Constitutive activation and oncogenicity are mediated by loss of helical structure at the cytosolic boundary of thrombopoietin receptor mutant dimers

Figures

• 

  Figure 1.. Activity and relative position of asparagine (Asn) mutants in murine thrombopoietin receptor (TpoR). The murine TpoR (mTpoR) wild type (WT) transmembrane domain sequence and numbering. The residues that make the KWQFP motif are highlighted in red (a). A helical wheel diagram shows the positions of L491, V494, S498, and G502 (blue) on the active interface (b). The relative tranlscriptional activity of mTpoR WT or mutants was assessed with STAT5 reporter Spi-Luc in HEK-293T cells. Shown are averages of separate experiments ± SEM (n = 3–4), each experiment being performed with three biological repeats for each condition (triplicates). Kruskal–Wallis non-parametric test with multiple-comparisons using Steel’s test with mTpoR WT as control (jmp pro12). ***p<0.0001; n.s., nonsignificant (c, upper panel). Cartoon illustrating how asparagine residues are predicted to promote crossing of transmembrane helices (c, lower panels). Short-term cell growth was measured using the Cell-Titer-Glo luminescent cell viability assay (Cell Titer Glo, Promega) with stable Ba/F3 cell lines expressing mTpoR WT or mutants with no ligand or increasing concentrations of Tpo ligand (0.1, 1, 10, and 100…

• 

  Figure 2.. In vivo data comparing asparagine (Asn) murine thrombopoietin receptor (TpoR) mutants. Red blood cell (RBC), platelets, white blood cells (WBC), and neutrophils were measured using a MS9 blood cell counter at day 40 post-transplantation. Values shown represent the average of at least six biological replicates ± SEM., Kruskal–Wallis nonparametric test with Dunn’s multiple-comparisons test (Prism 9.1.2); *p<0.05, **p<0.01 (a). Bone marrow/spleen cellularity and fibrosis have been evaluated by hematoxylin-eosin (HE) and reticulin (RET) staining, respectively. 3× zoom was performed on the .jpg file with Adobe Illustrator. (b). Liver and spleen were photographed and weighted (c). Values shown represent the average of two biological replicates ± SEM. Figure 2—source data 1.Raw data and statistics (Prism 9.1.2) for Figure 2a.Red blood cell, platelets, white blood cells, and neutrophils measured, using an MS9 blood cell counter, at day 40 after transplantation of C57BL/6 female mice with retrovirus-infected C57BL/6 male bone marrow for expression murine thrombopoietin receptor (mTpoR) wild type (WT) or mutant. Figure 2—source data 2.Raw data for Figure 2c.Liver and spleen weights (in…

• 

  Figure 2—figure supplement 1.. Mouse models expressing V494N, S498N, and G502N mutants of thrombopoietin receptor (TpoR) show weak, mild, and strong signs of myeloproliferation with a significant reduction of survival in the case of G502N mutants. Peripheral blood smears from mouse models expressing the indicated constructs were stained with May–Grunwald Giemsa (×10). 3× zoom was performed on the .jpg file with Adobe Illustrator. (a) Kaplan–Meier survival analysis was performed on the indicated mice. The G502N model shows significant decrease of survival with a median survival of 49.5 d in the G502N mice. Statistical analysis was performed using the Mantel–Cox Test. *p<0.05 (b). Figure 2—figure supplement 1—source data 1.Raw data and statistics (Prism 9.1.2) for Figure 2—figure supplement 1b.Survival after transplantation of C57BL/6 female mice with retrovirus-infected C57BL/6 male bone marrow for expression of murine thrombopoietin receptor (mTpoR) wild type (WT) or mutant.

• 

  Figure 3.. Secondary structure changes in the RWQFP insert upon activation. For solid-state NMR studies, peptides corresponding to the transmembrane and juxtamembrane (TM-JM) regions of thrombopoietin receptor (TpoR) were reconstituted into model membrane bilayers (a). 1D 13C MAS NMR spectra in the region of the 13C=O carbonyl resonances are shown of murine TpoR (mTpoR) wild type (WT) (b), mTpoR S498N (c), mTpoR G502N (d), human TpoR (hTpoR) WT (e), hTpoR W515K (f), hTpoR W515K+2 alanines (+2A) (g), mTpoR WT (h), mTpoR A499N (i), and mTpoR A499N+2A (j). The mTpoR peptides labeled with 1-13C L505 (murine) or 1-13C L512 (human) were reconstituted into DMPC:PG bilayers. A single MAS NMR spectrum of unlabeled mTpoR or hTpoR (dashed line) was obtained with a high signal:noise ratio for subtraction from the spectra of the reconstituted labeled peptides to obtain the resonance of the L505 (murine) or L512 (human) 13C=O in the lower panels. For the reconstituted 13C-labeled peptides, two independent technical replicates (reconstitutions and NMR data sets) were obtained for each experiment.

• 

  Figure 3—figure supplement 1.. Solution NMR comparison of wild-type and W515K thrombopoietin receptor (TpoR) transmembrane peptides. Solution NMR spectroscopy provides an alternative method to investigate structural changes in the TM-JM boundary of the TpoR upon mutation. We have probed the structural changes occurring between the wild-type hTpoR and the W515K mutant using 13C, 15N labeled TM-JM peptides expressed in E. coli and solubilized in detergent micelles (Brett, 2012). The construct sequence chosen (residues 481–520) stretches from the end of the β-sheet rich D2 domain on the extracellular side of the TM region of the receptor to two residues after Pro518 of the cytoplasmic RWQFP motif. (a) 1H-15N HSQC spectrum of wild-type human TpoR (residues 481–520) reconstituted into sodium dodecylsulfate (SDS) detergent micelles. Threonine and glycine resonances below 114 ppm are not shown in order to highlight the dispersion of the 1H-15N resonances. The concentration of the TpoR peptide was 1 mM in 10 mM sodium phosphate, pH 4.7, 85× CMC d25-SDS, and 10% D2O (v:v). Spectra were collected at 320 K, 32 scans each. (b) Calculated values of τc values for the WT and W515K human TpoR (481–520) in SDS micelles. The…

• 

  Figure 3—figure supplement 2.. AlphaFold predictions of thrombopoietin receptor (TpoR) W515X mutants. Predicted structures of key human TpoR mutants (purple) and wild type (blue). Residue 515 is shown in red. Predictions were performed with AlphaFold 2.0 (Jumper et al., 2021) running with mmSEQ2 for multiple sequence alignment (MSA).

• 

  Figure 3—figure supplement 3.. AlphaFold predictions of RRQFP, RKQFP, WRQFP, and WKQFP mutants. AlphaFold 2.0 (Jumper et al., 2021) predictions of human thrombopoietin receptor (TpoR) (residues 474–573) wild-type (blue), active mutants (purple), and inactive mutants (green). The residue W515 and residues that differ from the wild-type sequence (W515K/R and R514W) are shown in red.

• 

  Figure 4.. Localized substitution-induced loss of α-helicity of cytosolic juxtamembrane helix activates thrombopoietin receptor (TpoR) signaling in the absence of Tpo. Sequences of transmembrane (TM) and juxtamembrane (JM) sequences of murine TpoR (mTpoR) and position of substitutions to Gly-Pro, which interrupt secondary helical structure, and Ala substitutions, which are predicted to maintain secondary structure (a). The human TpoR (hTpoR) numbering adds +7 in the TM domain. For example, W515 in hTpoR is W508 in mTpoR. Models generated with AlphaFold2 (Jumper et al., 2021; Varadi et al., 2022) of the murine TM-JM domain for the wild-type receptor and with introduction of the GP mutation at position 512–513 and 513–514 with local disruption of the helix (b). STAT5 transcriptional activity in the absence or presence of thrombopoietin (Tpo) of the indicated mTpoR (c, d) or hTpoR mutants (e) in γ–2A cells. Shown are averages of three independent experiments each done with 2–3 biological replicates ± SEM, Kruskal–Wallis nonparametric test with multiple-comparisons Steel’s test with control (jmp pro12); *p<0.05, **p<0.01. Short-term proliferation assay (Cell Titer Glo, Promega) was performed on…

• 

  Figure 4—figure supplement 1.. W491A mutation inhibits activation by Gly-Pro substitutions after the RWQFP motif. STAT5 transcriptional activity reported by Spi-Luc in the absence or presence of thrombopoietin (Tpo) or eltrombopag (Elt) of human TpoR (hTpoR) wild type (WT) or mutants in HEK-239T cells. Shown are averages of three independent experiments each done with three biological replicates ± SEM, Kruskal–Wallis nonparametric test with multiple-comparisons Steel’s test with control (jmp pro12); *p<0.05, **p<0.01, ns, nonsignificant. Figure 4—figure supplement 1—source data 1.Raw data, scatter plot, and statistics (Prism 9.1.2, jmp pro12) for Figure 4—figure supplement 1.STAT5 transcriptional activity of the human thrombopoietin receptor (hTpoR) wild type (WT) or mutant assessed in HEK-293T cells by luciferase assay with STAT5 reporter Spi-Luc.

• 

  Figure 5.. Effects of +2 alanine insertions on asparagine (Asn) and W508K mutants. Alanine (Ala) insertion mutagenesis in the murine thrombopoietin receptor (mTpoR) at the indicated position is predicted to extend the helix and rotate clockwise by 206° (a). Model generated with AlphaFold2 (Jumper et al., 2021; Varadi et al., 2022) of wild type (WT) mTpoR transmembrane domain (TMD) (right panel) and with insertion of 2 Ala (right panel) (b). Helical wheel diagram showing the positions of residues I485 to F511 relative to active and inactive interfaces of mTpoR. In italic are indicated the positions of coiled coil (cc) fusion TpoR constructs (cc-mTpoR) each adopting a particular transmembrane dimeric interface as described in Staerk et al., 2011 (c). Spi-Luc STAT5 transcriptional activity in the absence or presence of thrombopoietin (Tpo) for the indicated mTpoR (d–f) or human TpoR (hTpoR) (g) WT or mutant constructs. Shown are averages of 3–10 independent experiments, each done with three biological replicates ± SEM in HEK293T cells and JAK2-deficient γ–2A cells; Kruskal–Wallis nonparametric test with multiple-comparisons Steel’s test with control (jmp pro12); *p<0.05, ***p<0.001, ns,…

• 

  Figure 6.. Thrombopoietin receptor (TpoR) transmembrane domain dimerization induces membrane binding of the C-terminal switch region and influence the proximity between JAK2 domains. Solid-state 13C NMR experiments using magic angle spinning and the INEPT pulse sequence were used to address membrane binding in the juxtamembrane (JM) region of the transmembrane and juxtamembrane (TM-JM) peptides of human TpoR (hTpoR). INEPT spectra were obtained at 37°C using hTpoR reconstituted into DMPC:PG bilayers. TpoR peptides were U-13C labeled at W529 at the C-terminus of the switch region. Two independent technical replicates (reconstitutions and NMR data sets) were obtained for each experiment (a–c). Dimerization of JAK2 bound to the indicated murine TpoRs (mTpoR) was assessed by Gaussia luciferase in HEK-293T cells (d–f). Shown are averages of separate experiments ± SEM (n = 8); each experiment being performed with three biological repeats for each condition (triplicates). Kruskal–Wallis nonparametric test with multiple-comparisons Steel’s test with controls (jmp pro12); **p<0.01, ****p<0.0001. Figure 6—source data 1.Raw data, scatter plot, and statistics (Prism 9.1.2, jmp pro12) for Figure 6e.JAK2…

• 

  Figure 6—figure supplement 1.. Dimerization of JAK2 is influenced by murine thrombopoietin receptor (TpoR) transmembrane mutations irrespectively of JAK2 activity. Dimerization of kinase-dead JAK2 K882D bound to the indicated receptors was assessed by split Gaussia luciferase in HEK-293T cells. Shown are averages of separate experiments ± SEM (n = 5); each experiment being performed with three biological repeats for each condition (triplicates). Kruskal–Wallis nonparametric test with multiple-comparisons Steel’s test with controls (jmp pro12); *p<0.05, ***p<0.001, ns, nonsignificant.