Table 1.
CXCR4 activation by SDF-1 mutantsa
| EC50 (nM)b | Foldedc | Fold Increase | p38 contact | |
|---|---|---|---|---|
| SDF-1 | 3.6 ± 1.4 | + | ||
| Arg20Ala | 4.3 ± 0.6 | + | 1.2 | + |
| Val23Ala | NA | − | NA | + |
| His25Arg | 5.1 ± 0.9 | + | 1.4 | − |
| Lys27Ala | 10.1 ± 2.9 | + | 2.8 | + |
| Lys27Glu | 16.8 ± 1.1 | + | 4.7 | + |
| Val39Ala | 27.1 ± 0.2 | + | 7.5 | + |
| Arg41Ala | 4.3 ± 0.9 | + | 1.2 | − |
| Arg47Ala | 14.1 ± 0.6 | + | 3.9 | + |
| Arg47Glu | 654 ± 93 | + | 181.7 | + |
| Val49Ala | 8.6 ± 2.4 | + | 2.4 | + |
| Glu60Ala | 4.1 ± 0.1 | + | 1.1 | − |
| Glu63Ala | 3.7 ± 0.8 | + | 1.0 | − |
| Lys64Ala | 5.0 ± 1.1 | + | 1.4 | − |
The current two-step, two-state model for CXCR4 activation implicates the SDF-1:p38 interaction in binding affinity and receptor specificity, but not in CXCR4 activation (12, 26). A peptide consisting of SDF-1 residues 1–8 fully activates CXCR4 at micromolar concentrations (49), and because each SDF-1 variant retained the native N-terminus, the EC50 value in the Ca2+ mobilization assay should reflect its apparent affinity for CXCR4. Consequently, an amino acid substitution that alters the EC50 for Ca2+ mobilization relative to that of wild-type SDF-1 has necessarily disrupted an interaction between the chemokine and the N-terminus or extracellular loops of CXCR4.
Values are reported as mean ± the standard deviation for two or more replicate measurements.
The 1H−15N HSQC spectrum displays chemical shifts similar to wild-type SDF-1 except for residues adjacent to the site of mutation.