H499 and W491 play opposite roles in regulating activation of TpoR by several activating mutations. (A) Effect of TpoR H499Y mutation on constitutive activity of the S505N mutant. (B) H499Y and H499Y-S505N increase TpoR cell-surface localization. (C) Effects of aromatic and aliphatic residues at H499 on TpoR activity. (D) Effects of H499Y/H499C on the constitutive activities of the indicated TpoR mutants (upper) and on their cell-surface localization (lower). (E) Model of the dimeric active S505N and inactive W491A TM domains of the TpoR. On the left, in the active S505N mutant, the N505 residues face one another. On the right, in the inactive W491A mutant, the H499 residues face one another and would therefore be inaccessible to eltrombopag (Elt). (F) Effects of W491A/W491K mutations on the indicated TpoR mutants and on stimulations by Tpo or Elt. In panels A, C, D, and F, HEK-293T cells were transiently transfected with the indicated hemagglutinin (HA)-TpoR variants, along with complementary DNAs coding for JAK2, STAT5, and SpiLuc Firefly luciferase reporter reflecting STAT5 transcriptional activity and normalized with a control reporter (pRL-TK) containing Renilla luciferase. Shown are averages of 3 to 4 independent experiments, each performed in triplicate + standard deviation (error bars). In panels B and D, HEK-293T cells were transiently transfected with the indicated HA-TpoR variants. TpoR cell-surface localization among fluorescein isothiocyanate–positive cells was assessed by flow cytometry using anti-TpoR antibody. Shown is 1 representative of 2 experiments. **P < .01, ***P < .001. EC, extracellular; eV, empty vector; ns, nonsignificant (nonparametric multiple-comparison Steel test with control); PE, phycoerythrin; PIG, pMX-IRES-GFP vector.