Structural Basis of Lipid Membrane Binding by Human Ferlins

Structural Basis of Lipid Membrane Binding by Human Ferlins | bioRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-M677548'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search New Results Structural Basis of Lipid Membrane Binding by Human Ferlins Constantin Cretu , Aleksandar Chernev , Csaba Zoltan Kibedi Szabo , Vladimir Pena , Henning Urlaub , View ORCID Profile Tobias Moser , Julia Preobraschenski doi: https://doi.org/10.1101/2025.01.25.634844 Constantin Cretu 1 Institute for Auditory Neuroscience and InnerEarLab, University Medical Center Göttingen , Göttingen, Germany 2 Biochemistry of Membrane Dynamics Group, Institute for Auditory Neuroscience, University Medical Center Göttingen , Göttingen, Germany 3 Cluster of Excellence "Multiscale Bioimaging: from Molecular Machines to Networks of Excitable Cells" (MBExC), University of Göttingen , Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site For correspondence: constantin.cretu{at}med.uni-goettingen.de julia.preobraschenski{at}med.uni-goettingen.de tmoser{at}gwdg.de Aleksandar Chernev 4 Bioanalytical Mass Spectrometry Group, Max Planck Institute for Multidisciplinary Sciences , Göttingen, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Csaba Zoltan Kibedi Szabo 5 Research Group Mechanisms and Regulation of Splicing, The Institute of Cancer Research , London, UK Find this author on Google Scholar Find this author on PubMed Search for this author on this site Vladimir Pena 5 Research Group Mechanisms and Regulation of Splicing, The Institute of Cancer Research , London, UK Find this author on Google Scholar Find this author on PubMed Search for this author on this site Henning Urlaub 3 Cluster of Excellence "Multiscale Bioimaging: from Molecular Machines to Networks of Excitable Cells" (MBExC), University of Göttingen , Germany 4 Bioanalytical Mass Spectrometry Group, Max Planck Institute for Multidisciplinary Sciences , Göttingen, Germany 6 Bioanalytics Group, Institute for Clinical Chemistry, University Medical Center Göttingen , Göttingen, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Tobias Moser 1 Institute for Auditory Neuroscience and InnerEarLab, University Medical Center Göttingen , Göttingen, Germany 2 Biochemistry of Membrane Dynamics Group, Institute for Auditory Neuroscience, University Medical Center Göttingen , Göttingen, Germany 3 Cluster of Excellence "Multiscale Bioimaging: from Molecular Machines to Networks of Excitable Cells" (MBExC), University of Göttingen , Germany 7 Auditory Neuroscience and Synaptic Nanophysiology Group, Max-Planck-Institute for Multidisciplinary Sciences , Göttingen, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Tobias Moser For correspondence: constantin.cretu{at}med.uni-goettingen.de julia.preobraschenski{at}med.uni-goettingen.de tmoser{at}gwdg.de Julia Preobraschenski 1 Institute for Auditory Neuroscience and InnerEarLab, University Medical Center Göttingen , Göttingen, Germany 2 Biochemistry of Membrane Dynamics Group, Institute for Auditory Neuroscience, University Medical Center Göttingen , Göttingen, Germany 3 Cluster of Excellence "Multiscale Bioimaging: from Molecular Machines to Networks of Excitable Cells" (MBExC), University of Göttingen , Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site For correspondence: constantin.cretu{at}med.uni-goettingen.de julia.preobraschenski{at}med.uni-goettingen.de tmoser{at}gwdg.de Abstract Full Text Info/History Metrics Preview PDF Abstract Ferlins, ancient membrane proteins with a unique architecture, are central to multiple essential, Ca 2+ -dependent vesicle fusion processes. Despite numerous functional studies and their link to burdening human diseases, a mechanistic understanding of how these multi-C 2 domain proteins interact with lipid membranes to promote their remodeling and fusion is currently lacking. Here, we elucidate the near-complete cryo-electron microscopy structures of human myoferlin and dysferlin in their Ca 2+ and lipid-bound states. We show that ferlins adopt compact, ring-like tertiary structures achieved upon membrane binding. The top arch of the ferlin ring, comprising the C 2 C-C 2 D region, is rigid and varies little across the observed functional states. In contrast, the N-terminal C 2 B and the C-terminal C 2 F-C 2 G domains cycle between alternative conformations and, in response to Ca 2+ , close the ferlin ring, promoting tight interaction with the target membrane. Probing key domain interfaces validates the observed architecture and informs a model of how ferlins engage lipid bilayers in a Ca 2+ - dependent manner. This work reveals the general principles of human ferlin structures and provides a framework for future analyses of ferlin-dependent cellular functions and disease mechanisms. Introduction Excitable cells rely on precisely timed Ca 2+ signals to trigger exocytosis in neural synapses and contraction in muscle cells ( Clapham, 2007 , Jahn & Fasshauer, 2012 , Kuo & Ehrlich, 2015 , Luan & Wang, 2021 ). Less is known how the uncontrolled influx of Ca 2+ through a large membrane lesion promotes a rapid acute response ( Cooper & McNeil, 2015 ). C 2 - domain proteins are key molecular players utilizing the incoming Ca 2+ signal to tether intracellular vesicles to their target membranes and ultimately promote their fusion by coordinated binding to phospholipids and other factors ( Rizo, 2022 , Rizo & Sudhof, 1998 ). Beyond the well-studied synaptotagmins, which feature only two consecutive C 2 domains ( Rizo & Sudhof, 1998 , Sudhof, 2013 ), ferlins, predicted to comprise up to eight C 2 domains ( Dominguez, McCord et al., 2022 ), play a pivotal role in mediating these processes ( Bansal & Campbell, 2004 , Cooper & Head, 2015 , Cooper & McNeil, 2015 , Pangrsic, Reisinger et al., 2012) and are critically needed at multiple steps (Chen, Monga et al., 2024, Cooper & McNeil, 2015 , Pangrsic et al., 2012 , Rizo, 2022 ). Ferlins, such as dysferlin (FER1L1), myoferlin (FER1L3), and otoferlin (FER1L2), form an ancient group of C 2 domain Ca 2+ -sensing proteins present in almost all eukaryotic lineages ( Bansal & Campbell, 2004 , Han & Campbell, 2007 , Pangrsic et al., 2012 , Petit, Bonnet et al., 2023 ). Dysferlin and myoferlin are highly expressed in skeletal and heart muscle cells, which are prone to membrane injuries during contractions ( Bansal & Campbell, 2004 , Cooper & McNeil, 2015 , Pramono, Tan et al., 2009 ). Distributed at the sarcolemma, its specialized internal structures and various endomembrane vesicles ( Paulke, Fleischhacker et al., 2024 ), dysferlin has been proposed to play major roles in Ca 2+ -dependent membrane resealing, in the biogenesis and maintenance of the transverse-tubules system ( Kerr, Ziman et al., 2013 , Paulke et al., 2024 ), and other trafficking pathways ( Cooper & McNeil, 2015 , Glover & Brown, 2007 , Han & Campbell, 2007 ). The central physiological role of dysferlin in maintaining the integrity of muscle cell membranes is evident from over 400 disease-causing DYSF mutations identified in limb-girdle muscle dystrophy type 2B (LGMD2B) and Miyoshi myopathy (MM) – rare autosomal recessive muscle wasting disorders characterized by deficiencies in sarcolemma repair (Bansal, Miyake et al., 2003, Cooper & Head, 2015 ). Similar to dysferlin, myoferlin ( Cooper & McNeil, 2015 , Davis, Delmonte et al., 2000, de Morree, Hensbergen et al., 2010), has been linked to diverse membrane remodelling and organelle repair events, including in other cell types, as well as to myoblast membrane fusion during myogenesis ( Davis et al., 2000 , Doherty, Cave et al., 2005). However, unlike dysferlin, myoferlin has recently also been found to be overexpressed in various human cancers and the altered vesicular trafficking and function of myoferlin has been linked to cancer cell proliferation, metastasis, and resistance to chemotherapy ( Cooper & McNeil, 2015 , Gupta, Yano et al., 2021 , Zhang, Li et al., 2018 ). The closely related otoferlin is mainly expressed in sensory hair cells of the inner ear, and several hundred pathogenic mutations cause the deafness DFNB9 ( Moser & Starr, 2016 , Petit et al., 2023 , Santarelli, del Castillo et al., 2015 , Vona, Rad et al., 2020 , Yasunaga, Grati et al., 1999 ). In mechanistic terms, it has been suggested that otoferlin is involved in Ca 2+ -sensing for synaptic vesicle fusion ( Johnson & Chapman, 2010 , Michalski, Goutman et al., 2017 , Roux, Safieddine et al., 2006 ) and replenishment ( Pangrsic, Lasarow et al., 2010 , Vogl, Panou et al., 2016 ), as well as exocytosis-endocytosis coupling ( Duncker, Franz et al., 2013 , Jung, Maritzen et al., 2015 , Kroll, Jaime Tobon et al., 2019 , Strenzke, Chakrabarti et al., 2016 ), and its mutations disrupt synaptic sound encoding, resulting in an auditory synaptopathy ( Moser & Starr, 2016 , Pangrsic et al., 2012 ). Despite the recent insights into the individual roles of ferlins and them representing targets for gene therapy or drug discovery ( Al-Moyed, Cepeda et al., 2019 , Gupta et al., 2021 , Llanga, Nagy et al., 2017, Moser, Chen et al., 2024 , Zhang et al., 2018 ), the underlying molecular mechanisms, especially concerning how they promote membrane remodelling and fusion through concerted Ca 2+ and phospholipid binding, have remained largely unresolved, in part, due to limited structural information on their active states. In structural terms, ferlins arguably possess the most unique and complex architecture among the known Ca 2+ -sensitive C 2 -domain proteins ( Lek, Evesson et al., 2012 , Pangrsic et al., 2012 ). The ferlin amino-terminal (N-terminal) cytoplasmic domain is predicted to contain up to eight β-sandwich C 2 domains ( Dominguez et al., 2022 ). The ferlin C 2 domains are thought to be connected by unstructured linker regions and are followed by a carboxy-terminal (C- terminal), single-pass transmembrane region, anchoring the proteins to cellular membranes. In addition, type-I ferlins, such as dysferlin and myoferlin, have two accessory domains with an unknown function, DysF and FerA, inserted between the third and fourth C 2 domains ( Dominguez et al., 2022 ). Generally, similar to synaptotagmins ( Chapman, 2008 , Rizo, 2022 , Rizo & Sudhof, 1998 ), the individual C 2 domains of ferlins have variable Ca 2+ and phospholipid binding activities (Abdullah, Padmanarayana et al., 2014 , Marty, Holman et al., 2013), with few exceptions (otoferlin’s C 2 A domain (Helfmann, Neumann et al., 2011, Johnson & Chapman, 2010 )). However, previous studies failed to clarify how full-length ferlins are precisely organized to interact with lipid membranes and which of their structural motifs are critical for membrane binding. Consequently, it remains poorly understood how ferlins act on lipid membranes to promote their remodelling and fusion, and how Ca 2+ - sensitive conformational rearrangements mediate these processes ( Lek et al., 2012 , Pangrsic et al., 2012 , Xu, Pallikkuth et al., 2011 ). Herein, we leverage structural biology approaches and functional analyses to obtain near complete and high-resolution cryo-EM models of the two largest human ferlins, myoferlin and dysferlin, in their Ca 2+ and lipid-bound states. Besides revealing the intricate organization of these essential vesicle trafficking factors, our ferlin structures shed light on how the C 2 and accessory motifs engage lipid bilayers in a coordinated manner. We further advance a model of how ferlins cycle between alternative conformational states to transiently bind lipid membranes and facilitate their remodelling and fusion. Results 2.4 Å cryo-EM structure of membrane-bound human myoferlin To obtain the first complete structure of a human ferlin, we established the heterologous expression and purification of full-length myoferlin and dysferlin ( Fig 1A and Appendix Fig S1A ). In addition to the membrane-anchored constructs, we expressed the entire cytosolic region of the ferlins, comprising all C 2 and accessory motifs ( Appendix Fig S1A ). Generally, the detergent or liposome reconstituted samples were homogenous and retained their ability to bind Ca 2+ and negatively charged lipid membranes, as previously reported for the individual ferlin domains ( Abdullah et al., 2014 , Marty et al., 2013 , Padmanarayana, Hams et al., 2014 ) ( Appendix Fig S1B and S1D-G ). However, in contrast to an early model of dysferlin ( Xu et al., 2011 ) and a recent report ( Huang, Grandinetti et al., 2024 ), we did not observe a significant tendency of the proteins to dimerize through the C 2 domains, in mass photometry measurements and size-exclusion chromatography, suggestive of the type-I ferlins being organized as a monomers in solution ( Appendix Fig S1B-C and S1H ). Download figure Open in new tab Figure 1. Cryo-EM structure of human myoferlin bound to a model lipid bilayer (nanodisc). A Schematic depicting the structure-based domain composition of human myoferlin and dysferlin. The C 2 domains of myoferlin and dysferlin are shown as coloured ellipses, whereas the DysF, FerA and the transmembrane region (TM) are displayed as boxes. The linker helix between C 2 A and C 2 B is depicted as a brown box, whereas the remaining linker regions are coloured grey. dsRBD-like: the double-stranded RNA-binding-like subdomain of C 2 F. B Schematic representation of the soluble myoferlin-lipid nanodisc cryo-EM structure. The transmembrane helix is depicted for orientation purpose and is absent from the soluble expression construct. The modelled Ca 2+ ions are indicated as black dots. C Typical top-view of lipid-bound soluble myoferlin. Myoferlin domains, which are visible in the 2D class average, are indicated. D The cryo-EM map of lipid-bound soluble myoferlin displayed in three different orientations. The individual ferlin domains are color-coded as in A. The soluble myoferlin map has been locally scaled and low-pass filtered to 2.8 Å, whereas the nanodisc density (middle panel, contoured in gray around the ferlin cytosolic region) has been low-passed to 8 Å resolution to allow the visualization of the ordered lipid regions (see also Appendix Fig S2-S6 ). The inset (middle panel) shows a typical side-view class average, consistent with an asymmetric membrane recognition mechanism. Initial attempts at single-particle cryo-EM imaging of lipid-free ferlins revealed the intrinsic flexibility of the N-terminal (C 2 A-C 2 B) and C-terminal domains (C 2 F-C 2 G), which limited the resolution of the maps. We, therefore, hypothesized that these more dynamic motifs, all predicted to engage lipid bilayers ( Abdullah et al., 2014 , Johnson & Chapman, 2010 , Kwok, Otto et al., 2023, Marty et al., 2013 ), would become organized upon membrane binding. In our attempts to identify an optimal membrane system, we observed that the cytosolic region of myoferlin (residues 1-1997) formed stable complexes with large MSP2N2 (membrane scaffold protein 2 N2)-based lipid nanodiscs, comprising anionic phospholipids and chosen to accommodate all interacting domains ( Cannon, Sarsam et al., 2023 ) ( Fig 1D and Appendix Fig S1B-C ). We further stabilized the myoferlin-nanodisc complexes through glutaraldehyde crosslinking ( Kastner, Fischer et al., 2008 ) and imaged four different protein-lipid samples, which were assembled on membranes containing PS (Phosphatidylserine) alone or combined PS and PI(4,5)P 2 (Phosphatidylinositol-4,5-bisphosphate) ( Appendix Fig S2-S5 ). Computational image sorting allowed us to identify intact particles and reconstruct near- complete cryo-EM maps of the lipid-bound myoferlin ( Fig EV1, Appendix Fig S2-S5 and Movie EV1 ). These well-resolved maps refined in 3D to a 2.4-2.9 Å global resolution, allowing us to confidently assign and build nearly all C 2 and accessory domains, apart from the flexible N-terminal C 2 A domain ( Fig 1D , Fig EV1, Appendix Fig S2-S5 and Appendix Tables S1 - S4 ). Myoferlin interacts with the MSP2N2 nanodisc through multiple binding motifs, all projecting on one side and forming well-defined protein-lipid contact sites ( Fig 1D , Fig EV1 and Fig EV2 ). As expected, the remaining lipid membrane is largely disordered in these structures ( Fig 1D and Fig EV2 ). View this table: View inline View popup Download powerpoint Table S1. Data collection, refinement, and validation statistics for the lipid-bound soluble myoferlin (1-1997): 25 mol% DOPS/5 mol% PI(4,5)P 2 lipid nanodisc complex. View this table: View inline View popup Download powerpoint Table S2. Data collection, refinement, and validation statistics for the lipid-bound soluble myoferlin (1-1997): 25 mol% DOPS/5 mol% Cholesterol/5 mol% PI(4,5)P 2 lipid nanodisc complex. View this table: View inline View popup Download powerpoint Table S3. Data collection, refinement, and validation statistics for the lipid-bound soluble myoferlin (1-1997): 15 mol% DOPS/5 mol% Cholesterol lipid nanodisc complex. View this table: View inline View popup Download powerpoint Table S4. Data collection, refinement, and validation statistics for the lipid-bound soluble myoferlin (1-1997): 15 mol% DOPS/2 mol% PI(4,5)P 2 lipid nanodisc complex. Complex tertiary interfaces organize the lipid-bound state of myoferlin Although predicted to adopt an extended, “beads on a string”-like topology ( Dominguez et al., 2022 , Leclère & Dulon, 2023 ), the overall cryo-EM map of the lipid-bound myoferlin revealed a surprisingly compact domain architecture ( Fig 1B and 1D ). As observed in 2D class averages ( Fig 1C , Appendix Fig S2B , Fig S4C , Fig S5B and S5E ), the individual structural motifs of myoferlin are distributed in an almost coplanar manner around a ∼30 Å central cavity, describing an elliptic ring that spans ∼150 Å and ∼90 Å along its long and short axes, respectively. One side of the composite ring engages the membrane bilayer at four distinct contact sites, covering the entire nanodisc perimeter ( Fig 1D and Fig EV2 ). In contrast, the solvent-facing side of myoferlin harbors no lipid recognition motifs, consistent with the presence of a single and nearly planar membrane-binding surface ( Fig 1D and Fig EV2 A-F ). The top region of the ferlin ring ( Fig 1A-B and 1D ) is formed by the structurally more rigid C 2 C (residues 358-544), C 2 CD-FerA (residues 571-842) and C 2 D (residues 1137-1271) domains, which we refer to as the ferlin core module (Fer core , Fig EV1D -E and Movie EV2 ). The inner and outer DysF domains ( Fig 1A and 1D ), each comprising two long β-strands, are closely linked to the Fer core , through C 2 D and C 2 CD, respectively, and separated by ∼14 Å ( Fig 1A , Fig 1D , Fig 2D and Fig EV1F ). Opposite of the Fer core , the N-terminal C 2 B (residues 198-324), the C-terminal C 2 F (residues 1551-1768), and the membrane-proximal C 2 G (residues 1786-1974) approach each other in 3D and pack closely, despite being separated by more than 1200 residues ( Fig 1D , Fig 2A-B , Fig EV1 and Movie EV2 ). Unlike the Fer core motifs, these C 2 domains are engaged in fewer and transient interdomain contacts, in part mediated by loop regions or unstructured elements ( Fig 2B-C and Movie EV2 ). C 2 B contributes to the largest number of contacts and is surrounded from three sides by the top loops of C 2 F (L1, L3 and L4), the long β6-β7 subdomain of C 2 G (residues 1906-1944, Fig EV1J -K ) and the C 2 C domain of Fer core ( Fig 2B-C ). Additional contacts are provided by the membrane-binding β-hairpin of C 2 G (residues 1859-1879), inserted between the β4-β5 strands, and the upstream linker helix (residues 175-190), stacked between C 2 B and C 2 C ( Fig 2B-C , Fig EV1A and EV1J-L ). The conserved but largely unstructured FerI motif (residues 325-357) appears to fold between C 2 F, C 2 G and C 2 B, before reaching C 2 C, likely also stabilizing the composite interface ( Fig 1A , Fig 2A-B and Fig EV1A ). Finally, the top and bottom arches of the ferlin ring are connected through the C 2 E domain, which interacts with both C 2 F and C 2 G through its extended anchor loop (residues 1447-1498), as well as with the upstream C 2 D through its loop 4 (L4, Fig 2A , Fig 2D-E and Fig EV1H ). Download figure Open in new tab Figure 2. The multipartite structural organization of lipid-bound myoferlin. A Overall cryo-EM map of lipid-bound soluble myoferlin (residues 1-1997), viewed from its N-terminal side. The map has been color-coded after the modelled subunits and the key structural motifs are indicated. The Fer core module of myoferlin (the C 2 C-C 2 D region) is contoured in light gray. B The multiple tertiary interfaces between the N-terminal C 2 B-C 2 C and the C-terminal C 2 F- C 2 G, observed in the lipid-bound myoferlin structure. Additional structural elements, such as the linker helix and the FerI motif, bridging the C 2 B and C 2 C domains, and the β-hairpin subdomain of C 2 G, are indicated. The C 2 domains, except C 2 B, are depicted as transparent solvent-excluded surfaces with the domain model fitted inside. C The C 2 B domain orientation in the lipid-bound state appears to be induced by its shared interfaces with C 2 C, C 2 F and C 2 G. C 2 B is shown in a surface representation. The top loops (L1-L4) of the interacting domains are indicated. D The peripheral inner and outer DysF motifs of myoferlin, viewed from the membrane- interacting side. The C 2 C, C 2 F and C 2 G domains are depicted as surfaces. Note that the two DysF motifs are connected to the Fer core module through C 2 D (orange) and C 2 CD (red). E C 2 E connects the Fer core module to the C-terminal C 2 F domain through a long insertion loop (the “anchor loop”). C 2 E is depicted as cartoon, whereas the C 2 B-C 2 C and C 2 F-C 2 G are shown as surfaces. The anchor loop inserts between the β6 and β7 strands of C 2 E (see also Appendix Fig S4h ). At the same time ( Fig 2D ), its L4 loop is oriented towards C 2 D. The rigid ferlin core module is stabilized by a new C 2 -like accessory domain Consistent with limited proteolysis experiments in cultured cells ( Woolger, Bournazos et al., 2017 ), the multipartite organization of the myoferlin ring appears to be centered around the tightly packed Fer core module ( Fig 3A-B and Movie EV2 ). Fer core covers almost one half of the ferlin ring and is distributed symmetrically at the two ends of the arch ( Fig 3A-B ). In between C 2 C and C 2 D, four distinctive motifs are observed ( Fig 3A-B ): (i) the FerA domain (residues 649-769), located in the proximity of C 2 C and folded as a four-helix bundle ( Harsini, Chebrolu et al., 2018 ); (ii) the outer DysF domain (also known as N-DysF, residues 856-915 and 1047-1132), occupying a more central location; (iii) the inner DysF domain (or C-DysF, residues 930-1027), inserted between the two β-strands of the outer DysF ( Patel, Harris et al., 2008 , Sula, Cole et al., 2014 ) and oriented towards the membrane surface; (iv) a previously undetected C 2 -like motif, which we denoted as the C 2 CD domain (residues 571- 648 and 770-842). Download figure Open in new tab Figure 3. Structure of the rigid Fer core module. A Structure of the Fer core module, spanning the top ferlin arch. Fer core comprises the C 2 C-C 2 D region (color-coded cryo-EM map after the modelled subunits) and includes a new C 2 -like structural domain, the C 2 CD domain (red). The neighbouring domains, C 2 B and C 2 E-C 2 F are shown in cartoon representation. The modelled linker regions, connecting the C 2 domains, are coloured grey. B Interaction interfaces between C 2 CD and its neighbouring FerA, C 2 C and C 2 D domains. C 2 CD and FerA are depicted as cryo-EM maps, whereas C 2 C, C 2 D, outer and inner DysF are shown as cartoons. C Structure of myoferlin’s C 2 CD and FerA domains. The seven β-strands of C 2 CD and the four α-helices of FerA are indicated. The newly identified C 2 CD domain is present in all ferlins, including dysferlin and otoferlin ( Fig 3C , Fig 6C-D , Appendix Fig S9A-D and S10A-D ), and spans the distance between C 2 C and C 2 D, thereby appearing to “glue” the Fer core module together. Like all other C 2 domains of myoferlin, C 2 CD consists of two β-sheets and has a type-II topology ( Dominguez et al., 2022 ). However, unlike typical C 2 domains, its second β-sheet comprises only three β- strands ( Fig 3C and Appendix S10B ), the top loops (L1-L4) are extended, lack conservation and engage in tertiary contacts with C 2 C, inner DysF and C 2 D, instead of being available for Ca 2+ and/or phospholipid binding (see also Appendix Fig S9E-G ). C 2 CD is closely associated with the four-helix bundle of FerA, inserted between its β4-β5 strands ( Fig 3C ). Interestingly, FerA does not engage the lipid nanodisc and exits the C 2 CD domain opposite from the membrane-binding surface ( Fig 3B-C , see also Appendix Fig S9E-G and Fig S11 ), partially closing the central cavity of myoferlin as it aligns diagonally along the C 2 C-C 2 E axis, without reaching C 2 E. We suggest that the large contact interfaces of C 2 CD with the remaining Fer core motifs (∼4000 Å 2 ) indicate a primary role as a repurposed C 2 domain packing platform ( Appendix Fig S11A-C ), contrasting the Ca 2+ -sensing and lipid binding functions of other myoferlin C 2 domains (see also Fig 4A and Fig. 5 ). Download figure Open in new tab Figure 4. The composite lipid membrane recognition interface of human myoferlin. A The lipid nanodisc contact interfaces of the C 2 B and C 2 C domains. The focused myoferlin (1-1997) cryo-EM map (color-coded after the modelled subunits) is shown together with the 8 Å low-passed overall map ( Appendix Fig S2C ). The ordered lipid nanodisc regions interact with myoferlin at multiple contact sites, defined by C 2 (C 2 B, C 2 C, C 2 F, C 2 G) and accessory (inner DysF, β-hairpin of C 2 G) domains. B C 2 B interacts with lipid bilayers independent of Ca 2+ binding. Although no bound Ca 2+ ions were observed and C 2 B lacks conservation of the typical aspartate residues involved in divalent cation coordination, the domain engages the lipid bilayer (nanodisc) through its L3 top loop. C 2 B residues directly interacting or located close to the lipid bilayer are shown as sticks. C Close-up of C 2 G’s lipid nanodisc contact interface. The top loops of C 2 G (L1 and L3) project close to C 2 B’s binding site. The two modelled Ca 2+ ions are indicated. D The peripheral inner DysF motif engages the lipid bilayer close to C 2 C’s binding site. The lipid nanodisc density is coloured in grey, whereas the cryo-EM map of myoferlin is coloured-coded as in A. E Close-up of the inner DysF domain interacting with the lipid nanodisc. Inner DysF’s loop residues located close to the nanodisc surface are shown as sticks, for the sake of orientation. Note that the outer DysF and Ca 2+ -bound C 2 D, flanking the inner DysF, do not bind the lipid nanodisc. F C 2 C establishes extensive, both Ca 2+ -dependent and independent, contacts with lipid bilayers. C 2 C’s membrane binding involves hydrophobic and basic residues of the L1, L3 and L4 loops, as well as direct Ca 2+ -mediated phospholipid headgroup coordination. The bound phosphatidylserine (PS) moiety and key interface residues are depicted as sticks. G-H The C-terminal C 2 F and C 2 G bind lipid bilayers through a composite interface. The key lipid-binding motifs (the L3 loops of C 2 F and C 2 G, the β-hairpin of C 2 G) are indicated and the membrane-facing residues are shown as sticks. I The top L1 loop of C 2 C interacts with the lipid bilayer through an amphipathic insertion helix. Note the virtually parallel orientation of L1’s amphipathic helix, with its hydrophobic residues engaging the lipid bilayer (see also Fig 4F ). The Ca 2+ -coordinating residues and the bound PS headgroup are shown as sticks. Download figure Open in new tab Figure 5. Ca 2+ and phospholipid-binding sites modelled in the lipid-bound structure of myoferlin. A Three Ca 2+ ions and a phosphatidylserine (PS) headgroup bind the acidic pocket of C 2 C. The normalized cryo-EM difference map (cyan, 6.5σ) is contoured around the modelled Ca 2+ ions. Ca1(I) and Ca2(II) have a hexadentate coordination. Ca3(III) is coordinated by both L3 loop residues and a PS coordinating oxygen. The coordination bonds are depicted as dashed lines. B The modelled PS headgroup contributes to C 2 C Ca 2+ ion coordination. The cryo-EM difference map (cyan, 6.5σ) is contoured around PS and two Ca 2+ ions (Ca2 and Ca3). See also Fig EV1M . C Molecular recognition of the PS headgroup by myoferlin’s C 2 C domain. The modelled PS is shown as sticks (dark red). Ca 2+ coordination bonds are depicted as black dashed lines, whereas the polar interactions of PS with the L3 and L4 loops are coloured cyan. D The Ca 2+ -binding sites of the C 2 D domain. The cryo-EM difference map (cyan, 6.5σ) is contoured around the modelled Ca 2+ ions. The two Ca 2+ ions of C 2 D (Ca1, Ca2) have a hexadentate coordination achieved through interaction with both L1 and L3 loop residues. E Ca 2+ and phospholipid binding sites of the C 2 F domain. The cryo-EM difference map (cyan, normalized, 5σ) is contoured around the three modelled Ca 2+ ions (Ca1-Ca3). Two Ca 2+ ions (Ca1 and Ca2), bound to C 2 F, have a hexadentate coordination. The Ca3 ion is coordinated by the L3 loop and PS oxygen atoms. C 2 F residues contributing to Ca 2+ coordination are depicted as sticks and coloured in teal. F Several hydrophobic (F1628, Y1625) and basic (R1630) residues, located at the tip of the L3 loop, engage the lipid bilayer, possibly independent of Ca 2+ coordination. The PS headgroup and the modelled Ca 2+ ions are represented as in E. Download figure Open in new tab Figure 6. Cryo-EM structures of human dysferlin and myoferlin in the lipid-free states. A Schematic representation of dysferlin’s cryo-EM structure. The transmembrane region, absent from the expressed cytosolic domain of dysferlin (residues 1-2017), is shown for the sake of orientation. B Two representative, reference-free 2D class averages of dysferlin. The visible dysferlin domains are indicated. C The cryo-EM map of the lipid-free and Ca 2+ -bound dysferlin (1-2017), rendered in three different orientations. The individual ferlin domains are colour-coded as in A . The map has been locally scaled and low-pass filtered to 5 Å (see also Appendix Fig S6-S8 ). D The overall model of dysferlin (1-2017) showing the observed tertiary interfaces between C 2 B, C 2 E, C 2 F, and C 2 G. The final model of dysferlin (1-2017) is fitted within the overall map. E The cryo-EM structure of lipid-free and Ca 2+ -bound myoferlin (1-1997). The individual ferlin domains are colour-coded and the overall map has been low-pass filtered to 10 Å (see also Fig EV3E ) to better visualize the dynamic C 2 G domain. F. Structural superposition between dysferlin (1-2017) and myoferlin (1-1997) in the lipid- free states. Consistent with it being rigid, the Fer core module is virtually identical in the two lipid-free ferlin structures. Note that, compared to the C-terminal C 2 F-C 2 G, C 2 B and C 2 E adopt similar poses in the lipid-free structures (see also Fig 6C ). G Circular plot displaying all the observed intramolecular crosslinks between dysferlin’s structural motifs. The BS3 chemical crosslinking data at FDR (false discovery rate) of 1% were not filtered by score or spectral count (see also Appendix Fig. S14A-C and Dataset EV1 ). H Selected interdomain BS3 crosslinks mapped onto the lipid-free state of dysferlin. The displayed Euclidean distances (coloured lines) were calculated between the Cα atoms of the crosslinked lysine residues and are all below the 35 Å theoretical distance threshold (see also Appendix Fig S14A-B and Dataset EV1 ). The composite membrane-binding interface of human myoferlin Despite numerous structural studies of other C 2 -domain proteins, such as synaptotagmins ( Rizo, 2022 , Rizo, Sari et al., 2022 , Schauder, Wu et al., 2014 , Seven, Brewer et al., 2013 ), there has not been direct experimental evidence for how a multi-C 2 domain protein interacts with lipid bilayers through the combined binding of Ca 2+ and phospholipids. In particular, the relative orientation of all the membrane-binding domains, their lipid-interacting motifs, and their sensitivity to Ca 2+ (or lack thereof), remained speculative ( Arac, Chen et al., 2006 , Honigmann, van den Bogaart et al., 2013 , Rizo et al., 2022 , Seven et al., 2013 ). The cryo-EM structure of human myoferlin elucidates the organization of its seven C 2 and accessory domains upon concerted membrane-binding, enabling us to identify the precise motifs involved in lipid recognition. In the nanodisc-bound structure ( Fig 4A , 4D and 4G ), the cytosolic region of myoferlin engages the lipid surface primarily through the C 2 B-C 2 G, C 2 C, the inner DysF and C 2 F-C 2 G domains, establishing a total of four membrane anchoring points. At the first binding site, we observe that the N-terminal C 2 B ( Fig 4B ) faces the lipid membrane with its concave surface at a ∼28° tilt angle. In this side-orientation, C 2 B projects the loop 3 (L3) nearly perpendicular to the membrane plane, binding the bilayer through several hydrophobic (L271, A273) and polar (S270, R272) residues located at the L3 tip. The remaining residues of the L1-L3 loops do not participate in Ca 2+ -coordination, as no ion densities were observed in the cryo-EM maps. Instead, they establish contacts with C 2 G, which contributes to the same contact site, suggestive of cooperative membrane binding ( Fig 4B-C ). Like C 2 B, the C-terminal C 2 G faces the nanodisc with its concave side tilted by ∼70° ( Fig 4C and Fig 4H ), directly interacting with a small membrane patch through polar and hydrophobic residues of L3 (K1892, F1893, L1895). However, in contrast to C 2 B, the acidic residues in L1 and L3 of C 2 G are conserved and, likely, involved in the coordination of two Ca 2+ ions ( Fig 4C and Fig EV1J -L ). Surprisingly, our structures revealed that an additional membrane anchoring point ( Fig 4D-E ) is provided by the inner DysF motif. Located between C 2 CD and C 2 D, the inner DysF forms direct membrane contacts ( Fig 4D-E and Fig EV1F ), likely, through hydrophobic (I997, I999, P1000, P1001), basic (K1013, K1016), and aromatic residues (Y1015), providing first evidence, to our knowledge, of its role as a generic class of lipid-binding motifs ( Kaur, de la Ballina et al., 2023 ). The positioning of the inner DysF at the membrane surface appears to be influenced by its interactions with the L2 of C 2 CD and the L1 of the neighbouring C 2 D ( Fig 2E and Fig 4D ). While engaging the inner DysF, C 2 D is positioned ∼20 Å away from the lipid nanodisc, where it still coordinates two Ca 2+ ions through its L1 and L3 ( Fig 4E , Fig 5D and Fig EV1G ) and interacts through the L3 loop with the L4 of the downstream C 2 E ( Fig 2D-E ). Overall, these observations suggest that the C 2 D domain might guide the inner DysF to the membrane and couple this binding event to the recruitment of C 2 F-C 2 G via C 2 E, possibly in a Ca 2+ -dependent manner (see also Fig 7E ). Download figure Open in new tab Figure 7. Large-scale conformational transition of ferlins upon lipid membrane binding. A Side-by-side comparison of lipid-free and membrane-bound myoferlin. The structures were superimposed based on the Fer core module and the individual domains are colour coded. B The large-scale conformational rearrangement of C 2 B upon lipid membrane binding. Myoferlin in the lipid-free state is shown as a transparent model. As a result of a ∼160° out- of-plane rotation and ∼11 Å translation, the top L3 loop of C 2 B projects towards the nanodisc surface in the lipid-bound state. Concomitantly, the linker helix and FerI motif are positioned between C 2 C and C 2 F, respectively, likely stabilizing the new pose of C 2 B. C The large-scale displacement of the C 2 F domain upon lipid bilayer binding. Compared to the lipid-free states of myoferlin and dysferlin, where the top loops are oriented towards C 2 B, C 2 F moves by ∼20 Å in the direction of Fer core and towards the membrane plane. D In-plane rotation of C 2 E which accompanies the lipid membrane recognition by human myoferlin. Because of the observed rearrangement, the long L4 loop of C 2 E establishes contacts with the Ca 2+ -bound L1 loop of C 2 D, likely fixing the new pose. As the anchor loop wraps around C 2 G and is displaced together with C 2 E, it is possible that C 2 E and C 2 G domain movements are coupled. E C 2 G translocates by ∼40 Å to engage the lipid bilayer together with C 2 B and C 2 F in the myoferlin structure. The β-hairpin subdomain of C 2 G is coloured in orange. Because in the lipid-free states of myoferlin and dysferlin C 2 G is dynamic and shares no interfaces with C 2 B and C 2 F, it is likely that the lipid-dependent movement of C 2 B and C 2 F precedes the recruitment of the membrane proximal C 2 domain. Myoferlin model in the lipid-free state is depicted as transparent cartoons. Therefore, consistent with our assays ( Appendix Fig S1D-G ), the established membrane contacts of C 2 B, C 2 G and inner DysF have a mixed nature: they are neither exclusively electrostatic nor hydrophobic, nor do they appear to confer strict specificity for acidic phospholipids (such as PS and PI(4,5)P 2 ) via Ca 2+ -binding. Such an interaction mechanism suggests that the positioning of C 2 B, C 2 G, and possibly, inner DysF might be modulated by neighbouring C 2 domains, rather than by bona fide Ca 2+ -driven membrane recruitment. Myoferlin’s Ca 2+ -and phospholipid binding sites Compared to the side-oriented C 2 B and C 2 G, the C 2 C domain establishes a major membrane contact site by engaging the lipid bilayer through its Ca 2+ -coordinating top loops ( Fig 4F , Fig 4I and Fig 5A-C ). The local resolution of the myoferlin map enabled accurate modelling of three bound Ca 2+ ions and a recruited PS lipid headgroup, revealing an intricate membrane recognition mechanism ( Fig 4F and Fig 5A-C ). The largest contact interface between C 2 C and the nanodisc is provided by the long L1. While coordinating two Ca 2+ ions (Ca1 and Ca2) through D374 and D396 at its base ( Fig 5A-C and Fig EV1B -C ), the distal end of C 2 C’s L1 comprises a conserved helical insertion (residues 377-382, Fig 4I ) of amphipathic nature that interacts with the membrane mainly with its hydrophobic side (F377, V381, K382). The L3 of C 2 C ( Fig 5A-C ) coordinates all three Ca 2+ ions (Ca1-Ca3) through D444, W445 (backbone carbonyl), D446, T449, D452, and simultaneously contacts the lipid bilayer ( Fig 4F ) through both hydrophobic and polar residues at its tip (R447, L448, T449, K450). Intriguingly, the modelled PS headgroup ( Fig 5B-C and Fig EV1M ), originating from the lipid nanodisc, binds at the interface between the L3 and the L4 loops of C 2 C. As observed in other C 2 domain structures ( Honigmann et al., 2013 , Rizo & Sudhof, 1998 , Verdaguer, Corbalan-Garcia et al., 1999 ), the bound PS headgroup completes the coordination shells of the second and third Ca 2+ sites (Ca2 and Ca3). At the same time, its serine moiety ( Fig 5B-C ) is recognized by L4 residues (Y513, G515), while L4 itself forms several additional membrane contacts (via F516 and P517). Like C 2 C, the C 2 F domain binds the nanodisc surface using its Ca 2+ -coordinating loops, while also recruiting the β-hairpin subdomain of C 2 G to the same site ( Fig 4G-H and Fig 5E-F ). In such a configuration, the L3 tip of C 2 F inserts deeply into the lipid bilayer (residues F1628, T1629, R1630), contributing, along with L1, to the coordination of three Ca 2+ ions ( Fig 5E - F ). The first Ca 2+ ion (Ca1) is coordinated by typical aspartate residues in L1 (D1575, D1569) and L3 (D1624, Y1625), while the second and third Ca 2+ (Ca2 and Ca3) involve a glutamate residue in L3 (E1632), a backbone carbonyl (R1630), and a recruited PS headgroup to complete their coordination shells. Interestingly, although not directly binding Ca 2+ or phospholipids, the L2 of C 2 F forms close contacts with the β-hairpin motif of C 2 G at the same membrane contact site ( Fig 4H ). The β-hairpin subdomain (residues 1860-1879), inserted between β4-β5 strands of C 2 G, is conserved among ferlins and comprises two antiparallel β-strands connected by a loop. Consistent with it representing a lipid-binding accessory motif ( Fig 4H and Fig EV1J ), the β-hairpin loop penetrates the membrane surface and interacts with the bilayer through hydrophobic (F1869, W1870, I1872) and basic residues (K1866, H1868), possibly alongside the L3 of C 2 F and the nearby C 2 G. In conclusion, membrane recognition by human myoferlin is achieved through both Ca 2+ - dependent and Ca 2+ -independent mechanisms, involving four C 2 domains (C 2 B, C 2 C, C 2 F, and C 2 G) and two accessory motifs (inner DysF and the β-hairpin of C 2 G). The lipid- interacting groups are arranged asymmetrically to one side of the ferlin ring and form multiple composite interfaces (C 2 B-C 2 G, C 2 F-β-hairpin), highlighting a unique binding mode by this large multi-C 2 domain protein. Lipid-free structures of human myoferlin and dysferlin To understand how ferlins are organized prior to their membrane recruitment and to clarify which conformational rearrangements accompany their membrane binding, we obtained additional cryo-EM structures of Ca 2+ -bound soluble myoferlin (residues 1-1997) and of the closely related dysferlin (residues 1-2017) at overall resolutions of ∼3.4 Å and ∼3.5 Å, respectively ( Fig 6A-F , Appendix Fig S6-S8 and Fig EV3 ). In the lipid-free state of both ferlins, the C 2 domains also form a generally similar ring-like structure consisting of only six C 2 motifs ( Fig 6A-F and Movie EV3 ). The rigid Fer core module is well-resolved ( Appendix Fig S9A-C and Appendix Fig S10-S11 ) and has a similar organization in both ferlin structures, including the C 2 C, C 2 CD-FerA, and C 2 D domains ( Appendix Fig S8A-G and Fig S9D-G ). Importantly, in both structures ( Fig EV3F and Fig S12A-C ), C 2 C and C 2 D coordinate two divalent cations each, and alanine substitution of seven key aspartates in their top loops completely abolished the Ca 2+ and liposome-binding activity of the Fer core module ( Appendix Fig S12D-G and Fig S13A-C ). The inner and outer DysF motifs are linked to the Fer core and have a generally similar architecture as observed in the lipid-bound myoferlin structures ( Fig 6C , Fig 6F , Fig EV3C -E and Appendix Fig S8 ). In contrast, the membrane- interacting C 2 B and C 2 F-C 2 G are more dynamic and adopt different poses in the lipid-free state of ferlins ( Fig 6E-F , Fig EV3A , Appendix Fig S6 and Fig S7G-J ). Compared to the nanodisc-bound structures ( Fig 1D and Fig EV2 A-F ), the lipid-free cryo- EM models of both dysferlin ( Fig 6C-E and Appendix Fig S7-S9 ) and myoferlin ( Fig 6E , Fig EV3A -E ) revealed that the N-terminal C 2 B and the C-terminal C 2 E domains have a virtually antiparallel orientation, with the top loops of C 2 B projecting away from the membrane-facing side of the ferlin ring. The C 2 B and C 2 E domains also establish several small tertiary interfaces with C 2 C and C 2 D of the Fer core module, respectively, as well as with the C 2 F domain (both C 2 B and C 2 E), aligned along the short axis of the ring ( Fig 6F and Appendix Fig S7G-J ). At the same time, like in the lipid-bound state, the anchor loop of C 2 E wraps around C 2 F, ∼40 Å away from C 2 E’s β-sheets ( Fig 6C and 6F ). As a result of these multiple contacts, the C 2 F domain appears surrounded on three sides by C 2 B, C 2 E, and the poorly resolved and dynamic C 2 G domain ( Fig 6D-F , Fig EV3 D-E and Appendix Fig S7G - J ). This conformation of C 2 F significantly differs from the membrane-bound state of myoferlin, where the domain engages the lipid bilayer through its Ca 2+ -binding loops ( Fig 4G and Fig 6F ). Consequently, in the lipid-free states of myoferlin and dysferlin ( Fig 6C-F and Fig EV3 D-E ), while still available for Ca 2+ -binding, the orientation of C 2 F’s L1-L3 loops seems incompatible with efficient membrane insertion, partly due to steric hindrance from the interacting C 2 B domain. To confirm the physiological relevance of the lipid-free ferlin state, resolved here at lower- resolution for myoferlin and dysferlin, we subjected soluble dysferlin (residues 1-2017) to chemical crosslinking in the presence of Ca 2+ and identified the crosslinked residues by tandem mass-spectrometry ( Cretu, Schmitzova et al., 2016 ). Mapping the crosslinked lysines onto the cryo-EM structure of dysferlin ( Fig 6G-H , Appendix Fig S14A-C and Dataset EV1 ) indicated that 66 out of 71 observed unique crosslinks (∼92.96%) occurred between residues less than 35 Å apart (48/52, ∼92.31%, when omitting the more dynamic C 2 G domain). Importantly, the existence of characteristic interfaces between C 2 B, C 2 E, C 2 F- dsRBD and C 2 G, as observed in the lipid-free dysferlin model ( Fig 6H ), was strongly supported by the crosslinking data, with only five outlier interdomain crosslinks being observed, all of them involving the flexible C 2 G and the anchor loop of C 2 E. Consistent with the dysferlin cryo-EM map ( Fig 6G-H , Appendix Fig S14B and S14D ), our crosslinking data placed the C 2 E domain between C 2 B, C 2 D and the C 2 F-dsRBD module. Furthermore, the location of C 2 F-dsRBD at the base of the dysferlin ring was supported by its crosslinks to both C 2 B and C 2 E, and to the neighbouring C 2 G, flanking C 2 F from both sides ( Fig 6H and Appendix Fig S14B ). From these analyses, we conclude that the lipid-free conformation is indeed populated in solution, possibly representing the default state of the ferlins in cells. The different conformational states of human ferlins Despite being comparable in general terms, the lipid-free and membrane-bound ferlin structures represent two distinct conformations ( Fig 7A and Movie EV4 ). Structure-based superposition and modelling of the transition between the two resolved states of myoferlin allowed us to decipher the possible sequence of events ( Fig 7B-E , Movies EV4 and EV5 ). We delineated major structural rearrangements occurring at several N-terminal (C 2 B) and C- terminal (C 2 E-C 2 G) sites of the myoferlin ring, indicating a multistep transition ( Fig 7B-E and Movie EV5 ). Compared to the rigid Fer core , our analyses show that the N-terminal C 2 B appears to rotate by ∼160° upon membrane binding and moves by ∼11 Å towards the membrane plane ( Fig 7B-E and Movie EV5 ). This dramatic change in the domain’s orientation entails both disruption (C 2 E), reconfiguration (C 2 C and C 2 F), and formation (with C 2 G and the linker helix) of tertiary interfaces and membrane contacts (through C 2 B’s L3). Concomitantly, our superposition indicates that the C 2 F-dsRBD module, which, in the lipid-free state, interacts with C 2 B, appears to translocate by ∼20 Å and rotate by an additional ∼38° from the ring’s periphery towards the Fer core ( Fig 7C ). The large-scale displacement of the C 2 F domain would enable the previously hindered Ca 2+ -binding loops to directly engage the lipid bilayer ( Fig 7C ). Accompanying this C 2 F repositioning ( Fig 7D and Movie EV5 ), the upstream C 2 E, interacting with C 2 F through the anchor loop, also rotates by ∼32° in the direction of the Fer core , while its long L4 contacts the Ca 2+ -bound L3 of C 2 D. Notably, the new poses of C 2 E and C 2 F appear to be stabilized initially by the repositioned C 2 B ( Fig 7C - D ) and the unstructured FerI motif, which covers the distance between C 2 B and C 2 C. Finally, the C 2 G domain, flexible in the lipid-free states, binds the reconfigured C 2 B-C 2 F interface at both of its sides to close the myoferlin ring. Besides establishing new interdomains contacts, the rearranged C 2 G simultaneously engages the lipid membrane with its Ca 2+ -binding loops and its β-harpin subdomain, inserting into the bilayer in the vicinity of C 2 B and C 2 F, respectively ( Fig 7E ). Interestingly, our structures also show that C 2 G binds more tightly phosphoinositide-containing bilayers, possibly because of additional interactions between its polybasic patch and exposed PI(4,5)P 2 headgroups, as previously suggested and/or observed in other C 2 domain structures ( Carpenter, Khuu et al., 2023 , Guerrero-Valero, Ferrer-Orta et al., 2009 , Guillen, Ferrer-Orta et al., 2013 , Kwok et al., 2023 , Padmanarayana et al., 2014 ) ( Fig EV4 A-C ). The conformational rearrangement of myoferlin (and possibly dysferlin) to its lipid-bound state is possible because the C 2 domains undergoing displacement share relatively small contact interfaces in the lipid-free conformation, and the connecting linker regions are long and partially unstructured ( Fig 1A , Fig EV2 B-C , Appendix Fig S13D ). For example, C 2 E’s contact interfaces with C 2 B (∼252 Å 2 ), C 2 D (∼347 Å 2 ) and C 2 F (∼235 Å 2 , excluding the anchor loop) are equally small. Likewise, in the lipid-free state, the C 2 B-C 2 F interface involves a relatively small number of residues (∼193 Å 2 ) and is not present in all the imaged particles ( Fig 6D-E , Fig EV3A and Appendix S6C ). Moreover, calpain-cleavage of myoferlin and alternative dysferlin isoforms appears to release of the C-terminal C 2 F-C 2 G domains, which indicates that the C 2 B-C 2 F and C 2 E-C 2 F interfaces are also dynamic in cells ( Piper, Ross et al., 2017 , Redpath, Woolger et al., 2014 ). In this respect, the accessory ferlin structural motifs – the β-hairpin of C 2 G, the linker helix and FerI – resolved exclusively in the lipid-bound state of myoferlin, are likely required to stabilize the reconfigured C 2 domain interfaces. Therefore, the tertiary interactions between the Fer core , C 2 B, C 2 E and C 2 F, observed in the lipid-free structures, seem to be generally weak, having possibly evolved to facilitate efficient membrane sampling on fast timescales and in a complex cellular environment. Formation of new and extended contacts between C 2 B, C 2 F, and the repositioned C 2 G not only increases the structure’s stability, but also allows their concomitant, in-plane binding as part of a composite and asymmetric interface. Because the C-terminal C 2 G is connected to the transmembrane helix through a relatively short linker (approximately 20 residues), our structures indicate that the large-scale movement of the domain towards the Fer core would shorten the distance between the ferlin vesicle and the target membrane, likely facilitating their close apposition. Indeed, supporting such a scenario, we observe that both myoferlin and dysferlin engage lipid bilayers in a Ca 2+ -dependent manner and promote tight vesicle-vesicle interaction ( i.e. , docking) in vitro, when both PS and PI(4,5)P 2 are present on the target membrane ( Fig EV4 D-E ) . Given the high degree of conservation of their lipid-free structures, it is likely that dysferlin and, possibly, otoferlin transition through a similar sequence of domain rearrangements and conformational states upon membrane binding ( Fig 8 ). Such a role of ferlins in progressing from loose vesicle tethering to tight docking of membranes has been suggested for otoferlin based on electron tomography ( Vogl, Cooper et al., 2015 ) and, like myoferlin ( Fig EV4 D-E ) and dysferlin ( Codding, Marty et al., 2016 ), otoferlin’s C 2 domains accelerate SNARE-dependent membrane fusion in vitro ( Johnson & Chapman, 2010 ). These functional similarities indicate that ferlins may indeed share common organization principles and a conserved membrane remodelling mechanism. Download figure Open in new tab Figure 8. Model of ferlins’ Ca 2+ -sensitive recruitment and binding to lipid membranes. Schematic depicting a structure-guided model for how ferlins facilitate Ca 2+ -dependent vesicle targeting, docking, and local membrane remodelling to modulate and possibly promote fusion of two lipid bilayers. The ferlin domains are color-coded and those interacting with the lipid membranes are outlined in black. PS (phosphatidylserine) and PI(4,5)P 2 (phosphatidylinositol-4,5-bisphosphate) are indicated as red and orange, respectively. The membrane-sampling loose state has been observed in the lipid-free dysferlin cryo-EM data ( Appendix Fig S6A and Fig S14A ). The hypothetical membrane-bound loose state, where the C-terminal C 2 G domain is yet to engage the lipid bilayer, is based on Alphafold2 predictions of myoferlin and dysferlin ( Appendix Fig S15B-D ). The lipid-free and membrane-bound closed states have been depicted based on the cryo-EM structures of lipid- free myoferlin/dysferlin and membrane-bound myoferlin, respectively. Discussion Ferlins are the largest and the most complex multi-C 2 domain Ca 2+ -sensitive factors ( Lek et al., 2012 ). In this study, we report the near-atomic cryo-EM structures of human myoferlin in its Ca 2+ and lipid-bound state. We also obtained cryo-EM models of both myoferlin and the related dysferlin in their lipid-free states and carried out supporting experiments to validate our structural models. These analyses provide first insights into how the unique structural features of ferlins are adapted to their many ascribed roles in cellular trafficking pathways, underlying the acute response to membrane and organelles injuries, muscle biogenesis and metabolism, or fast transmitter exocytosis at hair cell ribbon synapses. We propose that the multifaceted roles of ferlins in diverse Ca 2+ -dependent pathways in cells could stem from their ability to sample alternative conformational states ( Fig. 8 ), achieved by exploiting a unique structural organization. Contrary to previous models ( Dominguez et al., 2022 , Woolger et al., 2017 , Xu et al., 2011 ), but in agreement with Alphafold2 predictions, nanoscopic imaging ( Shaib, Chouaib et al., 2024 ) and a recent lipid-free dysferlin structure ( Huang et al., 2024 ), the ferlin C 2 domains are not trivially organized as “beads on a string” or as in other, well-studied multi-C 2 domain factors ( Schauder et al., 2014 , Shin, Han et al., 2005 ). Instead, the conserved structural motifs pack uniquely in 3D and form state-defining interfaces, bridging both neighbouring and sequence-distant domains. The central ferlin core module (Fer core ) comprises the region spanning the C 2 C and C 2 D domains. Fer core , likely, maintains part of the, otherwise dynamic, ferlin structure rigid and might support the ordered and reversible Ca 2+ /lipid-driven conformational transitions of ferlins. In this respect, the newly identified ferlin C 2 -like domain, the C 2 CD domain, observed in our myoferlin and dysferlin structures, and the closely linked four-helix bundle of FerA might have evolved to promote the tight packing of the Fer core module. Importantly, Alphafold predictions support the presence of a similarly organized Fer core in all remaining paralogs, including otoferlin, whose mutations cause the nonsyndromic deafness DFNB9 and are a target of the first inner ear gene-therapy trials ( Lv, Wang et al., 2024 , Moser et al., 2024 ) ( Appendix Fig S15A and S15D ). Although appearing defined and stable enough to be observed upon cryo-EM imaging, the interfaces between the membrane-proximal domains and the Fer core module are more dynamic in the lipid-free states of myoferlin and dysferlin. This notable structural feature appears to be a direct consequence of the highly specialized roles served by ferlins in cells. Having low energy barriers between their discrete conformations could, possibly, explain the transiently formed contacts between state-defining C 2 domains, such as C 2 B, C 2 F and C 2 G ( Fig. 8 ). This may constitute a significant advantage in protein-rich microenvironments, such as the active zones of hair cell ribbon synapses (otoferlin) ( Moser, Grabner et al., 2020 ) or caveolae-rich plasma membrane compartments (dysferlin) (Corrotte, Almeida et al., 2013), allowing them to sample conformations close to the ground state. As a result, switching between the different ferlin conformations could be accomplished by reversibly shifting the equilibrium towards a given, functionally relevant state without a significant free energy consumption ( Fig. 8 ). Consistently, all ferlin-dependent cellular processes occur on fast timescales, depend on Ca 2+ , and are highly dynamic and reversible, with the ferlins appearing to be needed for both the forward and reverse reactions ( e.g. , synaptic vesicle exocytosis and replenishment, vesicle docking and undocking). An important question raised by our structural analyses pertains to the possible connections between the observed structural changes upon membrane binding in vitro ( Fig. 8 ) and the documented roles of ferlins in promoting Ca 2+ -sensitive vesicle tethering and membrane fusion. As specialized trafficking factors highly expressed in muscle cells, the membrane- anchored myoferlin and dysferlin reside at the sarcolemma and shuttle to the endosomal compartment, without accumulating in a defined intracellular vesicle pool ( Cooper & McNeil, 2015 , Lek et al., 2012 ). The related otoferlin is enriched at the presynaptic active zone membranes and in the synaptic vesicles of inner hair cells, while also shuttling through the endosomal pathway during a synaptic vesicle release cycle ( Jung et al., 2015 , Moser et al., 2020 , Pangrsic et al., 2012 , Revelo, Kamin et al., 2014 ). In all cases, the ability of ferlins to react to an increase in intracellular Ca 2+ , following large sarcolemma injuries (dysferlin) or a sound-evoked receptor potential (otoferlin), appears to be determined by their cellular localization on vesicles, interactions with other factors, and the local phospholipid composition of the target membranes ( Chen, Fang et al., 2023 , Chen et al., 2024 ). Consistent with having such multimodal mechanisms of action, the cryo-EM models of myoferlin and dysferlin revealed a dynamic structural organization, apparently evolved to accommodate different membrane environments and binding modalities. Our structures show that the Ca 2+ - sensitive N-terminal C 2 A (in type-I ferlins) and the membrane proximal domains (C 2 F-C 2 G) are more dynamic in the lipid-free ferlin states, which could result in an increased membrane capture radius ( Fig. 8 ). We, therefore, propose that by employing their mobile C 2 domains as pioneer Ca 2+ and lipid-sensitive motifs, ferlins would be able to efficiently sample and approach distant target membranes. Consequently, membrane attachment of these pioneer C 2 domains would trigger, in the next step, the tight association of the remaining Fer core core domains (C 2 C-C 2 D and the inner DysF), altogether assembling an asymmetric ring-like structure that could bridge two cellular membranes (tethering) and, at the same time, promote local membrane buckling and remodeling through deep insertion into the bilayer at multiple contact sites (tight vesicle docking, Fig. 8 ). In our model, as only one side of the ring would engage the lipid bilayer, the available ferlin domains, such as FerA, could simultaneously interact with other pathway factors, such as SNAREs ( Codding et al., 2016 ) or other ferlin molecules ( Pangrsic et al., 2012 ), further narrowing the gap between the two membranes to facilitate their fusion ( Fig. 8 ). This structural adaptation could be particularly relevant for dysferlin, whose functions in plasma membrane repair have been linked to the Ca 2+ -sensitive clustering of highly heterogenous endomembrane vesicles to promote their homotypic and heterotypic fusion ( Cooper & McNeil, 2015 ). Future functional and structural studies are needed to probe the validity of these mechanisms, beyond simple nanodisc membranes, particularly within the context of pathway-defining, higher-order ferlin complexes, as pioneered for synaptotagmins ( Zhou, Lai et al., 2015 , Zhou, Zhou et al., 2017 ). These studies will likely entail the use of alternative imaging modalities (cryo-electron tomography and super-resolution imaging) and computational approaches (molecular dynamics simulations), applied to more complex membrane systems (such as liposomes or lipid nanotubes). Such analyses will help exclude improbable non-physiological myoferlin conformations induced by nanodiscs and further elucidate the roles of additional ferlin molecules and ferlin-SNAREs interactions in the vesicle-membrane fusion cycle. Materials and Methods Reagents and tools table View this table: View inline View popup Methods and protocols Plasmids and molecular cloning The codon-optimized full-length canonical human dysferlin (DYSF, FER1L1) expression construct (NM_003494.4, transcript variant 8, isoform 1, UniProt O75923-1) was constructed by GenScript and cloned into the pFastBac1 vector backbone by using the EcoRI and KpnI restriction sites in frame with an N-terminal twin-StrepII affinity tag, which can be cleaved with the HRV-3C (Human Rhinovirus 3C) protease. The full-length human myoferlin (MYOF, FER1L2, Uniprot Q9NZM1-1) was obtained from Addgene (pCDNA3.1-Myoferlin- HA, cat#22443, ( Bernatchez et al., 2007 )) and subcloned by LIC (ligation independent cloning) into a modified pFastBac backbone (first described in ( Schmitzova et al., 2023 )), in frame with an HRV-3C protease cleavable N-terminal twin-StrepII-tag. The soluble dysferlin (residues 1-2017) and myoferlin (residues 1-1997) constructs and the domain truncation mutants of dysferlin (Fer core (the C 2 B-C 2 D fragment, residues 214-1281) and minimal dysferlin (residues 1503-1560 and 1613-2056), encompassing the C 2 F-C 2 G domains) were obtained by “around-the-horn” PCR-based cloning using the full-length constructs as templates and retained the N-terminal twin-StrepII tag. Neutralizing substitutions in the Ca 2+ - binding motifs of the dysferlin core region (residues 214-1281) were introduced by sequential PCR-based site-directed mutagenesis ( Appendix Table S7 ). All constructs were verified by Sanger sequencing (Microsynth Seqlab GmbH, Göttingen) of the open reading frames (ORFs). All cloning primers were synthesized by Microsynth Seqlab GmbH (Göttingen). View this table: View inline View popup Download powerpoint Table S5. Data collection, refinement, and validation statistics for the lipid-free soluble myoferlin. View this table: View inline View popup Download powerpoint Table S6. Data collection, refinement, and validation statistics for the lipid-free soluble dysferlin. View this table: View inline View popup Download powerpoint Table S7. PCR primers used in this study. Expression and purification of human myoferlin and dysferlin To enable their expression in insect cells, all myoferlin and expression constructs were incorporated into bacmids through transformation of electro-competent DH10MultibacY cells, as previously described ( Cretu, Agrawal et al., 2018 , Cretu, Gee et al., 2021 , Cretu et al., 2016 , Schmitzova et al., 2023 ), extracted for transfection using the High Pure Plasmid Isolation kit (Roche, cat#11754785001), and precipitated with isopropanol. To produce V 0 recombinant baculoviruses, adherent Sf9 ( Spodoptera frugiperda ) cells, cultured in Sf-900 III SFM (Gibco, cat#12658019), were transfected with the prepared bacmids with FuGENE HD (Promega, cat#E2311) in 6-well plates. The initial V o viruses were amplified by infecting Sf9 cells at a 1:10-1:20 ratio to produce the V 1 generation of baculoviruses. For large-scale protein production, the V1 baculoviruses were used to infect Sf9 or Hi5 ( Trichoplusia Ni ) suspension cultures, grown in the ESF 921 protein free medium (Expression Systems, cat#96- 001-01), in titers sufficient to induce cell proliferation arrest after 24 hours ( Cretu et al., 2016 , Schmitzova et al., 2023 ). Typically, Hi5 and Sf9 cells were infected at a density of ∼1.0x10 6 cells/mL and harvested 60-72 hours after baculovirus infection, when cells’ viability dropped to ∼80-85%. Insect cell infection was monitored every 16-24 hours by observing the expression of the eYFP marker. Full-length and soluble dysferlin (1-2017) were overexpressed on a preparative scale in Hi5 cells. All myoferlin samples (full-length myoferlin (1-2061) and soluble myoferlin (1-1997)), the Fer core of dysferlin and its mutant variants were expressed and purified from Sf9 cells. All dysferlin and myoferlin constructs were purified by exploiting the highly specific twin- StrepII affinity tag and their binding to an anion-exchange resin ( Fig S1A ). Membrane- anchored full-length myoferlin and dysferlin were purified in the presence of DDM (n- Dodecyl-β-D-maltoside, Glycon Biochemicals, cat#D97002), whereas, after cell lysis, all soluble ferlin constructs were purified in the absence of detergent. Typically, insect cells (Sf9 or Hi5) from ∼1 L culture were resuspended in 10-15 mL of the Lysis buffer (50 mM HEPES- KOH, pH 7.5, 300 mM KCl, 10% (v/v) Glycerol, 4 mM DTT (Dithiothreitol), 2-2.5% (w/v) DDM (or 0.2% (v/v) Triton X-100 in the case of the soluble constructs), EDTA-free cOmplete protease inhibitors cocktail (Roche, cat#11836170001, 1 tablet per 50 mL buffer)) per gram of cell pellet and detergent-extracted by incubation for 90 min at 4°C while gently rotating on a roller mixer or lysed by sonication on ice (soluble constructs), using the Branson Ultrasonics Sonifier 250 (Duty cycle 30%, Output 3, for 2 min). The crude lysate was cleared by ultracentrifugation at 46,300 r.p.m. for 1 h at 4°C in a type 70Ti rotor or at 15,000 r.p.m. for 1 eh at 4°C in the JA-18 rotor (Beckman Coulter) and further filtered through a 0.8 μm Minisart membrane (Sartorius, cat#16592). The filtered lysate was incubated with 2-3 mL Strep-Tactin XT 4FLOW high-capacity resin (4-6 mL 50% slurry, IBA Lifesciences, cat#2- 5030-010) per liter of culture, for 1 h at 4-8°C on a roller mixer. The resin was pelleted by centrifugation at 2000 r.p.m., transferred to an Econo-Pac (Bio-Rad, cat#7321010) gravity- flow column, and thoroughly washed with the Wash buffer (50 mM HEPES-KOH, pH 7.5, 300-500 mM KCl, 5% (v/v) Glycerol, 2 mM DTT, 0.5% (w/v) DDM (membrane-anchored constructs)) and then with the Binding buffer (50 mM HEPES-KOH, pH 7.5, 150 mM KCl, 5% (v/v) Glycerol, 2 mM DTT, 0.1% (w/v) DDM (membrane-anchored constructs)). The bound proteins were eluted by competition with biotin added to the Elution buffer (50 mM HEPES-KOH, pH 7.5, 120-140 mM KCl, 5% (v/v) Glycerol, 2 mM DTT, 0.1% (w/v) DDM (in case of the membrane-anchored constructs), 1 mM EDTA (Ethylenediaminetetraacetic acid), 60 mM biotin). The affinity tag was not cleaved. To remove nucleic acid contaminants and minor impurities, the Strep-Tactin eluates were further purified by anion-exchange chromatography on a 5 mL HiTrap Q Sepharose HP column (Cytiva, cat#17-1154-01), mounted on an Äkta go system. Prior to sample application, the Q Sepharose HP column was equilibrated with Buffer A (20 mM HEPES- KOH, pH 7.5, 150 mM KCl, 5% (v/v) Glycerol, 2 mM DTT (or 1 mM TCEP (Tris(2- carboxyethyl)phosphine)), 0.03% (w/v) DDM (membrane-anchored constructs)). In the next step, the column was washed extensively with Buffer A and the sample eluted off the column using a linear, 0-30% gradient formed between Buffer A and Buffer B (20 mM HEPES-KOH, pH 7.5, 1 M KCl, 5% (v/v) Glycerol, 2 mM DTT (or 1 mM TCEP), 0.03% (w/v) DDM (membrane-anchored constructs)) over 60-80 mL. The peak ferlin fractions were pooled and concentrated by ultrafiltration to ∼3-3.5 mg/mL (membrane-anchored constructs), ∼4.0-5.5 mg/mL (soluble dysferlin (1-2017) and myoferlin (1-1997)) or ∼7.5-10 mg/mL (the domain truncation mutants of dysferlin), aliquoted, snap frozen in liquid nitrogen, and stored at - 80°C. The identities of the purified proteins were verified by mass-spectrometry (Proteomics Facility, Max-Planck-Institute for Multidisciplinary Sciences, Göttingen). NanoDSF-based characterization of myoferlin and dysferlin As means of protein quality control, all myoferlin and dysferlin preparations (both soluble and membrane-anchored) were subjected to nanoDSF (nano differential scanning fluorimetry) measurements. In a nanoDSF typical assay, the 5 µL protein sample (∼0.6 µM) was mixed with 5 µL assay buffer (25 mM HEPES-KOH, pH 7.5, 150 mM KCl, 0.03% (w/v) DDM (membrane-anchored constructs)), which was pretreated with Chelex 100 (Bio-Rad, cat#142-1253) and supplemented with increasing concentrations of CaCl 2 or MgCl 2 (0-40 mM). Following 10 min incubation at room temperature, the samples were loaded into capillaries and the emission intensity at 350 nm and 330 nm (due to the intrinsic fluorescence of tryptophan residues at 280 nm) was measured as a function of temperature using a Prometheus NT.48 instrument (NanoTemper Technologies). The temperature was increased from 20°C to 95°C by applying an unfolding ramp of 1°C/min and the excitation power was adjusted to yield at least 2000 integrated fluorescence counts at both wavelengths. The melting temperatures ( T m ) were estimated by plotting the first derivative of the 350 nm/330 nm ratio as a function of temperature in the PR.ThermControl v2.3.1 software. The [Me 2+ ] 1/2 values were estimated in GraphPad Prism 10 (v10.3.1) by nonlinear regression fitting to a modified Hill function: T m ([Me 2+ ])= T m i+( T m f- T m i)x([Me 2+ ] n /([Me 2+ ] 1/2 n +[Me 2+ ] n )), where T m i and T m f, represent the initial and final T m , respectively, and n – the Hill coefficient. Preparation of liposomes All lipids (phospholipids and cholesterol), used to prepare liposomes (Large unilamellar vesicles) and lipid nanodiscs, were obtained from Avanti Polar Lipids as powder (or chloroform stocks), further dissolved in chloroform to their working concentrations, and stored at -20°C. Porcine brain PI(4,5)P 2 (L-α-phosphatidylinositol-4,5-biphosphate) was dissolved in a chloroform:methanol:water (20:9:1) solution. Unlabeled LUVs devoid of anionic phospholipids (referred to as “DOPC-DOPE-only LUVs”) were prepared by mixing DOPC (1,2-dioleoyl-sn-glycero-3-phosphocholine, 18:1 (Δ9-cis) PC), DOPE (1,2-dioleoyl- sn-glycero-3-phosphoethanolamine, 18:1 (Δ9-cis) PE) and Cholesterol (ovine wool) in a 7:2:1 molar ratio (70:20:10 mol%) to a final 8 mM concentration. LUVs comprising in addition DOPS (1,2-dioleoyl-sn-glycero-3-phospho-L-serine, referred to as “15 mol% DOPS LUVs”) and both DOPS and PI(4,5)P 2 (referred to as, for example, “25 mol% DOPS/5 mol% PI(4,5)P 2 LUVs”) were prepared by substituting a part of DOPC in the lipid mixture with respective anionic phospholipids to obtain the desired lipids’ ratio. Dansyl-labeled LUVs were prepared by replacing a portion of DOPE in the lipid mixture with 5 mol% 18:1 Dansyl PE (1,2-dioleoyl-sn-glycero-3-phsophoethanolamine-N-(5-dimethylamino-1- naphtalenesulphonyl)). Similarly, Rhodamine B and NBD (Nitro-2,1,3-benzoxadiazole-4-yl)- labelled LUVs were obtained by replacing a part of DOPE in the lipid mixture with 1 mol% 18:1 Lissamine Rhodamine B PE (1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N- (lissamine rhodamine B sulfonyl)) and 1 mol% 18:1 NBD PE (1,2-dioleoyl-sn-glycero-3- phosphoethanolamine-N-(7-nitro-2-1,3-benzoxadiazol-4-yl). To obtain unilamellar liposomes, the mixed lipids were transferred to a glass vial, the solvent was evaporated under a nitrogen stream and the lipid film was then dried in a vacuum desiccator (∼200 mbar) for at least 3 hours to overnight. The lipid film was hydrated in reconstitution buffer (20 mM HEPES-KOH, pH 7.5, 150 mM KCl) by gently vortexing. The resulting multilamellar vesicles were extruded through a 0.4 µm Nuclepore track-etch membrane (Cytiva, cat#10417104) for at least 21x using a Mini-extruder (Avanti Polar Lipids) and further used in coflotation and Dansyl-based lipid binding assays. Vesicles used for proteoliposome reconstitution and in lipid mixing assays were additionally passed 21x through a 0.1 µm Nuclepore membrane (Cytiva, cat#10419504). For the reconstitution of MSP2N2-based lipid nanodiscs, lipid films were prepared as for liposome reconstitution, except omitting DOPE and cholesterol from the lipid mixture (in the case of the 25 mol% DOPS/5 mol% PI(4,5)P 2 and 15 mol% DOPS/2 mol% PI(4,5)P 2 nanodiscs) or using lower ratios of DOPE (10 mol% DOPE and 5 mol% 18:1 Dansyl PE) and Cholesterol (5 mol%) when assembling the 25 mol% DOPS/5 mol% PI(4,5)P 2 /5 mol% Cholesterol and 15 mol% DOPS/5 mol% Cholesterol nanodiscs. For nanodisc reconstitution, the dried lipid films were hydrated in the reconstitution buffer supplemented with 1.7 % (w/v) or 0.6 % (w/v) DDM, resulting in 13.2 mM or 5.6 mM final lipid concentrations, respectively, and briefly sonicated with the microtip (Branson Sonifier 250). Reconstitution of ferlins into proteoliposomes Dysferlin and myoferlin ( Fig EV4D -E ) were reconstituted into 100 nm LUVs by mixing DOPC/DOPE-only liposomes, OG (n-octyl-β-D-glucoside, Glycon Biochemicals, cat#D970001) and the purified ferlin (in DDM micelles) at a 1:3500 protein-to-lipids ratio, an R-value of 1, and a 4 mM final lipid concentration. The samples were then incubated for 20 min at room temperature and then transferred to Slide-A-Lyzer MINI dialysis devices (with a MWCO 10 kDa, Thermo Fisher Scientific, cat#69570) or Slide-A-Lyzer cassettes (MWCO 10 kDa, Thermo Fisher Scientific, cat#66383) for overnight dialysis at 4-8°C against 2 L of reconstitution buffer (20 mM HEPES-KOH, pH 7.5, 150 mM KCl), supplemented with 2.5 g/L Bio-Beads SM-2 (Bio-Rad, cat#1523920). Bio-Beads SM-2 were prepared fresh by sequential washing with methanol, ethanol, ddH 2 O, and reconstitution buffer. The successful reconstitution of ferlins into liposomes was verified by liposome flotation on a Nycodenz step gradient following ultracentrifugation for 90 min at 50,000 r.p.m. in a TLS-55 rotor (Beckman Coulter). Liposome binding assays To assess the ability of soluble myoferlin (1-1997) to interact with model lipid bilayers, 50 µM Dansyl-labeled LUVs (labeled with 5 mol% 18:1 Dansyl PE), of a varying anionic phospholipid composition, were mixed with ∼0.9 µM ferlin sample in the presence of Ca 2+ . The total reaction volume was 15 µL. Following incubation for 5 min at room temperature, the protein-lipid samples were transferred into an ultra-micro cuvette (QS 105.252, 1.5 mm optical path, Hellma, cat#105-252-15-40) and the emission spectra were taken between 450- 560 nm at a 284 nm excitation wavelength (3 nm slits, 0.1 s integration time) using a Fluorolog 3 spectrofluorometer (Horiba Jobin Yvon). The relative protein-to-membrane FRET (Förster Resonance Energy Transfer) efficiency was calculated as follows: rFRET=( I - I min )/( I max - I min ), where I represents the average emission intensity of the sample at 518-520 nm, I min – the intensity of the protein-free blank sample, and I max – the maximum Dansyl emission of the titration series. To estimate the [Ca 2+ ] 1/2 values, the liposome binding data were fitted to the Hill equation ( Brandt, Coffman et al., 2012 ) in GraphPad Prism 10. All experiments were repeated three times (technical replicates). Coflotation assays In typical liposome coflotation experiment, soluble ferlin constructs were mixed with LUVs in presence of 50 µM or 0.5 mM ( Appendix Fig S12F ) Ca 2+ /Mg 2+ and added to the bottom layer of a Nycodenz step gradient. The final protein and liposomes assay concentrations were ∼1 µM and 1 mM, respectively. The bottom Nycodenz layer (40% (w/v)) was then overlaid with equal volumes (40 µL) of a 30% (w/v) Nycodenz solution and reconstitution buffer (20 mM HEPES-KOH, pH 7.5, 150 mM KCl), which were supplemented with 50 µM CaCl 2 or MgCl 2 (or 0.5 mM CaCl 2 /MgCl 2 ). The Nycodenz gradients were centrifuged at 50,000 r.p.m. for 90 min in a TLS-55 rotor (Beckman Coulter), harvested from the top in 20 µL fractions and analyzed by SDS-PAGE (NuPAGE 4-12% Bis-Tris gels, Thermo Fisher Scientific, cat#NP0323BOX). The SDS-PAGE gels were stained with One-Step Blue (Biotium, cat#21003). Under these experimental conditions, liposomes float to the top two Nycodenz fractions and comigration of ferlins to the top fractions is indicative of their stable interaction with the lipid bilayer. All coflotation experiments were conducted at least three times (technical and biological replicates). Lipid mixing assays In a typical “bulk” lipid mixing assay ( Hernandez, Stein et al., 2012 , Hoekstra & Duzgunes, 1993 , Yavuz, Kattan et al., 2018 ), used to monitor the ability of full-length myoferlin and dysferlin to promote tight vesicle-vesicle aggregation or docking, 20 µL empty dual labelled 100 nm LUVs (comprising 1 mol% NBD PE and 1 mol% Rhodamine B PE) were mixed in a 1:1 ratio with unlabelled proteoliposomes in 1 mL reconstitution buffer (20 mM HEPES- KOH, pH 7.5, 150 mM KCl), supplemented with 0.1-1 mM CaCl 2 . The lipid mixing data ( Fig EV4 D-E ) was acquired at 37°C using a Fluorolog 3 spectrofluorometer (Horiba Jobin Yvon) and corrected for signal intensity variations (the S/R acquisition mode). The extent of NBD (donor) dequenching because of lipid mixing was monitored at 460 nm (3 nm slit) and 538 nm (3 nm slit) excitation and emission wavelengths, respectively. The lipid mixing reactions were stopped upon addition of 5 µL 10 % (v/v) Triton X-100 (in reconstitution buffer) and the maximal NBD dequenching signal, after detergent solubilization of liposomes, was considered as the maximal fluorescence (F max ). The normalized lipid mixing efficiency was calculated as: (F-F i )/(F max -F i ), where F i represents the initial fluorescence of the labelled liposomes and F max – the final fluorescence of the sample (i.e., after detergent addition). The dual labelled LUVs comprised 25 mol % DOPS/5 mol% PI(4,5)P 2 or 15 mol% DOPS, whereas full-length myoferlin and dysferlin were reconstituted in DOPC/DOPE-only LUVs. The control reactions included only protein-free liposomes. The lipid mixing data was analysed in OriginPro 2020 (v9.7.0.188). The lipid mixing assays were repeated at least three times (technical replicates) and at least two separate proteoliposome reconstitutions (biological replicates). Mass photometry characterization of full-length human dysferlin All measurements ( Fig S1H ) were performed with the OneMP mass photometer (Refeyn). Images were acquired with Refeyn AcquireMP and analysed using Refeyn DiscoverMP software. For mass photometry measurements, twin-StrepII-tagged full-length dysferlin (residues 1-2080), was reconstituted into LMNG (Lauryl maltose neopentyl glycol, Anatrace, cat#NG310) micelles (instead of DDM) and purified by anion-exchange chromatography. The dysferlin sample was concentrated to ∼3 mg/mL in the presence of 0.01% (w/v) LMNG and dialysed before measurements using a Slide-A-Lyzer MINI device (MWCO 20 kDa, Thermo Fisher Scientific, cat#69590) against the dialysis buffer (20 mM HEPES-KOH, pH 7.5, 150 mM KCl, 2 mM DTT). The sample was then 20-fold diluted in the dialysis buffer and centrifuged at 13,000 r.p.m. for 10 min at 4°C before the measurements. Prior to loading, for each measurement, 1 µL of sample was added to a droplet of 12 µL of dialysis buffer. For the experiments carried out in the presence of CaCl 2 , the sample after the dialysis was diluted in a dialysis buffer containing 2 mM or 4 mM CaCl 2 and incubated for 45 – 60 minutes, before performing the measurement. Mass calibration was achieved by adding 6 µL of NativeMark™ Unstained Protein Standard (Invitrogen, cat#LC0725) diluted 100 times (in the dialysis buffer), to a drop of 12 µL of the dialysis buffer and using the peaks corresponding to bovine serum albumin (66 kDa), lactate dehydrogenase (146 kDa) and apo- ferritin (480 kDa). Each experiment was repeated four times (technical replicates). Cryo-EM sample preparation To preserve the integrity of the samples in vitreous ice, lipid-free soluble myoferlin (1-1997) and dysferlin (1-2017) were stabilized through Glutaraldehyde (GA, Electron Microscopy Sciences, cat#16220) crosslinking upon gradient centrifugation (GraFix) ( Kastner et al., 2008 ). Briefly, prior to vitrification, 80 µL soluble dysferlin (1-2017) at a ∼5.5 mg/mL concentration were applied to a linear 5-40% (w/v) sucrose gradient prepared by mixing equal volumes of the light (20 mM HEPES-KOH, pH 7.5, 150 mM KCl, 1 mM CaCl 2 , 5% (w/v) Sucrose) and Glutaraldehyde (GA)-supplemented heavy gradient solution (20 mM HEPES-KOH, pH 7.5, 150 mM KCl, 1 mM CaCl 2 , 40% (w/v) Sucrose, 0.2% (v/v) GA) using the Gradient Master 108 (Biocomp). The gradient was then centrifuged at 4°C for 15 h at 29,100 r.p.m. in a SW40Ti rotor (Beckman Coulter). The gradient was harvested from the top in 500 μL fractions using the Piston Gradient fractionator (Biocomp) and the crosslinker was immediately quenched by the addition of 50 mM L-Lysine and L-Arginine (final concentration), from a 500 mM stock. Following incubation for 2 hours on ice, the monomeric 6-8 fractions were pooled and concentrated to ∼80 µL, transferred to a Slide-A- Lyzer MINI device (Thermo Fisher Scientific, MWCO 10 kDa), and then dialyzed overnight against 2 L of the minimal buffer (20 mM HEPES-KOH, pH 7.5, 150 mM KCl, 1 mM CaCl 2 , 1 mM DTT, 2.5% (v/v) Glycerol). Following dialysis for an additional 2 hours against 1 L minimal buffer, the concentration was adjusted to A280 ∼1.08 (absorbance at 280 nm) and the sample was used directly for cryo-EM grid preparation. Cryo-EM grids were prepared using a Vitrobot Mark IV plunger (Thermo Fisher Scientific), operated at 4°C and 100% humidity. Soluble dysferlin (1-2017) grids suitable for data collection were obtained through application of ∼3 μL crosslinked sample to one side of UltrAuFoil R1.2/1.3 300 gold mesh grids (Jena Bioscience, cat#X-201-AU300), which were pretreated with the Plasma Cleaner (Harrick Plasma, cat#PDC-32G) for 1 min at medium settings prior to their vitrification in liquid ethane, cooled by liquid nitrogen. For optimal sample vitrification the cryo-EM grids were blotted for 2-3 s using a blot force of 5 and stored in liquid nitrogen prior to screening/data collection. Like dysferlin, lipid-free myoferlin (1-1997) was stabilized through GraFix. However, in contrast to dysferlin (1-2017), myoferlin (1-1997) samples were assembled in a minimal buffer containing 200 mM KCl (25 mM HEPES-KOH, pH 7.5, 200 mM KCl, 2.5% (v/v) Glycerol, 1 mM TCEP), in the presence 0.5 mM CaCl 2 and 50 µM WJ460 (GlpBio, cat#GC38203), and then centrifuged for 90 min at 50,000 r.p.m. on 5-40% (w/v) sucrose Grafix gradients using a TLS-55 rotor, mounted in a Optima MAX-XP ultracentrifuge (Beckman Coulter). The lipid-free myoferlin (1-1997) gradients were harvested from the top in 100 µL fractions and the crosslinker was quenched with 50 mM L-Lysine and L-Arginine. The peak gradient fractions of myoferlin (1-1997) were concentrated and the buffer exchanged to the minimal buffer (containing 0.5 mM CaCl 2 ) by repeated ultrafiltration using a Vivaspin 500 spin concentrator (MWCO 50 kDa, cat#VS0131). Lipid-free myoferlin (1- 1997) was then vitrified in a mixture of liquid ethane and propane (37%:63%) cooled by liquid nitrogen, following the application of 3 µL sample at A280∼1.15 to a plasma-cleaned Quantifoil R1.2/1.3 200 copper mesh grid (Jena Bioscience, cat#X-101-CU200), which was blotted for 7.5 s using a blot force of 3. To assemble lipid-bound myoferlin (1-1997) complexes ( Appendix Fig S1B-C ), empty nanodiscs with the desired lipid composition were reconstituted in the presence of the MSP2N2 scaffold, as recently described ( Cannon et al., 2023 ). Purified His-tagged human MSP2N2 (∼2.9 mg/mL, obtained from Cube Biotech, cat# 26172) and DDM-solubilized lipids were mixed in ∼1:80-1:200 protein:lipids molar ratios, incubated at room temperature for 20 min, and dialyzed overnight at 4-8°C against 2L of reconstitution buffer (25 mM HEPES-KOH, pH 7.5, 150-200 mM KCl, 1 mM DTT), supplemented with 2.5 g/L Bio-Beads SM-2. Empty nanodiscs were subsequently purified by size-exclusion chromatography (SEC) on Superdex 200 Increase 10/300 GL (Cytiva, cat#28-9909-44), equilibrated in the SEC buffer (25 mM HEPES-KOH, pH 7.5, 150-200 mM KCl, 1.25%-2.5% (v/v) Glycerol, 1 mM DTT (or 0.5 mM TCEP), 0.5 mM CaCl 2 ). Peak MSP2N2 nanodisc fractions were concentrated to ∼2.2 mg/mL and added in ∼2-fold molar excess over purified soluble myoferlin (1-1997) in the SEC buffer, followed by incubation for 30 min at room temperature. The formation of stable myoferlin (1-1997)-nanodisc complexes ( Appendix Fig S1B-C ) was assessed by SEC on a Superose 6 Increase 10/300 GL (Cytiva, cat#29-0915-96) column, equilibrated in the SEC buffer. For cryo-EM grid preparation, the myoferlin-nanodisc complexes were crosslinked on ice in batch with 0.05-0.08% (v/v) GA for 10-30 min and the reaction was stopped by quenching with 50 mM L-Lysine and L-Arginine for 15 min on ice. Following centrifugation at 14,800 r.p.m. at 4°C for 5 min, the crosslinked complexes were applied to a Superose 6 column and myoferlin (1-1997)-nanodisc fractions were pooled and concentrated by ultrafiltration to A280∼0.7-0.8. Lipid-bound myoferlin (1- 1997) complexes were vitrified in liquid ethane-propane (37%:63%) following the application of 3 µL sample to plasma-cleaned Quantifoil R1.2/1.3 200 copper mesh grids, which were blotted for 6.5-8 s at a blot force of 3 (except for the 15% DOPS/2 mol% PI(4,5)P 2 myoferlin (1-1997)-nanodisc complex, frozen on an UltrAuFoil R1.2/1.3 300 gold mesh grid). Cryo-EM data collection and processing of lipid-bound myoferlin complexes Sample size calculation was not performed, and no randomization or blinding was required. All lipid-bound myoferlin cryo-EM datasets were acquired on a high-end Titan Krios G4 electron microscope (Collaborative Laboratory and User Facility for Electron Microscopy, Georg-August-Universität Göttingen), operated at an accelerating voltage of 300 kV and equipped with a Falcon4i direct electron detector and a Selectris X zero-loss energy filter ( Appendix Fig S2, S4 and S5, Tables S1-S4 ). All cryo-EM movies were recorded with EPU (Thermo Fisher Scientific) at a 165,000x nominal magnification at the specimen level (resulting in a 0.72 Å/pixel exposure sampling rate), using a slit width of 10 eV and a 50 μm C2 aperture (the objective aperture was not inserted). Cryo-EM movies were stored as raw camera frames (in EER format) following exposure over ∼3.16-3.46 s, which resulted in a total fluence of ∼38.03-39.89 e - /Å 2 ( Appendix Tables S1 - S4 ). The EER movies were fractionated in 40 EER fractions during on-the-fly preprocessing (patch motion correction, patch CTF estimation, dose-weighting) with cryoSPARC Live (v4.4-v4.5) and curated (contaminated or low-resolution exposures were removed from subsequent analyses). Myoferlin particles were picked first using the Blob Picker and then with the Template Picker (or with crYOLO ( Wagner, Merino et al., 2019 ), Appendix Fig S5 ), extracted in 360 pixel boxes and 2x binned, before being subjected to 3D and 2D classification in cryoSPARC. The initial particle sets were first cleaned by supervised 3D classification (Heterogeneous refinement) using one “good” ab initio generated reference volume and 4-5 “junk” classes. Particles assigned to the nanodisc-bound myoferlin classes were, additionally, cleaned in 2D, refined in 3D, and then re-extracted in a 360 pixel box (0.72 Å/pixel) before Non-uniform (NU) refinement in cryoSPARC. To resolve the peripheric C 2 G and the missing inner DysF domain (from some particles), the cryoSPARC-refined particles were next subjected to 3D classification in RELION-5.0-beta ( Kimanius et al., 2024 ). Briefly, the monomeric myoferlin-nanodisc particles were first classified in 3D without image alignment using 4 classes and soft masks applied to the inner DysF domain or to both C 2 G, the inner, and outer DysF ( Appendix Fig S2, S4, and S5 ). The subset of particles exhibiting stronger density for the inner DysF and C 2 G were then re-imported and refined in cryoSPARC ( Appendix Fig S2, S4 and S5 ). Prior to focused classification, the myoferlin particles of the 15 mol% DOPS/5 mol% Cholesterol nanodisc dataset were subjected to an additional round of 3D classification in 5 classes with image alignment ( Appendix Fig S5 ). For all lipid-bound myoferlin complexes, the final particle sets were subjected to sequential CTF refinement and reference- based motion correction in cryoSPARC to obtain the final maps. To facilitate model building, the lipid-bound myoferlin maps were sharpened using LocScale ( Jakobi et al., 2017 ) and DeepEMhancer ( Sanchez-Garcia et al., 2021 ), and the local resolution of the maps was estimated in cryoSPARC. The continuous flexibility and conformational space of membrane- bound myoferlin particles were evaluated in cryoSPARC using the 3D variability analysis (3DVA) tool. Cryo-EM data collection and image analysis of lipid-free myoferlin and dysferlin datasets As in the case of the membrane-bound samples, the lipid-free myoferlin and dysferlin datasets were acquired using EPU on the same microscope and detector (Falcon4i/Selectris), with energy filter set to slit widths of 10 eV or 15 eV ( Appendix Fig S6 and Fig EV3, Tables S5-S6 ). Lipid-free dysferlin cryo-EM movies were recorded at a nominal magnification of 165,000x (pixel size of 0.72 Å per pixel), from the same grid in two sessions over a ∼3.0 s exposure, resulting in a total fluence of ∼40.22 e - /Å 2 . Cryo-EM movies of the lipid-free myoferlin particles were collected in two sessions at total fluences of ∼39.91 e - /Å 2 and ∼39.85 e - /Å 2 , respectively, and saved as raw EER frames. Lipid-free ferlin datasets were preprocessed and curated on-the-fly in cryoSPARC Live (v4.1 (lipid-free dysferlin) and v4.4- v4.5 (lipid-free myoferlin)), including beam-induced motion correction, dose-weighting, patch CTF estimation, and particle picking (using the Blob Picker and Template Picker). Particles from “good” micrographs were extracted in a 360 pixel box (0.72 Å/pixel), 2x binned, split into subsets of ∼0.7-1 mln particles each ( Appendix Fig S6 and Fig EV3 ), before supervised 3D classification in cryoSPARC with 4 or 5 classes, which included an ab initio generated ferlin reference volume. Following an additional 2D classification, the lipid- free myoferlin and dysferlin particles were refined in 3D to obtain the consensus maps, which exhibited strong density of the Fer core region. The C 2 F and C 2 E domains could also be resolved in the lipid-free myoferlin consensus maps ( Fig EV3 ). As C 2 B and the C-terminal C 2 F-C 2 G appeared dynamic in the imaged lipid-free ferlin particles, multiple rounds of global and focused 3D classifications were employed, with the aim of finding more ordered conformations of the domains. Consequently, focused classification with soft mask applied to the C 2 F and C 2 G domain regions of the map, allowed us to identity more homogenous dysferlin particles, which refined in 3D to 3.5 Å (dysferlin (1-2017) map M2) and 4.8 Å (dysferlin (1-2017) map M3) resolutions. In these maps, C 2 F appears to establish a defined interface with the N-terminal C 2 B, whereas C 2 G occupies a location near C 2 F. By employing a similar 3D classification approach, we could also resolve the C 2 B-C 2 F-C 2 G interface in a subset of lipid-free myoferlin particles ( Fig EV3 ). These particles refined in 3D to 3.2 Å (myoferlin (1-1997) map M9) and 8.8 Å (myoferlin (1-1997) map M11) global resolutions and are reminiscent of the lipid-free dysferlin (1-2017) maps. Attempts to locate the N- terminal C 2 A domain by masked 3D classification were not successful, consistent with it not sharing significant contact interfaces with the remaining ferlin C 2 domains or with it being destabilized in vitreous ice. Model building and refinement Model building was initiated from Alphafold2 predictions of human myoferlin and dysferlin ( Jumper et al., 2021 ). The theoretical models were divided into individual domains (or structural modules, such as the Fer core ), fitted into the overall ferlin maps in ChimeraX (v1.6- v1.8) and UCSF Chimera (v1.16), and their fitting adjusted with ISOLDE ( Croll, 2018 ). All structural models were corrected and rebuilt in Coot (v0.9.8.5) ( Emsley & Cowtan, 2004 ) and iteratively refined with phenix.real_space_refine (Phenix v1.19.2-4158-v1.21.2-5419) ( Afonine et al., 2018 ) with nonbonded_weight set to 1000 for the final refinements. Modelling of bound Ca 2+ ions and phospholipid headgroups was enabled by F o - F c omit maps, calculated with Servalcat ( Yamashita, Palmer et al., 2021 ). Geometry restraints for phospholipid refinement were generated with Grade2 ( https://grade.globalphasing.org ). No cryo-EM density was observed for the WJ460 ligand ( Zhang et al., 2018 ), added to the lipid- free myoferlin (1-1997) and the 15 mol% DOPS/2 mol% PI(4,5)P 2 nanodisc-myoferlin complex. As the proposed binding site residues of C 2 D are not readily accessible in the lipid- free and lipid-bound myoferlin cryo-EM structures, further studies are likely needed to clarify whether myoferlin is indeed the cellular target of the small-molecular compound ( Zhang et al., 2018 ). All structural models were validated with MolProbity in Phenix. Structural figures were prepared with ChimeraX (v1.6-v1.8) and Pymol (v2.5.4, Schrödinger LLC). The contact interface areas were estimated with PISA (v1.5.2) ( Krissinel & Henrick, 2007 ). Data collection and refinement statistics are provided in Appendix Tables S1 - S6 . Chemical crosslinking mass spectrometry characterization of human dysferlin Prior to mass spectrometry analysis ( Appendix S14A-D ), purified soluble dysferlin (1-2017) was complexed with 1.5 mM CaCl 2 and crosslinked in batch with 0.3 mM BS3 (Thermo Fisher Scientific, cat#A39266) for 30 min at room temperature. The crosslinking reaction was stopped with 50 mM Tris-HCl, pH 8.0 and the sample was further purified by size-exclusion chromatography on Superose 6 Increase 10/300 GL (Cytiva), equilibrated in the sample buffer (20 mM HEPES-KOH, pH 7.5, 150 mM KCl, 5% (v/v) Glycerol, 2 mM DTT, 1.5 mM CaCl 2 ). Individual monomeric peak fractions (fraction 14 and 15) were ethanol precipitated and resuspended in 12.5 µl 8 M Urea. The samples were then diluted to 2 M Urea end concentration and reduced by the addition of 10 µl 50 mM DTT for 30 minutes at 37 °C. Alkylation was achieved by addition of 10 µl 200 mM iodoacetamide and incubation for 30 minutes at 25 °C. Unreacted iodoacetamide was quenched with additional 10 µl 50 mM DTT. Protein digestion was performed overnight at 37 °C with the addition of 1 µg of trypsin (Promega) in the presence of 1 M Urea and 50 mM Tris-HCl, pH 7.75. The samples were acidified with formic acid to 0.1% (v/v) end concentration and acetonitrile was added to 5% (v/v) end concentration. Digested peptides were desalted with C18 reversed-phase MicroSpin columns (Harvard Apparatus). Bound peptides were eluted with 50 % (v/v) acetonitrile, 0.1% (v/v) formic acid. The samples were subsequently dried under vacuum and resuspended in 75 µl 2% (v/v) acetonitrile, 0.05% (v/v) TFA and 5 µl were used for LC-MS analysis. Chromatographic separation was achieved with Dionex Ultimate 3000 UHPLC (Thermo Fischer Scientific) coupled with C18 column packed in-house (ReproSil-Pur 120 C18-AQ, 1.9 µm particle size, 75 µm inner diameter, 33 cm length, Dr. Maisch GmbH) over a 74 minute linear gradient from 8% - 46% mobile phase B (mobile phase A - 0.1% (v/v) FA and B - 80% (v/v) ACN, 0.08% (v/v) FA). Eluting peptides were analyzed with Orbitrap Exploris 480 (Thermo Fischer Scientific) and the following settings for survey scans: Resolution – 120 000; Scan range – 380-1600; AGC target – 300%; Maximum injection time set to “Auto”. Analytes with charge state 3 to 8 were selected for fragmentation with 28% normalized collision energy. Dynamic exclusion was set to 15 s. Fragment spectra were acquired with the following settings: Resolution – 30 000; Isolation window – 1.6 m/z ; AGC target – 100%; Maximum injection time – 128 ms. Resulting .raw files were analyzed with pLink ( Chen et al., 2019 ) (v 2.3.9) against a database containing the sequence of the protein. Carbamidomethyl on cysteines was selected as fixed modification and oxidation of methionines as variable modifications. BS3 was selected as crosslinker, peptide tolerance was set to 6 ppm and False discovery rate was set to 1%. The identified crosslinked residues ( Dataset EV1 ) were mapped onto the Alphafold2 prediction of dysferlin and the lipid-free dysferlin (1-2017) cryo-EM model using xiNet ( Combe, Fischer et al., 2015 ), analysed and visualized in PyMOL (v2.5.4). Expanded View for this article is available online. Author contributions CC, TM and JP designed the research. CC engineered and cloned constructs, prepared samples for structural studies, collected and processed cryo-EM data. CC built, refined and interpreted the ferlin structural models. CC designed and performed functional assays, under the supervision of JP and TM. CC and AC prepared mass spectrometry samples. AC collected and analysed the proteomics and chemical crosslinking mass spectrometry data under the supervision of HU. CZKS and VP carried out the mass photometry measurements and contributed to data analysis. CC wrote the paper under the supervision of TM and JP and with input from all the authors. CC, TM and JP acquired funding for the project. Disclosure and competing interests statement The authors of this study declare that they have no conflict of interest. Data availability The atomic coordinates of the lipid-bound myoferlin were deposited in the Protein Data Bank (PDB) and Electron Microscopy Data Bank (EMDB) with the following accessing codes: PDB 9H6X/EMD-51902 (lipid-bound human myoferlin, 25 mol% DOPS/5 mol% PI(4,5)P 2 nanodisc). Extended View Figure Legends Download figure Open in new tab Figure EV1. Cryo-EM density snapshots of the lipid-bound soluble myoferlin (1-1997). A Selected cryo-EM density snapshots of C 2 B and of its proximal motifs (the linker helix and the FerI motif) cryo-EM density. Myoferlin’s residues are depicted as sticks. B Cryo-EM density of myoferlin’s C 2 C domain. The three bound Ca 2+ -ions and the recruited phosphatidylserine (PS) are indicated. C Ca 2+ -binding sites of the C 2 C domain of myoferlin. The cryo-EM density is contoured around the Ca 2+ -binding residues. D Cryo-EM density of the FerA motif of myoferlin. Note that the four-helix bundle domain inserts between the β4 and β5 strands of the C 2 CD domain (red, shown in a cartoon representation). E Cryo-EM density of the C 2 CD domain of myoferlin. The seven β-strands of the domain are indicated. F Cryo-EM densities of the modelled inner and outer DysF motifs of myoferlin. G Cryo-EM density and the observed Ca 2+ -binding sites of the C 2 D domain of myoferlin. C 2 D’s cryo-EM density is contoured around the Ca 2+ -binding sites (bottom panel). H Density of the C 2 E domain. The domain comprises an extended insertion loop (denoted as the anchor loop), inserted between the β6-β7 strands. The loop establishes contacts with the downstream C 2 F domain. I Cryo-EM density of C 2 F and of its Ca 2+ -/lipid-binding sites. As in the case of C 2 C, three Ca 2+ ions and a PS headgroup could be assigned in the cryo-EM density map. J Cryo-EM density of the C-terminal C 2 G domain, originating from the 25 mol% DOPS/5 mol% PI(4,5)P 2 myoferlin-nanodisc complex, shown in two orientations. The lipid-binding β- hairpin motif of C 2 G is coloured in light orange. K Density of the C-terminal C 2 G domain, resolved in the 15 mol% DOPS/2 mol% PI(4,5)P 2 myoferlin-nanodisc complex. Note that the insertion loop of C 2 G (residues 1906-1944) is ordered in this myoferlin complex. L Modelled Ca 2+ -binding sites of C 2 G, observed in the lipid-bound myoferlin structure (15 mol% DOPS/2 mol% PI(4,5)P 2 ). Because of the lower local resolution, Ca 2+ modelling was initiated by Alphafold3 predictions ( Abramson, Adler et al., 2024 ), and the optimal sites were refined against the cryo-EM map. Our modelling (right panel) suggests that two Ca 2+ ions are coordinated the L1 and L3 loops of C 2 G, in a phospholipid-independent manner ( Rizo & Sudhof, 1998 ). M Cryo-EM densities of the Ca 2+ -bound phosphatidylserine (PS) moieties, resolved in the myoferlin-nanodisc complexes. The cryo-EM density is contoured around the modelled phospholipid headgroups, and the ligands are depicted as sticks. Download figure Open in new tab Figure EV2. Structural comparison between the different lipid-bound myoferlin complexes. A Overall cryo-EM map (map M6, ∼3.3 Å) of soluble myoferlin (1-1997) bound to an MSP2N2 nanodisc comprising 15 mol% DOPS and 5 mol% Cholesterol. The map is depicted in three different orientations and color-coded. A low-passed map (transparent surface) is superimposed to visualize the nanodisc density. B The map M6 of the myoferlin (1-1997)-MSP2N2 complex with the final model fitted inside. Note the well-resolved C 2 domain linker regions, the linker helix, and the anchor loop of C 2 E. C Overall cryo-EM map (map M8, ∼3.43 Å) of the soluble myoferlin (1-1997) bound to an MSP2N2 nanodisc comprising 15 mol% DOPS and 2 mol% PI(4,5)P 2 . The map is depicted in three different orientations as in A . D Side-views of the map M8 (15 mol% DOPS/2 mol% PI(4,5)P 2 ) of the lipid-bound myoferlin (1-1997) together with the fitted model. The map has been low-pass filtered to 8 Å and the ordered nanodisc density is indicated. E Cryo-EM map of the soluble myoferlin (1-1997)-lipid complex assembled onto a 25 mol% DOPS/5 mol% PI(4,5)P 2 MSP2N2 nanodisc (∼2.56 Å, map M3). F Overall map of lipid-bound myoferlin (1-1997) complex (map M3) with the fitted final model. Except the flexible N-terminal C 2 A, all the ordered myoferlin domains (C 2 B-C 2 G) and accessory motifs (FerA, the linker helix, the anchor loop) could be assigned, resulting in the near-complete structure of ferlin’s cytosolic region. G Map versus model Fourier Shell Correlation (FSC) plots for the myoferlin (1-1997)- nanodisc complexes. Download figure Open in new tab Figure EV3. Cryo-EM image analysis of the vitrified lipid-free soluble myoferlin. A Cryo-EM image processing schematic for the Ca 2+ -bound, lipid-free myoferlin (1-1997). Resolution of the final maps (M9, M10, and M11 maps) was estimated according to the gold- standard Fourier Shell Correlation (FSC) criterion of 0.143. The structural domains of myoferlin are color-coded as in Fig 1 . B Reference-free 2D class averages of lipid-free myoferlin (1-1997). Note the similarities between the dysferlin (1-2017) ( Appendix Fig S6C ) and myoferlin (1-1997) 2D class averages in their lipid-free states. C Cryo-EM density snapshots of the myoferlin (1-1997) C 2 C, C 2 D and C 2 CD-FerA domains. The final model is fitted inside and the Fer core map of myoferlin (map M10) is depicted as a transparent surface. D Cryo-EM density of the C 2 F (map M9) and C 2 G (map M11) domains as modelled in the lipid-free myoferlin (1-1997) structure. Note the presence of tertiary interfaces between the C 2 F-C 2 B and C 2 F-C 2 G, as also observed in the lipid-free dysferlin (1-2017) structure. E Side-by-side comparison between the membrane-bound (map M3) and lipid-free (map M11) myoferlin structures. The domains undergoing significant displacement upon membrane binding (C 2 B, C 2 F, and C 2 G) are indicated with dashed arrows. F Ca 2+ -binding sites observed in the lipid-free myoferlin structure. The cryo-EM density is coloured in cyan and depicted as a transparent surface. The two modelled Ca 2+ ions, bound to C 2 C and C 2 D, are coloured green. G Global resolution estimates for the lipid-free myoferlin (1-1997) cryo-EM maps using Fourier Shell Correction (FSC) between half maps. The resolution estimates at FSC=0.143 and FSC=0.5 are indicated. H Local resolution of the overall map of the lipid-free soluble myoferlin (map M9). The map regions coloured in red indicate a higher resolution. I Angular distribution of the myoferlin particles contributing to the M9 overall map. The M9 map is shown above the 3D distribution representation. Relative cylinder height and the red colour indicate a higher number of particle images. Download figure Open in new tab Figure EV4. Full-length myoferlin and dysferlin appear to promote tight binding between vesicles in lipid mixing assays. A Comparison between the local lipid bilayer structure observed in the myoferlin (1-1997)- nanodisc cryo-EM maps. Note that the lipid bilayer forms close contacts with the concave surface of C 2 G in the presence (top), but not in the absence of PI(4,5)P 2 (bottom) in the assembled nanodiscs. The cryo-EM maps have been low-passed to 10 Å and the myoferlin model is fitted inside. B The C 2 G-proximal cryo-EM density element, with a tentatively modelled PI(4,5)P 2 headgroup. C 2 G and C 2 B residues, located at the interface, are displayed as sticks. The modelled PI(4,5)P 2 is located in close proximity to three lysine residues (K1830, K1828 and K1816), projecting from the concave surface of C 2 G. Note the proximal location of the Ca 2+ - binding L3 loop of C 2 G. C Polar contacts between C 2 G residues of the polybasic patch and a tentative acidic phospholipid, possibly PI(4,5)P 2 , originating from the nanodisc bilayer (comprising 25 mol% DOPS and 5 mol% PI(4,5)P 2 ). D Lipid mixing assays between myoferlin proteoliposomes and DOPS/PI(4,5)P 2 -bearing vesicles in the presence of different concentrations of Ca 2+. The fluorescent liposomes (100 nm LUVs, 25 mol% DOPS and 5 mol % PI(4,5)P 2 ) were labelled with the Lissamine Rhodamine B-NBD (Nitro-2,1,3-benzoxadiazole-4-yl) FRET (Förster Resonance Energy Transfer) pair and the NBD dequenching signal, due to lipid dilution, was used to monitor the extent of tight vesicle-vesicle docking (and, possibly, fusion) to non-fluorescent proteoliposomes. LUVs lacking full-length myoferlin (1-2061) were used as a control. The lipid mixing fluorescent traces were smoothened (using the Savitzky-Golay method with 10 points) and normalized to the maximal dequenching signal to calculate the lipid mixing efficiency. The initial jump in fluorescence is likely due to partial, Ca 2+ -independent myoferlin-induced aggregation of proteoliposomes. The assays were repeated at least three times and representative fluorescent traces are shown. E Lipid mixing assays between dysferlin proteoliposomes and DOPS/PI(4,5)P 2 -containing vesicles. The ability of the full-length dysferlin (1-2080) to promote tight vesicle-vesicle docking (or, possibly, vesicle-vesicle fusion) was monitored as a function of Ca 2+ concentration. The lipid mixing activity of soluble dysferlin (1-2017) was also tested (deep blue). Empty LUVs were used as a control. The assays were repeated at least three times and representative fluorescent traces are shown. F-G Ca 2+ -dependent liposome binding by soluble myoferlin (1-1997) and dysferlin (1-2017) using a coflotation assay. The liposomes used in the assay had a similar lipid composition as in D-E . The Nycodenz gradients were harvested from the top and analysed by SDS-PAGE. Note the increased liposome binding by both myoferlin (1-1997) and dysferlin (1-2017) in the presence of PI(4,5)P 2 -containing LUVs. Appendix Download figure Open in new tab Figure S1. Biochemical characterization of human myoferlin and dysferlin. A SDS-PAGE analysis of the purified, insect cell expressed myoferlin and dysferlin constructs used in structural and functional analyses. B Size-exclusion chromatography (SEC) elution profile of the soluble myoferlin (residues 1-1997, sMYOF), soluble dysferlin (residues 1-2017, sDYSF), and the reconstituted myoferlin (1-1997)-MSP2N2 complexes. The MSP2N2-based nanodiscs, comprising 15 mol% DOPS or 25 mol% DOPS/5 mol% PI(4,5)P 2 anionic phospholipids, were purified through SEC prior to the assembly of their complexes with sMYOF. Gel filtration standards (dashed lines) were applied to the Superose 6 column in the same buffer and their molecular masses are indicated. C SDS-PAGE analysis of the size-exclusion chromatography peak fractions from B. To allow the compact visualization of the myoferlin (1-1997) and dysferlin (1-2017) fractions, the top two gels were cropped (see also S1A ). D Ca 2+ -sensitive liposome binding activity of soluble myoferlin (1-1997) using a coflotation assay. The myoferlin-liposome (LUVs) complexes were assembled in vitro in the presence of 50 µM Ca 2+ or Mg 2+ and subjected to ultracentrifugation on a Nycodenz step gradient, which was harvested from the top. The anionic phospholipid compositions of the tested LUVs are indicated. LUVs lacking anionic phospholipids (DOPC/DOPE-only LUVs) were used as a control. E Ca 2+ -binding activity of membrane-anchored, full-length myoferlin (residues 1-2061, in detergent micelles) and soluble myoferlin (residues 1-1997) using a nanoDSF-based assay. Error bars represent the standard error of the mean (s.e.m., soluble myoferlin (1-1997): n=4, full-length myoferlin (1-2061)-DDM: n=7). [Ca 2+ ] 1/2 was estimated by nonlinear regression curve fit and the Mg 2+ titration was used as a control. F Ca 2+ -binding activity of full-length dysferlin (residues 1-2080, in detergent micelles, DDM), of soluble dysferlin (residues 1-2017) and of dysferlin proteoliposomes (dysferlin (1-2080)-LUVs) using a nanoDSF-based assay. The dysferlin proteoliposomes were prepared by dialysis of the detergent reconstituted full-length sample in the presence of extruded LUVs (assembled from 70 mol% DOPC, 20 mol% DOPE and 10 mol% Cholesterol). The error bars represent the s.e.m. (dysferlin (1-2017): n=4, dysferlin (1-2080)-DDM: n=6, dysferlin (1-2080)-LUV: n=6). G Ca 2+ -sensitive liposome binding activity of soluble myoferlin (residues 1-1997, sMYOF) using a FRET-based lipid-binding assay. Dansyl-labelled LUVs, prepared with DOPS (15 mol% DOPS), DOPS and PI(4,5)P 2 (25 mol% DOPS and 5 mol% PI(4,5)P 2 ) or without anionic phospholipids (DOPC/DOPE-only), were used at a 50 µM concentration. Error bars represent s.e.m. (n=3). H Mass-photometry characterization of full-length dysferlin (residues 1-2080) reconstituted in detergent micelles (LMNG), showing a major population of monomeric dysferlin particles (molar mass of ∼283 kDa). The inset depicts the mass distributions of dysferlin’s major species at two Ca 2+ concentrations (2 mM CaCl 2 and 4 mM CaCl 2 , orange and green curves, respectively), compared to the apo state (blue). In all cases, the mass photometry measurements were carried out at a ∼50 nM dysferlin concentration and in a sample buffer devoid of detergent. A dysferlin domain composition schematic is shown above the mass photometry profile of the ferlin. The mass photometry measurements were repeated four times (n=4). Download figure Open in new tab Figure S2. Computational image analysis of the vitrified myoferlin-MSP2N2 nanodisc complex (25 mol% DOPS/5 mol% PI(4,5)P 2 cryo-EM datasets). A Cryo-EM micrograph of the vitrified myoferlin (1-1997)-MSP2N2 complex (25 mol% DOPS and 5 mol% PI(4,5)P 2 nanodisc). A characteristic, monomeric particle is circled in cyan, in both the raw (left) and denoised (right) exposure. B Reference-free 2D class averages of the myoferlin (1-1997)-MSP2N2 complex (comprising 25 mol% DOPS and 5 mol% PI(4,5)P 2 ). The visible myoferlin domains are indicated on a characteristic top view of the complex (bottom panel). C Cryo-EM data processing routine of the lipid-bound myoferlin, assembled on a 25 mol% DOPS/5 mol% PI(4,5)P 2 MSP2N2 nanodisc (see also Fig S1B-C ). The final maps are boxed and colored-coded after the modeled domains ( e.g. , as in Fig 1 ). Download figure Open in new tab Figure S3. Resolution and quality of the lipid-bound myoferlin cryo-EM maps (25 mol% DOPS/5 mol% PI(4,5)P 2 nanodisc). A Fourier shell correlation (FSC) between the final myoferlin (1-1997)-MSP2N2 cryo-EM half-maps (25 mol% DOPS/5 mol% PI(4,5)P 2 ) datasets). The gold-standard FSC criterion (FSC=0.143) indicates global resolution estimates of ∼2.39 Å (consensus map), ∼2.46 Å (after masked C 2 G classification), and ∼2.56 Å (after masked DysF classification) for the M1, M2, and M3 maps, respectively. B-D Directional resolution and Fourier sampling of the myoferlin (1-1997)-MSP2N2 cryo-EM maps. The conical FSC, the conical FSC area ratio (cFAR), and the sampling compensation factor (SCF) were calculated and plotted with cryoSPARC v4.5. E Local resolution of the myoferlin (1-1997)-MSP2N2 cryo-EM maps (25 mol% DOPS/5 mol% PI(4,5)P 2 nanodisc). Red color indicates a higher resolution. F Map versus model FSC plot for the myoferlin (1-1997)-MSP2N2 complex (map M3, 25 mol% DOPS/5 mol% PI(4,5)P 2 datasets). G 3D angular distribution of the myoferlin particles contributing to the overall (M3) map (25 mol% DOPS/5 mol% PI(4,5)P 2 ) datasets). The similarly oriented and color-coded maps are shown above the angular distribution plot. Red color and relative cylinder height indicates a higher number of particle images. Download figure Open in new tab Figure S4. Cryo-EM image analysis of the vitrified myoferlin (1-1997)-MSP2N2 nanodisc complex (25 mol% DOPS/5 mol% PI(4,5)P 2 /5 mol% Cholesterol cryo-EM datasets). A Cryo-EM processing schematic for the soluble myoferlin (1-1997), bound to a 25 mol% DOPS/5 mol% PI(4,5)P 2 /5 mol% Cholesterol MSP2N2 nanodisc. The resolution estimates (FSC=0.143 criterion) of the final maps are indicated and the final maps are colored-coded. B Cryo-EM micrograph (raw and denoised) of the imaged myoferlin (1-1997)-MSP2N2 complex (comprising 25 mol% DOPS, 5 mol% PI(4,5)P 2 and 5 mol% Cholesterol). A single vitrified particle is circled in cyan. C Reference-free 2D class averages of the myoferlin (1-1997)-MSP2N2 complex (25 mol% DOPS, 5 mol% PI(4,5)P 2 , and 5 mol% Cholesterol datasets). D Superposition between the consensus cryo-EM maps of nanodisc-bound myoferlin (1-1997), obtained in the absence (purple, 25 mol% DOPS/5 mol% PI(4,5)P 2 ) or presence of Cholesterol (color-coded, 25 mol% DOPS/5 mol% PI(4,5)P 2 /5 mol% Cholesterol). E Low-passed cryo-EM map of the lipid-bound myoferlin (25 mol% DOPS/5 mol% PI(4,5)P 2 /5 mol% Cholesterol nanodisc, map M4 C) showing ordered lipid headgroups, including a tentative PI(4,5)P 2 , located close to the interface between C 2 F, C 2 G and C 2 B. F-H Fourier shell correlation (FSC) plots for the myoferlin (1-1997)-nanodisc maps (25 mol% DOPS/5 mol% PI(4,5)P 2 /5 mol% Cholesterol datasets). Download figure Open in new tab Figure S5. Cryo-EM analysis of the myoferlin (1-1997)-lipid complexes formed on 15 mol% DOPS/5 mol% Cholesterol and 15 mol%/2 mol% PI(4,5)P 2 MSP2N2 nanodiscs. A-B Cryo-EM (raw and denoised) micrograph and 2D class averages of the myoferlin (1-1997)-MSP2N2 complex (comprising 15 mol% DOPS and 5 mol% Cholesterol). C Cryo-EM data processing flow-chart for the vitrified myoferlin (1-1997), bound to a 15 mol% DOPS/5 mol% Cholesterol nanodisc. The final maps are color-coded after the modeled domains and the resolution estimates (after the FSC=0.143 criterion) are indicated. D-F Cryo-EM image analysis of the soluble myoferlin (1-1997)-lipid complex (assembled onto a 15 mol% DOPS/2 mol% PI(4,5)P 2 nanodisc). G Superposition between the M6 (15 mol% DOPS/5 mol% Cholesterol) and M8 (15 mol% DOPS/2 mol% PI(4,5)P 2 ) nanodisc-bound myoferlin (1-1997) maps. The two cryo-EM maps of the complex are nearly identical (CC=0.96). H-I Fourier shell correlation (FSC) plots for the lipid-bound myoferlin (1-1997) maps (comprising 15 mol% DOPS/2 mol% PI(4,5)P2 and 15 mol% DOPS/5 mol% Cholesterol). J Local resolution of the myoferlin (1-1997)-MSP2N2 cryo-EM maps depicted in C and F . The map regions colored in red color indicate a higher resolution. Download figure Open in new tab Figure S6. Computational image analysis of the lipid-free soluble dysferlin (1-2017) cryo-EM datasets. A Cryo-EM data processing routine of the Ca 2+ -bound soluble dysferlin (sDYSF, residues 1-2017). Resolution of the final maps (M1-M4) was estimated according to the gold-standard Fourier Shell Correlation (FSC) cutoff of 0.143. The key structural domains of dysferlin (1-2017) are color-coded and indicated in the final maps. B Typical cryo-EM micrograph (raw and denoised) of the vitrified lipid-free soluble dysferlin (1-2017). An imaged dysferlin particle is indicated in cyan. C Reference-free 2D class averages of lipid-free dysferlin (1-2017). 2D class averages calculated from the locked and loose dysferlin particle images are shown in the bottom panel. The C 2 F domain is indicated with an asterisk. The C-terminal domain is more ordered in the locked (closed) dysferlin 2D classes (left panel) compared to the loose state (right panel). Download figure Open in new tab Figure S7. Resolution and quality of the lipid-free dysferlin (1-2017) cryo-EM maps. A Local resolution of the lipid-free Fer core map (map M1) of dysferlin (1-2017). Red colored map regions indicate a higher resolution. B Overall maps of Ca 2+ -bound soluble dysferlin (map M2 and M3), colored after their local resolution. Note the higher local resolution of the Fer core (C 2 C-C 2 CD-FerA-C 2 D) compared to the more dynamic C 2 B-C 2 F-C 2 E. C-E. Fourier shell correlation (FSC) between the final lipid-free dysferlin (1-2017) half maps. The gold-standard FSC criterion (FSC=0.143) indicates global resolution estimates of ∼2.9 Å, ∼3.5 Å, and ∼4.8 Å for the M1, M2 and M3 maps, respectively. F 3D angular distribution of the dysferlin particles contributing to the core (M1) and overall (M2) maps. The similarly oriented maps are shown above the angular distribution representation. Red color and relative cylinder height indicate a higher number of particle images. G Tertiary interfaces between C 2 F, C 2 B, and C 2 E observed in the dysferlin (1-2017) cryo-EM model. In the lipid- free state, the C 2 F domain is positioned at the base of the dysferlin ring by multiple tertiary interactions with the N- terminal C 2 B and C-terminal C 2 E. The final model of the soluble dysferlin (1-2017) is represented as cartoons and fitted inside the overall cryo-EM map (transparent surface). The L1-L4 loops of C 2 F are colored pink, blue, teal, and green, respectively. H Selected polar interactions between C 2 F’s L3-L4 loops and C 2 B. The polar contacts between interacting residues are depicted as dashed lines. I Polar and hydrophobic interactions between the bottom loops of C 2 F and the L2 loop of C 2 E. Residues involved in hydrophobic interactions are shown as transparent spheres. J Contact interfaces between C 2 F and the anchor loop of C 2 E (residues 1475-1514). C 2 B and the anchor loop are shown as transparent surfaces. Download figure Open in new tab Figure S8. Snapshots of the lipid-free dysferlin (1-2017) cryo-EM density. A Selected snapshots of C 2 C’s cryo-EM density. C 2 C’s β-strands as shown as sticks. B Ca 2+ -binding sites of dysferlin’s C 2 C and C 2 D domains. The cryo-EM density map is colored in cyan and contoured around each of the two Ca 2+ -binding sites. C Cryo-EM density of the FerA domain helices. The FerA α-helices are amphipathic and form a typical, up-down- up-down four-helix bundle. D Cryo-EM densities of the C 2 D’s β-strands. E Cryo-EM density of C 2 CD’s L4 loop that bridges the C 2 C and C 2 D domains. F Cryo-EM densities of C 2 CD’s β6-β7 and L2 loops, contacting the FerA and C 2 D, domains, respectively. G Densities of selected C 2 CD β-strands. H-I Modeling of C 2 B, C 2 E and C 2 F in the locked (closed) state cryo-EM map (map M2). Note that the top loops of C 2 B and C 2 E face opposite surfaces of the dysferlin plane, whereas the top loops of C 2 F adopt a lateral orientation and are located close to C 2 B. J Side-view of C 2 B’s density with the modeled domain fitted inside the cryo-EM map. K Cryo-EM densities of the two, outer and inner DysF motifs of dysferlin. The outer and inner DysF occupy a peripheric location, between C 2 CD and C 2 D, and share a dynamic contact interface ( i.e. , not present in all imaged dysferlin particles). L Densities of the C-terminal C 2 F and C 2 G domains. The cryo-EM map (map M3) has been low-passed to 10 Å. The additional density element, connected to C 2 B, might correspond to the flexible, N-terminal C 2 A domain, which likely samples many conformational states in solution. The long linker between C 2 A and C 2 B is, likely, largely unstructured as it is susceptible to protease digestion in cells. Note that cryo-EM density for the linker helix is absent in the lipid-free dysferlin state. Download figure Open in new tab Figure S9. Structural organization of dysferlin’s Fer core module. A Structure of dysferlin’s core module (Fer core ) resolved at 2.9 Å resolution. The focused cryo-EM map of the C 2 C-C 2 D region (map M1), shown in two different orientations, has been sharpened and colored after the modelled subunits. B-C Interaction interfaces between C 2 CD and its neighboring FerA, C 2 C and C 2 D domains. The cryo-EM maps of the dysferlin domains have been sharpened. D An insertion in C 2 CD’s β6-β7 appears to stabilize the pose of the FerA domain. Dysferlin’s domains are depicted in cartoon representation and the sharpened cryo-EM map is contoured around FerA and C 2 D. The key polar interactions between C 2 CD’s β6-β7 loop and FerA are shows as dashed blue lines. E C 2 CD’s top loops are engaged in multiple contacts and bridge the C 2 C and C 2 D domains. The L1-L4 loops of C 2 CD are colored green, light blue, purple and blue, respectively. C 2 CD and FerA are depicted as cryo-EM maps. F Selected polar interactions between C 2 CD’s top loops (L1-L4) and the bridged C 2 C and C 2 D domains. The polar interactions are shown as dashed blue lines. C 2 C and C 2 D are depicted as surfaces. G The extensive contact interfaces between C 2 CD-FerA and the Ca 2+ -binding C 2 C and C 2 D. C 2 CD and FerA are shown as cryo-EM maps, whereas C 2 C and C 2 D are depicted as cartoons. Download figure Open in new tab Figure S10. C 2 CD represents a new C 2 -like ferlin structural domain. A Structure of the C 2 CD-FerA module, using dysferlin as an example. Note that the FerA helical bundle inserts between C 2 CD’s β4-β5 strands and projects from the bottom side of the C 2 -like domain. B Topology diagram of the C 2 CD domain. As all the other ferlin C 2 domains, C 2 CD has a type-II topology ( i.e. , the two β-sheets are formed by β1-β4-β7-β8 and β2-β3-β5-β6, respectively). C Surface-charge distribution of the C 2 CD-FerA module of dysferlin. The electrostatic potentials were colored from red (negative, -3 kT/e) to blue (positive, +3 kT/e) and mapped onto the solvent excluded surface. D Cryo-EM density of dysferlin’s C 2 CD-FerA module oriented and colored as in A . E-F Selected intramolecular contacts between C 2 CD and FerA. The polar interactions are represented as dashed lines. G C 2 CD’s β6-β7 loop interacts with FerA’s bottom side. FerA’s α1 helix is shown as surface representation and the C 2 CD interacting residues are colored red. H A similarly organized C 2 CD-FerA module is likely present in all human ferlins. Alphafold2 prediction of otoferlin’s C 2 CD-FerA region (right) is depicted side-by-side with its structural superposition to the cryo-EM model of dysferlin’s C 2 CD-FerA (left). Note the predicted large insertion between C 2 CD’s β2-β4 strands observed in otoferlin. Download figure Open in new tab Figure S11. Structural conservation of ferlins’ C 2 CD, FerA and DysF motifs. A Structural superposition between C 2 CD and the Ca 2+ -bound C 2 C domain of dysferlin, as an exemplary type-I ferlin. C 2 C’s Ca 2+ ions are shown as green spheres. B Structural similarity between dysferlin’s FerA motif and the four-helix bundle of BCAR1 (p130Cas, PDB 3t6g). The direction of the four α-helices is indicated by arrows. C Alignments between the C 2 CD of dysferlin and the structurally similar C 2 domains of the SHIP2 phosphatase (PDB 5okm), Cep120 (PDB 6flj), Munc13-1 (PDB 5ue8) and NEDD4 (PDB 3m7f). The C 2 domains were aligned using Foldseek (van Kempen, Kim et al., 2024). D Structural organization of extended synaptotagmin 2 (ESYT2) and human MCTP2 (Alphafold2 prediction). SMP – synaptotagmin - like mitochondrial lipid-binding protein; TMR1/2 – transmembrane region 1 and 2. E Alphafold2 prediction of TECPR1 (Tectonin Beta-Propeller Repeat-containing Protein 1)’s structure. The tectonin repeat β-propellers are colored blue and cyan respectively (TR1, TR2). The two spatially separated DysF motifs of TECPR1 are colored in green (DysF1) and magenta (DysF2), respectively. F Superposition between dysferlin’s outer and inner DysF motifs, as modelled in the cryo-EM map, and the related DysF1 domain of TECPR1, predicted by Alphafold2 ( Jumper et al., 2021 ). G Structural superposition between the DysF motifs of dysferlin and TECPR1’s DysF1. H Stabilizing arginine/tryptophan (R/W) stacking interactions are present in both dysferlin/myoferlin’s DysF and TECPR1’s DysF motifs. In both cases, the R/W stacks (shown as sticks) are mainly observed on one side of the DysF β-sheet (Patel, Harris et al., 2008, Sula, Cole et al., 2014). Download figure Open in new tab Figure S12. Ca 2+ -binding sites of the Fer core module of dysferlin, modeled in the lipid-free state. A In dysferlin, two Ca 2+ ions bind C 2 C’s Ca 2+ -binding pocket. The cryo-EM map (cyan mesh) is contoured around the Ca 2+ ions (green spheres). Residues substituted to alanine (in D-E ) are marked with asterisks. B The Ca 2+ -binding sites of the C 2 D domain. Residues substituted for alanine in D-E are indicated with asterisks. C Comparison between Ca 2+ -coordination by dysferlin’s C 2 C and C 2 A (PDB 7jof), and the C 2 -domains of PKCα (PDB 1dsy) and perforin (PDB 4y1t). C 2 C’s Ca 2+ ions are shown as magenta spheres, whereas the Ca 2+ sites of C 2 A, PKCα and perforin are colored in cyan. D Ca 2+ -binding activities of the Fer core of dysferlin (residues 214-1281) and of its C 2 D-QDA (“quadruple D (aspartate) to A (alanine)”) and C 2 C-TDA (“triple D (aspartate) to A (alanine)”)/C 2 D-QDA mutants measured using a nanoDSF assay. Four Ca 2+ -binding aspartates were substituted in C 2 D-QDA (D1168A, D1174A, D1230A, D1232A). Three additional residues were substituted to alanine (D417A, D465A, D473A) in the C 2 C-TDA/C 2 D-QDA variant (see also Appendix Fig S13A-C ). The Ca 2+ -activity measurements were performed in triplicates (n=3) for the C 2 D-QDA and C 2 C-TDA/C 2 D-QDA samples and repeated five times for the wild-type Fer core sample (n=5). Error bars represent the standard deviation (s.d.). The melting temperatures (T m ) of the analyzed samples were plotted as a function of Ca 2+ /Mg 2+ concentration and the [Ca 2+ ] 1/2 values were estimated by nonlinear regression curve fitting. E Bar plot of the Ca 2+ -activity data shown in D . Error bars represent the s.d.. F Ca 2+ -sensitive liposome binding activity of the wild-type and C 2 C-TDA/C 2 D-QDA Fer core of dysferlin (residues 214-1281), assessed using coflotation assays. The liposomes (LUVs) comprised 25 mol% DOPS and 5 mol% PI(4,5)P 2 . The Nycodenz step gradients were harvested from the top and analyzed by SDS-PAGE. Note that substitution of C 2 C’s and C 2 D’s Ca 2+ -binding sites reduced significantly the ability of the Fer core module to interact with lipid membranes comprising acidic phospholipids. G Modeled Ca 2+ -binding sites in dysferlin’s C 2 C and the C 2 domain of phospholipase A2 (PLA2, PDB 6eij). The Ca 2+ ions are depicted as spheres and the coordinating side-chains are shown as sticks. Download figure Open in new tab Figure S13. Biochemical validation of dysferlin’s Ca 2+ -binding sites. A Size-exclusion chromatography profiles of C 2 C and C 2 D Ca 2+ -binding mutants. The purified mutants have similar retention volumes on Superdex 200, comparable to the wild-type Fer core sample (dysferlin (214-1281)). Therefore, substitution of two (C 2 D-DDA-1: D1168A/D1174A and C 2 D-DDA2:D1230A/D1232A) or four Ca 2+ -binding aspartates (C 2 D-QDA: D1168A/D1174A/D1230A/D1232A) in C 2 D does not have a significant impact on Fer core ’s structure ( i.e. , does not destabilize its overall fold). Consistently, combined substitutions in C 2 C (C 2 C-TDA: D417A/D465A/D473A) and C 2 D (C 2 D-QDA) preserve the structural integrity of the Fer core of dysferlin while abolishing its Ca 2+ -binding activity (see also Appendix Fig S12D-E ). SDS-PAGE gels of the Fer core dysferlin mutants carrying substitutions in C 2 C’s and C 2 D’s Ca 2+ -binding loops are shown as an inset. Three aspartate to alanine substitutions were introduced in C 2 C (D417A, D465A, D473A) and two or four substitutions were introduced in C 2 D (D1168A/D1174A, D1230A/D1232A and D1168A/D1174A/D1230A/D1232A). B Ca 2+ -binding activity of C 2 D’s DDA (“double D (aspartate) to A (alanine)”) substitutions measured using a nanoDSF assay. The melting temperatures (T m ) of C 2 D-DDA-1 (D1168A/D1174A) and C 2 D-DDA-2 (D1230A/D1232A) were measured as a function of [Ca 2+ ] or [Mg 2+ ]. Single representative nanoDSF T m estimates are shown (n=1). C Ca 2+ -activity measurements of the C 2 D-QDA (D1168A/D1174A/D1230A/D1232A) Fer core mutant. The estimated T m values were plotted as a function of [Ca 2+ ] or [Mg 2+ ]. Error bars represent the s.d. (C 2 D-QDA: n=3, wild-type: n=5). D Probing the interaction between C 2 B-C 2 D and C 2 F-C 2 G dysferlin fragments by size-exclusion chromatography (SEC). The purified samples were mixed in trans , in the presence or absence of 1 mM Ca 2+ , subjected to SEC on the Superdex 200 column and the peak fractions analyzed by SDS-PAGE. In these experiments, no complex formation between C 2 B-C 2 D (Fer core , residues 214-1281) and C 2 F-C 2 G ( min DYSF, residues 1503-1560 and 1613-2056) was observed (C 2 B-C 2 D peaked in fraction 11, whereas the C 2 F-C 2 G fragment in fraction 14). The chromatograms of the reconstitution experiments are shown on the top. Download figure Open in new tab Figure S14. Chemical-crosslinking mass spectrometry analysis of the Ca 2+ -bound soluble dysferlin (1-2017). A Distribution of the Cα-Cα Euclidean distances between crosslinked lysine residues in the locked and loose dysferlin states. Dysferlin’s chemical crosslinking data indicate that, when mapped onto the locked/closed state model, ∼93% of unique crosslinks are observed between lysine residues less than 35 Å apart. Mapping to a loose state model lacking the more dynamic C 2 G, results in ∼85% crosslinks satisfying the distance threshold, showing that the dysferlin conformations could co-exist in the presence of Ca 2+ . B Selected chemical crosslinks between DYSF’s structural motifs. The Euclidean distances were calculated between the Cα atoms of the crosslinked lysines in the locked/closed dysferlin state and depicted as colored lines. C Distribution of the crosslink distances in the locked/closed state of soluble dysferlin (residues 1-2017, sDYSF) and a full-length dysferlin Alpahfold2 prediction (model 1, Appendix Fig S15B ). Note that compared to the locked/closed state model, only ∼64.9% (37/57) unique crosslinks occur between residues less than 35 Å apart. The crosslink outliers map to the C 2 B-C 2 F, C 2 F-C 2 G and C 2 A-C 2 B interfaces, indicating that the relative orientation of these domains is not accurately predicted in the Alphafold2 model. D BS3 crosslinks involving the flexible C 2 A and the C 2 A-C 2 B linker. The crosslinking data indicate that, although generally flexible, C 2 A, likely, samples a restricted number of conformational states in the proximity of the inner DysF and C 2 B. Some of the poses of C 2 A could be accommodated by the unassigned density region which continues from C 2 B towards the center of dysferlin’s (and myoferlin’s) membrane-binding surface. Download figure Open in new tab Fig. S15. Common organization principles of human ferlins. A Alphafold2-based domain composition schematic of human otoferlin (Jumper, Evans et al., 2021). B Alphafold2 models of human dysferlin. While the Fer core module is similar in these predictions, C 2 B, C 2 E, C 2 F, C 2 G and TM adopt alternative poses (indicated with arrows), consistent with them being structurally dynamic. The long linkers connecting the C 2 A-C 2 B and C 2 E-C 2 F are likely unstructured, having pLDDT (predicted local distance difference test) scores <50. C Comparison between soluble dysferlin’s locked (closed) state and an Alphafold2 prediction (model 1) of full- length dysferlin. Note the significant different poses of the C 2 B, C 2 E, C 2 F, and C 2 G domains (marked with arrows, r.m.s.d.=4.85 over 1156 atoms). Minor differences are observed for the FerA domain. The Fer core module (except the inner DysF motif) is similarly organized in the soluble dysferlin (residues 1-2017, sDYSF) structure and Alphafold2 models (r.m.s.d.=1.79 over 618 atoms). D Comparison between soluble dysferlin’s cryo-EM model and the Alphafold2 prediction of human otoferlin. Note that otoferlin lacks the DysF motifs and the C 2 E anchor loop (magenta) is more extended. In addition, C 2 B, C 2 E and C 2 F are arranged in a near circular conformation, compared to the more elliptical appearance of dysferlin’s ring – a configuration which appears reminiscent of the lipid-bound myoferlin state. E Highly pathogenic patient mutations mapped onto dysferlin’s structure in the lipid-free state. Pathogenic and likely pathogenic dysferlin substitutions were extracted from the ClinVar database (Landrum, Chitipiralla et al., 2020) and manually curated. Note that some dysferlin substitutions appear to cluster at tertiary C 2 domain interfaces, such as between C 2 C and C 2 CD ( e.g ., R555W and D625Y). Acknowledgements We thank Christiane Senger-Freitag, Sandra Gerke, Sina Langer, Andy Schacht, and Patricia Räke-Kügler for excellent technical and administrative support. We are grateful to Dr. Tat Cheung Cheng and Prof. Dr. Ruben Fernandez-Busnadiego for supporting cryo-EM data acquisition and Dr. Eri Sakata for critically reading the manuscript. This work was supported and funded by Deutsche Forschungsgemeinschaft (DFG) through CR 937/2-1 (to CC) and the Cluster of Excellence (EXC2067) Multiscale Bioimaging EXC 2067/1-390729940 (CC, TM and JP). CZKS and VP were supported by an Wellcome Trust Investigator grant (220300Z/20/Z). HU was supported by the Max Planck Society. Cryo-EM instrumentation was jointly funded by the DFG Major Research Instrumentation program (448415290) and the Ministry of Science and Culture of the State of Lower Saxony. TM was also supported by the Leibniz Program of the DFG (MO896/5), the Ernst Jung Prize for Medicine, by Fondation Pour l’Audition (FPA RD-2020-10) References ↵ Abdullah N , Padmanarayana M , Marty NJ , Johnson CP ( 2014 ) Quantitation of the calcium and membrane binding properties of the C2 domains of dysferlin . Biophys J 106 : 382 – 9 OpenUrl CrossRef PubMed ↵ Abramson J , Adler J , Dunger J , Evans R , Green T , Pritzel A , Ronneberger O , Willmore L , Ballard AJ , Bambrick J , Bodenstein SW , Evans DA , Hung CC , O’Neill M , Reiman D , Tunyasuvunakool K , Wu Z , Zemgulyte A , Arvaniti E , Beattie C et al. ( 2024 ) Accurate structure prediction of biomolecular interactions with AlphaFold 3 . Nature 630 : 493 – 500 OpenUrl CrossRef PubMed ↵ Afonine PV , Poon BK , Read RJ , Sobolev OV , Terwilliger TC , Urzhumtsev A , Adams PD ( 2018 ) Real-space refinement in PHENIX for cryo-EM and crystallography . Acta Crystallogr D Struct Biol 74 : 531 – 544 OpenUrl CrossRef PubMed ↵ Al-Moyed H , Cepeda AP , Jung S , Moser T , Kugler S , Reisinger E ( 2019 ) A dual-AAV approach restores fast exocytosis and partially rescues auditory function in deaf otoferlin knock-out mice . EMBO Mol Med 11 ↵ Arac D , Chen X , Khant HA , Ubach J , Ludtke SJ , Kikkawa M , Johnson AE , Chiu W , Sudhof TC , Rizo J ( 2006 ) Close membrane-membrane proximity induced by Ca(2+)-dependent multivalent binding of synaptotagmin-1 to phospholipids . Nat Struct Mol Biol 13 : 209 – 17 OpenUrl CrossRef PubMed Web of Science ↵ Bansal D , Campbell KP ( 2004 ) Dysferlin and the plasma membrane repair in muscular dystrophy . Trends Cell Biol 14 : 206 – 13 OpenUrl CrossRef PubMed Web of Science Bansal D , Miyake K , Vogel SS , Groh S , Chen CC , Williamson R , McNeil PL , Campbell KP ( 2003 ) Defective membrane repair in dysferlin-deficient muscular dystrophy . Nature 423 : 168 – 72 OpenUrl CrossRef PubMed Web of Science ↵ Bernatchez PN , Acevedo L , Fernandez-Hernando C , Murata T , Chalouni C , Kim J , Erdjument-Bromage H , Shah V , Gratton JP , McNally EM , Tempst P , Sessa WC ( 2007 ) Myoferlin regulates vascular endothelial growth factor receptor-2 stability and function . J Biol Chem 282 : 30745 – 53 OpenUrl Abstract / FREE Full Text ↵ Brandt DS , Coffman MD , Falke JJ , Knight JD ( 2012 ) Hydrophobic contributions to the membrane docking of synaptotagmin 7 C2A domain: mechanistic contrast between isoforms 1 and 7 . Biochemistry 51 : 7654 – 64 OpenUrl CrossRef PubMed ↵ Cannon SK , Sarsam RD , Tedamrongwanish T , Zhang K , Baker RW ( 2023 ) Lipid nanodiscs as a template for high-resolution cryo-EM structures of peripheral membrane proteins . J Struct Biol 215 : 107989 OpenUrl CrossRef PubMed ↵ Carpenter AP , Khuu P , Weidner T , Johnson CP , Roeters SJ , Baio JE ( 2023 ) Orientation of the Dysferlin C2A Domain is Responsive to the Composition of Lipid Membranes . J Phys Chem B 127 : 577 – 589 OpenUrl ↵ Chapman ER ( 2008 ) How does synaptotagmin trigger neurotransmitter release? Annu Rev Biochem 77 : 615 – 41 OpenUrl CrossRef PubMed Web of Science ↵ Chen H , Fang Q , Benseler F , Brose N , Moser T ( 2023 ) Probing the role of the C(2)F domain of otoferlin . Front Mol Neurosci 16 : 1299509 OpenUrl CrossRef PubMed ↵ Chen H , Monga M , Fang Q , Slitin L , Neef J , Chepurwar SS , Netto RCM , Lezirovitz K , Tabith A , Jr. , Benseler F , Brose N , Kusch K , Wichmann C , Strenzke N , Vona B , Preobraschenski J , Moser T ( 2024 ) Ca2+ binding to the C2E domain of otoferlin is required for hair cell exocytosis and hearing . Protein Cell 15 : 305 – 312 OpenUrl CrossRef PubMed ↵ Chen ZL , Meng JM , Cao Y , Yin JL , Fang RQ , Fan SB , Liu C , Zeng WF , Ding YH , Tan D , Wu L , Zhou WJ , Chi H , Sun RX , Dong MQ , He SM ( 2019 ) A high-speed search engine pLink 2 with systematic evaluation for proteome-scale identification of cross-linked peptides . Nat Commun 10 : 3404 OpenUrl CrossRef PubMed ↵ Clapham DE ( 2007 ) Calcium signaling . Cell 131 : 1047 – 58 OpenUrl CrossRef PubMed Web of Science ↵ Codding SJ , Marty N , Abdullah N , Johnson CP ( 2016 ) Dysferlin Binds SNAREs (Soluble N- Ethylmaleimide-sensitive Factor (NSF) Attachment Protein Receptors) and Stimulates Membrane Fusion in a Calcium-sensitive Manner . J Biol Chem 291 : 14575 – 84 OpenUrl Abstract / FREE Full Text ↵ Combe CW , Fischer L , Rappsilber J ( 2015 ) xiNET: cross-link network maps with residue resolution . Mol Cell Proteomics 14 : 1137 – 47 OpenUrl Abstract / FREE Full Text ↵ Cooper ST , Head SI ( 2015 ) Membrane Injury and Repair in the Muscular Dystrophies . Neuroscientist 21 : 653 – 68 OpenUrl CrossRef PubMed ↵ Cooper ST , McNeil PL ( 2015 ) Membrane Repair: Mechanisms and Pathophysiology . Physiol Rev 95 : 1205 – 40 OpenUrl CrossRef PubMed Corrotte M , Almeida PE , Tam C , Castro-Gomes T , Fernandes MC , Millis BA , Cortez M , Miller H , Song W , Maugel TK , Andrews NW ( 2013 ) Caveolae internalization repairs wounded cells and muscle fibers . Elife 2 : e00926 OpenUrl CrossRef PubMed ↵ Cretu C , Agrawal AA , Cook A , Will CL , Fekkes P , Smith PG , Luhrmann R , Larsen N , Buonamici S , Pena V ( 2018 ) Structural Basis of Splicing Modulation by Antitumor Macrolide Compounds . Mol Cell 70 : 265 – 273 e8 OpenUrl CrossRef PubMed ↵ Cretu C , Gee P , Liu X , Agrawal A , Nguyen TV , Ghosh AK , Cook A , Jurica M , Larsen NA , Pena V ( 2021 ) Structural basis of intron selection by U2 snRNP in the presence of covalent inhibitors . Nat Commun 12 : 4491 OpenUrl CrossRef PubMed ↵ Cretu C , Schmitzova J , Ponce-Salvatierra A , Dybkov O , De Laurentiis EI , Sharma K , Will CL , Urlaub H , Luhrmann R , Pena V ( 2016 ) Molecular Architecture of SF3b and Structural Consequences of Its Cancer-Related Mutations . Mol Cell 64 : 307 – 319 OpenUrl CrossRef PubMed ↵ Croll TI ( 2018 ) ISOLDE: a physically realistic environment for model building into low- resolution electron-density maps . Acta Crystallogr D Struct Biol 74 : 519 – 530 OpenUrl PubMed ↵ Davis DB , Delmonte AJ , Ly CT , McNally EM ( 2000 ) Myoferlin, a candidate gene and potential modifier of muscular dystrophy . Hum Mol Genet 9 : 217 – 26 OpenUrl CrossRef PubMed Web of Science de Morree A , Hensbergen PJ , van Haagen HH , Dragan I , Deelder AM , t Hoen PA , Frants RR , van der Maarel SM ( 2010 ) Proteomic analysis of the dysferlin protein complex unveils its importance for sarcolemmal maintenance and integrity . PLoS One 5 : e13854 Doherty KR , Cave A , Davis DB , Delmonte AJ , Posey A , Earley JU , Hadhazy M , McNally EM ( 2005 ) Normal myoblast fusion requires myoferlin . Development 132 : 5565 – 75 OpenUrl Abstract / FREE Full Text ↵ Dominguez MJ , McCord JJ , Sutton RB ( 2022 ) Redefining the architecture of ferlin proteins: Insights into multi-domain protein structure and function . PLoS One 17 : e0270188 OpenUrl CrossRef PubMed ↵ Duncker SV , Franz C , Kuhn S , Schulte U , Campanelli D , Brandt N , Hirt B , Fakler B , Blin N , Ruth P , Engel J , Marcotti W , Zimmermann U , Knipper M ( 2013 ) Otoferlin couples to clathrin-mediated endocytosis in mature cochlear inner hair cells . J Neurosci 33 : 9508 – 19 OpenUrl Abstract / FREE Full Text ↵ Emsley P , Cowtan K ( 2004 ) Coot: model-building tools for molecular graphics . Acta Crystallogr D Biol Crystallogr 60 : 2126 – 32 OpenUrl CrossRef PubMed Web of Science Fitzgerald DJ , Berger P , Schaffitzel C , Yamada K , Richmond TJ , Berger I ( 2006 ) Protein complex expression by using multigene baculoviral vectors . Nat Methods 3 : 1021 – 32 OpenUrl CrossRef PubMed Web of Science ↵ Glover L , Brown RH , Jr. ( 2007 ) Dysferlin in membrane trafficking and patch repair . Traffic 8 : 785 – 94 OpenUrl CrossRef PubMed Web of Science ↵ Guerrero-Valero M , Ferrer-Orta C , Querol-Audi J , Marin-Vicente C , Fita I , Gomez- Fernandez JC , Verdaguer N , Corbalan-Garcia S ( 2009 ) Structural and mechanistic insights into the association of PKCalpha-C2 domain to PtdIns(4,5)P2 . Proc Natl Acad Sci U S A 106 : 6603 – 7 OpenUrl Abstract / FREE Full Text ↵ Guillen J , Ferrer-Orta C , Buxaderas M , Perez-Sanchez D , Guerrero-Valero M , Luengo-Gil G , Pous J , Guerra P , Gomez-Fernandez JC , Verdaguer N , Corbalan-Garcia S ( 2013 ) Structural insights into the Ca2+ and PI(4,5)P2 binding modes of the C2 domains of rabphilin 3A and synaptotagmin 1 . Proc Natl Acad Sci U S A 110 : 20503 – 8 OpenUrl Abstract / FREE Full Text ↵ Gupta S , Yano J , Mercier V , Htwe HH , Shin HR , Rademaker G , Cakir Z , Ituarte T , Wen KW , Kim GE , Zoncu R , Roux A , Dawson DW , Perera RM ( 2021 ) Lysosomal retargeting of Myoferlin mitigates membrane stress to enable pancreatic cancer growth . Nat Cell Biol 23 : 232 – 242 OpenUrl CrossRef PubMed ↵ Han R , Campbell KP ( 2007 ) Dysferlin and muscle membrane repair . Curr Opin Cell Biol 19 : 409 – 16 OpenUrl CrossRef PubMed Web of Science ↵ Harsini FM , Chebrolu S , Fuson KL , White MA , Rice AM , Sutton RB ( 2018 ) FerA is a Membrane-Associating Four-Helix Bundle Domain in the Ferlin Family of Membrane- Fusion Proteins . Sci Rep 8 : 10949 OpenUrl CrossRef PubMed Helfmann S , Neumann P , Tittmann K , Moser T , Ficner R , Reisinger E ( 2011 ) The crystal structure of the C(2)A domain of otoferlin reveals an unconventional top loop region . J Mol Biol 406 : 479 – 90 OpenUrl CrossRef PubMed ↵ Hernandez JM , Stein A , Behrmann E , Riedel D , Cypionka A , Farsi Z , Walla PJ , Raunser S , Jahn R ( 2012 ) Membrane fusion intermediates via directional and full assembly of the SNARE complex . Science 336 : 1581 – 4 OpenUrl Abstract / FREE Full Text ↵ Hoekstra D , Duzgunes N ( 1993 ) Lipid mixing assays to determine fusion in liposome systems . Methods Enzymol 220 : 15 – 32 OpenUrl CrossRef PubMed ↵ Honigmann A , van den Bogaart G , Iraheta E , Risselada HJ , Milovanovic D , Mueller V , Mullar S , Diederichsen U , Fasshauer D , Grubmuller H , Hell SW , Eggeling C , Kuhnel K , Jahn R ( 2013 ) Phosphatidylinositol 4,5-bisphosphate clusters act as molecular beacons for vesicle recruitment . Nat Struct Mol Biol 20 : 679 – 86 OpenUrl CrossRef PubMed ↵ Huang HL , Grandinetti G , Heissler SM , Chinthalapudi K ( 2024 ) Cryo-EM structures of the membrane repair protein dysferlin . Nat Commun 15 : 9650 OpenUrl CrossRef PubMed ↵ Jahn R , Fasshauer D ( 2012 ) Molecular machines governing exocytosis of synaptic vesicles . Nature 490 : 201 – 7 OpenUrl CrossRef PubMed Web of Science ↵ Jakobi AJ , Wilmanns M , Sachse C ( 2017 ) Model-based local density sharpening of cryo-EM maps . Elife 6 ↵ Johnson CP , Chapman ER ( 2010 ) Otoferlin is a calcium sensor that directly regulates SNARE-mediated membrane fusion . J Cell Biol 191 : 187 – 97 OpenUrl Abstract / FREE Full Text Jumper J , Evans R , Pritzel A , Green T , Figurnov M , Ronneberger O , Tunyasuvunakool K , Bates R , Zidek A , Potapenko A , Bridgland A , Meyer C , Kohl SAA , Ballard AJ , Cowie A , Romera-Paredes B , Nikolov S , Jain R , Adler J , Back T et al. ( 2021 ) Highly accurate protein structure prediction with AlphaFold . Nature 596 : 583 – 589 OpenUrl CrossRef PubMed ↵ Jung S , Maritzen T , Wichmann C , Jing Z , Neef A , Revelo NH , Al-Moyed H , Meese S , Wojcik SM , Panou I , Bulut H , Schu P , Ficner R , Reisinger E , Rizzoli SO , Neef J , Strenzke N , Haucke V , Moser T ( 2015 ) Disruption of adaptor protein 2mu (AP-2mu) in cochlear hair cells impairs vesicle reloading of synaptic release sites and hearing . EMBO J 34 : 2686 – 702 OpenUrl Abstract / FREE Full Text ↵ Kastner B , Fischer N , Golas MM , Sander B , Dube P , Boehringer D , Hartmuth K , Deckert J , Hauer F , Wolf E , Uchtenhagen H , Urlaub H , Herzog F , Peters JM , Poerschke D , Luhrmann R , Stark H ( 2008 ) GraFix: sample preparation for single-particle electron cryomicroscopy . Nat Methods 5 : 53 – 5 OpenUrl CrossRef PubMed Web of Science ↵ Kaur N , de la Ballina LR , Haukaas HS , Torgersen ML , Radulovic M , Munson MJ , Sabirsh A , Stenmark H , Simonsen A , Carlsson SR , Lystad AH ( 2023 ) TECPR1 is activated by damage- induced sphingomyelin exposure to mediate noncanonical autophagy . EMBO J: e 113105 ↵ Kerr JP , Ziman AP , Mueller AL , Muriel JM , Kleinhans-Welte E , Gumerson JD , Vogel SS , Ward CW , Roche JA , Bloch RJ ( 2013 ) Dysferlin stabilizes stress-induced Ca2+ signaling in the transverse tubule membrane . Proc Natl Acad Sci U S A 110 : 20831 – 6 OpenUrl Abstract / FREE Full Text ↵ Kimanius D , Jamali K , Wilkinson ME , Lovestam S , Velazhahan V , Nakane T , Scheres SHW ( 2024 ) Data-driven regularization lowers the size barrier of cryo-EM structure determination . Nat Methods 21 : 1216 – 1221 OpenUrl CrossRef PubMed ↵ Krissinel E , Henrick K ( 2007 ) Inference of macromolecular assemblies from crystalline state . J Mol Biol 372 : 774 – 97 OpenUrl CrossRef PubMed Web of Science ↵ Kroll J , Jaime Tobon LM , Vogl C , Neef J , Kondratiuk I , Konig M , Strenzke N , Wichmann C , Milosevic I , Moser T ( 2019 ) Endophilin-A regulates presynaptic Ca(2+) influx and synaptic vesicle recycling in auditory hair cells . EMBO J 38 ↵ Kuo IY , Ehrlich BE ( 2015 ) Signaling in muscle contraction . Cold Spring Harb Perspect Biol 7 : a006023 OpenUrl Abstract / FREE Full Text ↵ Kwok E , Otto SC , Khuu P , Carpenter AP , Codding SJ , Reardon PN , Vanegas J , Kumar TM , Kuykendall CJ , Mehl RA , Baio J , Johnson CP ( 2023 ) The Dysferlin C2A Domain Binds PI(4,5)P2 and Penetrates Membranes . J Mol Biol 435 : 168193 OpenUrl CrossRef PubMed ↵ Leclère JC , Dulon D ( 2023 ) Otoferlin as a multirole Ca signaling protein: from inner ear synapses to cancer pathways . Frontiers in Cellular Neuroscience 17 ↵ Lek A , Evesson FJ , Sutton RB , North KN , Cooper ST ( 2012 ) Ferlins: regulators of vesicle fusion for auditory neurotransmission, receptor trafficking and membrane repair . Traffic 13 : 185 – 94 OpenUrl CrossRef PubMed Web of Science Llanga T , Nagy N , Conatser L , Dial C , Sutton RB , Hirsch ML ( 2017 ) Structure-Based Designed Nano-Dysferlin Significantly Improves Dysferlinopathy in BLA/J Mice . Mol Ther 25 : 2150 – 2162 OpenUrl CrossRef PubMed ↵ Luan S , Wang C ( 2021 ) Calcium Signaling Mechanisms Across Kingdoms . Annu Rev Cell Dev Biol 37 : 311 – 340 OpenUrl CrossRef PubMed ↵ Lv J , Wang H , Cheng X , Chen Y , Wang D , Zhang L , Cao Q , Tang H , Hu S , Gao K , Xun M , Wang J , Wang Z , Zhu B , Cui C , Gao Z , Guo L , Yu S , Jiang L , Yin Y et al. ( 2024 ) AAV1- hOTOF gene therapy for autosomal recessive deafness 9: a single-arm trial . Lancet 403 : 2317 – 2325 OpenUrl CrossRef PubMed ↵ Marty NJ , Holman CL , Abdullah N , Johnson CP ( 2013 ) The C2 domains of otoferlin, dysferlin, and myoferlin alter the packing of lipid bilayers . Biochemistry 52 : 5585 – 92 OpenUrl CrossRef PubMed ↵ Michalski N , Goutman JD , Auclair SM , Boutet de Monvel J , Tertrais M , Emptoz A , Parrin A , Nouaille S , Guillon M , Sachse M , Ciric D , Bahloul A , Hardelin JP , Sutton RB , Avan P , Krishnakumar SS , Rothman JE , Dulon D , Safieddine S , Petit C ( 2017 ) Otoferlin acts as a Ca(2+) sensor for vesicle fusion and vesicle pool replenishment at auditory hair cell ribbon synapses . Elife 6 ↵ Moser T , Chen H , Kusch K , Behr R , Vona B ( 2024 ) Gene therapy for deafness: are we there now? EMBO Mol Med 16 : 675 – 677 OpenUrl CrossRef PubMed ↵ Moser T , Grabner CP , Schmitz F ( 2020 ) Sensory Processing at Ribbon Synapses in the Retina and the Cochlea . Physiol Rev 100 : 103 – 144 OpenUrl CrossRef PubMed ↵ Moser T , Starr A ( 2016 ) Auditory neuropathy--neural and synaptic mechanisms . Nat Rev Neurol 12 : 135 – 49 OpenUrl CrossRef PubMed ↵ Padmanarayana M , Hams N , Speight LC , Petersson EJ , Mehl RA , Johnson CP ( 2014 ) Characterization of the lipid binding properties of Otoferlin reveals specific interactions between PI(4,5)P2 and the C2C and C2F domains . Biochemistry 53 : 5023 – 33 OpenUrl CrossRef PubMed ↵ Pangrsic T , Lasarow L , Reuter K , Takago H , Schwander M , Riedel D , Frank T , Tarantino LM , Bailey JS , Strenzke N , Brose N , Muller U , Reisinger E , Moser T ( 2010 ) Hearing requires otoferlin-dependent efficient replenishment of synaptic vesicles in hair cells . Nat Neurosci 13 : 869 – 76 OpenUrl CrossRef PubMed Web of Science ↵ Pangrsic T , Reisinger E , Moser T ( 2012 ) Otoferlin: a multi-C2 domain protein essential for hearing . Trends Neurosci 35 : 671 – 80 OpenUrl CrossRef PubMed Web of Science ↵ Patel P , Harris R , Geddes SM , Strehle EM , Watson JD , Bashir R , Bushby K , Driscoll PC , Keep NH ( 2008 ) Solution structure of the inner DysF domain of myoferlin and implications for limb girdle muscular dystrophy type 2b . J Mol Biol 379 : 981 – 90 OpenUrl CrossRef PubMed ↵ Paulke NJ , Fleischhacker C , Wegener JB , Riedemann GC , Cretu C , Mushtaq M , Zaremba N , Mobius W , Zuhlke Y , Wedemeyer J , Liebmann L , Gorshkova AA , Kownatzki-Danger D , Wagner E , Kohl T , Wichmann C , Jahn O , Urlaub H , Toischer K , Hasenfuss G et al. ( 2024 ) Dysferlin Enables Tubular Membrane Proliferation in Cardiac Hypertrophy . Circ Res ↵ Petit C , Bonnet C , Safieddine S ( 2023 ) Deafness: from genetic architecture to gene therapy . Nat Rev Genet 24 : 665 – 686 OpenUrl CrossRef PubMed ↵ Piper AK , Ross SE , Redpath GM , Lemckert FA , Woolger N , Bournazos A , Greer PA , Sutton RB , Cooper ST ( 2017 ) Enzymatic cleavage of myoferlin releases a dual C2-domain module linked to ERK signalling . Cell Signal 33 : 30 – 40 OpenUrl CrossRef PubMed ↵ Pramono ZA , Tan CL , Seah IA , See JS , Kam SY , Lai PS , Yee WC ( 2009 ) Identification and characterisation of human dysferlin transcript variants: implications for dysferlin mutational screening and isoforms . Hum Genet 125 : 413 – 20 OpenUrl CrossRef PubMed ↵ Redpath GM , Woolger N , Piper AK , Lemckert FA , Lek A , Greer PA , North KN , Cooper ST ( 2014 ) Calpain cleavage within dysferlin exon 40a releases a synaptotagmin-like module for membrane repair . Mol Biol Cell 25 : 3037 – 48 OpenUrl Abstract / FREE Full Text ↵ Revelo NH , Kamin D , Truckenbrodt S , Wong AB , Reuter-Jessen K , Reisinger E , Moser T , Rizzoli SO ( 2014 ) A new probe for super-resolution imaging of membranes elucidates trafficking pathways . J Cell Biol 205 : 591 – 606 OpenUrl Abstract / FREE Full Text ↵ Rizo J ( 2022 ) Molecular Mechanisms Underlying Neurotransmitter Release . Annu Rev Biophys 51 : 377 – 408 OpenUrl CrossRef PubMed ↵ Rizo J , Sari L , Qi YE , Im W , Lin MM ( 2022 ) All-atom molecular dynamics simulations of Synaptotagmin-SNARE-complexin complexes bridging a vesicle and a flat lipid bilayer . Elife 11 ↵ Rizo J , Sudhof TC ( 1998 ) C2-domains, structure and function of a universal Ca2+-binding domain . J Biol Chem 273 : 15879 – 82 OpenUrl FREE Full Text ↵ Roux I , Safieddine S , Nouvian R , Grati M , Simmler MC , Bahloul A , Perfettini I , Le Gall M , Rostaing P , Hamard G , Triller A , Avan P , Moser T , Petit C ( 2006 ) Otoferlin, defective in a human deafness form, is essential for exocytosis at the auditory ribbon synapse . Cell 127 : 277 – 89 OpenUrl CrossRef PubMed Web of Science ↵ Sanchez-Garcia R , Gomez-Blanco J , Cuervo A , Carazo JM , Sorzano COS , Vargas J ( 2021 ) DeepEMhancer: a deep learning solution for cryo-EM volume post-processing . Commun Biol 4 : 874 OpenUrl CrossRef PubMed ↵ Santarelli R , del Castillo I , Cama E , Scimemi P , Starr A ( 2015 ) Audibility, speech perception and processing of temporal cues in ribbon synaptic disorders due to OTOF mutations . Hear Res 330 : 200 – 12 OpenUrl CrossRef PubMed ↵ Schauder CM , Wu X , Saheki Y , Narayanaswamy P , Torta F , Wenk MR , De Camilli P , Reinisch KM ( 2014 ) Structure of a lipid-bound extended synaptotagmin indicates a role in lipid transfer . Nature 510 : 552 – 5 OpenUrl CrossRef PubMed ↵ Schmitzova J , Cretu C , Dienemann C , Urlaub H , Pena V ( 2023 ) Structural basis of catalytic activation in human splicing . Nature 617 : 842 – 850 OpenUrl CrossRef PubMed ↵ Seven AB , Brewer KD , Shi L , Jiang QX , Rizo J ( 2013 ) Prevalent mechanism of membrane bridging by synaptotagmin-1 . Proc Natl Acad Sci U S A 110 : E3243 – 52 OpenUrl Abstract / FREE Full Text ↵ Shaib AH , Chouaib AA , Chowdhury R , Altendorf J , Mihaylov D , Zhang C , Krah D , Imani V , Spencer RKW , Georgiev SV , Mougios N , Monga M , Reshetniak S , Mimoso T , Chen H , Fatehbasharzad P , Crzan D , Saal KA , Alawieh MM , Alawar N et al. ( 2024 ) One-step nanoscale expansion microscopy reveals individual protein shapes . Nat Biotechnol ↵ Shin OH , Han W , Wang Y , Sudhof TC ( 2005 ) Evolutionarily conserved multiple C2 domain proteins with two transmembrane regions (MCTPs) and unusual Ca2+ binding properties . J Biol Chem 280 : 1641 – 51 OpenUrl Abstract / FREE Full Text ↵ Strenzke N , Chakrabarti R , Al-Moyed H , Muller A , Hoch G , Pangrsic T , Yamanbaeva G , Lenz C , Pan KT , Auge E , Geiss-Friedlander R , Urlaub H , Brose N , Wichmann C , Reisinger E ( 2016 ) Hair cell synaptic dysfunction, auditory fatigue and thermal sensitivity in otoferlin Ile515Thr mutants . EMBO J 35 : 2519 – 2535 OpenUrl Abstract / FREE Full Text ↵ Sudhof TC ( 2013 ) A molecular machine for neurotransmitter release: synaptotagmin and beyond . Nat Med 19 : 1227 – 31 OpenUrl CrossRef PubMed ↵ Sula A , Cole AR , Yeats C , Orengo C , Keep NH ( 2014 ) Crystal structures of the human Dysferlin inner DysF domain . BMC Struct Biol 14 : 3 OpenUrl CrossRef PubMed ↵ Verdaguer N , Corbalan-Garcia S , Ochoa WF , Fita I , Gomez-Fernandez JC ( 1999 ) Ca(2+) bridges the C2 membrane-binding domain of protein kinase Calpha directly to phosphatidylserine . EMBO J 18 : 6329 – 38 OpenUrl Abstract / FREE Full Text ↵ Vogl C , Cooper BH , Neef J , Wojcik SM , Reim K , Reisinger E , Brose N , Rhee JS , Moser T , Wichmann C ( 2015 ) Unconventional molecular regulation of synaptic vesicle replenishment in cochlear inner hair cells . J Cell Sci 128 : 638 – 44 OpenUrl Abstract / FREE Full Text ↵ Vogl C , Panou I , Yamanbaeva G , Wichmann C , Mangosing SJ , Vilardi F , Indzhykulian AA , Pangrsic T , Santarelli R , Rodriguez-Ballesteros M , Weber T , Jung S , Cardenas E , Wu X , Wojcik SM , Kwan KY , Del Castillo I , Schwappach B , Strenzke N , Corey DP et al. ( 2016 ) Tryptophan-rich basic protein (WRB) mediates insertion of the tail-anchored protein otoferlin and is required for hair cell exocytosis and hearing . EMBO J 35 : 2536 – 2552 OpenUrl Abstract / FREE Full Text ↵ Vona B , Rad A , Reisinger E ( 2020 ) The Many Faces of DFNB9: Relating OTOF Variants to Hearing Impairment . Genes (Basel ) 11 ↵ Wagner T , Merino F , Stabrin M , Moriya T , Antoni C , Apelbaum A , Hagel P , Sitsel O , Raisch T , Prumbaum D , Quentin D , Roderer D , Tacke S , Siebolds B , Schubert E , Shaikh TR , Lill P , Gatsogiannis C , Raunser S ( 2019 ) SPHIRE-crYOLO is a fast and accurate fully automated particle picker for cryo-EM . Commun Biol 2 : 218 OpenUrl CrossRef PubMed ↵ Woolger N , Bournazos A , Sophocleous RA , Evesson FJ , Lek A , Driemer B , Sutton RB , Cooper ST ( 2017 ) Limited proteolysis as a tool to probe the tertiary conformation of dysferlin and structural consequences of patient missense variant L344P . J Biol Chem 292 : 18577 – 18591 OpenUrl Abstract / FREE Full Text ↵ Xu L , Pallikkuth S , Hou Z , Mignery GA , Robia SL , Han R ( 2011 ) Dysferlin forms a dimer mediated by the C2 domains and the transmembrane domain in vitro and in living cells . PLoS One 6 : e27884 OpenUrl CrossRef PubMed ↵ Yamashita K , Palmer CM , Burnley T , Murshudov GN ( 2021 ) Cryo-EM single-particle structure refinement and map calculation using Servalcat . Acta Crystallogr D Struct Biol 77 : 1282 – 1291 OpenUrl CrossRef PubMed ↵ Yasunaga S , Grati M , Cohen-Salmon M , El-Amraoui A , Mustapha M , Salem N , El-Zir E , Loiselet J , Petit C ( 1999 ) A mutation in OTOF, encoding otoferlin, a FER-1-like protein, causes DFNB9, a nonsyndromic form of deafness . Nat Genet 21 : 363 – 9 OpenUrl CrossRef PubMed Web of Science ↵ Yavuz H , Kattan I , Hernandez JM , Hofnagel O , Witkowska A , Raunser S , Walla PJ , Jahn R ( 2018 ) Arrest of trans-SNARE zippering uncovers loosely and tightly docked intermediates in membrane fusion . J Biol Chem 293 : 8645 – 8655 OpenUrl Abstract / FREE Full Text ↵ Zhang T , Li J , He Y , Yang F , Hao Y , Jin W , Wu J , Sun Z , Li Y , Chen Y , Yi Z , Liu M ( 2018 ) A small molecule targeting myoferlin exerts promising anti-tumor effects on breast cancer . Nat Commun 9 : 3726 OpenUrl CrossRef PubMed ↵ Zhou Q , Lai Y , Bacaj T , Zhao M , Lyubimov AY , Uervirojnangkoorn M , Zeldin OB , Brewster AS , Sauter NK , Cohen AE , Soltis SM , Alonso-Mori R , Chollet M , Lemke HT , Pfuetzner RA , Choi UB , Weis WI , Diao J , Sudhof TC , Brunger AT ( 2015 ) Architecture of the synaptotagmin-SNARE machinery for neuronal exocytosis . Nature 525 : 62 – 7 OpenUrl CrossRef PubMed ↵ Zhou Q , Zhou P , Wang AL , Wu D , Zhao M , Sudhof TC , Brunger AT ( 2017 ) The primed SNARE-complexin-synaptotagmin complex for neuronal exocytosis . Nature 548 : 420 – 425 OpenUrl CrossRef PubMed References ↵ Jumper J , Evans R , Pritzel A , Green T , Figurnov M , Ronneberger O , Tunyasuvunakool K , Bates R , Zidek A , Potapenko A , Bridgland A , Meyer C , Kohl SAA , Ballard AJ , Cowie A , Romera-Paredes B , Nikolov S , Jain R , Adler J , Back T et al. ( 2021 ) Highly accurate protein structure prediction with AlphaFold . Nature 596 : 583 – 589 OpenUrl CrossRef PubMed Landrum MJ , Chitipiralla S , Brown GR , Chen C , Gu B , Hart J , Hoffman D , Jang W , Kaur K , Liu C , Lyoshin V , Maddipatla Z , Maiti R , Mitchell J , O’Leary N , Riley GR , Shi W , Zhou G , Schneider V , Maglott D et al. ( 2020 ) ClinVar: improvements to accessing data . Nucleic Acids Res 48 : D835 – D844 OpenUrl CrossRef PubMed Patel P , Harris R , Geddes SM , Strehle EM , Watson JD , Bashir R , Bushby K , Driscoll PC , Keep NH ( 2008 ) Solution structure of the inner DysF domain of myoferlin and implications for limb girdle muscular dystrophy type 2b . J Mol Biol 379 : 981 – 90 OpenUrl CrossRef PubMed Sula A , Cole AR , Yeats C , Orengo C , Keep NH ( 2014 ) Crystal structures of the human Dysferlin inner DysF domain . BMC Struct Biol 14 : 3 OpenUrl CrossRef PubMed van Kempen M , Kim SS , Tumescheit C , Mirdita M , Lee J , Gilchrist CLM , Soding J , Steinegger M ( 2024 ) Fast and accurate protein structure search with Foldseek . Nat Biotechnol 42 : 243 – 246 OpenUrl CrossRef PubMed View the discussion thread. Back to top Previous Next Posted January 27, 2025. Download PDF Email Thank you for your interest in spreading the word about bioRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Structural Basis of Lipid Membrane Binding by Human Ferlins Message Subject (Your Name) has forwarded a page to you from bioRxiv Message Body (Your Name) thought you would like to see this page from the bioRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Structural Basis of Lipid Membrane Binding by Human Ferlins Constantin Cretu , Aleksandar Chernev , Csaba Zoltan Kibedi Szabo , Vladimir Pena , Henning Urlaub , Tobias Moser , Julia Preobraschenski bioRxiv 2025.01.25.634844; doi: https://doi.org/10.1101/2025.01.25.634844 Share This Article: Copy Citation Tools Structural Basis of Lipid Membrane Binding by Human Ferlins Constantin Cretu , Aleksandar Chernev , Csaba Zoltan Kibedi Szabo , Vladimir Pena , Henning Urlaub , Tobias Moser , Julia Preobraschenski bioRxiv 2025.01.25.634844; doi: https://doi.org/10.1101/2025.01.25.634844 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Biochemistry Subject Areas All Articles Animal Behavior and Cognition (7637) Biochemistry (17705) Bioengineering (13899) Bioinformatics (41970) Biophysics (21463) Cancer Biology (18605) Cell Biology (25526) Clinical Trials (138) Developmental Biology (13385) Ecology (19911) Epidemiology (2067) Evolutionary Biology (24329) Genetics (15615) Genomics (22514) Immunology (17743) Microbiology (40424) Molecular Biology (17194) Neuroscience (88650) Paleontology (667) Pathology (2835) Pharmacology and Toxicology (4827) Physiology (7648) Plant Biology (15160) Scientific Communication and Education (2046) Synthetic Biology (4302) Systems Biology (9825) Zoology (2271)