Abstract 3396

The tertiary structure of cytokine-receptor complex provides information about molecular interactions trigger the receptor activation, the primary event of physiological response. Despite the low homology in primary structures, cytokines share a left-handed antiparallel four-helix bundle fold. Cytokine receptors are composed of a common domains and motifs and some cytokine receptors share functional and conservational subunits. We can discuss about the structural similarities and differences, however, the structural features caused a specific molecular recognition of individual cytokines are difficult to clarify. To further our understanding of structural basis of cytokines, cross-species comparison of a cytokine structure is an important clue. When we identified erythropoietin (EPO) in African clawed frog, Xenopus laevis, namely xlEPO, it was our surprise that xlEPO shares only 38% of the amino acid sequence with human EPO (hEPO) and lacks N-glycosylations. Dispite the low molecular similarity, xlEPO could stimulate proliferation and differentiation of erythrocytic cells transcending species. Meanwhile the activity of hEPO, sharing 45% homology with the functional N-terminal domain of human thrombopoietin (hTPO), is distinct from that of hTPO. These findings lead us to explore a particular topology specific to EPO-EPOR binding by solving the tertiary structure of xlEPO that enables us to compare that of EPO.

Since xl EPO has no carbohydrate structures, we purified biologically active recombinant xlEPO expressed in Escherichia coli to homogeneity. To examine the in vivo effects, frogs were administrated xlEPO (0.15–0.25 mg/kg/day B.W.) for consecutive 8 days. As a result, immature erythrocytes were appeared in the circulation on day 7, and gradually increased to account for 20% in total peripheral blood cells on day 11. Structural stability of xlEPO was measured by circular dichroism spectrometry, resulting that non-glycosylated xlEPO is much more stable against thermal denaturation than heavily-glycosylated hEPO. Furthermore, we acquired the structural information about xlEPO by X-ray crystal structure analysis. Unbound form of xlEPO was crystallized and diffracted to 2.9 Å. The structure was determined by molecular replacement method with hEPO complexed to its receptor (PDB ID: 1EER) as search model. This is the first crystal structure of unbound form of EPO. Overall tertiary structure of xlEPO shows a left-handed antiparallel four helical bundle fold. xlEPO is composed of four long α-helices (H1, P5-S28; H2, Q58-F80; H3, L86-L107 and H4, F128-R152) arranged in an “up-up-down-down” topology, connected by two long cross-over loops (H1-2 and H3-4) and one short loop (2-3). H1 and H4 are bridged by disulfide bond through C7 and C151. In addition to the four long α-helices, two much shorter helices are observed in the long connecting loops. One is located at between H1 and H2 (V48-K53), and one at between H3 and H4 (Q110-Q114). One short stretch of antiparallel β-sheet is formed within long cross-over loops: strand 1, I39-P42 lies on between H1 and H2; strand 2, T122-V125 between helices 3 and 4. The α-carbon RMSD (root mean square distance) between xlEPO and hEPO are 2.35 Å and 3.06 Å, with 1EER and the NMR structure (PDB ID: 1BUY), respectively. Residues involved in receptor binding were highly conserved with 0.78 Å for 12 residues of site 1, 0.65 Å for 12 residues of site 2, calculating with 1EER. These conformational similarities in receptor binding sites are thought to be responsible for cross-reactivity between human and Xenopus EPO. One distinct structural feature is that xlEPO has a shorter loop length than hEPO. Especially, long loop between H3 and H4 of xlEPO is consisted of 7 residues (T118-K124), much shorter than 14 residues (I119-T132) of hEPO. Eventually the secondary structure content of xlEPO showed 71.5%, higher than 63.3%, that of hEPO. In addition, the number of glycine in α-helix, which is known to destabilize the helix, is 6 residues in hEPO, larger than in xlEPO, 3 residues. These structural features would account for the large difference in structural stability between xlEPO and hEPO. Together with the crystal structure of unbound form of xlEPO, the tertiary structure of receptor-binding form of xlEPO would enable us to understand the conformational change of xlEPO, and to design EPO mimetics based on newer concepts.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution