Constitutive activation and oncogenicity are mediated by loss of helical structure at the cytosolic boundary of thrombopoietin receptor mutant dimers
Curation statements for this article:-
Curated by eLife
eLife assessment
This very well-written paper advances our understanding of the mechanism of activation of the thrombopoietin receptor (TpoR), a very important cytokine receptor that regulates megakaryocyte differentiation and platelet production. The authors supply an elegant combination of NMR and cell biology experiments to support their conclusions and the data are of high quality.
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Molecular Biology (eLife)
Abstract
Dimerization of the thrombopoietin receptor (TpoR) is necessary for receptor activation and downstream signaling through activated Janus kinase 2. We have shown previously that different orientations of the transmembrane (TM) helices within a receptor dimer can lead to different signaling outputs. Here we addressed the structural basis of activation for receptor mutations S505N and W515K that induce myeloproliferative neoplasms. We show using in vivo bone marrow reconstitution experiments that ligand-independent activation of TpoR by TM asparagine (Asn) substitutions is proportional to the proximity of the Asn mutation to the intracellular membrane surface. Solid-state NMR experiments on TM peptides indicate a progressive loss of helical structure in the juxtamembrane (JM) R/KWQFP motif with proximity of Asn substitutions to the cytosolic boundary. Mutational studies in the TpoR cytosolic JM region show that loss of the helical structure in the JM motif by itself can induce activation, but only when localized to a maximum of six amino acids downstream of W515, the helicity of the remaining region until Box 1 being required for receptor function. The constitutive activation of TpoR mutants S505N and W515K can be inhibited by rotation of TM helices within the TpoR dimer, which also restores helicity around W515. Together, these data allow us to develop a general model for activation of TpoR and explain the critical role of the JM W515 residue in the regulation of the activity of the receptor.
Article activity feed
-
-
Author Response
Reviewer #1 (Public Review):
The authors sought to define the molecular mechanism of activation of the thrombopoietin receptor (TpoR), a very important cytokine receptor that regulates megakaryocyte differentiation and platelet production. They conducted a thorough series of experiments combining mutagenesis experiments with sophistical biological assays and that also includes solid-state NMR structural measurements. This work builds on a body of previous studies of TpoR from this group and from others. They focused both on (1) the role and impact of W515 located in the juxtamembrane cytosolic domain and (2) the impact of introducing either Asn at sites in the transmembrane domain to induce various dimerization modes, or insertion of pairs of Ala residues to induce helical rotation to the TM domain. There is a lot of …
Author Response
Reviewer #1 (Public Review):
The authors sought to define the molecular mechanism of activation of the thrombopoietin receptor (TpoR), a very important cytokine receptor that regulates megakaryocyte differentiation and platelet production. They conducted a thorough series of experiments combining mutagenesis experiments with sophistical biological assays and that also includes solid-state NMR structural measurements. This work builds on a body of previous studies of TpoR from this group and from others. They focused both on (1) the role and impact of W515 located in the juxtamembrane cytosolic domain and (2) the impact of introducing either Asn at sites in the transmembrane domain to induce various dimerization modes, or insertion of pairs of Ala residues to induce helical rotation to the TM domain. There is a lot of nice data in this paper, which is fairly intricate - a tough read, but that's because it's a complicated system. The writing is excellent.
This paper presents a model for receptor activation in which the inactive receptor is the monomeric form of the receptor in which the juxtamembrane domain, including W515, maintains a helical structure. Activation of the receptor triggers dimerization of the transmembrane domain and loss of helicity of the juxtamembrane segment, which facilitates optimal interactions of the kinase domains with their JACK2 domain phosphorylation substrates.
There is a lot to like in this careful work and the resulting manuscript. There is one major shortcoming in this manuscript, which concerns W515. It is known that mutation of W515 to any of 17 of the canonical amino acids, including Phe, is sufficient to trigger homodimerization and receptor activation. The authors present some evidence that the phenomenon behind this is that mutation of W515 to almost any other residues disrupts the helical secondary structure of the critical juxtamembrane segment, which promotes dimerization and receptor activation. What I find puzzling is why a Trp at site 515 promotes helix formation, but nearly all other amino acids at this site disrupt helix formation. This strongly suggests the side chain of W515 must be interacting with another domain of the protein in the inactive state, in a manner that is responsible for how Trp stabilizes the juxtamembrane helix, which is a central feature that helps define that state. I think that for this paper, this dangling missing piece of their mechanistic model should be resolved.
We agree with the reviewers that the mechanism by which Trp515 stabilizes the TM helix is central to the mechanism of activation. More broadly, our studies over the past decade have sought to address the importance of the entire RWQFP insert in the TM domain. Our working model for this sequence has been that cation-π interactions are central to the role of the Trp and the accompanying amino acids.
Arginine and tryptophan both are over-represented at the cytoplasmic TM-JM boundaries of membrane proteins. Arginine is positively charged and part of the “positive-inside” rule for membrane protein insertion. Arginine and lysine define the cytoplasmic ends of TM helices and prefer to be accessible to the water-exposed membrane surface. In contrast, tryptophan residues prefer hydrophobic head-group or membrane interior locations. A revealing aspect of the RWQFP motif is that the arginine and tryptophan are located at the membrane to cytosolic border. As a result, in order to accommodate arginine in a more water-inaccessible membrane environment, it interacts with the surface of the tryptophan indole ring. Partitioning of the RWQF sequence in a more water-inaccessible environment also drives the formation of helical secondary structure as an unpaired backbone C=O...NH in a hydrophobic environment is estimated to cost 3-6 kcal/mol of energy.
We have taken two approaches in respond to this essential criticism of the reviewers: one structural and one computational. Additional NMR data (structural approach) has been included in the supporting information (see response to point 2 below). Computational approaches provide a second way to address whether a cation– interaction between Trp515 and the positively charged Arg514 is responsible for stabilizing the C-terminal TM helix. We have included a new supporting figure using Alpha-Fold 2.0 that probes the structural changes upon mutation of Trp515. In the wild-type receptor, Arg514 is predicted to form a cation– interaction with Trp515. In the W515K mutant, the helical secondary structure in the RKQFP sequence is disrupted and Arg514 forms a new cation– interaction with Trp529. Similar changes occur in other Trp515 mutants (e.g. W515A) highlighting the ability of Alpha-Fold to predict such interactions and the consequences of mutation. Overall, 15 out of 19 W515X mutants are predicted to be unfolded. Experimentally, 17 out of 19 mutations lead to activation. Importantly, W515C and W515P are the only two amino acid substitutions that do not cause constitutive activity experimentally (Defour, Chachoua, Pecquet, & Constantinescu, 2016). Computationally, these two sites do not predict helix unraveling. In short, the overall the predictions of Alpha-Fold agree with the unique nature of tryptophan at position 515.
In addition, we have expanded the arguments supporting the potential role of cation–π interactions by adding a new section entitled “Unfolding of the RWQF -helical motif is a common mechanism of receptor activation”.
These modifications are now in the revised manuscript starting with line 213:
Our working model for the mechanism of activation in the wild-type or mutant receptors is that the RWQF motif is stabilized in the inactive state as an -helix as a result of a cation- interaction between R514 and W515. This interaction allows the RWQF sequence to partition into the more hydrophobic head-group region of the bilayer. Both Arg and Trp are over-represented at the cytoplasmic ends of TM helices (von Heijne, 1992), but whereas Arg prefers a water-accessible environment, Trp prefers to be buried in a more hydrophobic environment (Yau, Wimley, Gawrisch, & White, 1998). Since Arg and Trp are located at the border between membrane and cytosolic domains and Arg precedes Trp in the sequence, partitioning into the membrane head-group region results in a favorable interaction of the positive charge associated with the guanidinium group of the R514 side chain with the partial negative charge associated with the aromatic surface of the W515 side chain. Partitioning of the RWQF sequence into the more water-inaccessible environment drives the formation of helical secondary structure as an unpaired backbone C=O...NH in a hydrophobic environment is estimated to cost 6 kcal/mol of energy (Engelman, Steitz, & Goldman, 1986). In this model, activation of the receptor results in or is caused by disruption of the R514-W515 cation-π interaction. In the W515 mutants, R514 is no longer stabilized in a membrane environment and the helix containing the RWQFP sequence unravels to allow the positively charged side chain to reach outside of the membrane. In the case of the Asn mutants and in the wild-type receptor with bound Tpo, dimerization of hTpoR (or rotation of the TM helices in mTpoR dimer), places W515 in the center of the helix-helix interface. The data suggest that a steric clash of the W515 side chains results in unraveling of the cytoplasmic end of the TM helix.
Computational and additional NMR data are provided in the supplementary figures to support the model of helix unraveling suggested by the solid-state NMR studies. Computationally, we used AlphaFold 2.0 (Jumper et al., 2021) calculations of hTpoR TM-JM peptides to predict the influence of all possible mutations at position 515 on the TM-JM helix structure. Remarkably, -helix unraveling was predicted for 15 out of 20 possible amino acids at 515 (supplement 2 to Figure 3). Importantly, two of the mutations that are not predicted to cause helix unraveling are W515C and W515P. Experimentally, these two amino acid substitutions are the only ones that do not induce constitutive activity among all possible amin oacid substitutions at W515 (Defour et al., 2016). Introducing a Trp at the preceding position 514 instead of R/K in W515K/R mutants reverses helix unfolding in AlphaFold simulations (supplement 3 to Figure 3). This result agrees with our previous data that the WRQFP mutant is inactive and is essentially monomeric (J. P. Defour et al., 2013). Structurally, we have undertaken solution-NMR studies of the wild-type hTpoR TM-JM peptide and its W515K mutant. Relaxation measurements of the backbone 15N resonances show that W515K mutation leads to association of the TM helices, and that it induces upfield chemical shift changes in the RWQF sequence consistent with helix unraveling (supplement 1 to Figure 3).Reviewer #2 (Public Review):
The thrombopoietin receptor (TpoR) regulates stem cell proliferation, platelet production, and megakaryocyte differentiation. Past cell biology and biophysical studies have established that ligand-induced dimerization constitutes the mechanism of activation of TpoR. Specifically, ligands bind to the extracellular domain of TpoR and generate an allosteric response that is transmitted to the transmembrane domain, activating downstream signaling. However, up to now the molecular details of how the allosteric signals are transmitted to the intramembrane domains have been elusive. In this manuscript, Constantinescu and co-workers combined NMR, in vitro, and in vivo assays to investigate the activation and oncogenicity of TpoR. The authors concluded that the unwinding of the juxtamembrane domain is the main structural event that determines TpoR activation and regulates oncogenicity. The solid-state NMR studies were carried out in lipid membranes with polypeptides spanning the juxtamembrane and transmembrane residues. The authors show a series of spectra of 13CO resonances that encompass the juxtamembrane domain that is diagnostic of a structural transition from a helical conformation to a partially disordered state. The unwinding of the helical juxtamembrane domain was confirmed by site-specific mutations in this region. The chemical shift changes clearly indicate the transition from order to disorder (and vice versa) for selected sites. These conclusions are compounded by INEPT-type experiments that detect the most dynamic region of polypeptides. To rationalize the molecular mechanism for activation, the authors also used Ala-Ala insertions at strategic positions along the transmembrane domain. These experiments showed that the specific orientation of the transmembrane residues is central for TpoR activation, and a slight rotation of the helix is critical for activation of the receptor. Transcriptional activity assays confirm the importance of the proper orientation of the transmembrane domain for receptor activation.
Overall, I believe the data are solid, and both biophysical and cell biology studies support the conclusions of the authors. These new findings represent a significant advancement in understanding cytokine receptor activation.
We thank the reviewer for these comments.
Reviewer #3 (Public Review):
The authors sought to propose a mechanism by which cancer-causing mutations in the thrombopoietin receptor (TpoR) activate the receptor. To do so, they used a systematic approach of introducing non-native and naturally occurring mutations into the receptor and use a combination of in-vivo and cell-based assays and solid-state NMR spectroscopy. They propose that the proximity of the asparagine mutations to the cytosolic boundary influences the secondary structure of the receptor and suggests that this structural change induces receptor activation.
The strengths of this work are the importance of the system being studied and tackling a problem that is not yet fully resolved. The authors acquired a large and convincing set of biological data, including in vivo experiments that support the gain-of-function/activating role of the mutations studied. The solid-state NMR data are of high quality as well. In particular, the INEPT data in figure 6a display very clear differences within one region of the wild-type compared to the mutants.
One significant weakness is the validity of the conclusions given the limited atomistic measurements presented. Namely, the authors make rather specific conclusions about protein folding based on a single set of 13C alanine carbonyl chemical shifts in the wild-type and mutant TM peptides. Essentially, the authors observe chemical shift perturbations at this carbonyl carbon when mutations are introduced into a protein and use this information to make conclusions about secondary structure. I am not convinced that the authors have presented sufficient evidence to justify the conclusion that the helix unwinds and that this is responsible for the mechanism of activation. While the other cell-based experiments in mutations are interesting, deciphering such a specific folding mechanism with limited atomistic data is not justified.
We added both computational data and solution NMR to support our conclusion.
-
eLife assessment
This very well-written paper advances our understanding of the mechanism of activation of the thrombopoietin receptor (TpoR), a very important cytokine receptor that regulates megakaryocyte differentiation and platelet production. The authors supply an elegant combination of NMR and cell biology experiments to support their conclusions and the data are of high quality.
-
Reviewer #1 (Public Review):
The authors sought to define the molecular mechanism of activation of the thrombopoietin receptor (TpoR), a very important cytokine receptor that regulates megakaryocyte differentiation and platelet production. They conducted a thorough series of experiments combining mutagenesis experiments with sophistical biological assays and that also includes solid-state NMR structural measurements. This work builds on a body of previous studies of TpoR from this group and from others. They focused both on (1) the role and impact of W515 located in the juxtamembrane cytosolic domain and (2) the impact of introducing either Asn at sites in the transmembrane domain to induce various dimerization modes, or insertion of pairs of Ala residues to induce helical rotation to the TM domain. There is a lot of nice data in this …
Reviewer #1 (Public Review):
The authors sought to define the molecular mechanism of activation of the thrombopoietin receptor (TpoR), a very important cytokine receptor that regulates megakaryocyte differentiation and platelet production. They conducted a thorough series of experiments combining mutagenesis experiments with sophistical biological assays and that also includes solid-state NMR structural measurements. This work builds on a body of previous studies of TpoR from this group and from others. They focused both on (1) the role and impact of W515 located in the juxtamembrane cytosolic domain and (2) the impact of introducing either Asn at sites in the transmembrane domain to induce various dimerization modes, or insertion of pairs of Ala residues to induce helical rotation to the TM domain. There is a lot of nice data in this paper, which is fairly intricate - a tough read, but that's because it's a complicated system. The writing is excellent.
This paper presents a model for receptor activation in which the inactive receptor is the monomeric form of the receptor in which the juxtamembrane domain, including W515, maintains a helical structure. Activation of the receptor triggers dimerization of the transmembrane domain and loss of helicity of the juxtamembrane segment, which facilitates optimal interactions of the kinase domains with their JACK2 domain phosphorylation substrates.
There is a lot to like in this careful work and the resulting manuscript. There is one major shortcoming in this manuscript, which concerns W515. It is known that mutation of W515 to any of 17 of the canonical amino acids, including Phe, is sufficient to trigger homodimerization and receptor activation. The authors present some evidence that the phenomenon behind this is that mutation of W515 to almost any other residues disrupts the helical secondary structure of the critical juxtamembrane segment, which promotes dimerization and receptor activation. What I find puzzling is why a Trp at site 515 promotes helix formation, but nearly all other amino acids at this site disrupt helix formation. This strongly suggests the side chain of W515 must be interacting with another domain of the protein in the inactive state, in a manner that is responsible for how Trp stabilizes the juxtamembrane helix which is a central feature that helps define that state. I think that for this paper, this dangling missing piece of their mechanistic model should be resolved.
-
Reviewer #2 (Public Review):
The thrombopoietin receptor (TpoR) regulates stem cell proliferation, platelet production, and megakaryocyte differentiation. Past cell biology and biophysical studies have established that ligand-induced dimerization constitutes the mechanism of activation of TpoR. Specifically, ligands bind to the extracellular domain of TpoR and generate an allosteric response that is transmitted to the transmembrane domain, activating downstream signaling. However, up to now the molecular details of how the allosteric signals are transmitted to the intramembrane domains have been elusive. In this manuscript, Constantinescu and co-workers combined NMR, in vitro, and in vivo assays to investigate the activation and oncogenicity of TpoR. The authors concluded that the unwinding of the juxtamembrane domain is the main …
Reviewer #2 (Public Review):
The thrombopoietin receptor (TpoR) regulates stem cell proliferation, platelet production, and megakaryocyte differentiation. Past cell biology and biophysical studies have established that ligand-induced dimerization constitutes the mechanism of activation of TpoR. Specifically, ligands bind to the extracellular domain of TpoR and generate an allosteric response that is transmitted to the transmembrane domain, activating downstream signaling. However, up to now the molecular details of how the allosteric signals are transmitted to the intramembrane domains have been elusive. In this manuscript, Constantinescu and co-workers combined NMR, in vitro, and in vivo assays to investigate the activation and oncogenicity of TpoR. The authors concluded that the unwinding of the juxtamembrane domain is the main structural event that determines TpoR activation and regulates oncogenicity. The solid-state NMR studies were carried out in lipid membranes with polypeptides spanning the juxtamembrane and transmembrane residues. The authors show a series of spectra of 13CO resonances that encompass the juxtamembrane domain that is diagnostic of a structural transition from a helical conformation to a partially disordered state. The unwinding of the helical juxtamembrane domain was confirmed by site-specific mutations in this region. The chemical shift changes clearly indicate the transition from order to disorder (and vice versa) for selected sites. These conclusions are compounded by INEPT-type experiments that detect the most dynamic region of polypeptides. To rationalize the molecular mechanism for activation, the authors also used Ala-Ala insertions at strategic positions along the transmembrane domain. These experiments showed that the specific orientation of the transmembrane residues is central for TpoR activation, and a slight rotation of the helix is critical for activation of the receptor. Transcriptional activity assays confirm the importance of the proper orientation of the transmembrane domain for receptor activation.
Overall, I believe the data are solid, and both biophysical and cell biology studies support the conclusions of the authors. These new findings represent a significant advancement in understanding cytokine receptor activation.
-
Reviewer #3 (Public Review):
The authors sought to propose a mechanism by which cancer-causing mutations in the thrombopoietin receptor (TpoR) activate the receptor. To do so, they used a systematic approach of introducing non-native and naturally occurring mutations into the receptor and use a combination of in-vivo and cell-based assays and solid-state NMR spectroscopy. They propose that the proximity of the asparagine mutations to the cytosolic boundary influences the secondary structure of the receptor and suggests that this structural change induces receptor activation.
The strengths of this work are the importance of the system being studied and tackling a problem that is not yet fully resolved. The authors acquired a large and convincing set of biological data, including in vivo experiments that support the …
Reviewer #3 (Public Review):
The authors sought to propose a mechanism by which cancer-causing mutations in the thrombopoietin receptor (TpoR) activate the receptor. To do so, they used a systematic approach of introducing non-native and naturally occurring mutations into the receptor and use a combination of in-vivo and cell-based assays and solid-state NMR spectroscopy. They propose that the proximity of the asparagine mutations to the cytosolic boundary influences the secondary structure of the receptor and suggests that this structural change induces receptor activation.
The strengths of this work are the importance of the system being studied and tackling a problem that is not yet fully resolved. The authors acquired a large and convincing set of biological data, including in vivo experiments that support the gain-of-function/activating role of the mutations studied. The solid-state NMR data are of high quality as well. In particular, the INEPT data in figure 6a display very clear differences within one region of the wild-type compared to the mutants.
One significant weakness is the validity of the conclusions given the limited atomistic measurements presented. Namely, the authors make rather specific conclusions about protein folding based on a single set of 13C alanine carbonyl chemical shifts in the wild-type and mutant TM peptides. Essentially, the authors observe chemical shift perturbations at this carbonyl carbon when mutations are introduced into a protein and use this information to make conclusions about secondary structure. I am not convinced that the authors have presented sufficient evidence to justify the conclusion that the helix unwinds and that this is responsible for the mechanism of activation. While the other cell-based experiments in mutations are interesting, deciphering such a specific folding mechanism with limited atomistic data is not justified.
-
