VOLUME 288 (2013) PAGES 10024–10034
The panels labeled R2B in Fig. 3 should have been labeled R2C and vice versa. Similarly, the Kd values listed in Table 2 for R2B and R2C should be reversed. The corrected data are inconsistent with the speculation (page 10030) that the chemokine may have evolved to be optimally responsive (in terms of dimer dissociation) to the biologically dominant form of the receptor but do not change the main conclusions of this work.