Spectrin is a membrane skeletal protein best known for its structural role in maintaining cell shape and protecting cells from mechanical damage. Here, we report that α/βH-spectrin (βH is also called karst) dynamically accumulates and dissolves at the fusogenic synapse between fusing Drosophila muscle cells, where an attacking fusion partner invades its receiving partner with actin-propelled protrusions to promote cell fusion. Using genetics, cell biology, biophysics and mathematical modelling, we demonstrate that spectrin exhibits a mechanosensitive accumulation in response to shear deformation, which is highly elevated at the fusogenic synapse. The transiently accumulated spectrin network functions as a cellular fence to restrict the diffusion of cell-adhesion molecules and a cellular sieve to constrict the invasive protrusions, thereby increasing the mechanical tension of the fusogenic synapse to promote cell membrane fusion. Our study reveals a function of spectrin as a mechanoresponsive protein and has general implications for understanding spectrin function in dynamic cellular processes.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
We thank the Bloomington Stock Center for fly stocks, B. Paterson for the MHC antibody, F. Li for technical assistance, G. Zhang for help with the high-pressure freezing and freeze substitution method, J. Nathans for sharing confocal microscopes and D. Pan for critically reading the manuscript. This work was supported by the NIH grants (R01 AR053173 and R01 GM098816), the American Heart Association Established Investigator Award and the HHMI Faculty Scholar Award to E.H.C.; the NIH grants (R01 GM66817 and R01 GM109863) to D.N.R.; the NIH grants (R01 GM074751 and R01 GM114671) and the Chan Zucherberg Biohub Investigator Award to D.A.F.; the NSF grant MCB-1122013 to C.T.; the NSFC grant 11572316 to T.L.; and the NSFC grant 31771256 to R.D. R.D. was supported by an American Heart Association postdoctoral fellowship, K.S. by an American Heart Association Scientist Development Grant, D.M.L. by a Canadian Institute of Health Research postdoctoral fellowship and S.S. by a Life Sciences Research Foundation postdoctoral fellowship.
Integrated supplementary information
(a) Myoblast fusion defects caused by β-spectrin expression in muscle cells. Stage 15 embryos labeled with anti-MHC. Ventral lateral muscles of three hemisegments are shown in each panel. Unfused myoblasts are indicated by arrowheads. Expression of a functional, amino-terminal-tagged full-length Myc-β-Spec in all muscle cells with twi-GAL4 exacerbated the fusion defect in the β H -spec-/- mutant (left panel) and caused a minor fusion defect in wild-type (right panel) embryos. For each genotype, 10 embryos were imaged with similar results. Scale bar, 20 μm. (b–d) β-spectrin, adducin and protein 4.1 are not enriched at the fusogenic synapse. Fusogenic synapses (arrowheads) in stage 13–14 wild-type embryos marked by anti-Sltr. β-Spec (b), adducin (known as hu-li tai shao in Drosophila1) (c), and protein 4.1 (known as coracle in Drosophila2) (d) were visualized by immunostaining with respective antibodies. In each case, 10 embryos were imaged with similar results. Note that none of these proteins was enriched at the fusogenic synapse. Scale bar, 5 μm. (e) Ectopically expressed β-spectrin in muscle cells is not enriched at the fusogenic synapse. Myc-β-Spec was ectopically expressed with a muscle-specific mef2-GAL4 driver. βH-Spec and β-Spec were visualized by anti-βH-Spec and anti-Myc, respectively. The fusogenic synapse (arrowhead) was marked by phalloidin staining (F-actin) and the FCM was outlined with white dotted line. Note that βH-Spec was highly enriched at the fusogenic synapse, whereas β-Spec exhibited a general cortical localization pattern. Ten fusogenic synapses were imaged with similar results. Scale bar, 5 μm. (f) Mini-βH-spectrin is enriched at the fusogenic synapse. mCherry-βH-Spec was expressed in muscle cells with twi-Gal4 and visualized by the mCherry fluorescent signal. The fusogenic synapse (arrowhead) was marked by phalloidin staining (F-actin) and the FCM is outlined with white dotted line. Note that mini-βH-Spec is highly enriched at the fusogenic synapse. Immunostainings were repeated three times. Ten fusogenic synapses were imaged with similar results. Scale bar, 5 μm.
FRAP of βH-Spec in Drosophila embryonic epithelial cells. mCherry-βH-Spec was ectopically expressed in epithelial cells with the 69B-GAL4 driver. (a) Time-lapse stills of representative FRAP experiment in a stage 14 embryo. Arrowhead indicates the photo-bleached region. Scale bar, 5 μm. (b) Recovery kinetics of the mCherry fluorescence after photobleaching. The representative curve shows the fluorescence recovery of mCherry-βH-Spec from (a). The recovery half-time (t1/2) and percentage were quantified from multiple experiments. Each dot represents a fusogenic synapse; n = 31 fusogenic synapses were analyzed. The horizontal bars represent median value. The average t1/2 was 116 ± 38 sec (median: 112 sec) and percentage recovery was 64 ± 11% (median: 66%). Note that the half-time is longer than that at the fusogenic synapse (Fig. 2f).
(a, b) Actin and β-spectrin do not show mechanosensitive accumulation. (a) Representative DIC and fluorescent images of aspirated cells. Arrowheads indicate the base areas of aspirated cells. (i) No mechanosensitive accumulation of mCherry-α-Spec and GFP-actin when co-expressed. (ii) mCherry-α-Spec accumulation at the base area when co-expressed with βH-Spec. (iii) mCherry-β-Spec did not show mechanosensitive accumulation as did βH-Spec when the two spectrin proteins were co-expressed. Sample sizes ranged from n = 8 to 10 as reflected in the dot plots in (b). Scale bars, 5 μm. (b) Quantification of protein accumulation at the base areas of aspirated cells. Each data point represents an independent MPA experiment; n = 8, 8, 10, 10, 9, 9 (from left to right). The horizontal bars represent the mean ratio. An ANOVA with Fisher’s least significant difference test was applied to determine statistical significance. (c, d) The mechanosensitive accumulation of βH-spectrin is time- and force-dependent. (c) βH-Spec accumulation over time shown by MPA experiments. The data points were collected at a fixed pressure of 0.4 nN/μm2, and plotted every 3s for a total of 2.5 min. Different color codes indicate traces of individual cells examined (n = 12). Each raw data point was normalized to the initial value to remove variability caused by differences in protein distribution at the start of the experiment. The average values of the 12 cells measured at each time point are shown as black dots with error bars (SEM). βH-Spec accumulation reached its peak at around 80–90s after the onset of aspiration. (d) βH-Spec accumulation depends on applied force. Each value is the peak accumulation reached during the experiment normalized to the initial Ib/Io value. Note that βH-Spec accumulation increases with increased pressure.
(a) Domain structure of full length and mutant βH-Spec used in MPA and F-actin co-sedimentation assays. Actin-binding domains and the tetramerization site are indicated. Each distinct segment from the N-terminus to the C-terminus of βH-Spec is designated by a number. CH: calponin homology; SH3: Src homology 3; PH: pleckstrin homology; Spec: spectrin repeat. (b) F-actin co-sedimentation assay with purified βH-Spec fragments. The numbers (1, 29–31, 34) indicate βH-Spec fragments depicted in (a). S: supernatant; P: pellet. Note that βH-Spec fragment 1 precipitated with F-actin, whereas βH-Spec fragments 29–31 and 34 remained in the supernatant, confirming that the CH domains of βH-Spec bind F-actin. Experiments were repeated three times for each spectrin fragment with similar results.
Supplementary Figure 6 Working models on the function of the mechanoresponsive protein α/βH-spectrin at the fusogenic synapse.
(a) The mechanosensitive accumulation of α/βH-spectrin. Based on coarse-grained modeling, an invasive membrane protrusion generate by a pushing force causes maximal shear deformation (shape change) of the actin network in the receiving cell at the base area of the protrusion. The extensibility and flexibility of α/βH-spectrin heterotetramers enable them to accommodate a range of angle/distance changes triggered by shear stress and stay bound to the shear-deformed actin network, resulting in their mechanosensitive accumulation. Actin polymers and spectrin heterotetramers are not drawn to scale. Actin-binding domains at the ends of the spectrin heterotetramer are colored in blue, and the PH domains in the middle of the heterotetramer are indicated as a short stub. (b) α/βH-spectrin functions as a cellular fence and a cellular sieve. The top diagram shows a pair of a founder cell (FC; now a binucleated myotube) and an FCM engaging at the fusogenic synapse marked by the actin focus (green). The relative localization domains of α/βH-spectrin (Spec; red), Duf (blue) and F-actin (green) are shown at the early and late stages of the fusogenic synapse from different viewing angles (bottom four square panels). The plasma membrane is shaded in gray, and F-actin underneath the plasma membrane in the top view panels is shaded in light green. At the early stage, α/βH-spectrin accumulates at the base of invasive protrusions and is not closely associated with Duf due to their different modes of recruitment. After multiple rounds of mechanoresponsive feedback between the two fusion partners, α/βH-spectrin accumulates in large areas of the fusogenic synapse to (1) restrict Duf diffusion via biochemical interactions and/or molecular collision, and (2) constrict the diameter of invasive protrusions with spectrin-free microdomains within the spectrin network. See text for details.
Supplementary Figure 7 The contour plots of the maxima of dilation and shear deformations induced by invasive protrusions.
The maxima of dilation (a) and shear (b) deformations induced by invasive protrusions with different lengths (vertical axis) and radii (horizontal axis) were plotted. The values of the physical parameters can be found in the description of the coarse-grained molecular mechanics modeling in the Methods section. Note that the dilation and shear deformations increase with the protrusion length and decrease with the protrusion radius.
Supplementary Figures 1–7, Supplementary Video legends and Supplementary References.
Spectrin is dynamically associated with the F-actin focus at the fusogenic synapse
Spectrin accumulation correlates with the F-actin foci intensity
Dynamics of Spectrin at the fusogenic synapse
Dynamics of Spectrin at the cortex of epithelial cells
Mechanosensitive accumulation of Spectrin and MyoII induced by lateral indentation using AFM
Duf dispersal at the fusogenic synapse in α/β H -spectrin double-mutant embryos
Duf resides in a tight cluster at the fusogenic synapse in wild type embryos
F-actin dispersal in the absence of chemical signalling from Duf
βH-spectrin dispersal in duf,rst mutant embryos expressing DufΔintra
3D reconstruction of F-actin and βH-spectrin at the fusogenic synapse based on SIM imaging