Physiologic Concentrations of HMGB1 Have No Impact on Cytokine-Mediated Eosinophil Survival or Chemotaxis in Response to Eotaxin-2 (CCL24)

HMGB1 is an alarmin that can stimulate the innate immune system alone or in a complex with other inflammatory mediators. Given the recent interest in HMGB1 with respect to the pathogenesis of eosinophil-associated disorders, including asthmatic inflammation and chronic rhinosinusitis, we have explored the role of this mediator and in promoting eosinophil activation. HMGB1 receptors RAGE and TLR4 but not TLR2 were detected on freshly isolated human eosinophils from healthy donors. Physiologic and relevant pathophysiologic levels of biologically-active HMGB1 had no effect on survival of human eosinophils alone or in combination with pro-survival cytokines IL-5, IL-3, or GM-CSF, and increasing concentrations of HMGB1 had no impact on surface expression of RAGE, TLR2 or TLR4. Similarly, HMGB1 did not elicit chemotaxis of human eosinophils alone and had no effect in combination with the eosinophil chemotactic agent, eotaxin-2 (CCL24). However, surface expression of TLR2 and TLR4 increased in response to cell stress, notably on eosinophils that remain viable after 48 hours without IL-5. As such, HMGB1 signaling on eosinophils may be substantially more detailed, and may involve complex immunostimulatory pathways other than or in addition to those evaluated here.


Introduction
High mobility group box 1 protein (HMGB1) is an evolutionarily ancient protein that was originally characterized as a chromatin stabilizing nuclear DNA-binding protein. Wang and colleagues [1] were the first to identify an extracellular role for HMGB1, specifically its participation in cellular activation and pro-inflammatory responses (reviewed in [2][3][4][5][6][7][8][9][10]). Containing 215 amino acids comprising three distinct protein domains, HMGB1 is expressed ubiquitously, is released from dead and dying cells, and serves as an alarmin or damage-associated molecular pattern (DAMP) molecule, stimulating the innate immune system by itself or via immunostimulatory complexes with endotoxin, nucleic acids, or proinflammatory cytokines and chemokines [3,11]. Additionally, activated immune cells (macrophages, monocytes, dendritic cells and natural killer cells) and endothelial cells secrete HMGB1 in response to pro-inflammatory stimuli [2]. HMGB1 signals through multiple surface receptors; TLR2, TLR4, and RAGE, receptor for advanced glycation end product, are the best characterized [2,12] but HMGB1 can also signal through human CD24 / Siglec-10 [13]. A series of recent studies has revealed a role for HMGB1 in sensing and responding to exogenous and endogenous nucleic acids (doublestranded RNA, single-stranded RNA, CpG-containing oligodeoxynucleotides) and amplifying the responses of these ligands to pattern recognition receptors TLR3, TLR7, and TLR9 (reviewed in [9]). Interestingly, while HMGB1 gene-deleted mice die in infancy [14], mice with conditional ablation of HMGB1 in myeloid cells develop normally, although they are more sensitive to endotoxin shock compared with control mice [15].
There is substantial interest in HMGB1 signaling and inflammation associated acute and chronic disease, notably in diseases associated with eosinophilic inflammation [7,16]. Elevated levels of HMGB1 have been detected in sputum, plasma and nasal lavage of eosinophilic asthmatics as compared to normal controls, with levels of HMGB1 correlating with both sputum levels of IL-5, IL-13, and eosinophil counts [17][18][19]. Additionally, there is a negative correlation between HMGB1 levels and pulmonary function [20]. Similarly, HMGB1 has been implicated in the pathogenesis of chronic rhinosinusitis, an asthma co-morbidity characterized by eosinophils in nasal polyps and in mucous drainage [21]. Expression of HMGB1 was detected in paranasal sinus mucosae of individuals with this condition [22,23], with levels correlating directly with those of serum IL-5 and blood eosinophil counts [24]. HMGB1 expression is ubiquitous and serum levels in normal individuals are on the order of <5-30 ng/ml but can rise 3-fold or more under conditions associated with eosinophil activation and recruitment ( Table 1). Lotfi et al. [25] demonstrated that human eosinophils are mobilized and activated in response to supraphysiologic concentrations of HMGB1 (10 4 ng/ml) and proposed a role for this DAMP in the induction of eosinophilic inflammation. At these supraphysiologic concentrations, which are similar to those described by Ueno et al. [26] in lung epithelial lining cells responding to acute lung injury in response to sepsis, HMGB1 promoted survival, chemotaxis and degranulation of eosinophils. We were interested in expanding the findings of Lotfi et al. [25] and in exploring the effects of HMGB1 on human eosinophils at concentrations similar to those identified in acute and chronic states directly associated with eosinophilic disease. Additionally, we wanted to explore the role of HMGB1 in combination with eosinophil active cytokines as it is known that HMGB1 can form immunostimulatory complexes with other soluble mediators. Here, we examine the expression of HMGB1 receptors on human eosinophils isolated from peripheral blood and explore the impact of HMGB1 in promoting eosinophil chemotaxis and survival.

Isolation of eosinophils from normal donors
One hundred milliliter (100 mL) samples of peripheral blood from normal donors were obtained on the Laboratory of Allergic Diseases normal donor protocol (NIAID Protocol Number: NIH 09-I-0049). This protocol is designed to assure adequate and complete informed consent, counseling, and protection of the study subjects according the NIAID Institutional Review Board, Office of Human Subjects Research (OHSR), Office for Human Research Protections (OHRP) and other applicable Federal regulatory standards. As part of this protocol, written consent is obtained from each normal volunteer allowing their sample to be used in this study. Eosinophils were isolated from the blood according to the protocol of Percopo et al. [32]. Essentially granulocytes were separated on a 67% Percoll gradient (GE #17-0891-02) and eosinophils were isolated by negative selection using the Miltenyi eosinophil isolation kit as per manufacturer's directions. The eosinophils were either used immediately or cultured for 48 hours in the presence or absence of the anti-apoptotic cytokine, IL-5 (100 pg/mL) with or without HMGB1 as indicated. The monocytic cell fraction from the Percoll gradient was collected and used to determine the titer of the anti-human TLR2 (CD282) and anti-human TLR4 (CD284) antibodies. Additionally, the granulocytes, prior to isolation of eosinophils, were fixed in 4% paraformaldehyde (EMS, Hatfield, PA) and used to determine the titer of the antihuman RAGE antibody.

Flow cytometry
Non-specific antibody binding was blocked by incubating human eosinophils with FcR blocking reagent (Miltenyi). Antibody directed against the indicated HMGB1 receptor-RAGE, TLR2 or TLR4 is shown in Table 2. After incubation with antibodies for 30 minutes at 4°C, the cells were washed with 3 ml of 0.1% BSA/PBS and then fixed in 4% paraformaldehyde. The samples were stored at 4°C in the dark until analyzed. A minimum of 100,000 events was collected on an LSRII flow cytometer (BD Biosciences) and data was analyzed in FlowJo (Tree Star, Inc). Gating was performed by comparison to isotype control for each antibody. Isotype controls (fluorescence minus one) were subtracted from each data point.

HMGB1 activity assay
This assay was modified from that described by Gero et al. [33] RAW264.7 (ATCC, TIB71) were plated at 100,000 cells per well in a 96-well plate in 200 μL of media containing RPMI1640 (Life Technologies), 10% FBS (Lonza), 100 IU/mL penicillin, and 10 μg/mL streptomycin (Life Technologies) and grown overnight at 37°C in a cell culture incubator. The culture media was removed and 100 μL Optimem (Life Technologies) containing 0-100 ng/mL HMGB1 (R&D Systems) was added. Lipopolysaccharide (LPS; Sigma) was used at 50 ng/mL as a positive control. After 7 hours, the plate was centrifuged at 300 x g for 5 minutes at room temperature to remove the cells and the cell free supernatant was stored at −80°C. The samples were diluted 1:10 in Optimem prior to determining TNFα concentrations by ELISA (R&D) according the manufacturer's recommendations. We determined that recombinant HMGB1 used in our studies was biologically active, with 100 ng/mL eliciting 177 pg/mL TNF-alpha, *3-fold over background levels at t = 7 hrs.

Chemotaxis Assay
The chemotaxis assay was performed as previously described [34]. Briefly, eosinophils were suspended at 10 6 /mL in chemotaxis buffer and 100 μL was placed in the top well of a 5-μm 96-well Transwell plate (Corning). The 100 μL of the chemoattractant, diluted in chemotaxis buffer, was placed in the bottom well. The plate was incubated for 3 hours at 37°C in a humidified CO 2 incubator. Cells that migrate to the lower well through the membrane are collected and enumerated by collecting events on LSR II at high flow rate for 60 seconds. Data is reported as cell number or chemotactic index (CI = cells migrating in response to chemoattractant / cells migrating in response to vehicle). Recombinant human eotaxin (rhE2, R&D) was used as a positive control for eosinophil chemotaxis.

Expression of HMGB1 receptors on eosinophils
Human eosinophils were isolated from the peripheral blood of normal donors and surface expression of RAGE, TLR2 and TLR4 was examined. RAGE expression was variable among normal donors and on average 30% of the eosinophils isolated from normal human donors express RAGE (range 11-69%, n = 6 donors, Fig. 1A and 1B). The expression of pattern recognition receptors on eosinophils has recently been reviewed [4]. Our results are consistent with those of others who have shown documented expression of RAGE on human eosinophils [25,35]; however, this is the first study documenting the heterogeneous nature of RAGE expression on eosinophils isolated from multiple normal donors. TLR4 was detected on a small population (0.3-5%) of eosinophils from normal donors (Fig. 1C). Our findings are consistent with those of Wong et al. [36] who detected TLR4 by both Western blot and flow cytometry. In contrast, we did not detect TLR2 on freshly isolated normal human eosinophils. Our findings are consistent with those of Nagase et al. [37] and Sabroe et al. [38] who were unable to detect surface expression of TLR2 by flow cytometry. Interestingly Wong et al. [36] reported detection of TLR2 on and in human eosinophils with intracellular expression being higher then surface expression. Likewise Wong et al. [36] reported that peptidoglycan (PGN), a TLR2 ligand, signaled through this receptor; in contrast, even though TLR4 was present, its ligand LPS did not elicit a response. Similarly, Sabroe et al. [26] failed to elicit a response from eosinophils with LPS stimulation.
Eosinophils were isolated from peripheral blood of normal donors and cultured for 48 hrs either in the presence or absence of IL-5; living cells (87.9 ± 2.0% of total in response to 100 pg/mL IL-5 culture and 42.1 ± 5.9% in the absence of IL-5) were evaluated for the   expression of RAGE, TLR4, and TLR2. At the same time, both cultures were treated with increasing concentrations of HMGB1 (from 0 to 100 ng/mL). As shown, HMGB1 had no independent effect on the cell surface expression of RAGE, TLR4 or TLR2. In contrast, IL-5 had a clear and differential impact on receptor expression. While IL-5 had little to no effect on the surface expression of RAGE on live eosinophils ( Fig. 2A), this was not the case for surface expression of TLR4 (Fig. 2B) or TLR2 (Fig. 2C). Detection was minimal (TLR2) to non-existent (TLR4) after 48 hrs in culture with 100 pg / mL IL-5. In contrast, immunoreactive TLR2 and TLR4 was detected on *20% and *10% of the live cells, respectively, after 48 hrs in cultures devoid of IL-5. This finding is intriguing and requires further study.

Recombinant human HMGB1 does not enhance viability or chemotaxis of isolated human eosinophils
As noted in the Introduction, very high concentrations (10 4 ng/mL), were utilized in an earlier publication documenting HMGB1-mediated eosinophil activation [25]. This concentration exceeds physiologic serum levels (typically measured at 0-30 ng/mL), and likewise exceeds levels measured in serum, plasma and tissue fluids of individuals with eosinophil-associated disorders ( Table 1).
Likewise, as shown in Fig. 4A, HMGB1 at 10 ng/ml elicited no chemotaxis of human eosinophils and had no impact on chemotaxis elicited by the eosinophil chemotactic agent, eotaxin-2 (CCL24). Eosinophils incubated for 30 minutes with HMGB1 prior to introduction of an eotaxin-2 gradient migrated to the same extent as did the eosinophils pre-incubated with buffer alone. Furthermore, varying concentrations of HMGB1, including supraphysiologic concentrations (1-5000 ng/ml) had no impact on chemotaxis elicited by 100 nM eotaxin (Fig. 4B). Interestingly, this is in direct contrast to the inhibitory effect of HMGB1 observed when measuring neutrophil chemotaxis mediated via RAGE both at baseline at moderate concentrations (50-100 ng/mL) and in response to these concentrations in the presence of gradients of neutrophil-specific chemoattractants fMLP-and IL-8 [40].

Summary
At physiologic and pathophysiologic concentrations related to eosinophil-associated disease, recombinant HMGB1 has no impact on survival of freshly isolated human eosinophils, or on the expression of its receptors, RAGE, TLR2 or TLR4, either alone or in conjunction with IL-3, IL-5, or GM-CSF, in experiments performed ex vivo, although a greater fraction of viable cells express TLR2 and TLR4 in cultures undergoing pro-apoptotic stress. Additionally, HMGB1 has no apparent impact on eotaxin 2-mediated chemotaxis, in direct contrast to what has been reported for HMGB1 and human neutrophils [40]. HMGB1 also did not have an impact on cytokine-mediated survival over a wide range of concentrations when used in combination with known eosinophil-active cytokines. From our findings, we must consider the possibility that HMGB1 may be generating complexes in vivo with other cytokines or with nucleic acid ligands that modulate chemotaxis and prolong survival and/or that may have an impact on eosinophil biology via outcomes other than those explored here.

Author Contributions
Conceived and designed the experiments: KDD HFR. Performed the experiments: KDD. Analyzed the data: KDD HFR. Contributed reagents/materials/analysis tools: KDD HFR. Wrote the paper: KDD HFR.