Sialic acid–modified der p 2 allergen exerts immunomodulatory effects on human PBMCs

Background House dust mite extract–based allergen immunotherapy (AIT) to treat house dust mite allergy is substantially effective but still presents some safety and efficacy concerns that warrant improvement. Several major allergen-based approaches to increase safety and efficacy of AIT have been proposed. One of them is the use of the group 2 allergen, Der p 2. Objective We sought to investigate the immunomodulatory effects of sialic acid–modified major allergen recombinant Der p 2 (sia-rDer p 2) on PBMCs from healthy volunteers. Methods We activated PBMCs with anti-CD3/CD28 antibodies and incubated them at 37°C for 6 days in the presence or absence of either native rDer p 2 or α2-3 sialic acid–modified rDer p 2 (sia-rDer p 2). We assessed the changes in CD4+ T-cell activation and proliferation by flow cytometry and changes in T-lymphocyte cytokine production in cell culture supernatant by ELISA. Results We observed that PBMCs treated with sia-rDer p 2 presented with a markedly decreased expression of CD69 and an increased abundance of LAG-3+ lymphocytes compared with cells treated with rDer p 2. Moreover, PBMCs treated with sia-rDer p 2 showed a reduced production of IL-4, IL-13, and IL-5 and displayed a higher IL-10/IL-5 ratio compared with rDer p 2–treated PBMCs. Conclusions We demonstrate that sia-rDer p 2 might be a safer option than native rDer p 2 for Der p 2–specific AIT. This is most relevant in the early phase of AIT that is often characterized by heightened TH2 responses, because sia-rDer p 2 does not enhance the production of TH2 cytokines.

The standard therapeutic strategies for managing immune dysregulation, such as the excessive immune response in allergic reactions, mostly involve the use of immunosuppressive drugs, such as corticosteroids, Janus kinase, and calcineurin inhibitors, which lead to generalized immunosuppression. 1 Because of the multiple side effects associated with immunosuppressive drugs, preventing or attenuating immune dysregulation in a specific manner is highly desirable.3][4] These strategies generally aim to alter antigenpresenting cells (APCs) to suppress inflammatory T cells and induce regulatory T (Treg) cells. 5irectly targeting APCs including dendritic cells (DCs), monocytes, and B cells is an attractive strategy to modulate T-cell function and to induce antigen-specific tolerance. 5In particular, inhibitory receptors expressed by APCs, such as the Siglec (sialic acid-binding immunoglobulin-like lectin) family of carbohydrate-binding proteins, are a good target.Siglecs recognize sialic acid-containing glycoproteins and glycolipids and can be targeted to induce immunosuppressive responses. 5,6ost of the Siglecs, including Siglec-9 and Siglec-10, discussed in this study possess immunoreceptor tyrosine-based inhibitory motifs, which transmit immune deactivating signals on recognition of their ligand (sialic acids) and can thus be therapeutically exploited for the management of inflammatory diseases. 6e previously demonstrated that ex vivo and in vivo DC-targeting of a2-3 sialic acid-modified antigen to the mouse homolog of human Siglec-9 (ie, Siglec-E) drove naive CD4 1 T-cell differentiation into antigen-specific Treg cells. 7Moreover, DCs treated with sialic acid-modified antigen dampened T-cell differentiation to effector T cells even in the presence of native antigen-loaded DCs. 7Also, exposing LPS-stimulated human monocyte-derived DCs to Siglec-9 ligands, a2-3 sialic acids, or anti-Siglec-9 antibodies suppressed the production of IL-6 and IL-12. 8,9Together, these reports show that targeting Siglec-9 inhibits T H 1 responses.However, very little is known about whether Siglec-9-mediated interferences can also alter allergen-associated T H 2 responses, characterized by the secretion of IL-4, IL-5, and IL-13.
We therefore evaluated the immunomodulatory effects of a2-3 sialic acid-modified antigen on the T H 2 arm of T H cells in human PBMCs using Der p 2 as a model antigen.Der p 2 is one of the established major allergens from house dust mite (HDM) and is an important risk factor for the development of allergic rhinitis and asthma. 10,11Consequently, research into the use of Der p 2 for allergen immunotherapy (AIT) against Der p 2-HDM allergy instead of HDM extracts has grown in the past decades because AIT with HDM still presents some safety and efficacy concerns that warrant improvement. 4Exploiting the sialic acid-Siglec axis by modifying Der p 2 via its 12 free lysine residues 12 with sialic acids could provide Siglec-mediated crosslinking and signaling to attenuate allergen-specific T H 2-mediated immune responses. 13THODS a2-3 Sialic acid-recombinant Der p 2 binding to Siglec-human Fc chimeras NUNC MaxiSorp plates (Greiner Bio-One, Fredensborg, Denmark) were coated with 10 mg/mL of recombinant Der p 2 (rDer p 2) or a2-3 sialic acid-rDer p 2 (sia-rDer p 2) overnight at room temperature.After washing with HBSS (Gibco, New York, NY), wells were blocked with carbo-free blocking buffer-glycoprotein-free blocking agent (Vector Laboratories, Newark, Calif) diluted 1:10 in HBSS for 1 hour at room temperature and subsequently incubated with Siglec-Fc chimeras and goat anti-human Fc-peroxidase antibody (Jackson Laboratory, Bar Harbor, Me).The following Siglec-Fc chimeras (R&D Systems, Minneapolis, Minn) were used: Siglec-2-hFc (1968-SL-050), Siglec-3-hFc (1137-SL-050), Siglec-7-hFc (1138-SL-050), Siglec-9-hFc (1139-SL-050), and Siglec-10-hFc (2130-SL-050).
For all colorimetric assays, plates were developed using 3,39,5,59-tetramethylbenzidine (Merck, Germany) as a substrate and H 2 SO 4 as the stop solution.The iMark microplate absorbance reader (Bio-Rad) was used to measure absorbance at 450 nm.

Cell Isolation and culture
Blood from healthy volunteers for the isolation of PBMCs was supplied by the Sanquin Blood Bank (Sanquin, Amsterdam, The Netherlands).Healthy donors (nonallergic) gave their written consent for the use of blood donation for research purposes.We performed an ImmunoCAP assay to confirm the allergic status of donors.PBMCs were isolated from whole blood by density gradient centrifugation (at 2000 rpm for 30 minutes) using Lymphoprep (Serumwerk, Serumwerk, Germany).After isolation, PBMCs were frozen and stored at 2808C (or in liquid nitrogen) until use.PBMCs were thawed and plated in 96-well round-bottomed plates (Greiner Bio-One, Alphen aan den Rijn, The Netherlands) at a density of 10 3 10 6 cells/mL in RPMI 1640 (Thermo Fisher, Waltham, Mass) supplemented with 10% inactivated FBS (Biowhittaker, Switzerland), 1% glutamine (Thermo Fisher), and 1% penicillin-streptomycin (Lonza, Basel, Switzerland).The PBMCs were left inactivated or activated with plate-bound anti-CD3 (10 mg/mL) (clone SPV-T3b, Monoclonal Antibody Facility Department of Molecular Cell Biology and Immunology, Amsterdam, The Netherlands) and soluble anti-CD28 (2 mg/mL) (Sanquin).The activated PBMCs were subsequently either left untreated (control) or treated with rDer p 2 (10 mg/mL) or sia-rDer p 2 (10 mg/mL) for 6 days at 378C and 5% CO 2 in humidified air.Supernatants were harvested and stored at 2208C for cytokine measurements and cells were harvested for flow cytometry.

Prevention or reversion of PBMC responses
To investigate whether sia-rDer p 2 can revert an existing inflammation involving rDer p 2, anti-CD3-/CD28-activated PBMCs (10 3 10 6 cells/mL) were first treated with rDer p 2 (10 mg/mL) for 2 days, after which cells were washed with RPMI medium and sia-rDer p 2 (10 mg/mL) was added to the cells and cultured for another 4 days.To investigate whether sia-rDer p 2 can prevent inflammation involving rDer p 2, anti-CD3-/CD28activated PBMCs were first treated with sia-rDer p 2 (10 mg/mL) for 2 days, cells were washed with RPMI medium, and rDer p 2 (10 mg/mL) was added to the cells and cultured for another 4 days.This setup was adapted from that previously described. 14

Flow cytometry
For the characterization of the different cell populations within PBMCs and the evaluation of Siglec expression, cells (1 3 10 6 /well) were stained with mAbs (Table I) for 30 minutes on ice in a 96-well V-bottomed plate (Greiner Bio-One).Antibodies were diluted in PBS containing 0.1% BSA (Roche, Rotkreuz, Switzerland), 0.02% sodium azide, Fc block (in-house), and True-Stain monocyte blocker (BioLegend, Amsterdam, The Netherlands).Fixable Viability stain, Zombie NIR (BioLegend), or LIVE DEAD Blue (Thermo Fisher) was used to stain dead cells for 15 minutes before staining surface markers.For intracellular Ki67, IL-10, and FoxP3 detection, cells were incubated with 0.5 mg/mL of phorbol 12-myristate 13-acetate (Sigma-Aldrich, Sofia, Bulgaria) and 1 mg/mL of ionomycin (Sigma-Aldrich) in the presence of GolgiStop and Golgiplug (BD Biosciences, Erembodegem, Belgium) for 4 hours.After staining cell surface markers, cells were fixed-permeabilized using FoxP3/Transcription Factor Staining Buffer Set (Thermo Fisher, Waltham, Mass) as directed by the manufacturer and stained with intracellular antibodies (Table I).Labeled cells were fixed with 1% paraformaldehyde (Electron Microscopy Sciences, Hatfield, Pa) and stored at 48C until acquisition.Fluorescence minus 1 controls were prepared for each cell marker and used for gating.Stained samples were acquired with the 4-or 5laser Aurora flow cytometer (Cytek Biosciences, Amsterdam, The Netherlands).FCS files were analyzed using FlowJo software v10.8 (BD Biosciences).

Binding of sia-rDer p 2 to PBMCs
PBMCs (1 3 10 6 /well) were washed with 0.5% BSA in HBSS (ice-cold) in a 96-well V-bottomed plate.rDer p 2 or sia-rDer p 2 (10 mg/mL) or 10 mg/mL polyacrylamide-a2-3 biotin (positive control) was added to the cells and incubated for 1 hour at 378C.Subsequently, 10 mg/mL of biotinylated anti-Der p 2 antibody (Absolute Antibody, Amsterdam, The Netherlands) was added and incubated for 45 minutes at 48C.Cells were then stained with mAbs to characterize PBMC subsets (Table I) and streptavidin-phycoerythrin to detect binding (Jackson Laboratories) for 30 minutes at 48C.Stained cells were fixed in 1% paraformaldehyde and acquired with the 5-laser Aurora flow cytometer (Cytek Biosciences).

Statistical analysis
Data are presented as mean 6 SD (as indicated in figure legends).P values were determined by the Wilcoxon paired test using GraphPad Prism version 9 (GraphPad, San Diego, Calif) (as indicated in figure legends).Differences in values were considered significant at a P value of less than .05.

Sia-rDer p 2 binds to APCs
a2-3 sialic acids were chemically conjugated to rDer p 2 through a maleimide-thiol reaction to produce glyco-allergen conjugates (sia-rDer p 2) (see the Methods section in this article's Online Repository at www.jaci-global.org)(Fig 1, A).We performed a lectinbinding ELISA to confirm the presence of a2-3 sialic acids on sia-Der p 2 (see the Methods section and Fig E1 in this article's Online Repository at www.jaci-global.org).We then assessed whether sia-rDer p 2 would bind to selected Siglec-Fc chimeras and found that sia-rDer p 2 but not rDer p 2 interacted primarily with Siglec-9-Fc (Fig E1

sia-rDer p 2 modulates T H 2 cytokine production but not T H 1 or T H 17 cytokines
Given that sia-rDer p 2 equally modulated both T H 1-and T H 2-cell activation, we investigated whether a similar effect would occur in their cytokine profiles.When compared with baseline, rDer p 2 upregulated the production of IL-4 (Fig 3 E).However, sia-rDer p 2 downregulated the production of IL-5 (Fig 3, C) but did not affect IL-13, IL-4, IL-7, and IFN-g (Fig 3, A, B, D, and E).Moreover, compared with rDer p 2, sia-rDer p 2 induced lower production of IL-13 and IL-4 (Fig 3, A  and B).We then investigated whether the effects of sia-rDer p 2 on IL-5 were a result of direct binding to CD4 1 T cells or mediated by APCs present in PBMCs by coincubating rDer p 2 or sia-rDer p 2 with pure CD4 1 T cells activated with anti-CD3/CD28.The production of IL-5 was not affected (Fig 3 , F).We can therefore conclude that sia-rDer p 2downregulates IL-5 secretion and that this may be via APC-T-cell interactions.
The Treg/T H 2 cytokine ratio is higher with sia-rDer p 2 than with native rDer p 2 Successful AIT can be characterized by a high IL-10/IL-5 ratio. 12We therefore analyzed culture supernatants of anti-CD3-/ CD28-activated PBMCs to determine the impact of rDer p 2 or sia-rDer p 2 on IL-10 production.IL-10 secretion was lower in PBMCs treated with rDer p 2 compared with the control.Sia-rDer p 2 also attenuated IL-10 secretion but to a lesser extent than rDer p 2 (Fig 4 , A).We then determined the Treg/T H 2 cytokine balance by calculating the ratio of IL-10 to IL-4, IL-13, and IL-5, respectively.We observed that the IL-10/IL-4, IL-10/IL-13, and IL-10/IL-5 ratios were higher in sia-rDer p 2-treated PBMCs than in rDer p 2-treated PBMCs (Fig 4,  B-D).To further determine the role of IL-10 in the reduced expression of T H 2-cell cytokines, we stimulated PBMCs in the presence of an IL-10-blocking antibody and analyzed the IL-5 and IL-13 production.In the presence of an IL-10-blocking antibody, Sia-rDer p 2-treated PBMCs produced more IL-5 and IL-13 (Fig 4 , E and F).These data suggest that sia-rDer p 2, unlike rDer p 2, prevents the disruption of the Treg/T H 2 cytokine balance and that IL-10 plays a role in the downregulation of IL-5.
6][17] We therefore assessed whether sia-rDer p 2 could expand Treg cells.Given the vast phenotypic heterogeneity of human Treg cells, we analyzed only the following subsets: To do this, anti-CD3-/CD28-activated PBMCs were exposed to sia-rDer p 2 either 48 hours before (prevention of response) or after (reversal of response) coincubation with rDer p 2 (Fig 6, A).We observed that treating PBMCs with sia-rDer p 2 before treating with rDer p 2 resulted in the downregulation of both IL-5 and IL-13 (Fig 6 , B).However, the percentages of both CD69 1 CD4 1 and Ki67 1 CD4 1 T cells (Fig 6 , C) were unaltered.In contrast, treating PBMCs with sia-rDer p 2 after pretreating them with rDer p 2 resulted in an increase in IL-5 and IL-13 (Fig 6 , D).Nonetheless, the proportions of CD69 1 CD4 1 and Ki67 1 CD4 1 T cells appeared to be reduced (Fig 6, E).Together, these data demonstrate that sia-rDer p 2 can prevent the activation of T H 2 cytokine responses on challenge with rDer p 2.

DISCUSSION
This study aimed to investigate the immunomodulatory capacity of sialic acid-modified recombinant Der p 2 on the T H 2 arm of CD4 1 T cells in human PBMCs from nonallergic individuals ex vivo.We also evaluated whether sia-rDer p 2 could induce the expansion of Treg cells.We noted that sia-rDer p 2 moderately suppressed the activation of CD4 1 T cells, evidenced by a reduction in the expression of CD69 in both T H 1 and T H 2 cells and downregulated the production of T H 2 cytokines IL-5 and IL-13 but not the T H 1 cytokine IFN-g or the T H 17 cytokine IL-17.We also showed that sia-rDer p 2 binds to Siglec-9 present on monocytes, DCs, and B cells, and that these cells may be involved in the suppression of CD4 1 T-cell activation and in the diminished production of T H 2 cytokines.
9][20] This is because there is still ongoing debates about properly standardizing the components of the crude HDM extracts that are currently used in clinics, 21 and the risk of unwanted side effects is still quite high. 22Moreover, 79.2% of people with asthma have high IgE titers specific for Der p 2, 23 and Der p 2 itself is a strong risk factor for the development of asthma. 10Moreover, Der p 2 is a good model antigen for investigating the effects of sialic acid-protein modification on immune responses because it contains 12 free lysine residues on which sialic acid molecules could be conjugated.In sialic acid-Siglec interactions, the use of multivalent ligands is necessary to achieve sufficient avidity to stably bind Siglecs and subsequently transduce strong immunomodulatory signals. 24We hypothesized using Der p 2 decorated with sialic acids instead of native Der p 2 would prevent the initial spike in T H 2 responses 25 observed during HDM AIT and potentially decrease the time needed to develop full tolerance because of the induction of immunomodulatory responses mediated by the sialic acid-Siglec axis.
We observed that when PBMCs from nonallergic individuals were treated with rDer p 2, they produced more IL-4 and IL-13, features that are typical of an allergic phenotype.Indeed, nonatopic individuals have been reported to possess allergen- specific cells, 26,27 although the response to the allergen is of a lower intensity compared with those of atopic individuals. 28his could explain the observed increase in IL-4 and IL-13.Notably, however, the production of IL-4 and IL-13 from sia-rDer p 2-treated PBMCs from the same nonallergic individuals was unaltered, whereas the production of IL-5 was downregulated.This indicates that sia-rDer p 2 may be safer than native Der p 2 because it does not enhance T H 2 cytokine production.In addition, sia-rDer p 2 did not affect the production of both IL-17 and IFN-g, which can enhance the severity of allergic diseases and asthma. 29,30In an in vitro prophylactic setting, where we treated PBMCs with sia-rDer p 2 before treating them with rDer p 2, sia-rDer p 2 downregulated the production of IL-5 and IL-13.Collectively, these data show that sia-rDer p 2 inhibits T H 2 cytokine production, especially IL-5, and may therefore be a good candidate for Der p 2-specific AIT or be used for prophylactic vaccination against Der p 2-HDM allergy.
When tested in an ex vivo therapeutic setting (PBMCs were first treated with rDer p 2 and then with sia-rDer p 2), sia-rDer p 2 did not produce the suppressive effects described earlier.We hypothesize that the activation signal produced by a combination of anti-CD3/CD28 and rDer p 2 was too strong to be overcome by sia-rDer p 2. The fact that lowering the amounts of anti-CD3/CD28 did not result in a different outcome (data not shown) might indicate that the changes provoked in the PBMCs after activation with anti-CD3/CD28 cannot be reversed by the immunosuppressive signals generated by sialic acids.Indeed, anti-CD3-mediated T-cell activation can be overridden only when the suppressive compound is administered simultaneously with anti-CD3. 31Notably, the fact that sia-Der p 2 modulated only T H 2 cytokines, and not IFN-g or IL-17, as opposed to previous reports with ovalbumin as antigen, 7 indicates that the activity of sia-Der p 2 may be driven by APCs.Moreover, the involvement of APCs is also substantiated by our finding that sia-rDer p 2 does not affect T H 2 cytokine secretion when incubated with pure CD4 1 T cells.
Beyond the T H 2 cytokine response, we measured the effect of sia-rDer p 2 on the activation of CD4 1 T cells by the expression of CD25 and CD69.CD25 is widely recognized as the primary marker for cellular activation, 32 whereas a prolonged increase in the expression of CD69 is associated with allergic inflammation. 33Our findings show that sia-Der p 2 markedly suppressed the expression of CD69.The effects of sia-rDer p 2 on CD69 were similar in both T H 1 and T H 2 cells.Shinoda et al 34 reported that CD69-deficient CD4 1 T cells cannot help B cells to produce high-affinity antibodies and fail to mediate the generation of longlived plasma cells.Therefore, immunotherapy with sia-rDer p 2 would indirectly inhibit the generation of long-lived allergic plasma cells.
Another important mechanism that has been proposed to be involved in successful AIT is the induction of Treg cells. 12,22The 2 most commonly described subsets of Treg cells that play a key role in allergen tolerance are FoxP3 1 Treg cells 35,36 and IL-10 1 Tr1 cells. 16,37We showed that, compared with rDer p 2, treatment of PBMCs with sia-rDer p 2 did not affect the frequency of FoxP3 1 Treg cells but expanded the frequency of Tr1 cells.Of note, Tr1 cells, having strong suppressive activity, are expanded after AIT, and they correlate with a decrease in clinical scores. 16ecause LAG-3 is increasingly implicated in the downregulation of T-cell responses, 38,39 we measured the frequency of LAG-3 1 Treg-cell subsets.We observed that the frequencies of CD25 1 LAG-3 1 Treg cells, notably the FoxP3 2 subset and PD-1 1 LAG-3 1 , were higher in the sia-rDer p 2 than in the rDer p 2 condition.Dawicki et al 40 have reported that CD25 1 LAG-3 1 FoxP3 2 Treg cells can contribute to tolerance induction.These data suggest that AIT with sia-rDer p 2 instead of Der p 2 might maintain the abundance of necessary Treg-cell populations, which may accelerate the development of tolerance to Der p 2.
Overall, we have demonstrated that a2-3 sialic acid-modified rDer p 2 is able to suppress the activation of both T H 1 and T H 2 cells and downregulate the production of T H 2 cytokines IL-5 and IL-13 in PBMCs from nonallergic individuals.Moreover, sia-rDer p 2, unlike rDer p 2, does not enhance the activation and proliferation of CD4 1 T cells, does not alter the Treg/T H 2 cytokine balance, and does not alter the frequency of Treg cells.All these findings illustrate that the use of sia-rDer p 2 instead of rDer p 2 for Der p 2-specific AIT would potentially be more beneficial especially in the early phase of AIT, which is often characterized by a heightened T H 2 response and corresponding allergic side effects, although this needs to be confirmed using cells from subjects with established HDM allergy.Further studies, especially in cells from HDM-allergic donors, are needed to broaden the knowledge of the present findings and to confirm their possible use in routine clinical practice.

DISCLOSURE STATEMENT
This study was funded by HEALTH HOLLAND (HH LSHM19073) and DC4U Technologies.
Disclosure of potential conflict of interest: E. R. J. Li and Y. van Kooyk are involved in DC4U Technologies, which develops glycan-based technologies that enable steering the human immune response.R. van Ree receives consultancy fees from HAL Allergy BV, Citeq BV, Angany, Inc, Reacta Healthcare Ltd, AB Enzymes, Mission MightyMe, and The Protein Brewery; receives speaker fees from HAL Allergy BV, ALK, and Thermo Fisher Scientific; and possesses stock options at Angany, Inc.The rest of the authors declare that they have no relevant conflicts of interest.

FIG 1 .
FIG 1. sia-rDer p 2 binds to Siglec-9 and/or Siglec-10 expressed on APCs.A, Schematic representation of the thiol-malemide reaction for the conjugation of a2-3 sialic acids to rDer p 2. B, Representative histograms showing the binding of rDer p 2 and sia-rDer p 2 on classical monocytes (CD14 1 CD16 1 ), CD11c 1 DCs, and B cells.C, The percentage of cells among monocytes, CD11c 1 DCs, and B cells that bound polyacrylamide-a2-3 (positive control [PC]), rDer p 2, and sia-rDer p 2 (n 5 3).Data are shown as mean 6 SD.D, Representative histograms showing the expression of Siglec-9 on classical monocytes after incubation of cells with either rDer p 2 or sia-rDer p 2.
).To investigate whether sia-rDer p 2 would bind to Siglecs expressed on immune cells, we first measured the expression of Siglec-9 and Siglec-10 on the different immune cells within PBMCs (gating strategy; FigE1).Monocytes and DCs but not B cells expressed high levels of Siglec-9, whereas CD14 2 CD16 1

TABLE I .
mAbs used for flow cytometry, including clone, manufacturer, and catalog number FITC, Fluorescein isothiocyanate; MCBI, Molecular Cell Biology and Immunology; NA, not applicable; PE, phycoerythrin.