Research ArticleCytokine Signaling

Instructive roles for cytokine-receptor binding parameters in determining signaling and functional potency

See allHide authors and affiliations

Science Signaling  10 Nov 2015:
Vol. 8, Issue 402, pp. ra114
DOI: 10.1126/scisignal.aab2677

Modeling cytokine behavior

The use of cytokines, such as interleukin-2 (IL-2) or IL-13, as therapies has been hampered by the fact that many cytokines share receptor subunits on different cell types. Moraga et al. generated recombinant variants of IL-13 with a wide range of binding affinities for the IL-13 receptor. Mathematical modeling of the correlation between the receptor binding affinities of the variants and the extents to which they differentially stimulated early and late cellular responses highlighted aspects of receptor-ligand binding properties that should aid in the development of more effective cytokine therapies.


Cytokines dimerize cell surface receptors to activate signaling and regulate many facets of the immune response. Many cytokines have pleiotropic effects, inducing a spectrum of redundant and distinct effects on different cell types. This pleiotropy has hampered cytokine-based therapies, and the high doses required for treatment often lead to off-target effects, highlighting the need for a more detailed understanding of the parameters controlling cytokine-induced signaling and bioactivities. Using the prototypical cytokine interleukin-13 (IL-13), we explored the interrelationships between receptor binding and a wide range of downstream cellular responses. We applied structure-based engineering to generate IL-13 variants that covered a spectrum of binding strengths for the receptor subunit IL-13Rα1. Engineered IL-13 variants representing a broad range of affinities for the receptor exhibited similar potencies in stimulating the phosphorylation of STAT6 (signal transducer and activator of transcription 6). Delays in the phosphorylation and nuclear translocation of STAT6 were only apparent for those IL-13 variants with markedly reduced affinities for the receptor. From these data, we developed a mechanistic model that quantitatively reproduced the kinetics of STAT6 phosphorylation for the entire spectrum of binding affinities. Receptor endocytosis played a key role in modulating STAT6 activation, whereas the lifetime of receptor-ligand complexes at the plasma membrane determined the potency of the variant for inducing more distal responses. This complex interrelationship between extracellular ligand binding and receptor function provides the foundation for new mechanism-based strategies that determine the optimal cytokine dose to enhance therapeutic efficacy.


Cytokines exert their effects through receptor dimers on the cell surface that activate the Janus kinase (JAK)–signal transducer and activator of transcription (STAT) pathway (15). Although there are many variations on this theme, cytokine engagement generally occurs through a two-site binding paradigm that orients the receptors into homo- or heterodimeric signaling complexes (6). Signaling does not appear to be an “all or none” phenomenon and is influenced by parameters including the structure of the dimeric complex (7) as well as the interaction parameters between the cytokine and the receptors (5). For example, cytokine-receptor systems that use shared receptors to engage multiple cytokines can elicit differential cellular responses to different ligands. Examples of such functional receptor plasticity include the type I interferon (IFN) family (810), the interleukin-2 (IL-2)–IL-15 system (1113), and the IL-4–IL-13 system (1416). Moreover, viral cytokines that differentially activate the IL-10 receptor (IL-10R) have been reported (17). Differential cellular responses by these cytokines have been linked to variations in the affinity and stability of ligand interactions with the receptor subunits (15, 1822) and to cell surface receptor densities (23, 24), suggesting that the spatiotemporal dynamics of signaling complex formation plays a central role in regulating signaling specificity.

A notable feature of these systems is the sometimes poor correlation between receptor binding affinity and signal activation potencies (5, 9, 25). This phenomenon was highlighted by engineering cytokines with increased receptor binding affinities, specificities, or both (6). For example, an IFNα2 mutant (IFNα2-YNS) binds to IFN receptor 1 (IFNAR1) more than 60-fold better than does IFNα2, and it exhibits a much stronger antitumor response (26); however, IFNα2-YNS activates the JAK-STAT pathway to the same extent as does wild-type IFNα2 (26). In addition, IL-4 variants were engineered to bind to either the type I IL-4 receptor (consisting of the IL-4Rα and γc subunits) or the type II IL-4 receptor (consisting of the IL-4Rα and IL-13Rα1 subunits) with high affinity and specificity (27). Although these mutants bound with higher affinities to their specific receptors, they exhibited similar signal activation potencies (27). These observations suggest that cells can mount a robust signaling response that is elicited by low-affinity ligands. This issue has important practical implications for immune cytokine therapy, which historically has been limited by dose-dependent, off-target toxicity. A more granular understanding of the mechanisms governing cytokine potency at lower, less toxic doses could potentially rescue such approaches. For example, low-dose IL-2 has shown great promise as a therapeutic for cancer, diabetes, and other autoimmune diseases (2830). It is important to determine the parameters relating cytokine affinity to function to understand whether low-dose approaches that preserve efficacy, but limit toxicity, could be generalized to other cytokine systems.

Here, we systematically explored the role of ligand-receptor affinity and complex stability in cytokine receptor signaling using IL-13 as a model system. IL-13 forms a type II receptor complex with the IL-13Rα1 and IL-4Rα subunits, activates STAT6, and contributes to the regulation of the innate immune response by promoting the differentiation of dendritic cells and macrophages (3, 3134). Deregulation of IL-13 leads to the development of asthma and fibrosis, which makes this cytokine a very attractive pharmacological target (32, 35, 36). Using yeast surface display, we isolated a series of IL-13 variants with different binding affinities for IL-13Rα1, spanning a range of nearly six orders of magnitude. This broad bandwidth of binding affinities enabled us to study signal activation, including receptor dimerization and endocytosis, STAT6 phosphorylation and nuclear translocation, and long-term cellular responses. On the basis of correlations between the binding constants and potencies of IL-13 variants at different levels of signaling, we developed a quantitative model that matched the experimental data. Our results highlight a complex interplay between ligand binding kinetics and feedback by receptor endocytosis, and begin to explain the apparent nonlinear correlations between the extracellular cytokine-receptor binding parameters and functional output. The practical implication of these results is that the toxicity that is commonly seen in cytokine therapies could be a by-product of dosing strategies that are much higher than necessary, given that substantially lower-affinity cytokine variants retain full biological activity. Our results are also supportive of strategies to engineer cytokines for cell type selectivity through fine-tuning their receptor affinity and to differentially target cell populations.


Engineering IL-13 variants with a range of receptor binding affinities

IL-13 engages the type II receptor first through binding to the IL-13Rα1 subunit, which is followed by binding of the IL-13–IL-13Rα1 complex to IL-4Rα. To study the role that cytokine-receptor complex stability plays in signaling and function, we used yeast surface display to engineer a series of IL-13 variants with a wide range of binding affinities for IL-13Rα1. IL-13 was displayed on the yeast surface, and biotinylated ectodomains of IL-13Rα1 and IL-4Rα coupled to Alexa Fluor 647–conjugated streptavidin were used as selection reagents. We first confirmed that IL-13 displayed on the yeast surface was functional. Yeast-displayed IL-13 could bind to IL-13Rα1 (Fig. 1A, second panel) but could only bind to IL-4Rα in the presence of IL-13Rα1 (Fig. 1A, third and fourth panels), in accordance with the previously described cooperativity exhibited by these two receptor subunits (15).

Fig. 1 Structure-based engineering of IL-13 variants.

(A) Surface staining of IL-13–displaying yeast with biotinylated IL-13Rα1 and IL-4Rα ectodomains (ECDs). IL-13Rα1 binds to IL-13 displayed on yeast independently of IL-4Rα (second panel), whereas IL-4Rα requires the presence of IL-13Rα1 to bind to IL-13, recapitulating the cooperativity that has been observed previously (15) (third and fourth panels). These data are from a single experiment and are representative of three independent experiments. FACS, fluorescence-activated cell sorting. (B) Crystal structure of the IL-13 ternary complex (IL-13Rα1 in green, IL-4Rα in blue, and IL-13 in gold). Magnifications show the site II (helices A to D) and site III (C-D loop) binding interfaces highlighting the amino acid residues in IL-13 that are involved in binding to IL-13Rα1. (C) Sequential enrichment of IL-13 variants. The site-specifically mutagenized library was selected against decreasing concentrations of IL-13Rα1 for five rounds. Clones were selected from each round to isolate variants with a range of IL-13Rα1 binding affinities.

The binding interface between IL-13 and IL-13Rα1 is composed of two binding sites: one in the receptor D2D3 domains, which is called site II, and the other between the top of IL-13 and the receptor D1 domain, which is called site III (Fig. 1B). Helices A and D (site II) and the loop that connects helices C and D (C-D loop, site III) on IL-13 contribute to the formation of these two binding interfaces. We generated a site II/site III–focused library and introduced a “soft randomization” on 13 amino acid residues in helices A and D and the C-D loop (Fig. 1B). The theoretical diversity of this library was 108 variants. We performed selections by enriching the library for IL-13Rα1–binding variants through five rounds of selection in which the IL-13Rα1 concentration was gradually decreased from 100 to 1 nM (Fig. 1C). Eighteen clones were selected on the basis of their on-yeast binding titration curves for IL-13Rα1 (Fig. 2A), which were subsequently confirmed by surface plasmon resonance (SPR) experiments (Fig. 2B). The binding affinity constant (KD) values for the A11, B2, A7, A6, A8, A5, B4, and B6 variants and IL-13 were calculated with the kinetic model, whereas the KD values for the C10, C11, C12, C2, C3, C4, C7, C9, and D7 variants were calculated with the steady-state equilibrium model because of their extremely fast on and off binding rates (fig. S1). These engineered agonists exhibited IL-13Rα1 binding affinities ranging from 80-fold more to 10,000-fold less than that of wild-type IL-13 (Fig. 2, B and C). The increase in binding affinity was mainly achieved by a decrease in the dissociation rate constant (koff), that is, by increasing the half-life of the ligand-receptor complex, whereas the association rate constants (kon) were similar among the different IL-13 variants. For those IL-13 variants with binding affinities in the higher nanomolar to micromolar range, koff could not be quantified because of the limited time resolution of SPR; however, we can assume that their koff values were further increased proportionally to the change in KD. Sequencing of these IL-13 variants showed that whereas the loss of their binding affinities for IL-13Rα1 occurred through mutations in either the site II or site III binding interface, gains in affinity mainly occurred through mutations in site III (Fig. 2C). IL-13 variants with increased binding affinity for IL-13Rα1 shared common mutations in site III (for example, D87G or D87S as well as T88S), suggesting that in all cases, the increase in IL-13Rα1 binding affinity was achieved through reorganization of the site III binding interface.

Fig. 2 Biophysical characterization of selected IL-13 variants.

(A) Binding isotherms of the IL-13 variants displayed on the surface of yeast. Yeast-displayed IL-13 variants were incubated with the indicated concentrations of biotinylated IL-13Rα1 ectodomain, stained with Alexa Fluor 647–conjugated streptavidin, and assessed by flow cytometry. Sigmoidal curves were fitted with Prism software. These data are from a single experiment and are representative of three independent experiments. (B) Normalized KD binding affinities of the recombinant IL-13 variants were determined by SPR analysis. The KD of IL-13 was set to 1, and other variants were normalized accordingly. (C) Summary of the amino acid residues mutated in the IL-13 variants as well as the kon, koff, and KD values obtained from SPR experiments. Colors represent amino acid positions in IL-13 where the amino acid sequence converged after selection.

Structure of a high-affinity IL-13 variant complexed with IL-13Rα1 and IL-4Rα

The overall receptor-ligand binding geometry of the IL-13 variants is an important factor to consider when interpreting differences in receptor activation (7, 37). To address this, we solved the structure of the ternary complex formed by the high-affinity IL-13 variant A11 and the IL-13Rα1 and IL-4Rα subunits to 3.0-Å resolution (Fig. 3A). Superposition of the ternary complexes formed by the A11 variant and wild-type IL-13 showed no major perturbations in cytokine-receptor architecture [root mean square deviation (RMSD), 0.738] (Fig. 3B). In the IL-13 A11–IL-13Rα1 binding interface, side-chain densities were clear for both helix D and the C-D loop of A11 (fig. S2). The site II binding interface on A11 closely resembled that of wild-type IL-13, with the amino acid residues Phe107 and Arg108 predominantly mediating binding (Fig. 3, C and D). The site III binding interfaces, however, differed substantially between the IL-13– and A11-containing ternary complexes. The gain in binding affinity achieved by A11 seems to have occurred through a new network of interactions that were generated in the site III binding interface (Fig. 3, C and D). Mutation of the amino acid residues Asp87 and Thr88 to the smaller amino acids glycine or serine enabled Trp65 to assume a different rotamer on IL-13Rα1, enabling this residue to form extensive contacts with Trp35 and Arg86 on IL-13 (Fig. 3D). Overall, these structural data suggest that the A11 variant had increased binding affinity by optimizing the site III, rather than the site II, binding interface network. Because the IL-13 variants have highly related sequences, it is reasonable to assume that they recruit IL-13Rα1 and IL-4Rα with similar geometries, and therefore that any signaling changes observed during our study can most likely be attributed to changes in binding affinities and the stabilities of receptor-ligand complexes.

Fig. 3 Structural characterization of the high-affinity IL-13 variant A11.

(A) Crystal structure of the IL-13 variant A11 (brown) in complex with IL-13Rα1 (green) and IL-4Rα (blue). (B) Overlay of the ternary complex crystal structures of wild-type (WT) IL-13 and the IL-13 A11 variant. Note that IL-13 A11 engages the two receptor chains with identical geometry to that of IL-13 WT (RMSD, 0.738). (C and D) Schematic representation of the ligand-receptor contact interaction of IL-13Rα1 with IL-13 (top half) versus IL-13 A11 (bottom half). Black lines and arrows represent van der Waals interactions, blue lines represent hydrogen bonds, and red lines designate aromatic amino acid interactions.

Receptor dimerization and spatiotemporal dynamics

Initiation of signaling by IL-13 requires simultaneous interaction of the ligand with IL-13Rα1 and IL-4Rα, but the spatial and temporal organization of these receptor subunits in the plasma membrane before and after ligand binding remains ill-defined. To interrogate the effects of ligand binding affinity on the spatiotemporal dynamics of the IL-13 signaling complex at physiologically relevant receptor concentrations on the cell surface, we devised quantitative ligand binding and receptor dimerization assays based on single-molecule fluorescence imaging (fig. S3). Using total internal reflection fluorescence (TIRF) microscopy, we probed the binding of IL-13 A11 variant, which was site-specifically labeled with the photostable fluorophore DY647 (DY647IL-13 A11) with a degree of labeling >90%. Specific binding of DY647IL-13 A11 to endogenous receptor at the surface of HeLa cells was detected at a concentration of 2 nM, which was sufficient to saturate all IL-13Rα1 subunits (movie S1 and fig. S3A). At moderate excitation power, individual diffraction-limited signals were observed, which were randomly distributed and continuously diffusing in the membrane (movie S1). Single-step photobleaching at enhanced laser power confirmed that each signal corresponded to individual IL-13 A11 molecules rather than to clusters (movie S2 and fig. S3B). An average density of 0.14 IL-13Rα1 molecules/μm2 was determined from binding experiments with DY647IL-13 A11, whereas a statistically significantly ↓ number of 0.07 IL-4Rα molecules/μm2 was obtained for binding experiments with DY647IL-4 (fig. S3C).

To quantify receptor dimerization at these relatively low abundances of cell surface receptor subunits, we used dual-color, single-molecule imaging. HeLa cells were transfected with plasmids encoding the IL-13Rα1 and IL-4Rα extracellular domains fused to the HaloTag and the SNAPf tag, respectively (Fig. 4A). These tags were used for selective posttranslational labeling with tetramethylrhodamine (TMR) and DY647, respectively, which ensure increased brightness and photostability. The integrity of ligand recognition by HaloTag–IL-13Rα1 and SNAPf–IL-4Rα with respect to receptor-ligand complex stability was confirmed by binding experiments with wild-type DY647IL-13 and DY647IL-4, respectively, and chasing with unlabeled ligand (fig. S3, D and E). Through dual-color TIRF microscopy, individual IL-13Rα1 and IL-4Rα subunits at the plasma membrane were detected and followed in real time for extended periods (movie S3). Single-step photobleaching at enhanced laser power confirmed the random distribution of IL-13Rα1 and IL-4Rα in the plasma membrane rather than their co-organization into clusters (movie S4). Individual receptor subunits were localized beyond the diffraction limit with an average precision of 20 nm, which enabled quantification of receptor dimerization on the single-molecule level by complementary image analysis techniques (fig. S3, F to H). Single-molecule co-locomotion analysis revealed that substantial receptor dimerization could only be discerned after the addition of ligand (fig. S3G), consistent with a two-step mechanism of ternary complex assembly. Co-trajectories corresponding to individual ternary complexes were observed (Fig. 4, A to C, and movie S5), with IL-13Rα1 and IL-4Rα exhibiting co-locomotion, for more than 500 frames (≈16 s).

Fig. 4 Assembly and dynamics of the ternary complexes formed by IL-13 variants.

(A to C) IL-13 receptor dimerization as detected by dual-color, single-molecule imaging and co-locomotion analysis. Representative images identifying an individual dimer containing HaloTag–IL-13Rα1 and SNAPf–IL-4Rα in the plasma membrane of HeLa cells in the presence of 200 nM IL-13 at different time points of a 500-frame (16-s) movie (A), the distance between the two molecules in each frame (B), and an overlay of the individual trajectories of HaloTag–IL-13Rα1 (magenta) and SNAPf–IL-4Rα (green) (C) are presented. Scale bars, 1 μm in (A) and (C). (D) Diffusion properties represented as step-length distribution obtained for HaloTag–IL-13Rα1 in the absence (red) and presence (orange) of IL-13. Individual step lengths from single-molecule trajectories were determined for a time lapse of 32 ms (1 frame), and the histograms (fig. S5) were fitted by a sum of two-dimensional Gaussian probability distributions (lines shown in the plot). For comparison, the step-length distribution obtained for receptor dimers identified by co-locomotion analysis is shown (blue). Histograms were obtained from >2500 trajectories (dimers from 2433 trajectories). (E and F) Quantitative analysis of receptor dimerization by PICCS. (E) Representative analysis depicting the cumulative probability of detecting a HaloTag–IL-13Rα1 molecule in the vicinity of a SNAPf–IL-4Rα molecule as a function of the squared distance, corrected for the statistical probability, in the presence or absence of WT IL-13. For comparison, the cumulative correlation probability obtained in the presence of IL-13DN is shown. (F) Correlated fraction α for individual HaloTag–IL-13Rα1 and SNAPf–IL-4Rα subunits obtained from PICCS analysis as a function of IL-13 variant affinity. (G) Assembly and dissociation of an individual dimer of HaloTag–IL-13Rα1 and SNAPf–IL-4Rα in the presence of the IL-13 variant D7. Overlaying co-locomotion trajectories are shown in white. Scale bar, 0.5 μm. (H) Average lifetime of IL-13Rα1–IL-4Rα dimers as a function of agonist affinity. Mean values and SDs were obtained from fitting exponential decay functions to lifetime histograms built from >400 trajectories for each mutant.

For all IL-13 variants that were subjected to these studies (wild-type, A11, C10, D7, and C4), substantial ligand-induced receptor subunit co-locomotion was confirmed (fig. S4). The local diffusion properties of the receptor subunits in the absence and in the presence of ligand were quantified by a step-length distribution analysis (fig. S5), which revealed a small yet statistically significant decrease in receptor mobility upon formation of the ternary complex (Fig. 4D). The decreased diffusion constant of dimerized receptors may have been caused by increased friction because of the increased diameter of the dimerized receptor or by further intracellular interactions of the receptor because of downstream signaling. A deconvolution of the step-length distribution observed for the receptor subunits after the binding of IL-13 (fig. S5D) suggested that ≈50% of the available IL-13Rα1 was recruited into ternary complexes at saturating concentrations of IL-13. In contrast, no co-locomotion of IL-13Rα1 and IL-4Rα was detectable in the presence of a dominant-negative IL-13 mutant (IL-13DN), which binds to IL-13Rα1 with high affinity but does not bind to IL-4Rα (Fig. 2C and fig. S3G). Accordingly, stimulation of A549 cells with IL-13DN did not result in STAT6 phosphorylation (fig. S6A).

Similar dimerization efficiencies over a broad spectrum of receptor binding affinities

To quantify the spatial co-organization of IL-13Rα1 and IL-4Rα in the absence and presence of ligand, we used single-molecule image analysis by particle image cross-correlation spectroscopy (PICCS) (fig. 3F) (38). For this purpose, cells expressing similar densities of IL-13Rα1 and IL-4Rα (~1 molecule/μm2) were chosen. No substantial cross-correlation of IL-13Rα1 and IL-4Rα molecules was observed in the absence of agonist or in the presence of IL-13DN (correlated fraction α ≈ 0), which excluded the possibility of there being transient receptor preorganization before ligand binding (Fig. 4E). Upon stimulation with wild-type IL-13 at a saturating concentration, cross-correlation was detectable with a correlated fraction α ≈ 0.1 (Fig. 4E), confirming ligand-induced dimerization of IL-13Rα1 and IL-4Rα. On the basis of control experiments with a transmembrane helix fused to both the HaloTag and the SNAPf tag, the correlated fraction α that was observed after stimulation was consistent with ~50% of IL-13Rα1 being in complex with IL-4Rα, as estimated from the diffusion analysis (fig. S5D). This highly efficient dimerization by IL-13 is in stark contrast to the findings of a study that proposed that the dimerization of IL-13Rα1 and IL-4Rα at the plasma membrane was very inefficient (39).

Taking the correlated fraction α as a measure of receptor dimerization, we studied the effects of stabilizing and destabilizing the interaction of IL-13 with IL-13Rα1 (Fig. 4F). The efficiency of receptor dimerization did not change for either the complex-stabilizing A11 variant or the moderately destabilizing C10 variant; however, it decreased substantially for the low-affinity D7 and C4 variants, probably because saturated receptor occupancy was not achieved (Fig. 4F). These results suggest that the recruitment of IL-13Rα1 and IL-4Rα by IL-13 on the plasma membrane is very robust across a wide range of binding affinities, tolerating a decrease in affinity by almost two orders of magnitude. Because the engineered IL-13 variants bound to IL-13Rα1 with markedly altered stabilities, we investigated the lifetime of the individual complexes, a property that has been conjectured to play a critical role in the functional plasticity of other cytokine receptors. Indeed, formation and dissociation of individual complexes was observed (movie S6 and Fig. 4G), which is consistent with reversible ligand binding to the cell surface receptor. Although the lengths of the co-trajectories were limited by tracking fidelity and photobleaching, substantial differences in the lifetimes of the complexes were observed for the low-affinity IL-13 variants D7 and C4 (Fig. 4H). The mean lifetime of complexes paralleled the increased dissociation kinetics of these IL-13 variants as measured by SPR (Fig. 2B), corroborating the relevance of reversible ligand binding for the spatiotemporal dynamics of ternary complexes. These data highlight the potential to control the signal delivered to a responsive cell by two distinct mechanisms: (i) by regulating the maximum number of complexes formed on the cell surface by a given cytokine, which is determined by the receptor density, and (ii) by regulating the lifetime of individual complexes, which heavily relies on the koff of the ligand.

STAT6 activation profiles elicited by the IL-13 variants

We next studied the signaling signatures exhibited by the different IL-13 variants that we characterized by affinity measurements, structural analysis, and single-molecule microscopy. We measured the concentration-dependent generation of phosphorylated STAT6 (pSTAT6) as well as the kinetics of signal activation in the IL-13–responsive cell line A549. A549 cells were stimulated for 15 min with various concentrations of IL-13 or its variants, and the amounts of pSTAT6 that were generated were analyzed by flow cytometry. We observed two groups of variants on the basis of their abilities to activate STAT6. One group of variants (C2, C3, C4, C7, C9, C11, C12, and D7) activated STAT6 to a substantially lower extent than did wild-type IL-13 (Fig. 5A). The other group (A11, B2, A5, A6, A7, A8, B4, B6, and C10) activated STAT6 with very similar potencies to each other and to IL-13 (Fig. 5A). These observations were confirmed by Western blotting analysis (fig. S6B). The range of IL-13Rα1 binding affinities among the second group of agonists varied from 80 pM (A11) to 100 nM (C10), suggesting the existence of a large “buffering region” that enables A549 cells to respond very similarly to agonists that exhibit a wide range of binding affinities (Fig. 5B). Indeed, about a 1000-fold decrease in binding affinity compared to that of IL-13 was required for an IL-13 variant to result in substantially decreased amounts of pSTAT6. When the binding affinities of the IL-13 variants for IL-13Rα1 were plotted against their STAT6 activation potencies, we observed a region consisting of IL-13 variants from A11 to C10 in which the activation of STAT6 did not substantially differ (Fig. 5B).

Fig. 5 STAT6 activation profile in response to IL-13 variants.

(A) A549 cells were stimulated with the indicated concentrations of WT IL-13 or IL-13 variants, and the abundance of pSTAT6 in the cells was analyzed by flow cytometry with pSTAT6-specific antibodies coupled to fluorescent dyes. Sigmoidal curves were fitted with Prism software. Data are means ± SD from three independent replicates. (B) Dot plot in which the normalized EC50 values of each IL-13 variant for pSTAT6 generation are plotted against their normalized KD values. The EC50 values of WT IL-13 for pSTAT6 generation and its KD values were set to 1, and the parameter values of the IL-13 variants were normalized accordingly. Variants with binding parameters within 100-fold of that of WT IL-13 in either direction led to similar profiles of STAT6 activation, generating the denoted buffering region. Data are means ± SD from three independent experiments. (C) A549 cells were stimulated with 200 nM WT IL-13 or the indicated IL-13 variants, and the kinetics of STAT6 phosphorylation were analyzed by flow cytometry. Data are means ± SD from three independent experiments. (D) Total amounts of pSTAT6, calculated by measuring the area under the curve (AUC) from the kinetics studies shown in fig. S6. Data are means ± SD from three independent experiments. (E) HeLa cells were stimulated for the indicated times with 200 nM WT IL-13 or the indicated IL-13 variants, and the translocation of mEGFP-STAT6 to the nucleus was measured by fluorescence microscopy. Mean values and SDs were obtained from >20 cells in two independent experiments for each IL-13 variant. (F) A549 cells were transfected with different amounts of IL-13Rα1–specific of control siRNA to reduce IL-13Rα1 abundance to the indicated percentages compared to control cells. Cells were then stimulated with WT IL-13 or the indicated IL-13 variants, and pSTAT6 abundance was quantified by flow cytometry. Data are means ± SD from three independent experiments. (G) Median effective concentration (EC50) values for the induction of TF-1 cell proliferation by the indicated IL-13 variants were obtained from fitting sigmoidal profiles to the dose-response curves shown in fig. S10B. Data are means ± SD from three independent experiments. (H) Analysis of the effects of the indicated concentrations of IL-13 variants on the cell surface abundance of CD86 on monocytes. Data are means ± SD from three independent experiments.

Because changes in binding affinity affected the lifetime of cytokine-receptor complexes formed by different IL-13 variants, we next asked whether the kinetics of STAT6 activation would be modulated by these changes in half-life. We performed a series of STAT6 activation kinetics studies in A549 cells in which the times of generation and decay of pSTAT6 were measured. Agonists that bound very weakly to IL-13Rα1 (D7 and C4) exhibited a delay in STAT6 activation (Fig. 5C). The C4 variant, which binds to IL-13Rα1 8000 times weaker than does wild-type IL-13, stimulated 50% of STAT6 activation when compared to that stimulated by IL-13 after 1 hour of stimulation. This delayed kinetics of STAT6 activation exhibited by the D7 and C4 variants was also observed for wild-type IL-13 and the A11 and C10 variants when sub-saturating doses of ligand were used (fig. S6, C to G). The number of STAT6 molecules activated per unit of time, measured as the AUC, by all of the IL-13 variants tested was minimally altered by changes in STAT6 activation kinetics or ligand concentrations (Fig. 5D), which suggests that cells may have evolved to activate a finite number of STAT molecules.

A functional consequence of STAT6 activation is its translocation to the nucleus where it regulates gene transcription. We next studied how receptor-ligand complex stability affected the kinetics of STAT6 nuclear translocation in HeLa cells stably transfected with plasmid encoding STAT6 fused to monomeric enhanced green fluorescence protein (mEGFP) (fig. S7A). Similar to the kinetics of STAT6 activation, the nuclear translocation of STAT6 was further delayed for agonists that weakly bound to IL-13Rα1 (Fig. 5E). Note that the nuclear translocation of STAT6 induced by C10 was slower than that induced by IL-13; however, the two ligands generated very similar amounts of pSTAT6. Overall, these data indicate that although the amount of STAT6 activated by IL-13 is very resistant to changes in binding affinity, the duration of STAT6 activation is more sensitive to changes in the affinity of the ligand for the receptor and closely correlates with the stability of the ternary complex.

In the case of the IFN system, surface receptor density contributes to signal activation potency (23). Thus, we next studied whether changes in the IL-13Rα1 surface density would modulate STAT6 activation by the different IL-13 variants. To this end, we compared the STAT6 activation profile induced by IL-13 variants in cells in which the abundance of IL-13Rα1 was decreased by short inhibitory RNA (siRNA). Whereas minimal differences in STAT6 activation profiles between agonists were observed in cells treated with control siRNA (Fig. 5F), when IL-13Rα1 abundance was reduced to 40% of that in control cells, the IL-13 variants B4 and A8, which bound more weakly than wild-type IL-13 to IL-13Rα1, activated STAT6 less potently than did the other IL-13 variants. These differences in STAT6 activation profile were further increased when IL-13Rα1 abundance was reduced to 20% of that in control cells (Fig. 5F). These data suggest that there was a close correlation between binding affinities, cell surface receptor abundance, and the extent of signaling.

Correlation of ligand binding parameters with signaling and functional potencies

We next studied the effect of receptor-ligand complex stability on the biological activities induced by IL-13. The TF-1 cell system is commonly used to assess the potency of many cytokines, including IL-13. In response to IL-13, these cells proliferate in a concentration-dependent manner, which readily enables evaluation of the potency of IL-13 variants (40). We stimulated TF-1 cells with different concentrations of the various IL-13 variants for 96 hours and then measured cell numbers by flow cytometry. We observed a broad range of proliferation potencies (Fig. 5G), which contrasted with the similar STAT6 activation profiles induced by the different IL-13 variants (Fig. 5A). Whereas A11 was virtually identical to wild-type IL-13 in its ability to activate STAT6 (Fig. 5A), it was the more potent inducer of TF-1 cell proliferation (Fig. 5G and fig. S7B). The different extents of cell proliferation induced by the different IL-13 variants more closely matched their binding affinities than their respective abilities to activate STAT6.

To broaden our evaluation of the biological activities of the IL-13 variants, we analyzed the differentiation of monocytes into dendritic cells in vitro. We stimulated monocytes with IL-13 or our engineered variants for 7 days and then used flow cytometric analysis to determine the cell surface abundances of CD86 and CD209, which are classical dendritic cell markers. The potency of the different IL-13 variants to stimulate dendritic cell generation (Fig. 5H and fig. S7, C to E) correlated with their binding affinities, consistent with our findings for the proliferation response of TF-1 cells. As before, we observed that A11 was more potent than IL-13 in inducing distal responses, that is, the generation of dendritic cells, despite eliciting very similar profiles of STAT6 activation (fig. S7F). We could not correlate the EC50 values of the IL-13 variants for dendritic cell generation to the kon and koff constants because we could not accurately estimate these kinetic parameters for several of the weakly binding agonists (C4, C2, and D7) tested in this experiment. The lack of correlation between receptor-ligand complex stability and the signaling potencies exhibited by the IL-13 variants could not be ascribed to cell surface receptor abundances in the cells used for the experiments or to differential receptor down-regulation induced by the different IL-13 variants (fig. S8).

We asked whether other parameters involved in the binding affinity of IL-13 for IL-13Rα1 contributed to the observed bioactivities. The KD of a given ligand is derived from two kinetic parameters: kon, which reflects the rate at which the ligand binds to its cognate receptor, and koff, which reflects the length of time the ligand remains bound to the receptor. Therefore, we studied how these two components correlated with the activation of membrane-proximal signaling events, that is, STAT6 activation, as well as with the potency of bioactivities induced by IL-13, that is, TF-1 cell proliferation. We generated a series of linear correlation plots in which the EC50 values of the different IL-13 variants for the generation of pSTAT6 and the induction of TF-1 cell proliferation were plotted against the kon, koff, or KD values of these agonists (Fig. 6). The extent of STAT6 activation induced by the IL-13 variants correlated reasonably well with their kon values (R2 = 0.52) but correlated poorly with their koff values (R2 = 0.021) (Fig. 6, B and C). For example, whereas the KD of wild-type IL-13 was about 80-fold lower than that of the A11 variant, their kon values were very similar (Fig. 2C). Consistent with their similar kon values, both cytokines similarly elicited STAT6 activation. In contrast, the potency of induction of TF-1 cell proliferation by the different IL-13 variants correlated more closely with their koff values (R2 = 0.54) and very poorly with their kon values (R2 = 2.8 × 10−5) (Fig. 6, E and F).

Fig. 6 Correlation of receptor binding kinetics with signaling and bioactivities induced by IL-13 variants.

(A to C) Correlation of the EC50 values of WT IL-13 and the indicated IL-13 variants for the induction of STAT6 phosphorylation with their binding parameters KD (top), kon (middle), and koff (bottom). Data are means ± SD from three independent experiments and were obtained with A549 cells. (D to F) Correlation of the EC50 values of WT IL-13 and the indicated IL-13 variants for the induction of TF-1 cell proliferation with their binding parameters KD (top), kon (middle), and koff (bottom). Data are means ± SD from three independent experiments.

Role of receptor endocytosis in signaling and functional potencies of IL-13 variants

Analysis of kinetic binding parameter correlations suggested that downstream cellular processes were involved in regulating the phosphorylation of STAT6. Because endocytosis was proposed to play an important role in IL-4 signaling (39), we hypothesized that the unexpectedly high activity of low-affinity agonists may be caused by endocytosis of functional signaling complexes. We confirmed the efficient endosomal uptake of both wild-type IL-13 and the D7 variant bound to endogenous receptors in HeLa cells (Fig. 7A). We also observed the colocalization of HaloTag–IL-13Rα1 and SNAPf–IL-4Rα in intracellular vesicular structures, which provided further evidence that entire signaling complexes were endocytosed (fig. S9A). Endocytosis of IL-13 was substantially reduced upon addition of EHT 1864 (Fig. 7B), an inhibitor of Rho family guanosine triphosphatases, which have previously been implicated in the endocytosis of IL-2Rs (41). In cells treated with the highest concentration of EHT 1864 that could be applied without causing loss of cell surface receptors (100 μM), we observed a twofold reduction in endocytosis as quantified from the intensity of staining of intracellular endosomes and the number of ligands bound to the cell surface (Fig. 7C). The functional role of endocytosis was investigated by time-lapse STAT6 phosphorylation assays in the absence or presence of the endocytosis inhibitor (Fig. 7D). We observed a correlation between agonist binding affinity and the role of endocytosis; whereas no statistically significant reduction in STAT6 phosphorylation kinetics was observed for A11 in the presence of EHT 1864, STAT6 phosphorylation in response to D7, which is a low-affinity agonist, was slowed (Fig.7D). Furthermore, EHT 1864 did not affect the EC50 values of any of the IL-13 variants for STAT6 activation, which was suggestive of a specific effect on the kinetics of STAT6 activation (fig. S9B).

Fig. 7 Receptor endocytosis and its role in STAT6 phosphorylation.

(A) Epifluorescence images showing the endocytosis of DY647IL-13 WT (left) and DY647IL-13 D7 (right) bound to endogenous IL-13Rα1 in HeLa cells. (B) Epifluorescence images showing the endocytosis of DY647IL-13 WT (left) and DY647IL-13 D7 (right) bound to endogenous IL-13Rα1 in HeLa cells after treatment with 100 μM EHT 1864. (C) Left: Quantification of the amounts of DY647IL-13 WT and DY647IL-13 D7 in endosomal compartments (left) and at the plasma membrane (right) in the absence and presence of EHT 1864. Each data point represents the quantification from a single cell, and the box plots indicate the data distribution of the second and third quartiles (box), median (line), mean (closed squares), and whiskers (1.5× interquartile range). (D) Kinetics of STAT6 phosphorylation in HeLa cells treated with WT IL-13 (top), IL-13 D7 (middle), or IL-13 A11 (bottom) in the absence and presence of EHT 1864. Data are means ± SD from three independent experiments.

A mechanistic model correlating cytokine binding to functional output

These results suggested that endocytosis was a key determinant of STAT6 phosphorylation. On the basis of these results, we conceived a simple steady-state model for receptor activation consisting of reversible cell surface binding followed by irreversible uptake of functional signaling complexes into endosomes and then their sorting into degradative and recycling pathways (Fig. 8A). Note that this model assumes that signaling complexes remain active with respect to their ability to induce STAT6 phosphorylation in endosomes for some time. As a consequence of endosomal uptake, the ligand is considered to remain bound to the receptor because of the high local concentration (0.1 to 1 mM) of individual ligand molecules in the very small volume of the endosome. On the basis of this model, we simulated the kinetics of pSTAT6 formation for different IL-13 variants (see Fig. 5C), assuming that the measured concentration of pSTAT6 was proportional to the total number of active signaling complexes in the plasma membrane and in endosomes. Using the experimental binding constants of the different IL-13 variants, we obtained global, realistic rate constants of endocytosis, degradation, and recycling by fitting the experimental data. This simple model reproduced the observed pattern of STAT6 phosphorylation kinetics as well as the selective effect of inhibited endocytosis on the activity of low-affinity agonists (Fig. 8, B and C, and fig. S10A). Furthermore, this model confirmed the correlation between pSTAT6 abundance and kon value (Fig. 6B and Fig. 8D) as well as the characteristic STAT6 phosphorylation kinetics at decreased concentrations of IL-13 variants (Fig. 8, E and F, and fig. S10B). These results highlight the idea that the receptor-proximal signaling events are mostly regulated by the interplay of ligand binding with receptor endocytosis kinetics.

Fig. 8 Quantitative model of IL-13 signaling.

(A) Steady-state model of ligand binding, receptor endocytosis, and further endocytic trafficking used for simulations of STAT6 phosphorylation kinetics and dose-response curves. (B) Simulated kinetics of STAT6 phosphorylation obtained for different IL-13 variants. (C) Comparison of simulated STAT6 phosphorylation in absence (top) or presence (bottom) of an endocytosis inhibitor. (D) Correlation of EC50 values from dose-response curves simulated for different IL-13 variants with their kon binding parameters. (E) Kinetics of STAT6 phosphorylation obtained for the indicated different nanomolar concentrations of WT IL-13 (left) and the IL-13 D7 variant (right). (F) Comparison of simulated STAT6 phosphorylation kinetics for different IL-13 variants at a concentration of 10 nM.


There are several instances of cytokines, including type I IFNs, IL-4 and IL-13, and IL-10 and viral IL-10, among others, sharing common cell surface receptors and yet eliciting a range of both redundant and differential activities (5, 9, 25, 32). Evidence has accumulated suggesting that the stability of the cytokine-receptor complexes formed by the different members of these families correlates with their potency in activating long-term cellular responses, for example, proliferation, differentiation, and apoptosis (19, 22, 27, 42). Long-lived complexes can promote more potent long-term responses than can short-lived complexes (19, 27, 42). However, the same correlation does not hold true for membrane-proximal signaling events, for which low- and high-affinity ligands activate downstream signaling to the same extent (19, 27, 42). This apparent disconnect between cytokine-receptor complex stability and signal activation has raised the question of how cells integrate different receptor binding affinities into similar patterns of STAT activation while preserving functional diversity. Our data support a model in which the fine-tuning of cytokine receptor binding kinetics enables cytokines to exhibit robust signal activation at a wide range of binding affinities and yet preserve differential responses. We found that the kon and koff values of cytokines correlated with distinct functions in determining the signaling output of a given cytokine: kon correlated with the amount of pSTAT that was generated by controlling the number of ligand-receptor complexes formed in the plasma membrane, whereas koff correlated with the kinetics of STAT activation through the modulation of the half-life of ligand-receptor complexes as a function of endocytosis.

An observation that emerges from our data is the presence of a region in the receptor binding affinity space, herein referred as the “signal buffering region,” in which the extent of STAT6 activation appeared unaffected by large changes in binding affinity: ligands with 100-fold increased (A11) and 100-fold decreased (C10) binding affinities for the IL-13Rα1 subunit compared to that of wild-type IL-13 elicited STAT6 activation to similar extents to that of wild-type IL-13. Similar observations have been made in other systems in which affinity maturation of the cognate ligands did not yield stronger signal activation (19, 26, 27, 43). Prominently, the observed 10,000-fold difference in the binding affinities of the IL-13 variants resulted from changes in their koff values. All IL-13 variants exhibited comparable kon values, which suggested that STAT activation remained largely unaffected by changes in koff above a certain threshold. Moreover, all IL-13 variants yielded cell surface complex formation to similar extents, arguing in favor of kon, but not koff, as the main factor that determines the number of complexes formed by a given cytokine and generates the signal buffering region that we observed in our experimental data. A caveat to this interpretation is that because our IL-13 variants have rather similar kon values, we could not test the effect that large changes in kon could have in cytokine receptor complex formation and signal activation. More systematic engineering of kon by manipulating the electrostatic potential (44) will be required to explore whether a further increase in STAT activation potency could be achieved by increasing complex association (the speed of association of the cytokine to the receptor). Collectively, our data suggest that cytokines can effectively signal given a sufficient kon value largely independent of their affinity. In nature, cytokine signaling would benefit from loose affinity constraints by enabling cells to sense and respond to a wide range of binding affinities and complex stabilities.

How cells translate virtually identical STAT activation profiles into different extents of cell proliferation and differentiation remains unclear. Our results showed that IL-13 agonists with slower koff rates persisted longer as receptor complexes, promoted faster kinetics of STAT activation, and induced more potent effector functions (that is, increased proliferative responses and enhanced monocyte differentiation) than did low-affinity ligands, which suggests that cytokines fine-tune their distal responses by controlling the half-lives of their complexes with receptors and the kinetics of downstream signaling. Consistent with this notion, we showed that the total amounts of STAT6 molecules activated by a cytokine-receptor complex remained unaltered through a wide range of cytokine-receptor complex stabilities and ligand concentrations. Although short-lived complexes and reduced concentrations of ligand resulted in delayed kinetics of STAT6 activation, the absolute number of molecules activated per unit of time did not change substantially. These observations are consistent with previous studies by our laboratory in the IL-4 system. We described a series of IL-4 mutants with enhanced receptor binding affinity and specificity when compared to wild-type IL-4, including super-4, a type I receptor–specific mutant, and KFR, a type II receptor–specific mutant (27). Despite there being a more than 10,000-fold difference in the receptor binding affinities of these mutants, they activated STAT6 to roughly similar extents. However, here again, the potencies of their distal responses correlated with the stabilities of their respective ligand-receptor complexes. In light of these data, we propose that the different responses elicited by super-4 and KFR most likely result from the different kinetics of signal activation induced by these mutants.

Our data also suggest that the cell surface abundance of cytokine receptors plays a major role, through mass action, in “titrating” high- and low-potency ligands with respect to signaling and function. Cells with increased amounts of cytokine receptors are relatively insensitive to differences in ligand affinity with respect to the membrane-proximal activation of STATs. In contrast, in cells that have decreased amounts of cytokine receptors, ligands with slower koff rates gain a functional advantage. Indeed, we observed that the siRNA-mediated reduction in the cell surface abundance of IL-13Rα1 resulted in low-affinity IL-13 variants signaling more weakly than high-affinity IL-13 variants. Similarly, cell surface receptor density determined the differential antiproliferative responses exhibited by low- and high-affinity IFN subtypes (23, 24). Although we observed a clear correlation between koff, receptor complex half-life, signaling kinetics, and the potency of distal signaling, we still do not fully understand how all of these parameters interconnect together and translate into more potent and diverse responses. The effect of negative feedback mechanisms in signal activation remains largely unexplored. Indeed, a previously uncharacterized feedback mechanism acting at the level of ligand-receptor complex formation was described for IFN (45). Low-affinity IFN subtypes were more sensitive than were high-affinity IFN subtypes to this form of regulation. It is thus tempting to speculate that high-affinity ligands could withstand the negative regulation of these feedback mechanisms by forming more stable complexes, thus eliciting more potent or more diverse distal signaling responses.

Our results challenge the equilibrium model for cytokine-mediated signal activation. In that model, cytokines recruit their receptor subunits through a two-step mechanism, reaching a dynamic equilibrium between binary and ternary complexes on the plasma membrane (9, 46, 47). Modulation of the on and off binding rates for any of the two receptors subunits is predicted to produce a parallel alteration in the number of complexes formed by a given cytokine and in its signaling potency; however, our data showed that 10,000-fold differences in binding affinity that resulted from changes in koff only marginally altered STAT activation. This is not a unique feature of IL-13; it is also found in other cytokines, such as those of the type I IFN family, of which there are more than 15 members that share the same cell surface receptor but bind with very different affinities and still activate STAT with very similar potencies (48). By extending the equilibrium ligand binding model into a steady-state model including endocytic trafficking, we reproduced with high fidelity the kinetics and the potencies of STAT phosphorylation by IL-13 variants that exhibited a large range of kon and koff rates. The intricate interplay of cytokine-receptor complex stability and receptor trafficking has been appreciated for explaining the potencies of cytokines and other hormones (49, 50). Here, we applied a minimal model comprising reversible ligand binding to the cell surface receptor followed by irreversible uptake of intact signaling complexes into endosomes and further sorting into degradative and recycling pathways. In this model, increasing the binding affinity of the ligand by increasing the stability of the complex (that is, by decreasing ligand dissociation from the cell surface) enhanced signaling only until the koff rate was exceeded by the rate of endocytosis (depicted as ke in Fig. 8A). Once ke dominates over koff, the final number of signaling complexes is determined by the kon rate of the ligand, and any further increases in affinity by changes in koff will not have an effect on signal activation. Thus, this model explains the existence of a signal buffering region in which variants with similar kon values and koff < ke will promote signal activation to comparable extents. Moreover, this model also explains the delayed kinetics of signal activation that we observed for low-affinity IL-13 variants. Assuming that downstream signaling is maintained by complexes in early endosomes (because irreversible endocytosis constantly removes active signaling complexes from the equilibrium), the number of active signaling complexes increases with ongoing endocytosis. A study proposed a critical role for receptor endocytosis in efficient receptor dimerization and signaling exclusively from endosomes (39), which is not supported by our dimerization assays that were performed with full-length receptors at physiologically relevant amounts. In contrast, our studies revealed a critical role for endocytosis in controlling ligand binding to the receptor, thus buffering signaling activity over a large range of ligand binding affinities.

Together, our results provide additional insight into how cytokines exhibit such functional plasticity despite activating a limited set of signaling molecules (four JAKs and seven STATs). From our analysis, there appears to be a mechanistic rationale for purposeful modulation of on and off binding rates by protein engineering to design engineered cytokine variants that preserve efficacy yet are less toxic therapeutics. Low-dose IL-2 is showing great promise as a therapeutic for cancer, diabetes, and other autoimmune diseases (2830), and it appears likely that similar low-dose approaches with other cytokines or higher-dose treatments with affinity-impaired, rather than affinity-enhanced, cytokines could rescue cytokine therapies that were previously limited by dose-dependent toxicity.


Protein expression and purification

The complementary DNAs (cDNAs) encoding human IL-13, the IL-13Rα1 ectodomain (amino acid residues 1 to 310), and the IL-4Rα ectodomain (amino acid residues 1 to 202) were subcloned into the pAcGP67-A vector (BD Biosciences) in frame with an N-terminal gp67 signal sequence and a C-terminal hexahistidine tag, and tagged proteins were produced using the baculovirus expression system, as described previously (15). Baculovirus stocks were prepared by transfection and amplification in Spodoptera frugiperda (Sf9) cells grown in SF900II media (Invitrogen), and protein expression was performed in Trichoplusia ni (High-Five) suspension cells grown in Insect-Xpress media (Lonza). After their expression was induced, proteins were captured from High-Five cell culture medium after 60 hours by nickel–nitrilotriacetic acid (Ni-NTA) agarose (Qiagen) affinity chromatography, concentrated, and purified by size-exclusion chromatography on a Superdex 200 column (GE Healthcare). The proteins were then equilibrated in 10 mM Hepes (pH 7.2) containing 150 mM NaCl. Recombinant cytokines were purified to >98% homogeneity. IL-13 variants and the IL-13Rα1 subunit used in SPR measurements and cell-based assays were expressed fully glycosylated, as estimated from Coomassie-stained SDS–polyacrylamide gel electrophoresis analysis. For site-specific fluorescence labeling by enzymatic phosphopantetheinyl transfer, a ybbR tag (DSLEFIASKLA) (51) was fused to the N terminus of wild-type IL-13 and its variants as well as to IL-4. Protein labeling with a DY647–coenzyme A conjugate in the presence of the phosphopantetheinyl transferase (Sfp) was performed as described previously (52). The uncompromised interaction of fluorescently labeled IL-13 and IL-4 with the extracellular domains of IL-13Rα1 and IL-4Rα, respectively, was confirmed by simultaneous TIRF spectroscopy and reflectance interference detection as described previously (53). To generate biotinylated receptors, the cDNAs encoding the ectodomains of IL-4Rα and IL-13Rα1 were subcloned into the pAcGP67-A vector with a C-terminal biotin acceptor peptide (BAP)–LNDIFEAQKIEWHW followed by a hexahistidine tag. Receptors were coexpressed with BirA ligase in the presence of excess biotin (10 μM). For crystallization studies, the IL-13 A11 variant was coexpressed with IL-13Rα1 in the presence of 10 μM tunicamycin to inhibit glycosylation, and IL-4Rα was coexpressed with endoglycosidase H in the presence of 5 μM Kifunensin to prevent its glycosylation. After Ni-NTA purification, the binary complex consisting of IL-13Rα1 and the IL-13 A11 variant was incubated together with IL-4Rα to form a ternary complex. The proteins were then treated overnight with carboxypeptidase A (Sigma) and B (Calbiochem) at a ratio of 1:100 (w/w) and were subsequently purified by size-exclusion chromatography. Protein was concentrated to 8 to 20 mg/ml for crystallization. Protein concentrations were quantified by ultraviolet spectroscopy at 280 nm with a NanoDrop 2000 spectrometer (Thermo Scientific).

Cell lines and media

The IL-13–responsive cell lines A549 and TF-1 were cultured in RPMI containing 10% (v/v) fetal bovine serum (FBS), penicillin-streptomycin, and 2mM l-glutamine. TF-1 cells were cultured in the presence of granulocyte macrophage colony-stimulating factor (GM-CSF) to promote their proliferation and survival. HeLa cells were cultured in Eagle’s minimum essential medium containing 10% (v/v) FBS, NaHPO3 (2 g/liter), 2 mM l-glutamine, 10 mM Hepes, nonessential amino acids, and penicillin-streptomycin. All cell lines were maintained at 37°C with 5% CO2. HeLa cells stably transfected with mEGFP-STAT6 were cultured in the presence of G-418 (0.8 mg/ml, Invitrogen) for selection and to maintain stable transfectants.

Flow cytometric analysis and antibodies

For cell surface staining of IL-13Rs, A549 cells were incubated with antibodies against IL-13Rα1 (1:300, BD Biosciences) coupled to fluorescein isothiocyanate for 1 hour at 4°C. Cells were then washed, and the cell surface abundance of IL-13Rα1 was measured with an Accuri C6 flow cytometer. Intracellular staining of pSTAT6 was performed after permeabilization of cells with ice-cold methanol (100%, v/v). Alexa Fluor 488–conjugated antibodies specific for pSTAT6 were purchased from BD Biosciences and used at a 1:50 dilution. The extent of STAT6 phosphorylation was calculated by subtracting the mean fluorescence intensity (MFI) of pSTAT6 of the stimulated samples from that of the unstimulated sample. The normalized values were plotted against cytokine concentration to yield dose-response curves from which the EC50 values were calculated on the basis of nonlinear, least squares regression fit to a sigmoidal curve.

Yeast display of IL-13

General yeast display methodologies were modified from previously described protocols (54). The cDNA encoding human IL-13 was subcloned into the yeast display vector pCT302. The Saccharomyces cerevisiae strain EBY100 was transformed with the pCT302_IL-13 vector and grown for 2 days at 30°C on selective dextrose casamino acids (SDCAA) plates. Individual colonies of IL-13–displaying yeast were grown overnight at 30°C in SDCAA liquid medium (pH 4.5), followed by induction in selective galactose casamino acids (SGCAA) medium (pH 4.5) for 2 days at 20°C. Yeast were incubated with biotinylated IL-13Rα1, tetramerized biotinylated IL-4Rα, or biotinylated IL-4Rα in the presence of biotinylated IL-13Rα1. IL-4Rα tetramers were formed by incubating 500 nM biotinylated IL-4Rα with 125 nM Alexa Fluor 647–conjugated streptavidin for 15 min on ice. Fluorescence was analyzed on an Accuri C6 flow cytometer.

Assembly, transformation, and selection of an IL-13 site-directed library

Assembly polymerase chain reaction (PCR) was performed with six overlapping primers, three of which contained the randomized codon sequences used for mutation [L10 (LFIV), R11 (RSNHLI), I14 (LFIV), V18 (LFIV), R86 (RKTM), D87 (EDKR), T88 (ITKR), K89 (RKTM), L101 (LFIYHN), K104 (RKTM), K105 (KTAE), F107 (LFIV), and R108 (RKTM)]. Numbers represent the amino acid residue mutated in IL-13, and the letters in parentheses represent the mutations introduced at those sites. The PCR product was further amplified with the primers 5′-GTAGCGGTGGGGGCGGTTCTCTGGAAGTTCTGTTCCAGGGTCCGAGCGGCGGATCCCCAGGCCCTGTGCCTC-3′ and 5′-AGATCTCGAGCAAGTCTTCTTCGGAGATAAGCTTTTGTTCGCCACCAGAAGCGGCCGCGTTGAACTGTCCCTC-3′. These primers also contained the necessary homology to the pCT302 vector sequence that is a requisite for homologous recombination. Insert DNA was combined with linearized vector backbone pCT302, and electrocompetent S. cerevisiae EBY100 cells were electroporated and rescued, as previously described (54), to form a library of 3 × 108 transformants. Selections were performed on this library by magnetic-activated cell sorting (Miltenyi). The first round of selection was performed with 2 × 109 cells from the yeast library, which provided about 10-fold coverage relative to the number of transformants. Subsequent rounds of selection used 1 × 107 yeast cells (greater than 10-fold coverage in each round). Naïve and sorted IL-13 libraries were grown fresh overnight at 30°C in SDCAA liquid medium (pH 4.5), followed by induction in SGCAA liquid medium (pH 4.5) for 2 days at 20°C. Monomeric selection with IL-13Rα1 was performed through sequential binding of IL-13Rα1, streptavidin–Alexa Fluor 647 (2 μg/ml), and 50 μl of Miltenyi anti–Alexa Fluor 647 microbeads. Fluorescence analysis was performed on an Acurri C6 flow cytometer.

SPR analysis

SPR experiments were performed on a Biacore T100 instrument with a Biacore streptavidin sensor chip (GE Healthcare). Biotinylated IL-13Rα1 was captured at a low density [50 to 100 response units (RU)], and kinetics measurements were performed at 30 μl/min. An unrelated biotinylated protein was immobilized as a reference surface for the streptavidin sensor chip with an RU that matched that of the experimental surface. All measurements were made with threefold serial dilutions of IL-13 variants in the running buffer [1× Hepes-buffered saline–surfactant P20 (GE Healthcare) and 0.1% bovine serum albumin]. The IL-13Rα1 bound to the chip surface was regenerated with 7 mM glycine (pH 3.0) and 250 mM NaCl. Kinetic parameters were determined with 120 to 190 s of IL-13 variant association time and 20 to 1200 s of dissociation time. All data fitting was performed with the Biacore T100 evaluation software version 2.0 with a 1:1 Langmuir binding model.

Crystallization and data collection

Crystals of the ternary complex of IL-13 A11, IL-13Rα1, and IL-4Rα were grown in sitting drops at 25°C by mixing 0.1 μl of protein solution [12 mg/ml in 10 mM Hepes (pH 7.2) and 150 mM NaCl] with an equal volume of 0.2 M lithium sulfate, 0.1 M sodium phosphate–citrate (pH 4.2), and 20% polyethylene glycol (PEG) 1000. Crystals grew in 2 to 5 days and were subsequently flash-frozen in liquid nitrogen with mother liquor containing 27% glycerol as a cryoprotectant. A 3.0-Å data set was collected at beamline 8.2.2, Advanced Light Source (ALS), University of California, Berkeley. The data set was indexed, integrated, and scaled with the HKL2000 software package (55). Data processing statistics are presented in table S1.

Structure determination and refinement

The IL-13 A11–containing ternary crystal structure was solved by molecular replacement with the program PHASER using the coordinates of IL-13, IL-13Rα1, and IL-4Rα separately [Protein Data Bank (PDB) accession no. 3BPN]. After all the domains were placed, the sequence of wild-type IL-13 was converted to that of the IL-13 A11 variant, and iterative rounds of refinement with PHENIX and model adjustment with COOT were used to refine the structures. Ramachandran analysis was performed with MolProbity ( All structural figures and overlays were prepared with PyMOL.

Live-cell, dual-color, single-molecule imaging studies

Single-molecule imaging of cytokine-receptor complexes in living cells was performed as described previously (7, 56). HeLa cells were grown to 50% confluence on glass cover slides coated with arginylglycylaspartic acid (RGD)–functionalized poly-l-lysine graft-PEG (PLL-PEG-RGD), which was prepared as described previously (57). For ligand binding studies, DY647IL-13 variants or DY647IL-4 was added to a final concentration of 2 nM and incubated for at least 5 min at room temperature before imaging. To investigate receptor dimerization, IL-13Rα1 and IL-4Rα were fused to N-terminal HaloTag and SNAPf tag, respectively. To this end, cDNAs encoding full-length IL-13Rα1 and IL-4Rα, respectively, without the N-terminal signal sequences were subcloned into the pDisplay vector (Invitrogen) at the Bgl II and Pst I sites (for IL-13Rα1) or the Xho I site (for IL-4Rα). Subsequently, cDNAs encoding the HaloTag and SNAPf tag, respectively, were inserted at the Bgl II site. The constructs including the signal sequence of the pDisplay vector (immunoglobulin κ) were transferred by restriction digest with Eco RI and Xho I into the pSems-26m vector (Covalys Biosciences) to generate pSems–HaloTag–IL-13Rα1 and pSems–SNAPf–IL-4Rα. HeLa cells seeded on PLL-PEG-RGD–coated glass cover slides were transfected with pSems–HaloTag–IL-13Rα1 and pSems–SNAPf–IL-4Rα by calcium phosphate precipitation. Four to five days later, the cover slides were mounted into home-built microscopy chambers, and the receptor molecules were labeled by their respective tags with fluorescent dyes through incubation with 20 nM HaloTag-TMR (Promega) and 40 nM SNAP–Surface 647 (NEB) for 15 min followed by five washes with phosphate-buffered saline. Labeled receptors are denoted as TMRIL-13Rα1 and DY647IL-4Rα henceforth. Imaging experiments were started immediately after labeling in cell culture medium supplemented with 1 mM ascorbic acid, catalase (40 μg/ml), 1 mM methylviologen, glucose (4.5 mg/ml; all from Sigma-Aldrich), and glucose-oxidase (0.5 mg/ml; Roche) to optimize the photostability of DY647 (58). Ternary complex formation was induced by adding the designated IL-13 variant to a final concentration of 200 nM, and images of the same cover slide were acquired for a maximum of 30 min after stimulation to reduce artifacts that may have arisen from cellular feedback mechanisms, such as receptor endocytosis or stress responses. Individual receptor molecules were simultaneously imaged for 50 to 300 frames at a frame rate of 31 Hz with an inverse IX71 microscope that was equipped with a triple-line total internal reflection illumination condenser (Olympus), a 150× TIRF objective (UAPO 150×/1.45 TIRFM, Olympus), a UNIBLITZ VMM-D4 shutter (Vincent Associates), an iXon DU-897 electron-multiplying charge-coupled device camera (512 × 512 pixels, Andor Technology), and a spectral image splitter (DV2, Optical Insights). The fluorescent dyes were simultaneously excited with a 200-mW 561-nm solid-state laser (CL561-200, CrystaLaser) and a 140-mW 642-nm diode laser (LuxX, Omicron-Laserage). Appropriate dichroic and emission filters were purchased from Semrock and Chroma Technology. The image splitter was calibrated with fluorescent beads that have pixel accuracy and therefore generated images that were further used to calculate a correction matrix that later enabled the evaluation algorithms to precisely colocalize the particles in both channels with subpixel resolution.

Single-particle tracking of receptors, ternary complex quantification, and diffusion analysis

Dual-color images were separated into their two respective channels, and localization of single molecules with submicrometer precision (10 to 40 nm), as well as tracking of their movement and filtering of immobile particles (nonspecifically bound fluorophores), was performed by in-house software on the basis of algorithms described previously (59). Individual TMRIL-13Rα1–DY647IL-4Rα dimers were identified by co-locomotion analysis (fig. S3F): after frame-by-frame colocalization of molecules in both channels with a distance threshold of 2 pixels (214 nm), colocalized molecules were subjected to tracking with a minimum trajectory length of 10 consecutive steps as cutoff (56). Co-trajectory length histograms were fitted to an exponential decay function to determine the average lifetimes of TMRIL-13Rα1–DY647IL-4Rα dimers. The fraction of receptors in a ternary complex was determined by PICCS, as described previously (38, 56). To this end, an in-house MATLAB script was used to calculate the correlated fraction of the coordinates of localized particles in both channels in a 25-μm2 region of interest in the first 20 frames (to minimize the reduction of cross-correlation by photobleaching) of each image stack. Because only those cells that showed similar amounts of both receptor subunits were analyzed, a single correlated fraction always refers to the channel with the limiting number of molecules. For each IL-13 variant, all correlated fraction values α from at least 10 cells were plotted as histograms and fitted by a Gaussian distribution to obtain a mean value for the correlated fraction α. The diffusion constants of TMRIL-13Rα1 in the absence of ligand and of TMRIL-13Rα1–DY647IL-4Rα dimers observed in the presence of wild-type IL-13 were determined from a collection of trajectories of the same cell by fitting a step-length histogram (t = 1, frame = 0.032 s) to a two-dimensional Gaussian probability distribution:P(r,t)=14πDter24Dt2πr(1)where P is the probability of finding a molecule with a diffusion constant D (in μm2/s) in a distance r (in μm) from its origin at time t0 = 0 after time t (in s). A similar histogram of TMRIL-13Rα1 trajectories of the same cell after the addition of wild-type IL-13 was fitted with a linear combination of the previously obtained distributions with their parameters as fixed constants and an additional parameter f (Eq. 2) to get an estimate of the fraction of IL-13Rα1 molecules moving in ternary complexes:P(r,t)=f·14πD1ter24D1t·2πr+(1f)·14πD2ter24D2t·2πr(2)

siRNA-mediated knockdown of IL-13Rα1

A549 cells were plated at 3 × 105 cells per well in 60-mm plates and incubated for 16 hours before being transfected. Cells were transfected with 25, 75, or 125 nM IL-13Rα1–specific siRNA (Sigma-Proligo) or control siRNA (Sigma-Proligo) per plate with Lipofectamine RNAiMAX (Invitrogen). Forty-eight hours after transfection, cells were plated in a 96-well plate and used for IL-13 stimulation and surface IL-13Rα1 determination.

TF-1 cell proliferation assays

Two thousand TF-1 cells per well were seeded in a 96-well plate and stimulated with the concentrations of wild-type or variant IL-13 indicated in the figure legends. After 96 hours of stimulation, the cells were harvested, and cell number was determined by flow cytometry–based counting on an Accuri C6 flow cytometer. The number of cells obtained for each agonist was plotted against the cytokine concentration to obtain sigmoidal dose-response curves, from which the EC50 values for TF-1 cell proliferation were calculated.

Dendritic cell studies

CD14+ monocytes were isolated (>97% purity) from peripheral blood mononuclear cells obtained from healthy blood donors (Stanford Blood Center) by density centrifugation with the RosetteSep Human Monocyte Enrichment Cocktail (STEMCELL Technologies), which was followed by magnetic separation with microbeads conjugated to antibodies against CD14 (Miltenyi Biotec). We subsequently cultured 0.5 to 1 × 106 CD14+ monocytes with GM-CSF (50 ng/ml) alone or with the indicated concentrations of the IL-13 variants in 12-well plates (Corning) containing Iscove’s modified Dulbecco’s medium (Gibco) supplemented with 10% (v/v) human AB serum, penicillin (100 U/ml), streptomycin (100 μg/ml), 2 mM l-glutamine, sodium pyruvate, nonessential amino acids, and 50 μM 2-mercaptoethanol. Fresh cytokines were added on days 2 and 4 of culture. Cells were processed between days 6 and 7 with 5 mM EDTA and subsequently stained with 4′,6-diamidino-2-phenylindole (Invitrogen), fluorescently labeled antibodies against CD86 (no. 555660) and CD209 (no. 551265), or appropriate fluorescently labeled isotype control antibodies. All antibodies were used at a 1:50 dilution. Differentiation of monocytes into dendritic cells was assessed by flow cytometry with a BD LSRII flow cytometer, and the MFIs of the cell surface markers CD86 and CD209 were determined with FlowJo analysis software (Tree Star).

Quantification of STAT6 nuclear translocation

HeLa cells stably expressing STAT6 N-terminally fused to mEGFP were imaged at 37°C with an inverted confocal laser scanning microscope FV1000 (Olympus) that was equipped with a 60× oil immersion objective (numerical aperture, 1.35; Olympus) and a temperature-controlled microscope stage system BC-110 (20/20 Technology). An Argon multiline laser GLG3135 (Showa Optronics) was used for the excitation of mEGFP at 488 nm. Appropriate dichroic and emission filters were obtained from Olympus. The diameter of the pinhole was set to 80 to 100 μm, and images were captured with FV1000 application software (FV10-ASW, Olympus). The nuclear translocation of mEGFP-STAT6 was initiated by the addition of 200 nM of each IL-13 variant, and single-section images were captured every 2 min for a total duration of 80 to 100 min at 37°C. The nuclear and cytoplasmic MFIs of 23 to 45 cells were measured, and background was subtracted with ImageJ software [National Institutes of Health (NIH)] for each time point. The nuclear/cytoplasmic fluorescence intensity ratios with respect to time t (in min) were calculated relative to the mean initial ratio at t = 0 and t = 2 min and averaged. The resulting sigmoidal curves were fitted tonc(t)=1+(ncend1)tnt1/2n+tn(3)to obtain the amplitude n/cend and half-time t1/2 (in min) of STAT6 translocation for each IL-13 variant. The Hill coefficient n was fixed at 2 because of the bimolecular homodimerization of STAT6 after activation.

Imaging of receptors in endosomes

HeLa cells were grown on PLL-PEG-RGD–coated glass cover slides and incubated with 2 nM DY647IL-13 wild-type or DY647IL-13 D7. After an incubation time of 15 min at 37°C followed by five washes, endosomes containing large amounts of labeled ligand were imaged in a highly inclined and laminated optical sheet (HILO) mode (60) with the TIRF microscopy apparatus described earlier. Endocytosis was blocked or decreased in some experiments by incubating the cells with 100 μM EHT 1864 (Santa Cruz Biotechnology) for 1 hour. To quantify endocytosed ligands, the integral intensity was quantified after background subtraction with a rolling cylinder algorithm (61). For direct co-visualization of receptors in endosomes, HeLa–TMRIL-13Rα1–DY647IL-4Rα cells that were used for dual-color, single-molecule imaging were imaged by laser scanning microscopy 15 min after receptor labeling and incubation at 37°C.

Statistical analysis

Data are expressed as means ± SD. Statistical differences between two groups were determined by Student’s t test.



Fig. S1. SPR binding sensograms for select IL-13 variants.

Fig. S2. Electron density of the site II and site III cytokine binding interfaces in the IL-13 A11–IL-13Rα1–IL-4Rα ternary complex.

Fig. S3. Visualization of IL-13–containing ternary complex assembly in live cells.

Fig. S4. Co-locomotion trajectories of HaloTag–IL-13Rα1 and SNAPf–IL-4Rα in response to IL-13.

Fig. S5. Diffusion properties of HaloTag–IL-13Rα1, SNAPf–IL-4Rα, and the dimerized complex.

Fig. S6. STAT6 activation induced by different concentrations of IL-13 variants.

Fig. S7. Functional characterization of IL-13 variants.

Fig. S8. Analysis of IL-13Rα1 and IL-4Rα abundances in different IL-13–responsive cells.

Fig. S9. Endocytosis of IL-13 receptor subunits.

Fig. S10. Simulations of STAT6 phosphorylation kinetics and dose-response curves.

Table S1. Data collection and refinement statistics (molecular replacement) for the IL-13 A11–IL-13Rα1–IL-4Rα ternary complex.

Movie S1. Quantification of receptor density and spatial organization by single-molecule imaging.

Movie S2. Single-molecule photobleaching experiments.

Movie S3. Single-molecule imaging of TMRIL-13Rα1 and DY647IL-4Rα.

Movie S4. Single-molecule photobleaching of TMRIL-13Rα1 and DY647IL-4Rα.

Movie S5. Single-molecule photobleaching of IL-13Rα1 and IL-4Rα.

Movie S6. Co-locomotion of individual IL-13Rα1–IL-4Rα dimers.


Acknowledgments: We thank members of the Garcia and Piehler laboratories for helpful advice and discussion, G. Hikade and H. Kenneweg for technical assistance, and P. Selenschik for support with PICCS analyses. Funding: This work was supported by the R01-AI51321 from the NIH (K.C.G.), the Howard Hughes Medical Institute (K.C.G.), the American Asthma Foundation (K.C.G.), and SFB 944 from the Deutsche Forschungsgemeinschaft (J.P.). Author contributions: I.M. and K.C.G. conceived the project; I.M., K.C.G., D.R., and J.P. wrote the manuscript; I.M. performed signaling, biophysical, and structural studies; J.P., D.R., S.W., and H.W. designed and performed single-particle microscopy experiments; J.P. and D.R. performed MATLAB simulation studies; K.J. and C.T. performed analysis of the crystal structure; M.M.S. and E.G.E. performed dendritic cell differentiation studies. Competing interests: The authors declare that they have no competing interests. Data and materials availability: The PDB accession number for the IL-13 A11–IL-4Rα–IL-13Rα1 complex reported here is 5E4E.
View Abstract

Stay Connected to Science Signaling

Navigate This Article