CD62L-and CD11b crosslinking did not further potentiate this effect.CD49d crosslinking followed by IL-3 increased CD63 expression following anti-IgE.IL-3 potentiated the effect of fMLP on AND while adhesion molecule crosslinking did not.IL-33 had impact on PMD only when followed by adhesion molecule crosslinking but did not potentiate neither IgE-dependent nor IgEindependent degranulation.Our data indicate that sequential interactions between basophils, cytokines and adhesion molecule ligands have a decisive effect on basophil degranulation and that these interactions are operational for finetuning the activity of tissue dwelling basophils.These data should be considered when the effect of different pharmaceutical on basophil function is studied.

INTRODUCTION
Basophils are rare granulocytes originating from stem cell progenitors in the bone marrow.They express a wide range of receptors for cytokines (e.g., IL-3, IL-33) [1,2], growth factors (e.g., GM-CSF) and immunoglobulins (e.g., IgE and IgG) [3,4].After entering the circulation, basophils can be triggered by different factors via their receptors and infiltrate the tissues.By migrating into extravascular tissues, basophils recruit other immune cells such as eosinophils and neutrophils, induce Th2 immune responses and activate B-cell production of IgE [5][6][7].Studies have suggested that circulatory basophils in blood have different profile than the ones found in tissues [8,9].
Basophil trafficking from circulation to tissues involves different steps; rolling via selectins, for example, L-selectin (CD62L), tight adhesion by complement receptor 3 (CR3, CD11b/CD18) and very late antigen 4 (VLA-4, CD49d/CD29) and migration regulated by various factors such as IL-3 [10].CD62L has a stable expression on basophils upon activation compared to other immune cells such as neutrophils [11] while CR3 can be upregulated by cytokines such as IL-33 [12].IL-3 is both released from, and involved in, basophil development and activation [13,14] and basophil priming by IL-3 has been observed in diseases such as allergy.IL-33 is an alarmin cytokine which is released from tissue-derived cells and involved in tissue haemostasis, damage, inflammation and repair.Studies have shown the involvement of IL-33 in allergic, infectious-and chronic inflammatory diseases [15,16].IL-3 has important impact in circulation while IL-33 is more involved at the tissue level [17].
Basophils hold small granules containing potent inflammatory mediators (e.g., histamine, leukotrienes) [18] which can be released in two regulated ways upon stimulation; piecemeal (PMD) and anaphylactic degranulation (AND).In PMD, small amounts of contents are gradually released with help of intermediate vesicles at a slow rate.PMD is assumed to be involved both in haemostasis, for example, regulation of microvasculature tone [19], and chronic diseases and cancers [18,20].PMD is often identified by upregulation of CD203c expression.On the contrary, in AND basophils undergo a rapid explosive release of the entire content of their granules which in turn results in clinical manifestations [21].AND is identified by an upregulation of CD63 on basophil surface and the mode of degranulation is dictated by the stimulus type [22].It has been shown that basophil degranulation in tissues is caused by a range of mediators including IL-3 and their responses to cytokines can be affected by cell adhesion for example via integrins [23].
It is not fully known whether basophils become activated during migration and how the different steps of adhesion and migration process affect basophil function and responsiveness.In a previous study, we imitated different steps in cell recruitment, such as cell adhesion and cytokine priming.Crosslinking of CD62L and CD49d had a regulatory impact on basophil IgE-mediated degranulation and priming of basophils using IL-3 and IL-33 induced more IgE-independent degranulation using fMLP.In this study, the primary aim was to gain further knowledge on how the migration process affects both the slow and fast-paced degranulation of basophils.A secondary aim was to investigate these processes in individuals with or without allergic asthma.

Study population
Blood donors (n = 13) (median age 26 years) were recruited from the BAMSE cohort study at Södersjukhuset, Stockholm, Sweden [24,25].Six of the blood donors were non-allergic without asthma and seven blood donors had a history of allergic asthma in combination with being sensitised to airborne allergens.The demographic characteristics of participants are shown in Table 1.Venous blood samples (9 mL) were collected in sodium-heparin vacutainer tubes (Vacutainer, Becton Dickinson, USA).All samples were stored at 4 C and experiments were performed within 4 h from blood sampling.Informed consent was obtained from all participants.

Flow cytometric analysis of basophils
Piecemeal degranulation (PMD) of basophils following crosslinking of either CD62L, CD49d or CD11b and IL-3 or IL-33 stimulation A sequential stimulation with either cytokine priming followed by crosslinking of adhesion molecules or adhesion molecules crosslinking followed by cytokine priming was applied to induce piecemeal degranulation in basophils.
Whole blood was divided into tubes representing negative control (RPMI only), cytokine stimulation (either IL-3 or IL-33) and cytokine stimulation plus adhesion molecule crosslinking in different sequential order.Dosedependent stimulation with cytokines IL-3 and IL-33 have previously been shown [9].
All the tubes were then placed on ice and stained with anti-CD203c-PE antibody (Beckman Coulter, France) for 25 min.Later, red blood cells (RBCs) were lysed with 2 mL cold lysis buffer (containing 154 mM NH4Cl, 10 mM KHCO3 supplemented with 0.1 mM EDTA, pH 7.2) and were washed with cold PBS for 5 min at 300 Â g at 4 C. Cells were re-suspended in 300 μL of cold PBS and analysed using flow cytometry (Navios, Beckman Coulter, France).PMD was analysed as mean fluorescence intensity (MFI) for CD203c.

Anaphylactic degranulation (AND) induced by IgE-dependent basophil activation following crosslinking of either CD62L, CD49d or CD11b and cytokine stimulation
Basophils were activated in an IgE-dependent manner following prior sequential stimulation with either cytokine priming followed by crosslinking of adhesion molecules or crosslinking of adhesion molecules followed by cytokine priming.
Whole blood was divided into tubes representing negative control (RPMI only), cytokine stimulation (either IL-3 or IL-33), cytokine stimulation plus adhesion molecule crosslinking in different sequential order as well as cytokine priming plus adhesion molecule crosslinking plus anti-IgE stimulation in different sequential order.
The same setting was applied as described in the previous part.In one set of tubes, whole blood was incubated with cytokines.The cytokine-primed cells were then incubated with unconjugated primary antibody towards adhesion molecules followed by incubation with unconjugated anti-mouse secondary antibody.
In the other set of tubes, whole blood was incubated with unconjugated primary antibody towards adhesion molecules, followed by incubation with unconjugated anti-mouse secondary antibody.The cells were then incubated with 100 ng/mL IL-3 or IL-33-containing RPMI.
The relevant tubes were then incubated with either RPMI or 1 μg/mL anti-IgE antibody (Beckman Coulter, France) for 30 min at 37 C.The tubes were thereafter placed on ice and incubated with anti-CD63-FITC (Beckman coulter, France) for 25 min.Later, RBCs were lysed, washed, and analysed using flow cytometry (Navios, Beckman Coulter).AND was analysed as percent CD63 positive basophils.
Anaphylactic degranulation (AND) induced by IgE-independent basophil activation following crosslinking of either CD62L, CD49d or CD11b and cytokine stimulation Basophils were activated in an IgE-independent manner following prior sequential stimulation with either cytokine priming followed by crosslinking of adhesion molecules or crosslinking of adhesion molecules followed by cytokine priming.
The same setting was applied as described in previous part except the stimulator which was fMLP.Following priming with cytokine and/or adhesion molecule crosslinking, the relevant tubes were incubated with either RPMI or 5 Â 10 À7 M fMLP for 30 min at 37 C.The tubes were then placed on ice and incubated with anti-CD63-FITC (Beckman coulter) for 25 min.The red blood cells (RBCs) were lysed, washed and analysed using flow cytometry (Navios, Beckman Coulter).AND was analysed as percent CD63 positive basophils.
Comparison of basophil surface expression of CD203c/CD63 in non-sensitised subjects without asthma and sensitised subjects with asthma following stimulation We divided our study population into two groups: nonsensitised subjects without asthma and sensitised subjects with asthma (allergic asthma) to investigate whether the differences between various treatments are disease dependent.

Statistical analyses
Statistical analysis and scatter plots were done in Graph-Pad Prism 8.3.1 (GraphPad Software, Inc., USA).In scatter plots, whiskers represent 25%-75% interquartile range (IQR) with the median shown by a line.Due to a low number of study participants the results were not normally distributed, hence non-parametric tests were used.Comparisons between treatments were performed using Friedman test with Dunn's multiple comparisons test and comparisons between two groups were performed using Wilcoxon matched-pairs signed rank test.A p-value <0.05 was considered statistically significant.

PMD of basophils following crosslinking of either CD62L, CD49d or CD11b and cytokine stimulation
We compared basophil surface expression of CD203c (marker for PMD) after stimulation with either IL-3 or IL-33 or in combination with adhesion molecule crosslinking (Figure 1; all samples included).
Priming of basophils with IL-3 alone induced PMD (p = 0.0002).This effect was potentiated when combined with CD62L-or CD11b crosslinking either before (p = 0.005, p = 0.02) or after (p = 0.0002, p = 0.001) IL-3 stimulation.CD49d crosslinking followed by IL-3 stimulation potentiated PMD ( p = 0.002) contrary to when IL-3 priming was followed by CD49d crosslinking which did not have additive effect on PMD.All comparisons between treatments are shown in Supplementary Table 1.

AND induced by IgE-dependent basophil activation following CD62L-, CD49d-or CD11b crosslinking and cytokine stimulation
We compared basophil surface expression of CD63 (marker for AND) after IL-3 and adhesion molecule crosslinking in different sequential combinations followed by anti-IgE stimulation (Figure 2).IL-3 alone did not impact the percentage of CD63+ cells whereas IL-3 priming followed by CD62L-or CD11b crosslinking induced AND (p = 0.005, 0.004).IL-3 potentiated the effect of anti-IgE on AND ( p = 0.0002) but addition of CD62L-or CD11b crosslinking did not further impact the IL-3 plus anti-IgE effect regardless of sequential order.IL-3 priming followed by CD49d crosslinking, but not the reverse order, induced AND ( p = 0.03).Anti-IgE activation after CD49d crosslinking followed by IL-3 stimulation, but not the reverse sequence, showed an increased CD63 expression (p = 0.05) compared to IL-3 only followed by anti-IgE stimulation (Supplementary Table 2).
Priming with IL-33 did not potentiate the IgEdependent basophil degranulation neither alone nor in combination with CD62L-, CD49d-or CD11b crosslinking (Supplementary Figure 1 and Supplementary Table 3).

AND induced by IgE-independent basophil activation following crosslinking of either CD62L, CD49d or CD11b and cytokine stimulation
We compared basophil surface expression of CD63 (marker for AND) after stimulation with IL-3 and adhesion molecule crosslinking in different sequential combinations followed by fMLP (Figure 3).IL-3 potentiated the impact of fMLP on AND (p = 0.002).However, addition of CD62L-, CD49d-or CD11b crosslinking, regardless of sequential order, did not further impact the effect of IL-3 and fMLP on AND (Supplementary Table 2).

Comparison of basophil surface expression of CD203c/CD63 in non-sensitised subjects without asthma and sensitised subjects with asthma, following crosslinking of CD62L, CD49d or CD11b and cytokine stimulation
We divided the study subjects into two groups: nonsensitised subjects without asthma and sensitised subjects with asthma (allergic asthma) to investigate whether the     1.
observed differences in basophil response to cytokines and/or adhesion molecule crosslinking associate with having allergic asthma.We observed that basophil CD203c expression (PMD) was similar between the two groups when the cells were only stimulated with IL-3 or IL-33.A combination of IL-3 priming and CD62L-, CD49d-or CD11b crosslinking resulted in equal degranulation comparing the groups, on the contrary a combination of IL-33 priming, and adhesion molecule crosslinking showed differences between healthy individuals and asthmatic patients (Supplementary Figure 3).
The basophil CD63 expression (AND) was also similar between groups when basophils were only treated with either IL-3 or IL-33 (Supplementary Figure 4).The combination of IL-3 priming and CD62L-or CD49d crosslinking did not show differences in degranulation between the groups.

DISCUSSION
In this study we provide data that indicate that sequential interactions between basophils, IL-3, IL-33 and adhesion molecule ligands have a profound effect on basophil degranulation.This study expands our previous findings where we investigated basophil degranulation after stimulation with either cytokines or cross-linking of adhesion molecules.Herein we investigate further how the sequential stimulation with both cytokines and adhesion molecule crosslinking affect the basophil degranulation process.
Basophils predominantly circulate in the blood, but also frequently migrate to the sites of inflammation, for example, airways and lungs, so called homing.Most studies on basophils have been conducted on peripheral cells and consequently the knowledge of the impact of the homing process on the basophil tissue signature is scanty.Therefore, we have focused on the events that precedes basophil activation through the IgE-dependent and IgEindependent pathways, for example, exposure to cytokines and adhesion molecule ligands, and the impact of these events on the subsequent activation stages.We have used PMD and AND as read-outs.During PMD, small amounts of contents are released due to gradual translocation of intermediate vesicles to the cell surface accompanied with an up-regulation of CD203c.In the course of AND, basophils undergo a rapid explosive release of the entire content of their granules and is paralleled by up-regulation of CD63 and manifested in clinical symptoms [21].
We show that crosslinking of CD62L or CD11b, regardless of sequential order, potentiates the effect of IL-3 on PMD.We could also show that CD49d crosslinking potentiates effect of IL-3 on PMD when crosslinking occurs prior to cytokine exposure.Unlike IL-3, IL-33 exposure alone does not have any impact on PMD.However, IL-33-primed basophils do up-regulate CD203c when followed by CD62L-, CD49d-or CD11b crosslinking but not in the reverse mode.This observation is of importance since IL-33 is primarily regarded as a tissueresiding cytokine that is secreted by epithelial and endothelial cells in response to tissue damage.In a previous study, we noticed a slight IL-33 dependent up-regulation of CD203c that was not reproduced in this study and could be referred to different donors and cohort size [9].Together these data indicate that IL-3 exposure impacts  1.
the degree of PMD in tissue-dwelling basophils after adhesion molecules have been engaged during or after the transmigration process.Data also indicate that IL-33 priming when followed by crosslinking of adhesion molecules, for example, due to tissue-bound matrix protein interactions, may cause PMD.
We demonstrate that IL-3 exposure followed by either CD62L, CD49d or CD11b crosslinking, but not the reverse order, induces AND determined by CD63 upregulation.This illustrates that peripheral IL-3-primed basophils once transmigrated have initiated a significant degranulation process involving CD63 translocation to the cell surface.However, the level of AND induced by IL-3 priming and adhesion molecule crosslinking is moderate which may indicate an existing low level of AND.Hypothetically, this level of activation may continue for an extended period during which the basophil population continues to release mediators and hence have a role in the regulation of the immune response in the tissue.
In the present study, we show that IL-3 potentiates the effect of anti-IgE on the magnitude of AND.However, engagement of adhesion molecules does not have any additional effect, despite a minor effect when CD49d crosslinking precedes IL-3 and anti-IgE exposures.This suggests that when basophils are activated through the IgE receptor pathway, the outcome is not influenced by prior engagement of adhesion molecules that occurs during the transmigration.Anti-IgE induced AND results in a rapid and pronounced degranulation which is operational in, for example, allergic asthma.This degranulation process is believed to be short-lived and once completed, the cell is incapable to further mediator synthesis and release.
It is well recognised that basophils can be activated through an IgE-independent pathway, and we have recently shown that IL-3 exposure, but not CD62L-or CD49d crosslinking, amplifies the effect of fMLP activation on CD63 expression [9].Herein we extend this observation and show that CD62L-, CD11b-, CD49d crosslinking does not impact the effect of IL-3 only.We therefor propose that IgE-independent AND in IL-3-primed basophils is not influenced by adhesion molecule interactions during homing.
A secondary aim of this study was to investigate if subjects with allergic asthma and controls without allergic asthma differ in regulation of basophil degranulation.We did not observe any major differences.However, the number of subjects was too low to allow any firm conclusions Even so, we find our data valuable for generating hypothesis that can be focus for future studies with larger sample sizes and other patient groups.
This project is designed as a human whole blood study with the purpose to investigate how the basophil behaves in an environment that mimics the in vivo situation, since interactions with other cells may affect the biological outcome.There is a possibility that there is an indirect effect of other immune cells and soluble mediators, and it would be interesting to study the effect on purified basophils.However, in the present study the incubation time is rather short and does probably not allow generation and release of newly synthesised cytokines from other immune cells.Also, we have in previous in experiments with purified basophils in microfluidic chips identified that the isolation process results in preactivation and spontaneous degranulation followed by CD63-expression [26,27] The same type of pre-activation was noticed also for basophils after Ficoll-separation (inhouse data), which encouraged us to continue with our whole blood model.Also, one must consider that a significant number of blood donations would be needed to generate an enough number of purified basophils to complete the suggested studies.Another interesting discussion is the use of animal models.Mice models are important and useful tools to study cell migration and activation in vivo.However, there are known differences in the characteristics of mice and human basophils, including pathways that regulate activation and release of inflammatory mediators [28].Our focus has been human basophil biology, and we focus on selected key factors involved in basophil migration such as crosslinking and cytokine exposure We have focused on two cytokines, IL-3 and IL-33, which have main impact in the circulation and at the tissue level, respectively.However, one ought to consider that a variety of other cytokines are operational in tuning basophil characteristics, for example, gamma Interferon [29].
Our data indicate that sequential interactions between basophils, cytokines and adhesion molecule ligands have a decisive effect on basophil degranulation and that these interactions are operational for fine-tuning the activity of tissue dwelling basophils.These data should be considered when the effect of different pharmaceutical on basophil function is studied.

K
E Y W O R D S adhesion molecules, basophils, cell activation, cell trafficking, cytokines

1
Piecemeal degranulation of basophils, represented by surface expression of CD203c (MFI), following crosslinking of adhesion molecules and cytokine stimulation.Selected significant differences are shown in the figure by lines.Detailed comparisons are shown in Supplementary Table

F
I G U R E 2 IgE-dependent anaphylactic degranulation of basophils, represented by the percentage of CD63+ cells (CD63%), following crosslinking of adhesion molecules and/or cytokine stimulation.Selected significant differences are shown in the figure by lines.Detailed comparisons are shown in Supplementary Table U R E 3 fMLPdependent anaphylactic degranulation of basophils, represented by the percentage of CD63+ cells (CD63%) following crosslinking of adhesion molecules and/or cytokine stimulation.Selected significant differences are shown in the figure by lines.Detailed comparisons are shown in Supplementary Table Demographic characteristics of study population.
T A B L E 1 a SABA: short-acting beta-agonists.b LABA: long-acting beta-agonists.