Galectin-3 is Essential for IgE-Dependent Activation of Human Basophils by A549 Lung Epithelial Cells

Capsule Summary:

Using novel co-culture model systems, galectin-3 is shown to facilitate basophil secretion of IL-4/IL-13, with implications that this unique mode of activation may be linked to a variety of diseases ranging from asthma to cancer.

To the Editor:

We recently reported that human basophils are activated to release robust levels of IL-4/IL-13 when co-cultured with A459 epithelial cells (EC) –a lung adenocarcinoma. ¹ These responses proved dependent on cell-to-cell contact and basophil-associated IgE. They were also inhibited by compounds targeting FcεRI-dependent signaling and by n-acetyllactosamine. Involvement of galectin-3 (Gal-3) was hypothesized, given its initial description as an IgE-binding lectin. ² Moreover, Gal-3 expression is common to EC, especially those of cancer origin, and is unique among galectins in possessing multivalent binding capacity. ³ Therefore, we theorized that A549-associated Gal-3 non-specifically binds IgE on basophils, triggering a unique activation similar to standard IgE-dependent stimulation, but in a manner independent of allergen.

We address herein the direct involvement of Gal-3 in basophil activation (see Methods section of this article’s online repository at www.jacionline.org.). First, we use shRNA to suppress Gal-3 expression in A549, creating stable clones whereby Gal-3 protein is nearly ablated, as is their capacity to activate basophils for IL-4/IL-13. Secondly, we couple rhGal-3 onto microspheres (MS), showing that Gal-3, when immobilized on this matrix, activates basophils much like A549 EC.

As shown in Fig. 1A, two clones (IE2 and 1B10) demonstrated >95% suppression of total Gal-3 protein, as determined by Western Blot. In contrast, Gal-3 was only partially knocked-down in a 3^rd clone, 3G8, showing ~60% reduction, compared to wild-type A549 (A549-WT). Flow cytometry validated these findings, showing that Gal-3 expression by clones 1E2, 1B10, and 3G8 was similarly reduced by 90%, 88% and 49%, respectively, when compared to A549-WT cells (Fig. 1C). No phenotypic features distinguished the clones from wild-type cells (Fig. 1D).

Fig. 1.

A, Western Blot analysis of lysates from A549-WT and clones after utilizing shRNA to repress Gal-3 protein (Three experiments, Exp, conducted at weekly intervals), with B, densitometry analyses. *, p=0.047; **, p<0.009 vs. A549-WT. C, Gal-3 on A549-WT and clones, as indicated using flow cytometry. D, Light-field images as indicated (original magnification: 400×). Box-whisker plots of E, IL-4 (n=6–7) and F, IL-13 (n=11) secreted in basophil/EC co-cultures containing IL-3 (10 ng/ml), anti-IgE (20 ng/ml) or medium alone. *, p≤0.006; **, p=0.00001 vs. corresponding conditions with basophils co-cultured with A549-WT.

As shown in Fig. 1E, basophils alone secreted low levels of IL-4 (8±2 pg/10⁶, n=6) to IL-3. This response was augmented ~10-fold when basophils were co-cultured with A549-WT, confirming our previous findings. ¹ However, these augmented responses were ablated (to basophil-only levels) when co-cultured with the Gal-3-deficient clones (1E2 and 1B10). In contrast, the Gal-3⁺ clone, 3G8, retained the capacity to activate basophils for IL-4, with levels nearly equal to those induced by A549-WT.

Basophils co-cultured with A549-WT also produced ~30–50% more IL-4 in response to stimulation with an activating anti-IgE antibody (Fig. 1D). In contrast, these enhanced responses were not evident when basophils were co-cultured with the Gal-3-deficient clones (p=0.100 & p=0.190, respectively, n=6–7). In contrast, clone 3G8 retained near full capacity to augment IgE-dependent IL-4 production, much like A549-WT EC.

Comparable patterns were observed for IL-13 secretion (Fig. 1E) and were substantially greater than those for IL-4 –an observation consistent with our initial findings. ¹ For example, IL-13 averaged 69±14 pg/10⁶ (n=11) when basophils were cultured with IL-3 alone. These increased ~30-fold to 2001±268 pg/10⁶ when basophils were co-cultured with A549-WT. In contrast, IL-13 levels were markedly suppressed (82% and 79%) when basophils were co-cultured with clones 1E2 and 1B10, respectively (p=0.00001, n=11). This contrasts to just 14% suppression when basophils were co-cultured with clone, 3G8 (p=0.541, n=11).

Basophils also secreted ~9-fold greater levels of IL-13 in response to anti-IgE when additionally co-cultured with A549-WT. In contrast, the Gal-3-deficient clones (1E2 and 1B10) only slightly augmented this response. (p≤0.006, n=11). Again, clone 3G8 retained near capacity to co-stimulate basophils for IL-13, inducing responses (220±57 pg/10⁶ basophils, n=11) nearly equal to those seen in co-cultures with A549-WT.

To further support the concept that Gal-3 mediates basophil activation, we conducted experiments using recombinant protein. Importantly, there was no indication that A549-WT secreted Gal-3 into the culture medium, as determined by ELISA (data not shown). We therefore added rhGal-3 to basophils alone and to A549-WT/basophil co-cultures. As shown in Fig. 2A, rhGal-3 mediated no effect on basophil IL-4 production until used at 10 μg/ml, which only marginally increased IL-3-dependent secretion ~2-fold (~25 pg/10⁶ to ~50 pg/10⁶ basophils). Once again, basophils co-cultured with A549-WT produced ~10-fold greater levels of IL-4, yet the addition of rhGal-3 failed to augment this response. In fact, at 10 μg/ml rhGal-3 actually suppressed this IL-4 (~50%), hinting that this high concentration competes with A549-associated Gal-3, hindering its capacity to activate basophils.

Fig. 2.

A, Effects of Gal-3 protein on IL-4 secreted by basophils alone and in co-culture with A549-WT in the presence of IL-3 (10 ng/ml) (mean±SEM, n=3). B, IL-4 and C, IL-13 secreted by basophils co-cultured with MS (5 μm-sized) pre-coated with BSA, Gal-3, or Gal-9 in the absence and presence of IL-3 (10 ng/ml). *p<0.05, **p<0.01. D, Current conception of how EC-associated Gal-3 might crosslink IgE on basophils, resulting in activation for IL-4/IL-13.

Observing that rhGal-3 only marginally activated basophils, and at concentrations ~1000-fold greater than even the serological levels associated with disease ⁴, we considered that immobilized Gal-3 might better mimic A549-dependent basophil activation. We therefore coupled rhGal-3 to polycarbonate MS, testing whether these activate basophils. MS coupled with BSA (MS-BSA) and those with Gal-9 (MS-Gal-9) were also tested. When co-cultured with basophils at a 5:1 ratio (MS:basophil) and in the presence of IL-3, only the MS-Gal-3 significantly activated basophils, both for IL-4 (Fig. 2B) and IL-13 (Fig. 2C). Cytokine levels from basophils co-stimulated with MS-BSA and MS-Gal-9 were not significantly different from those induced using IL-3 alone.

Overall, these data support our hypothesis that Gal-3 is essential for A549-dependent basophil activation. Whether this mode of activation will translate to other cell lines and/or primary EC (normal and/or diseased) requires further investigation, as do the mechanisms controlling Gal-3 expression. Instead, the significance of these data currently lies with the fact that Gal-3 is now implicated as a biomarker and/or factor contributing to the fibrosis associated with and/or pathogenesis of a wide range of diseases/conditions, including asthma andatopic dermatitis, but also in cancer, autoimmunity (lupus), cardiovascular disease, and wound healing. Indeed, publications now link Gal-3 to disorders beyond the realm of allergic disease, with the overall number having grown exponentially in recent years ⁵. Ironically, most of this work seldom focuses on Gal-3’s interaction with IgE, despite its initial description as an IgE-binding lectin. ² We predict that the data herein will spark renewed interest in Gal-3/IgE interactions by providing novel insight into how this lectin, when expressed on matrix cells (e.g. EC), activates basophils (Fig. 2D) and possibly other IgE-bearing cells. Such findings could provide foresight regarding how these basophils participate in diseases not typically driven by specific IgE/allergen. For example, IL-4-producing basophils have been identified in lupus nephritis ⁶ and recently in cancer. ⁷ It’s intriguing to contemplate that the increased Gal-3 also seen in the affected tissues of these diseases might play a role in activating infiltrating basophils, and the consequences thereof.

Abbreviations:

EC

epithelial cells

Gal

galectin

MS

microspheres