Phosphatidylserine orchestrates Myomerger membrane insertions to drive myoblast fusion

Abstract

Muscle cell fusion is a multistep process where the final step of the reaction drives progression beyond early hemifusion events to complete fusion. This step requires activity of the muscle-specific fusogen Myomerger, a single-pass transmembrane protein containing 84 amino acids with an ectodomain that includes two α-helices. Previous studies have demonstrated that Myomerger acts by destabilizing membranes through generation of elastic stresses in the outer leaflet of the plasma membrane. An obvious question is how such destabilizing activity might be regulated to avoid membrane and cellular damage, and how the two juxtaposed helices cooperate in fusion. Using cellular fusion assays and in vitro liposome assays, we report that the two helices possess unique characteristics, both of which are needed for full activity of the protein. We demonstrate that externalized phosphatidylserine (PS), a lipid previously implicated in myoblast fusion, has a determinant role in the regulation of Myomerger activity. The membrane-proximal, amphipathic Helix-1 is normally disordered and its α-helical structure is induced by PS, making membrane interactions more efficacious. The distal, more hydrophobic Helix-2 is intrinsically ordered, possesses an ability to insert into membranes, and augments the membrane-stressing effects of Helix-1. These data reveal that Myomerger fusogenic activity is an exquisitely orchestrated event involving its two ectodomain helices, which are controlled by membrane lipid composition, providing an explanation as to how its membrane-stressing activity is spatially and temporally regulated during the final step of myoblast fusion.

Keywords: Myomerger/Myomixer; membrane fusion; myoblast fusion; phosphatidylserine.

Fig. 1.

Myomerger ectodomain requires Helix-1 and Helix-2 for fusogenic activity. (A) Helical wheel projections of Helix-1 and Helix-2 were generated using https://pss.sjtu.edu.cn/cgi-bin/wheel.cgi. (B) Coomassie-stained SDS-PAGE of recombinantly purified Myomerger ectodomain constructs WT-Ecto, Helix-1, and Helix-2. (C) Treatment of Myomerger^−/− myoblasts with 0.5 µM WT-Ecto or Helix-1 for 20 hours on day 3 of differentiation induced myotube formation (arrows) but Helix-2 had no effect. Cells were immunostained with myosin antibodies and Hoechst dye was used to stain nuclei. (D) Percentage of myosin⁺ cells that contain 3 or ≥4 nuclei in Myomerger^−/− myoblasts after incubation with ectodomain proteins. (D) Statistical analyses and data presentation. Data are presented as mean ± SD; one-way ANOVA with a Tukey’s post hoc test; NS, not significant; ***P < 0.001, ****P < 0.0001. Results shown are from two independent protein purifications that were tested on Myomerger^−/− myoblasts. (Scale bar, 50 µm.)

Fig. 2.

Helix-1 and Helix-2 contribute to the membrane-stressing activities of Myomerger. (A) Schematic representation of the liposome leakage assay. Upon disruption of the liposomal membrane, the self-quenched sulforhodamine B is released, generating an increase in fluorescence signal. (B) Sulforhodamine B fluorescence after addition of various concentrations of recombinant proteins. Results are plotted as a percentage of complete liposome disruption by SDS. (C) Diagram of the lipid mixing assay. One set of liposomes (PC:PS 70:30) was labeled with a FRET donor (DiI) and an acceptor (DiD), and mixed with unlabeled liposomes (PC:PS 70:30). Lipid mixing was detected as an increase in the donor emission. (D) WT-Ecto induced lipid mixing but Helix-1 or Helix-2 did not. (B) Statistical analyses and data presentation. Data are presented as mean ± SD; one-way ANOVA with a Tukey’s post hoc test; *P < 0.05, **P < 0.01, ****P < 0.0001. For liposome leakage and lipid mixing assays, two independent protein purifications were assayed with two independent liposome preparations.

Fig. 3.

PS enhances membrane interactions of the amphipathic Helix-1 through induction of helical structure. (A) A schematic of the liposome flotation assay using a Ficoll gradient to assess binding of protein with liposomes. (B) Supernatant and pellet fractions were separated with SDS-PAGE and protein was detected using Myomerger antibodies. The proteins present in different fractions were quantified and are represented as a percentage of protein bound (Right). (C) WT-Ecto did not induce lipid mixing in the absence of PS (Left). Lipid mixing was observed only at high concentrations when labeled PC:PS liposomes were mixed with unlabeled PC liposomes (Right). (D) CD spectra of Helix-1, Helix-2, and WT-Ecto synthetic peptides (without the C-terminal tail) in the absence or presence of lipids (PC or PC:PS). Apo is the solution without peptide. (B) Statistical analyses and data presentation. Data are presented as mean ± SD; one-way ANOVA with a Tukey’s post hoc test; **P < 0.01, ****P < 0.0001. For liposome flotation and lipid mixing, two independent protein purifications were assayed with two independent liposome preparations.

Fig. 4.

Myomerger activity is controlled by cell-surface PS. (A) A permeabilization assay to determine the effects of PS on Myomerger function. Myomerger^−/− myoblasts were differentiated and treated with WT-Ecto with or without TopFluor PS, and then treated with water (to induce osmotic shock) and PI (a cell-impermeant DNA-binding dye). The presence of PI indicates permeabilization. (B) Representative images show an increase in PI-labeled nuclei (yellow arrows) with WT-Ecto–treated cells when cells were incubated with TopFluor PS (Left). Quantification of PI⁺ cells (Right). (C) Cells were treated with A01 (an inhibitor of the Tmem16 family) prior to incubation with WT-Ecto to reduce outer-leaflet PS. Images show that A01 reduced the number of PI⁺ cells (yellow arrows). Quantification of the effect of A01 treatment (Bottom Right). (B and C) Statistical analyses and data presentation. Data are presented as mean ± SD; one-way ANOVA with a Tukey’s post hoc test; ****P < 0.0001. For experiments on Myomerger^−/− myoblasts, two independent protein purifications were assayed with two independent liposome preparations. (Scale bars, 25 µm.) KO, knockout.

Fig. 5.

PS regulates Helix-1 of Myomerger insertion into lipid heads but not lipid tails. (A) A tryptophan residue was added at the end of Helix-1 (Myomerger^Leu40Trp). Changes in the fluorescence intensity at 332 nm were plotted against different PC or PC:PS (70:30) liposome concentrations. (B) Schematic of the FRET insertion assay. PC or PC:PS (70:30) liposomes were prepared with either a FRET pair of head-placed probes or FRET pair of acyl chain–placed probes. Efficiency of FRET for probes at the lipid heads or at the lipid tails was plotted against the protein concentration. (C) Slopes for WT ectodomain with lipid head probes and lipid acyl probes with and without PS. (D) FRET in lipid heads and lipid tails for Helix-1 peptide. (C and D) Statistical analyses and data presentation. Data are presented as mean ± SD; Welch’s t test; NS, not significant; *P < 0.05, ****P < 0.0001. Slopes in C and D are from three or four independent titrations.

Fig. 6.

Amphipathic characteristic of Myomerger Helix-1 and membrane interactions of Helix-2 are required for activity. (A) ΔArg Helix-1 expression in Myomerger^−/− myoblasts was not able to induce fusion as assayed by immunostaining with myosin antibodies. Nuclei were labeled with Hoechst. Quantification of fused myosin⁺ cells (Right). (B) Fused myotubes were observed in Myomerger^−/− cells expressing ΔLeu Helix-2, although quantification of fusion showed a reduction compared with WT (Right). (C) FRET insertion assay for Ecto^WT and Ecto^{ΔLeu Helix-2} in both head and tail regions. (D) Schematic of Myomerger-induced pore formation during myoblast fusion. In a prefusion stage with low levels of PS on the outer leaflet of the plasma membrane, Helix-1 remains disordered whereas Helix-2 is inserted into the acyl chains of the bilayer. Accumulation of PS on the outer leaflet triggers structuring of Helix-1 causing disruption of lipid head groups driving positive curvature needed for pore formation and expansion. (A–C) Statistical analyses and data presentation. Data are presented as mean ± SD; two-way ANOVA with a Tukey’s post hoc test (A and B) and Welch’s t test (C); **P < 0.01, ***P < 0.001, ****P < 0.0001. All data shown are representative of at least two independent experiments performed in duplicate. Slopes in C are from five to seven independent titrations. (Scale bars, 50 µm.)