Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 Jan 1.
Published in final edited form as: J Allergy Clin Immunol. 2013 Jul 2;133(1):10.1016/j.jaci.2013.05.018. doi: 10.1016/j.jaci.2013.05.018

Mice deficient in the St3gal3 gene product α2,3 sialyltransferase (ST3Gal-III) exhibit enhanced allergic eosinophilic airway inflammation

Takumi Kiwamoto a, Mary E Brummet a, Fan Wu a, Mary G Motari b, David F Smith c, Ronald L Schnaar b, Zhou Zhu a, Bruce S Bochner a
PMCID: PMC3874253  NIHMSID: NIHMS502478  PMID: 23830412

Abstract

Background

Sialic acid–binding immunoglobulin-like lectin (Siglec)-F is a proapoptotic receptor on mouse eosinophils, but little is known about its natural tissue ligand.

Objective

We previously reported that the St3gal3 gene product α2,3 sialyltransferase (ST3Gal-III) is required for constitutive Siglec-F lung ligand synthesis. We therefore hypothesized that attenuation of ST3Gal-III will decrease Siglec-F ligand levels and enhance allergic eosinophilic airway inflammation.

Methods

C57BL/6 wild-type mice and St3gal3 heterozygous or homozygous deficient (St3gal3+/− and St3gal3−/−) mice were used. Eosinophilic airway inflammation was induced through sensitization to ovalbumin (OVA) and repeated airway OVA challenge. Siglec-F human IgG1 fusion protein (Siglec-F-Fc) was used to detect Siglec-F ligands. Lung tissue and bronchoalveolar lavage fluid (BALF) were analyzed for inflammation, as well as various cytokines and chemokines. Serum was analyzed for allergen-specific immunoglobulin levels.

Results

Western blotting with Siglec-F-Fc detected approximately 500-kDa and approximately 200-kDa candidate Siglec-F ligands that were less abundant in St3gal3+/− lung extracts and nearly absent in St3gal3−/− lung extracts. After OVA sensitization and challenge, Siglec-F ligands were increased in wild-type mouse lungs but less so in St3gal3 mutants, whereas peribronchial and BALF eosinophil numbers were greater in the mutants, with the following rank order: St3gal3−/− ≥ St3gal3+/− > wild-type mice. Levels of various cytokines and chemokines in BALF were not significantly different among these 3 types of mice, although OVA-specific serum IgG1 levels were increased in St3gal3−/− mice.

Conclusions

After OVA sensitization and challenge, St3gal3+/− and St3gal3−/− mice have more intense allergic eosinophilic airway inflammation and less sialylated Siglec-F ligands in their airways. One possible explanation for these findings is that levels of sialylated airway ligands for Siglec-F might be diminished in mice with attenuated levels of ST3Gal-III, resulting in a reduction in a natural proapoptotic pathway for controlling airway eosinophilia.

Keywords: Eosinophils, asthma, Siglec-F, 6′-sulfated sialyl Lewis X, 6′-sulfated sialyl N-acetyl-D-lactosamine, apoptosis, glycan ligands, lung, St3gal3

Sialic acid–binding immunoglobulin-like lectins (Siglecs) are a family of single-pass, transmembrane, cell-surface receptors found predominantly on leukocytes.1-3 Among them, Siglec-F in mice and Siglec-8 in human subjects are functional paralog inhibitory receptors expressed by eosinophils.4 Because engagement of Siglec-F or Siglec-8 induces eosinophil apoptosis,5-10 targeting these molecules might help to fill an unmet need in the treatment of various eosinophilic diseases.11-13

Previous studies have reported that both Siglec-F and Siglec-8 uniquely and preferentially recognize the same sialoside ligand 6′-sulfo sialyl Lewis X (6′-su-sLex; NeuAcα2-3[6-SO3]Galβ1-4 [Fucα1-3]GlcNAc).14,15 For Siglec-F, constitutive and cytokine-inducible lung epithelial ligands have been detected.9,16-18 These ligands contain α2,3-linked terminal sialic acid residues and in mouse lung require the St3gal3 gene product α2,3 sialyltransferase (ST3Gal-III) for their constitutive synthesis.17,18 We therefore hypothesized that attenuation of ST3Gal-III will decrease Siglec-F airway ligand levels and selectively enhance eosinophilic airway inflammation. Using St3gal3 heterozygous and homozygous deficient (St3gal3+/− and St3gal3−/−) mice, we found that they exhibit a more intense allergic eosinophilic airway inflammatory reaction and produce less sialylated high molecular weight Siglec-F ligands in their airways.

Methods

Glycan array analysis of Siglec-F and Siglec-8 ligand binding

The sugar-binding specificities of Siglec-F and Siglec-8 were reanalyzed on the printed glycan microarray, which was developed by the Consortium for Functional Glycomics. Previously reported analyses were carried out with an earlier array format, the microtiter plate array (Version 2.0), which only had 184 biotinylated glycosides and controls immobilized in streptavidin-coated micro titer plates.14,15 Version 4.1 of the printed glycan microarray was composed of 465 defined, amino-functionalized glycan structures printed as replicates of 6 on an NHS-activated microscope slide. Siglec-F human IgG1 Fc chimera (Siglec-F-Fc; R&D Systems, Minneapolis, Minn) or Siglec-8 human IgG1 Fc chimera (Siglec-8-Fc)19 were applied in 70 μL of Tris-buffered saline at several different concentrations to the microarray under a cover slip and incubated at room temperature in a humidified chamber. After 1 hour, the cover slip was carefully removed, and the slides were washed in buffer to remove excess protein. The protein-glycan complexes were detected with an Alexa Fluor 488–conjugated goat anti-human IgG (Invitrogen, Carlsbad, Calif) by using the same procedure. After removing the cover slip, the slides were washed, air dried, and scanned with a PerkinElmer ProScanArray fluorescence scanner (PerkinElmer, Waltham, Mass) set at an excitation of 485 nm and an emission of 535 nm, and the data were processed with Imagene Software (BioDiscovery, El Segundo, Calif). After removing the high and low measurements of each set of replicates, the averages and SDs of the remaining 4 values were used to determine the average relative fluorescence units. Data are reported as a histogram of relative fluorescence units versus glycan structure.

Animals, ovalbumin sensitization, and airway challenge

Mice used in these experiments were 8 to 10 weeks old and included C57BL/6 mice (wild-type [WT]) and St3gal3 heterozygous deficient (St3gal3+/−) and homozygous deficient (St3gal3−/−) mice. Breeding sets of St3gal3−/− mice were kindly provided by Dr Jamey Marth (University of California, San Diego, Calif), as previously described, but because of breeding limitations of these mutant mice, all St3gal3−/− mice used in these studies came from the approximately 25% of St3gal3+/− × St3gal3+/− mice off-spring that were homozygous null.17,20 Mice were intraperitoneally sensitized with 500 μg of ovalbumin (OVA; Sigma-Aldrich, St Louis, Mo) in 1 mg of alum on days 1 and 8 and then challenged intranasally with 200 μg of OVA on days 17, 19, and 21 to induce allergic airway inflammation. All mice were killed 24 hours after final challenge. Control mice were injected and challenged with PBS. All procedures performed on mice were in accordance with the National Institutes of Health's guidelines for humane treatment and approved by the Johns Hopkins University Institutional Animal Use and Care Committees.

Siglec-F ligand expression

Expression levels of Siglec-F binding protein in lung-derived samples were determined by means of Western blotting with a 4% to 15% polyacrylamide gel (Bio-Rad Laboratories, Hercules, Calif). Protein extracted from lung tissue homogenates was generated as previously described.17 After electrophoresis, proteins were transferred to polyvinylidene difluoride membranes (Bio-Rad Laboratories) and incubated overnight with 4% BSA (Sigma-Aldrich). The membranes were incubated with Siglec-F-Fc (0.5 mg/mL, 1 hour, room temperature; R&D Systems). After washing, the membranes were incubated with horseradish peroxidase–linked polyclonal sheep anti-human IgG polyclonal antibody (45 minutes, room temperature; GE Healthcare, Piscataway, NJ). Bands were visualized with the ECL Western blotting detection system (GE Healthcare UK Ltd, Buckinghamshire, United Kingdom). Some of these samples were pretreated with sialidase (Clostridium perfringens, 1.6 mU/mL, 2 hours, 37°C; Sigma-Aldrich) to confirm sialidase sensitivity.

Tissue distribution of potential Siglec-F ligands was studied by means of immunohistochemistry, as described previously.17 Briefly, the 10-μm cryostat sections of lung and tracheal tissues were incubated with Siglec-F-Fc (1 μg/mL, 1 hour, 37°C) and detected by using an alkaline phosphatase imaging method. Isotype-matched humanized IgG1 mAb (omalizumab, 1 μg/mL, 1 hour, 37°C; Genentech, South San Francisco, Calif) was used as a negative control. In some experiments samples were pretreated with sialidase (10 mU/mL, 24 hours, 37°C; Sigma-Aldrich).

Bronchoalveolar lavage and peripheral blood cell counting

Mouse lungs were lavaged by using 5 repeated instillations of 0.6 mL of saline each through the tracheal cannula. The first 1.2 mL of bronchoalveolar lavage fluid (BALF) was centrifuged, and the supernatant was used for the measurements of various cytokines and chemokines. Cells were counted with a hemocytometer. Differential cell counts in BALF and blood sample smears were performed under standard light microscopy after Diff-Quik staining.

Peribronchial eosinophil enumeration

Major basic protein (MBP) immunohistochemistry was performed on 5-μm paraffin sections, as previously described.9 Rat anti-mouse MBP mAb was obtained from the laboratory of Dr James J. Lee (Mayo Clinic, Scottsdale, Ariz). Numbers of MBP+ cells in the peribronchial area were counted by a blinded observer to quantitate the severity of eosinophilic airway inflammation, as previously described.16 The number of cells staining positive was counted in 10 randomly selected peribronchial regions to a depth of 30 μm beneath the epithelium. Results are expressed as the number of cells staining positive for MBP per 150- to 200-μm-sized internal diameter bronchiole.

Measurement of immunoglobulin levels

Serum total IgE (1:100 dilution; detection range, 0.156-20 μg/mL) and OVA-specific IgE (1.56-200 ng/mL), IgG1 (1:20,000 dilution, 31.3-4000 μg/mL), and IgG2a (1:100 dilution, 0.156-20 μg/mL) levels were determined by using commercial ELISA kits, per the manufacturer's instructions (R&D Systems).

Measurement of cytokines and chemokines

IL-3 (0.2-21,632 pg/mL), IL-4 (2.1-9,372 pg/mL), IL-5 (0.3-13,315 pg/mL), IL-10 (1.0-12,066 pg/mL), IL-17A (0.8-43,337 pg/mL), GM-CSF (5.6-3,401 pg/mL), IFN-γ (1.2-30,164 pg/mL), and TNF-α (1.4-59,626 pg/mL) concentrations in BALF were determined with a Bio-Plex kit (Bio-Rad Laboratories), according to the manufacturer's instructions. IL-13 (1.5-500 pg/mL), IL-33 (14.3-2000 pg/mL), eotaxin (CCL11; 3-1000 pg/mL), and eotaxin-2 (CCL24; 1:4 dilution, 0.0156-12 ng/mL) concentrations were determined by using commercial ELISA kits, according to the manufacturer's instructions (R&D Systems).

Statistical analysis

Data are shown as means ± SEMs or individual dot plots with medians. Statistical significance between groups was evaluated by means of ANOVA and the Tukey multiple comparison test. P values of less than .05 were considered statistically significant.

Results

Siglec-F recognizes 6′-su-sLex and a closely related nonfucosylated structure, 6′-sulfated sialyl N-acetyl-D-lactosamine

Previous studies have shown that both Siglec-8 and Siglec-F uniquely and specificallyrecognize the same glycan structure (6′-su-sLex), but only 184 glycans wereavailable as part of this original screening process.14,15 Because of the availability of an expanded panel of glycans subsequently made available through the Consortium for Functional Glycomics (Version 4.1, http://www.functionalglycomics.org), repeat screening was initiated. By using Siglec-F-Fc and Siglec-8-Fc as probes, 465 carbohydrate-based structures were tested for specific Siglec-F and Siglec-8 binding. As expected, both Siglec-F-Fc and Siglec-8-Fc once again had unique affinity for 6′-su-sLex, but both also recognized its nonfucosylated form, 6′-sulfated sialyl N-acetyl-D-lactosamine (6′-su-sLacNAc; not present on the glycan panel Version 2.0 when previously tested) both of which contain an α2,3-linked terminal sialic acid on a galactose that also carries a 6-sulfate group (Fig 1). Twenty other ligands with terminal NeuAc α2,3 linked to Gal did not bind either Siglec, including closely related glycans, such as 6-su-sLex and sLex.14, 15 These results support the hypothesis that unique α2,3-linked sialic acid–containing airway glycans recognize Siglec-8 and Siglec-F and that sialyltransferases capable of creating terminal α2,3-linked sialic acid residues will be required for natural Siglec-F and Siglec-8 ligand synthesis.18

Fig 1.

Fig 1

Open in a new tab

A, Glycan array screening with Version 4.2 of the Consortium for Functional Glycomics glycan array (http://www.functionalglycomics.org) reveals 2 shared ligands for Siglec-F-Fc (1 μg/mL) and Siglec-8-Fc (200 μg/mL) among 465 glycans tested, namely 6′-su-sLex and 6′-su-sLacNAc. Values represent means ± SDs of replicates from a single assay. Structures are displayed with nomenclature used in the Essentials of Glycobiology textbook, second edition (http://www.ncbi.nlm.nih.gov/books/NBK1908/). Diamonds, N-acetylneuraminic acid (sialic acid); circles, galactose; squares, N-acetylglucosamine; triangles, fucose. B, Chemical structures of the 2 shared ligands, 6′-su-sLex and 6′-su-sLacNAc, that differ only with respect to the presence or absence, respectively, of the fucose residue.

Attenuation of ST3Gal-III decreases levels of sialic acid containing Siglec-F ligands in mouse airways

Western blotting of whole-lung extracts was performed to explore possible endogenous lung ligands for Siglec-F. By using Siglec-F-Fc as a probe, approximately 500-kDa and approximately 200-kDa sialidase-sensitive bands were detected in resting whole-lung extracts (Fig 2, A). These bands were less abundant in St3gal3+/− lung extracts and almost absent in St3gal3−/− lung extracts under nonsensitized and challenged condition. Levels of the approximately 200-kDa candidate ligands increased after OVA sensitization and challenge in WT mouse whole-lung extracts (Fig 2, B) and remained reduced or absent in OVA-sensitized and OVA-challenged St3gal3+/− and St3gal3−/− lung extracts. However, levels of the approximately 500-kDa candidate ligands were increased after OVA sensitization and challenge in WT mouse lung extracts, as well as in St3gal3+/− and St3gal3−/− lung extracts (Fig 2, B). In summary, data in Fig 2, A, show (1) the presence of at least 2 sialidase-sensitive ligands from resting mouse lung, one approximately 200 kDa and another broader band at approximately 500 kDa; (2) reduction but not elimination of the approximately 200-kDa material in the St3gal3+/− and St3gal3−/− mice; and (3) reduction of the approximately 500-kDa material in the St3gal3+/− mice and its elimination in the St3gal3−/− mice. Data in Fig 2, B, show (1) an increase in the approximately 500-kDa Siglec-F ligand in whole-lung extracts after OVA sensitization and challenge compared with PBS control, even in the St3gal3+/− and St3gal3−/− mice (and based on Fig 2, C, data; this material might be coming from airway glandular sources) and (2) an increase in the approximately 200-kDa Siglec-F ligand in whole-lung extracts after OVA sensitization and challenge compared with PBS control in the WT mice but less so in the St3gal3+/− and St3gal3−/− mice.

Fig 2.

Fig 2

Open in a new tab

St3gal3+/− and St3gal3−/− mice demonstrate decreased Siglec-F ligand expression in the lung and airway. A, Constitutive ligand characterization in WT, St3gal3+/−, and St3gal3−/− mouse whole-lung paren-chymal extracts with or without sialidase pretreatment probed by means of Western blotting with Siglec-F-Fc and for β-actin as a positive control. For each lane, 150 μg of lung extract protein was loaded. Data are representative of 3 experiments with similar results. B, Western blotting of mouse whole-lung parenchymal extracts in control and OVA-sensitized and challenged mice probed with Siglec-F-Fc or β-actin. Data are representative of 3 experiments with similar results. C, Representative detection and localization of Siglec-F ligand by using Siglec-F-Fc in immunohistochemistry from lung tissues of normal WT, St3gal3+/−, and St3gal3−/− mice in tracheal epithelium (panels a-f) and submucosal tracheal glands (panels g-l). All samples were obtained 24 hours after final challenge with PBS or OVA. Data are representative of 3 experiments with similar results.

Immunohistochemistry was performed on tracheal and bronchial tissue sections using Siglec-F-Fc to determine the localization of endogenous Siglec-F tissue ligands in the airway and effects of ST3Gal-III attenuation (Fig 2, C). As previous studies reported,9,16,17 expression levels of Siglec-F ligands were constitutive (Fig 2, panel a) and increased in airway epithelium in OVA-challenged WT mice compared with that of PBS-challenged mice (Fig 2, C; compare panels a vs d). In airway epithelial cells after OVA challenge, these inducible ligands were less abundant in St3gal3+/− mice (Fig 2, C; compare panels d vs e) and absent in St3gal3−/− mice (Fig 2, C; compare panels d and f). In addition, ligands were studied and detected for the first time in tracheal submucosal glands, and there were no obvious differences in Siglec-F ligand expression before or after OVA challenge between WT mice (Fig 2, C; panels g and j), St3gal3+/− mice (Fig 2, C; panels h and k), and St3gal3−/− mice (Fig 2, C; panels i and l). Further analysis of these tracheal submucosal ligands for Siglec-F revealed that they were completely sialidase sensitive (see Fig E1 in this article's Online Repository at www.jacionline.org). These results indicate that there are at least 2 sources of sialidase-sensitive lung ligands for Siglec-F: one constitutively produced by airway epithelium that manifest ST3gal-III–dependent increases under allergic inflammatory conditions and a second one constitutively produced in tracheal submucosal glands in which production is ST3gal-III independent and is also increased under allergic inflammatory conditions.

St3gal3−/− mice show a selective enhancement in eosinophilic airway inflammation in the OVA-induced allergic lung inflammation model

OVA sensitization and repeated intranasal challenge was performed on WT, St3gal3+/− and St3gal3−/− mice to test the hypothesis that diminished ST3Gal-III–dependent Siglec-F ligand production will lead to enhancement of lung eosinophilia. As shown in Fig 3, A and B, all 3 types of mice had lung inflammation after OVA sensitization and challenge, with increases in BALF total leukocyte and eosinophil counts significantly greater in St3gal3−/− mice than in the other mice (Fig 3, A, panels a and e), as well as a higher percentage of eosinophils in BALF (see Fig E2, A, in this article's Online Repository at www.jacionline.org). However, when lung sections labeled with anti-MBP mAb (Fig 3, B) were quantified (Fig 3, C), a similar statistically significant increase in tissue eosinophil counts was seen for St3gal3+/− and St3gal3−/− mice compared with WT mice after OVA sensitization and challenge. There was a nonstatistically significant increase in blood eosinophilia in the St3gal3−/− mice that was unaffected by OVA sensitization and challenge (see Fig E2, B). These data demonstrate a preferential increase in peribronchial and luminal eosinophil counts for St3gal3+/− and St3gal3−/− mice compared with WT mice after OVA sensitization and challenge.

Fig 3.

Fig 3

Open in a new tab

Attenuation of St3gal3 is associated with enhanced OVA-induced allergic eosinophilic airway inflammation. A, BALF cytology of each indicated cell type using samples obtained 24 hours after PBS or OVA final challenge from WT, St3gal3+/−, and St3gal3−/− mice (n = 8-12). B, Representative lung tissue sections from WT (panels a and b), St3gal3+/− (panels c and d), and St3gal3−/− (panels e and f) mice obtained 24 hours after PBS (panels a, c, and e) or OVA (panels b, d, and f) final challenge analyzed by using immunohistochemistry after being labeled with anti-MBP mAb. Representative photomicrographs are provided for each group (magnification ×200). C, Peribronchial eosinophils (MBP+ cells) in lung tissue sections (like those in Fig 3, B) were enumerated. Values represent means ± SDs (n = 5-7). For Fig 3, A and C, horizontal bars indicate direct statistical comparisons between OVA-challenged WT, St3gal3+/−, or St3gal3−/− mice. NS, Not significant.

St3gal3−/− mice show increases in OVA-specific IgG1 levels in the OVA-induced allergic lung inflammation model

Given the enhanced eosinophilic inflammation mentioned above, it was important to determine whether the St3gal3+/− and St3gal3−/− mice displayed any other evidence of enhanced allergic responses. Serum total IgE levels were significantly increased in St3gal3−/− mice after OVA sensitization and challenge compared with those seen in PBS control animals (Fig 4, A). Although there was a higher mean level of total IgE in the St3gal3−/− mice, this change was not statistically significant compared with WT or St3gal3+/− OVA mice and was due to 1 high outlier in this group of mice (not shown). All 3 types of OVA-sensitized and challenged mice showed increases in serum OVA-specific IgE, IgG1, and IgG2a levels compared with PBS control animals (Fig 4, B-D). When levels were directly compared from OVA-sensitized and challenged WT, St3gal3+/−, and St3gal3−/− mice, a significant increase was only seen in OVA-specific IgG1 levels (Fig 4, C).

Fig 4.

Fig 4

Open in a new tab

Effect of attenuation of St3gal3 on OVA-induced serologic markers associatedwith allergic sensitization. Total IgE (A), OVA-specific IgE (B), OVA-specific IgG1 (C), and OVA-specific IgG2a (D) levels in serum obtained 24 hours after PBS or OVA final challenge (n = 9-13). *P < .05, **P < .01, and ***P < .001 relative to the WT-PBS control mice. Horizontal bars indicate direct statistical comparisons between OVA-challenged WT, St3gal3+/−, or St3gal3−/− mice. NS, Not significant.

Levels of various cytokines and chemokines in the BALF of St3gal3−/− mice are not different than those in WT mice in the OVA-induced allergic lung inflammation model

BALF was tested for a variety of potentially relevant mediators to test the hypothesis that the exaggerated lung eosinophilia in the St3gal3+/− and St3gal3−/− mice was dueto an altered allergic cytokine and chemokine response. As shown in Fig E3 in this article's Online Repository at www.jacionline.org, IL-3, IL-4, IL-10, IL-13, IL-17A, IL-33, eotaxin, eotaxin-2, GM-CSF, IFN-g, and TNF-α levels were not statistically different between OVA-sensitized and challenged WT, St3gal3+/−, and St3gal3−/− mouse BALF. Unexpectedly, IL-5 levels were higher in BALF of WT-OVA mice than OVA-challenged St3gal3−/− mice, even though eosinophil levels in BALF were higher in OVA-challenged St3gal3−/− mice. Overall, there were only 2 situations in which there was a significant increase in cytokine levels between the BALF of OVA- versus PBS-treated mice: (1) IL-5 levels were significantly increased in WT-OVA–challenged mice compared with WT-PBS control animals and (2) IL-13 levels were significantly increased in St3gal3+/− and St3gal3−/− OVA-challenged mice compared with WT-PBS control animals, even though IL-13 levels were not different between WT, St3gal3+/−, and St3gal3−/− OVA-challenged mice. Taken together, these data suggest that the increased lung eosinophilia is not readily explained by an increase in levels of any of the 10 mediators measured 24 hours after the last intranasal OVA challenge and that the increased eosinophilia seen in St3gal3+/− and St3gal3−/− OVA-challenged mice occurs along with an unexpected blunting of IL-5 production compared with that seen in WT mice.

Discussion

This article reports several novel findings. First, the data not only confirm the binding specificity of 6′-su-sLex for both Siglec-8 and Siglec-F14,15 but also demonstrate that the fucose residue in 6′-su-sLex is not required tomaintain this ligand-binding specificity (Fig 1). This is important because both Siglecs still recognize the same unique glycans, suggesting that it is relevant to use mouse models involving Siglec-F for the development of glycan-based therapeutics targeting Siglec-8. Second, the results confirm and extend the findings of Guo et al17 by supporting the critical role of St3gal3 in generating sialidase-sensitive mouse glycan ligands for Siglec-F on the lung epithelium. Interestingly, newly described submucosal ligands do not appear to require St3gal3 to generate ligands (Fig 2). By using Western blotting of whole-lung extracts, 2 candidate ligands of approximately 500 kDa and approximately 200 kDa have been identified, with complete attenuation of the approximately 500-kDa ligand seen in St3gal3−/− unchallenged mouse lungs and reduced, but not absent, levels of the approximately 200-kDa ligands in St3gal3−/− unchallenged mouse lungs (Fig 2, A). After challenge, increases in levels of the approximately 500-kDa ligand are seen regardless of St3gal3 attenuation, whereas levels of the approximately 200-kDa ligands remain reduced in St3gal3-attenuated mice. Although the exact nature of these 2 candidate ligands is still under investigation,18 it appears likely that the approximately 500-kDa ligand might be primarily glandular in origin.

With respect to the biological significance of Siglec-F ligands, this article is the first to demonstrate that a specific enzyme responsible for decorating glycans with sialic acid, namely ST3Gal-III, can influence lung eosinophilic inflammation. Increased lung eosinophilia in the St3gal3+/− and St3gal3−/− mice was selective among infiltrating leukocytes in both BALF and peribronchial tissues (Fig 3). Although the increase in BALF eosinophil counts was greatest in the St3gal3−/− mice, eosinophil counts were equally and significantly increased in peribronchial tissues in both the St3gal3+/− and St3gal3−/− mice (Fig 3, C). Even though Siglec-F is also expressed by alveolar macrophages,21 attenuation of St3gal3 had no effect on their numbers, which is consistent with a lack of biological effect seen through Siglec-F targeting on these cells compared with eosinophils.22 The selective lung eosinophilia was not associated with any detectable alteration in levels of a panel of potentially relevant BALF cytokines or chemokines (see Fig E3). One possible explanation is that Siglec-F regulation of lung eosinophilia is upstream of these proinflammatory cytokines in allergen-induced responses. The dissociation between the numbers of eosinophils (increased) and the levels of BALF IL-5 (decreased) in St3gal3−/− mice is intriguing and might result from the lack of eosinophil apoptosis in St3gal3−/− mice. Many cytokines contain amino acids that would permit N- or O-linked glycosylation, but little is known as to whether alterations in sialylation of cytokines or chemokines affect their biology or half-life. Although an increase in OVA-specific IgG1 levels was seen, there were no other changes in serologic markers of sensitization observed in the St3gal3−/− mice after sensitization and challenge (Fig 4). The reason for the increase in OVA-specific IgG1 levels is unknown but could be due to changes in the function or longevity of B cells or plasma cells as a result of an alteration of the glycans that are produced by the ST3Gal-III enzyme. Taken together, these data are consistent with previous reports that mice deficient in Siglec-F have selectively exaggerated eosinophilic inflammation9,16 and that targeting Siglec-F with antibodies selectively reduces eosinophilic inflammation.10,23-25

There are several potential shortcomings of the current work. Analysis of airways hyperreactivity was not performed, in part because it was not directly relevant to the hypotheses being tested and in part because prior studies involving alteration of Siglec-F did not show any significant effects on airways hyperreactiv ity.12,26 Despite the highly selective nature of the eosinophilic inflammation observed in St3gal3-attenuated mice, further work is needed to unequivocally prove that this is due to exclusively reduced amounts of sialylated airway glycan ligands acting through Siglec-F to induce eosinophil apoptosis. Indeed, it is possible that deletion of St3gal3 might have an effect on these mice in other ways. Future experiments are needed, for example, to determine whether the St3gal3 gene deletion has an effect directly on eosinophils per se and whether reconstitution of the ST3Gal-III enzyme into the airway abrogates the lung eosinophilia to conclude that eosinophilic inflammation is more intense because and only because of elimination of Siglec-F ligands in the lung. Whether these ligands contain 6′-su-sLex, 6′-su-sLacNAc, or both remains to be determined. Furthermore, the relative roles for the approximately 500-kDa versus the approximately 200-kDa ligands and the relative roles of gland-derived versus epithelium-derived ligands in mediating Siglec-F-dependent apoptosis are not yet known. Efforts are underway to isolate enough of each of these candidate ligands for glycomic and proteomic characterization. Nevertheless, the fact that St3gal3+/− and St3gal3−/− mice have more intense allergic eosinophilic airway inflammation and less sialylated Siglec-F ligands in their airways strongly suggest that reducing levels of airway glycan ligands for Siglec-F diminishes a natural proapoptotic pathway for controlling airway eosinophilia and that disorders associated with reductions in the function of ST3Gal-III in the lung might manifest exaggerated eosinophilic airway inflammation. These data also support the notion that specific Siglecs and their sialylated glycan ligands are important for controlling lung eosinophilia and could potentially be exploited for therapeutic benefit.

Supplementary Material

Fig E1. Localization of Siglec-F ligand in OVA-challenged mouse trachea (A) and lung (B). Fig E1, A, Sialidase-sensitive Siglec-F ligands were expressed in airway epithelium and submucosal glands (panels b-g). Fig E1, B, Siglec-F ligands were observed in bronchial epithelium in lung tissue (panels b-d).

Fig E2. Attenuation of St3gal3 enhances the percentages of eosinophils in OVA-induced allergic eosinophilic airway inflammation (A, n = 8-12) but has no effect on peripheral blood (PB) eosinophil numbers (B, n = 9-14) when examined 24 hours after PBS or OVA final challenge. *P < .05 and ***P < .001. NS, Not significant.

Fig E3. Effect of attenuation of St3gal3 on levels of various cytokines and chemokines in BALF. Levels of the indicated mediators (A-L, with median values also shown) were quantified in BALF obtained 24 hours after PBS or OVA final challenge (n = 6-15). *P < .05 and **P < .01 relative to WT-PBS control mice. Horizontal bars above each set of data indicate direct statistical comparisons between OVA-challenged WT, St3gal3+/−, or St3gal3−/− mice. Dashed red lines indicate the level of detection. NS, Not significant.

Key messages.

  • This is the first report of sialic acid biosynthesis affecting eosinophilic inflammation in that mice deficient in a single sialyltransferase that adds sialic acid to defined substrates, such as glycoproteins, have exaggerated allergic pulmonary eosinophilic responses.

  • This enhanced lung eosinophilia might be due to a change in the sialic acid content of airway glycoproteins, including those that would normally be present and have a proapoptotic effect on these infiltrating eosinophils through their cell-surface Siglec-F.

  • These data support the notion that specific Siglecs and their sialylated glycan ligands are important for controlling lung eosinophilia and could be exploited for therapeutic benefit.

Acknowledgments

Supported by grants AI72265 and HL107151 from the National Institutes of Health.

We thank Drs James Lee and Jamey Marth for providing critical reagents and mice and Dr Allen Myers for assistance with histologic analyses. We also thank the Consortium for Functional Glycomics funded by the NIGMS GM62116 and GM98791 for services provided by the Glycan Array Synthesis Core (Scripps Research Institute, La Jolla, Calif), which produced the mammalian glycan microarray and the Protein-Glycan Interaction Core (Emory University School of Medicine, Atlanta, Ga) that assisted with analysis of samples on the array.

Abbreviations

BALF

Bronchoalveolar lavage fluid

MBP

Major basic protein

OVA

Ovalbumin

Siglec

Sialic acid–binding immunoglobulin-like lectin

Siglec-8-Fc

Siglec-8 human IgG1 Fc chimera

Siglec-F-Fc

Siglec-F human IgG1 Fc chimera

ST3Gal-III

St3gal3 gene product α2,3 sialyltransferase

6′-su-sLacNAc

6′-Sulfated sialyl N-acetyl-D-lactosamine

6′-su-sLex

6′-Sulfated sialyl Lewis X

WT

Wild-type

Footnotes

Disclosure of potential conflict of interest: T. Kiwamoto, F. Wu, M. G. Motari, and Z. Zhu have received grants from the National Institutes of Health (NIH). D. F Smith has received a grant from Emory University School of Medicine. R. L. Schnaar has received grants and travel support from the NIH, has received royalties from Oxford University Press and Johns Hopkins, has stock/stock options in Zacharon Pharmaceuticals, and has received travel support from Academia Sinica and the University of Alberta. B. S. Bochner has received a grant from the NIH, has consultant arrangements with Allakos, has patents through Johns Hopkins University regarding Siglec-8 and its ligand, receives royalties from Johns Hopkins University, and has an equity position in Allakos. M. E. Brummet declares that she has no relevant conflicts of interest.

Note added in proof: during the review of a revised version of this manuscript, a related paper published online before print on May 15, 2013 (Suzukawa et al doi:10.4049/jimmunol.1203455).

References

  • 1.Crocker PR, Paulson JC, Varki A. Siglecs and their roles in the immune system. Nat Rev Immunol. 2007;7:255–66. doi: 10.1038/nri2056. [DOI] [PubMed] [Google Scholar]
  • 2.Varki A, Angata T. Siglecs—the major sub-family of I-type lectins. Glycobiology. 2006;16:1R–27R. doi: 10.1093/glycob/cwj008. [DOI] [PubMed] [Google Scholar]
  • 3.von Gunten S, Bochner BS. Basic and clinical immunology of Siglecs. Ann N Y Acad Sci. 2008;1143:61–82. doi: 10.1196/annals.1443.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Bochner BS. Siglec-8 on human eosinophils and mast cells, and Siglec-F on murine eosinophils, are functionally related inhibitory receptors. Clin Exp Allergy. 2009;39:317–24. doi: 10.1111/j.1365-2222.2008.03173.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Na HJ, Hudson SA, Bochner BS. IL-33 enhances Siglec-8 mediated apoptosis of human eosinophils. Cytokine. 2012;57:169–74. doi: 10.1016/j.cyto.2011.10.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Nutku E, Aizawa H, Hudson SA, Bochner BS. Ligation of Siglec-8: a selective mechanism for induction of human eosinophil apoptosis. Blood. 2003;101:5014–20. doi: 10.1182/blood-2002-10-3058. [DOI] [PubMed] [Google Scholar]
  • 7.Nutku E, Hudson SA, Bochner BS. Mechanism of Siglec-8-induced human eosinophil apoptosis: role of caspases and mitochondrial injury. Biochem Biophys Res Commun. 2005;336:918–24. doi: 10.1016/j.bbrc.2005.08.202. [DOI] [PubMed] [Google Scholar]
  • 8.Nutku-Bilir E, Hudson SA, Bochner BS. Interleukin-5 priming of human eosinophils alters Siglec-8 mediated apoptosis pathways. Am J Respir Cell Mol Biol. 2008;38:121–4. doi: 10.1165/rcmb.2007-0154OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Zhang M, Angata T, Cho JY, Miller M, Broide DH, Varki A. Defining the in vivo function of Siglec-F, a CD33-related Siglec expressed on mouse eosinophils. Blood. 2007;109:4280–7. doi: 10.1182/blood-2006-08-039255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Zimmermann N, McBride ML, Yamada Y, Hudson SA, Jones C, Cromie KD, et al. Siglec-F antibody administration to mice selectively reduces blood and tissue eosinophils. Allergy. 2008;63:1156–63. doi: 10.1111/j.1398-9995.2008.01709.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Bochner BS, Gleich GJ. What targeting eosinophils has taught us about their role in diseases. J Allergy Clin Immunol. 2010;126:16–25. doi: 10.1016/j.jaci.2010.02.026. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Kiwamoto T, Kawasaki N, Paulson JC, Bochner BS. Siglec-8 as a drugable target to treat eosinophil and mast cell-associated conditions. Pharmacol Ther. 2012;135:327–36. doi: 10.1016/j.pharmthera.2012.06.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Bochner BS, Book W, Busse WW, Butterfield J, Furuta GT, Gleich GJ, et al. Workshop report from the National Institutes of Health Taskforce on the Research Needs of Eosinophil-Associated Diseases (TREAD) J Allergy Clin Immunol. 2012;130:587–96. doi: 10.1016/j.jaci.2012.07.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Bochner BS, Alvarez RA, Mehta P, Bovin NV, Blixt O, White JR, et al. Glycan array screening reveals a candidate ligand for Siglec-8. J Biol Chem. 2005;280:4307–12. doi: 10.1074/jbc.M412378200. [DOI] [PubMed] [Google Scholar]
  • 15.Tateno H, Crocker PR, Paulson JC. Mouse Siglec-F and human Siglec-8 are functionally convergent paralogs that are selectively expressed on eosinophils and recognize 6′-sulfo-sialyl Lewis X as a preferred glycan ligand. Glycobiology. 2005;15:1125–35. doi: 10.1093/glycob/cwi097. [DOI] [PubMed] [Google Scholar]
  • 16.Cho JY, Song DJ, Pham A, Rosenthal P, Miller M, Dayan S, et al. Chronic OVA allergen challenged Siglec-F deficient mice have increased mucus, remodeling, and epithelial Siglec-F ligands which are up-regulated by IL-4 and IL-13. Respir Res. 2010;11:154. doi: 10.1186/1465-9921-11-154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Guo JP, Brummet ME, Myers AC, Na HJ, Rowland E, Schnaar RL, et al. Characterization of expression of glycan ligands for Siglec-F in normal mouse lungs. Am J Respir Cell Mol Biol. 2011;44:238–43. doi: 10.1165/rcmb.2010-0007OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Kiwamoto T, Katoh T, Tiemeyer M, Bochner BS. The role of lung epithelial ligands for Siglec-8 and Siglec-F in eosinophilic inflammation. Curr Opin Allergy Clin Immunol. 2013;13:106–11. doi: 10.1097/ACI.0b013e32835b594a. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Kikly KK, Bochner BS, Freeman S, Tan KB, Gallagher KT, D'Alessio K, et al. Identification of SAF-2, a novel siglec expressed on eosinophils, mast cells and basophils. J Allergy Clin Immunol. 2000;105:1093–100. doi: 10.1067/mai.2000.107127. [DOI] [PubMed] [Google Scholar]
  • 20.Ellies LG, Sperandio M, Underhill GH, Yousif J, Smith M, Priatel JJ, et al. Sialyltransferase specificity in selectin ligand formation. Blood. 2002;100:3618–25. doi: 10.1182/blood-2002-04-1007. [DOI] [PubMed] [Google Scholar]
  • 21.Stevens WW, Kim TS, Pujanauski LM, Hao X, Braciale TJ. Detection and quantitation of eosinophils in the murine respiratory tract by flow cytometry. J Immunol Methods. 2007;327:63–74. doi: 10.1016/j.jim.2007.07.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Feng YH, Mao H. Expression and preliminary functional analysis of Siglec-F on mouse macrophages. J Zhejiang Univ Sci B. 2012;13:386–94. doi: 10.1631/jzus.B1100218. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Rubinstein E, Cho JY, Rosenthal P, Chao J, Miller M, Pham A, et al. Siglec-F inhibition reduces esophageal eosinophilia and angiogenesis in a mouse model of eosinophilic esophagitis. J Pediatr Gastroenterol Nutr. 2011;53:409–16. doi: 10.1097/MPG.0b013e3182182ff8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Song DJ, Cho JY, Miller M, Strangman W, Zhang M, Varki A, et al. Anti-Siglec-F antibody inhibits oral egg allergen induced intestinal eosinophilic inflammation in a mouse model. Clin Immunol. 2009;131:157–69. doi: 10.1016/j.clim.2008.11.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Song DJ, Cho JY, Lee SY, Miller M, Rosenthal P, Soroosh P, et al. Anti-Siglec-F antibody reduces allergen-induced eosinophilic inflammation and airway remodeling. J Immunol. 2009;183:5333–41. doi: 10.4049/jimmunol.0801421. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Shin MK, Bochner BS. 15.4 Therapeutic approaches targeting siglecs. In: Rosenberg HF, Lee JJ, editors. Eosinophils in health and disease. Amsterdam: Elsevier; 2012. pp. 592–9. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Fig E1. Localization of Siglec-F ligand in OVA-challenged mouse trachea (A) and lung (B). Fig E1, A, Sialidase-sensitive Siglec-F ligands were expressed in airway epithelium and submucosal glands (panels b-g). Fig E1, B, Siglec-F ligands were observed in bronchial epithelium in lung tissue (panels b-d).

Fig E2. Attenuation of St3gal3 enhances the percentages of eosinophils in OVA-induced allergic eosinophilic airway inflammation (A, n = 8-12) but has no effect on peripheral blood (PB) eosinophil numbers (B, n = 9-14) when examined 24 hours after PBS or OVA final challenge. *P < .05 and ***P < .001. NS, Not significant.

Fig E3. Effect of attenuation of St3gal3 on levels of various cytokines and chemokines in BALF. Levels of the indicated mediators (A-L, with median values also shown) were quantified in BALF obtained 24 hours after PBS or OVA final challenge (n = 6-15). *P < .05 and **P < .01 relative to WT-PBS control mice. Horizontal bars above each set of data indicate direct statistical comparisons between OVA-challenged WT, St3gal3+/−, or St3gal3−/− mice. Dashed red lines indicate the level of detection. NS, Not significant.

NIHMS502478-supplement-supplement_1.pdf (1,022KB, pdf)

RESOURCES