Through direct selection of peptide libraries in patients 
, we isolated a ligand peptide mimicking IL-11 from the prostate vasculature 
and proposed the IL-11Rα as a target during the progression of prostate cancer 
. Here we elected to explore the structural and functional attributes of this putative protein-protein interaction because the currently known binding sites within human IL-11 
do not encompass the selected peptide sequence within the native protein ().
Residues identified within known human IL-11 binding sites.
We used complementary strategies such as site-directed mutagenesis and NMR-based studies to establish that all six residues of the RRAGGS motif are likely required for binding to IL-11Rα. Because NMR analysis did not unambiguously reveal which of the arginine residues (Arg3 or Arg4) participates in receptor binding, studies based on targeted phage alanine scanning and IL-11 site-directed mutagenesis were indicated. In both cases, Arg4 turned out to be the key residue. All three methods also suggested that Gly and Ser residues contribute to the level of IL-11Rα binding by the ligands. Indeed, the Ser8 mutation abolished the binding of CGRRAGGSC to IL-11Rα; the Δδ observed in Ser8 induced by binding to IL-11Rα corroborated this result.
We have attempted to generate and interpret molecular models of IL-11 based on high-resolution structural studies available for other gp130-type cytokines, including IL-6 
and CNTF 
and on mutagenesis of IL-6, CNTF and LIF 
. In the case of the binding site for IL-11Rα or site I, directed mutagenesis of specific residues at the C-terminus of helix D and the AB loop established that these regions participate in receptor binding and cytokine-mediated bioactivity 
, analogous to IL-6 
. Both locations are positioned close to each other in these models and occupy the C-terminus of the four-helix bundle structure, which is situated opposite the BC loop. Similar structural and mutagenesis-based studies on gp130-type cytokines (vIL-6, CNTF and LIF) implicated residues within the BC loop in receptor recognition and bioactivity 
. However, such residues were part of site II or site III (for interaction with gp130 or gp130/LIF receptor), and not site I. Notably, the crystal structure of the IL-6 receptor complex hexamer demonstrated no role in IL-6 receptor binding for the BC loop, which is facing the N-terminal region of D2 in the second gp130 molecule 
. Our data show that CGRRAGGSC is a peptide mimic of IL-11, and as such is capable of recognizing and binding to IL-11Rα to activate cell signaling and proliferation. Although it is generally assumed that the model for the cell membrane-bound IL-11 receptor complex would be similar to that for IL-6 
, certain structural characteristics in the complex subunits may actually diverge significantly. For instance, we noticed the existence of four unreported leucine-zippers within IL-11 (three of them extending from the BC loop through the helix C, ). They have no clear parallel among most other gp130-type cytokines (only OSM and CT-1 appear to contain one of such regions). Moreover, there are two IL-11Rα membrane-bound isoforms that differ by the presence or absence of a cytoplasmic domain, whereas IL-6Rα has only one form with a longer cytoplasmic tail 
. Such isoforms could participate in the formation of a receptor complex unlike those observed in the IL-6Rα complex formation. Future detailed X-ray crystallography of the IL-11 receptor complex will further elucidate these fine structural features.
Functionally, the IL-11 mimic peptide CGRRAGGSC showed biological effects mediated by IL-11Rα (concentration-dependent stimulation of cell proliferation) both in the presence and in the absence IL-11. Furthermore, CGRRAGGSC-mediated cell proliferation in the presence of sIL-11Rα was reduced, and to a greater degree in the presence of native IL-11, but was not affected by an unrelated control receptor. Based on these findings, one might speculate on an interaction between the peptide CGRRAGGSC and the complex IL-11/IL-11Rα. Indeed, the soluble peptide CGRRAGGSC was found to induce STAT3 phosphorylation in the same manner as the IL-11 native cytokine. These results have biological precedent in other systems. Wrighton et al. 
isolated peptides that bind to and activate the erythropoietin receptor (EPOR); crystallographic analysis of the receptor complex revealed a peptide that generated a functional symmetric arrangement of the EPOR 
. Cwirla et al. 
selected peptide agonists for the thrombopoietin receptor and proposed an activation mechanism. Recent studies have also shown that orientation and residence time of ligand-receptors must be considered for activation 
. Moreover, peptides selected for binding to certain receptors bind to several sites on the receptor and activation might occur through conformational changes rather than multimerization 
From a supra-molecular viewpoint, IL-11 receptor complex formation is in all probability intricate: the ligand cytokine (IL-11) may need to bind first to a presenting receptor subunit (IL-11Rα) and subsequently recruit a dimer of signaling subunits (i.e., gp130). Studies of the related IL-6 receptor complex support the contention that preformed inactive dimers of receptor subunits exist in cell membranes 
. Previous reports determined that gp130-dimerization is not sufficient for receptor activation and that active conformational adjustment is required for a biological response 
. Therefore, it is possible that the soluble peptide CGRRAGGSC in complex with IL-11Rα might strengthen gp130-dimerization and/or induce a signaling-competent conformation.
The strength of combining various functional assays (site-directed mutagenesis of native proteins in tandem with ELISA plus peptide-alanine scanning in tandem with targeted phage binding assays) with structural studies (NMR-based spectroscopy of the peptide-receptor interaction) may overcome some of the limitations of mutational studies in the determination of binding sites of ligand-receptor interactions (such as difficult gene/protein expression, and mutant protein misfolding or steric hindrance).
In summary, we have shown that (i) the RRAGGS sequence (corresponding to a site within human IL-11, residues 112–117) mimics a candidate protein-binding site within IL-11, (ii) the molecular interaction between the peptide CGRRAGGSC and IL-11Rα is functional, and (iii) the IL-11-like motif induces proliferation of responsive cells through IL-11Rα-mediated STAT3 activation.