Structure‐Function analysis of human IL‐6: Identification of two distinct regions that are important for receptor binding

Abstract

Interleukin-6 (IL-6) is a multifunctional cytokine that plays an important role in host defense. It has been predicted that IL-6 may fold as a 4 a-helix bundle structure with up-up-down-down topology. Despite a high degree of sequence similarity (42%) the human and mouse IL-6 polypeptides display distinct species-specific activities. Although human IL-6 (hIL-6) is active in both human and mouse cell assays, mouse IL-6 (mIL-6) is not active on human cells. Previously, we demonstrated that the 5 C-terminal residues of mIL-6 are important for activity, conformation, and stability , Protein Sci 2:1472-1481). To further probe the structurefunction relationship of this cytokine, we have constructed several human/mouse IL-6 hybrid molecules. Restriction endonuclease sites were introduced and used to ligate the human and mouse sequences at junction points situated at Leu-62 (Lys-65 in mIL-6) in the putative connecting loop AB between helices A and B, at Arg-113 (Val-117 in mIL-6) at the N-terminal end of helix C, at Lys-150 (Asp-152 in mIL-6) in the connecting loop CD between helices C and D, and at Leu-178 (Thr-180 in mIL-6) in helix D. Hybrid molecules consisting of various combinations of these fragments were constructed, expressed, and purified to homogeneity. The conformational integrity of the IL-6 hybrids was assessed by far-UV CD. Analysis of their biological activity in a human bioassay (using the HepG2 cell line), a mouse bioassay (using the 7TD1 cell line), and receptor binding properties indicates that at least 2 regions of hIL-6, residues 178-184 in helix D and residues 63-113 in the region incorporating part of the putative connecting loop AB through to the beginning of helix C, are critical for efficient binding to the human IL-6 receptor. For human IL-6, it would appear that interactions between residues Ala-180, Leu-181, and Met-184 and residues in the N-terminal region may be critical for maintaining the structure of the molecule; replacement of these residues with the corresponding 3 residues in mouse IL-6 correlated with a significant loss of a-helical content and a 200-fold reduction in activity in the mouse bioassay. A homology model of mIL-6 based on the X-ray structure of human granulocyte colony-stimulating factor is presented.

Interleukin‐6: Structure‐function relationships

“…IL-6 has been expressed in rabbit reticulocyte lysate (Geiger et al, 1988), Escherichia coli (Brakenhoff et al, 1987;Tonouchi et al, 1988;Kriittgen et al, 1990a;Yasueda et al, 1990;Arcone et al, 1991a;Grennet et al, 1991;Zhang et al, 1992;Hammacher et al, 1994), Saccharomyces cerevisiae (Guisez et al, 1991), Pichia pastoris (this laboratory, unpublished data), Aspergillus nidulans , Xenopus laevis (Fontaine et al, 1991), baculovirus insect cells and Chinese hamster ovary (CHO) cells (Eisenthal et al, 1993;Orita et al, 1994).…”

“…These similarities between IL-6 and G-CSF led three groups to use the structure of G-CSF (Hill et al, 1993) as a template to construct three-dimensional models of hIL-6 (Ehlers et al, 1994;Savino et al, 1994b) and mIL-6 (Hammacher et al, 1994). All these models show the expected four-helix bundle topology of the helical cytokines with a small helix in the AB-loop, though the model of Savino et al (1994b) has a second helix in the AB-loop similar to that found in hGH (de Vos et al, 1992).…”

“…Each group's alignments of the helices of G-CSF and IL-6 are identical except for helix D. Ehlers et al (1994) align the C-tenninii of the molecules while the other groups align Trp 157 (conserved in all IL-6 species) with the conserved phenylalanine seen in the "Dl motif" (Bazan, 1991) of the other members of the gp130 family of cytokines. This shifting of the register of helix D by three residues between the models will affect the packing of the helix bundle, and how helix D interacts with residues 52-62 in the putative AB-loop helix of IL-6 (Hammacher et al, 1994). The sequence of IL-6 aligned to G-CSF, as used by Savino et al (1994b) and Hammacher et al (1994), is shown in Figure 3 along with the structural alignment of G-CSF, LlF, CNTF, and GH.…”

“…This shifting of the register of helix D by three residues between the models will affect the packing of the helix bundle, and how helix D interacts with residues 52-62 in the putative AB-loop helix of IL-6 (Hammacher et al, 1994). The sequence of IL-6 aligned to G-CSF, as used by Savino et al (1994b) and Hammacher et al (1994), is shown in Figure 3 along with the structural alignment of G-CSF, LlF, CNTF, and GH.…”

“…G-CSF (Hill et ai., 1993). LIF (Robinson et al, 1994), CNTF , and GH (de Vos et al, 1992) are aligned based on optimal superpositions of their respective crystal structures, and IL-6 is aligned to G-CSF as used in the modeling of Hammacher et al (1994). The LIF secondary structure is taken from the crystal structure of murine LIF (Robinson et al, 1994).…”