Ubiquitin-dependent signal amplification in lipid saturation sensing

Ubiquitin-dependent signal amplification in lipid saturation sensing | 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 Ubiquitin-dependent signal amplification in lipid saturation sensing View ORCID Profile Jona Causemann , Barbara Schmidt , Daniel Granz , View ORCID Profile Thorsten Mosler , View ORCID Profile Martin Jung , View ORCID Profile Ivan Dikic , Heiko Rieger , View ORCID Profile Robert Ernst doi: https://doi.org/10.1101/2025.11.05.686737 Jona Causemann 1 Medical Biochemistry and Molecular Biology, Medical Faculty, Saarland University 2 Preclinical Center for Molecular Signaling, Medical Faculty, Saarland University 4 Center of Biophysics (ZBP), Saarland University Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Jona Causemann Barbara Schmidt 3 Department of Theoretical Physics and Center for Biophysics, Faculty of Natural Sciences, Saarland University 4 Center of Biophysics (ZBP), Saarland University Find this author on Google Scholar Find this author on PubMed Search for this author on this site Daniel Granz 1 Medical Biochemistry and Molecular Biology, Medical Faculty, Saarland University 2 Preclinical Center for Molecular Signaling, Medical Faculty, Saarland University 4 Center of Biophysics (ZBP), Saarland University Find this author on Google Scholar Find this author on PubMed Search for this author on this site Thorsten Mosler 5 Institute of Biochemistry II, University Hospital Frankfurt, Goethe University Frankfurt am Main Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Thorsten Mosler Martin Jung 1 Medical Biochemistry and Molecular Biology, Medical Faculty, Saarland University Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Martin Jung Ivan Dikic 5 Institute of Biochemistry II, University Hospital Frankfurt, Goethe University Frankfurt am Main 6 Max-Planck-Institute of Biophysics, Frankfurt am Main Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Ivan Dikic Heiko Rieger 3 Department of Theoretical Physics and Center for Biophysics, Faculty of Natural Sciences, Saarland University 4 Center of Biophysics (ZBP), Saarland University Find this author on Google Scholar Find this author on PubMed Search for this author on this site Robert Ernst 1 Medical Biochemistry and Molecular Biology, Medical Faculty, Saarland University 2 Preclinical Center for Molecular Signaling, Medical Faculty, Saarland University 4 Center of Biophysics (ZBP), Saarland University Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Robert Ernst For correspondence: robert.ernst{at}uni-saarland.de Abstract Full Text Info/History Metrics Preview PDF Abstract Cellular membranes are dynamic platforms whose composition and biophysical properties are surveyed by sensor proteins to maintain homeostasis. How these sensors convert weak physical cues into robust biochemical outputs remains unclear. Failure to preserve membrane homeostasis, however, results in cellular stress and organelle dysfunction. Here, we investigate the prototypical yeast lipid saturation sensor Mga2 to reveal how lipid-controlled rotational movements in the transmembrane region are decoded into robust transcriptional responses. Using a fully reconstituted in vitro ubiquitylation system with quantitative fluorescence-based readouts, we uncover how Mga2 amplifies subtle, fluctuating membrane signals into pronounced differences in ubiquitylation. Kinetic modeling demonstrates that this amplification does not require deubiquitylating enzymes. Instead, negative feedback arises from a diversion of ubiquitin flux toward inhibitory autoubiquitylation of the cognate E3 ligase Rsp5. Lipid saturation shifts this ubiquitin flux, thereby enabling several-fold signal amplification. Our findings provide a mechanistic framework for how membrane property sensors convert weak, fluctuating physical inputs into robust biochemical outcomes, and establish ubiquitin flux partitioning as a general principle of signal amplification in cellular surveillance mechanisms. Introduction Biological membranes are dynamic structures whose physical properties — such as thickness, compressibility, surface charge, and lipid packing — are critical for cellular function( Bigay & Antonny, 2012 ; Cybulski et al , 2010 ; Holthuis & Menon, 2014 ; Renne & Ernst, 2023 ). Eukaryotic cells regulate these properties across subcellular compartments, ensuring that each membrane maintains its characteristic compositions and properties suited to its role( Harayama & Riezman, 2018 ; Renne & Ernst, 2023 ). This balance, known as biophysical membrane homeostasis, is maintained by sensor proteins that detect deviations in membrane properties and trigger adaptive responses with a wide impact on lipid metabolism and transport, protein quality control, and membrane trafficking( Ernst et al , 2018 ; Covino et al , 2018 ). Membrane property sensors are diverse in form and function, yet they share two defining features: they are highly sensitive to specific physical parameters of the lipid bilayer, and they are structurally metastable, allowing them to respond to subtle fluctuations in the membrane environment( Antonny, 2011 ; Covino et al , 2018 ). Examples include sensors of curvature elastic stress( Boumann et al , 2006 ; Haider et al , 2018 ), lipid saturation( Covino et al , 2016 ; Ballweg et al , 2020 ), and reduced membrane compressibility( Halbleib et al , 2017 ; Alsayyah et al , 2025 ). Intriguingly, lipid saturation sensors have been implicated in controlling the cellular response to low temperature and hypoxia( Zhang et al , 1999 ; Jiang et al , 2001 ; Nakagawa et al , 2002 ; Jiang et al , 2002 ), while a membrane compressibility sensor coordinates the synthesis of membrane lipids and proteins via the unfolded protein response (UPR) of the endoplasmic reticulum (ER)( Ernst et al , 2024 ; Renne & Ernst, 2023 ). Even though the biological functions of these pathways are well established, the mechanisms by which small changes in membrane properties are translated into decisive signaling remain poorly understood. One of the best-characterized pathways for biophysical membrane homeostasis is the OLE pathway in Saccharomyces cerevisiae ( Hoppe et al , 2000 ; Rape et al , 2001 ; Shcherbik et al , 2003 ; Ballweg & Ernst, 2017 ). It controls lipid packing in the ER membrane by regulating the production of unsaturated fatty acids (UFAs) and operates through two ER-membrane–anchored transcription factors, Mga2 and Spt23, which sense membrane lipid saturation to regulate expression of the essential fatty acid desaturase gene OLE1 ( Stukey et al , 1989 ; Zhang et al , 1999 ; Hoppe et al , 2000 ; Covino et al , 2016 ). Among the two, Mga2 plays the dominant role: deletion of MGA2 leads to reduced OLE1 expression, accumulation of saturated lipids, and lipid bilayer stress with pronounced morphological changes of the ER membrane( Chellappa et al , 2001 ; Jiang et al , 2001 ; Surma et al , 2013 ). To become active, Mga2 and Spt23 must be released from their membrane anchor before they can translocate to the nucleus( Hoppe et al , 2000 ). This release is supported by ubiquitylation near the membrane interface and carried out by the E3 ubiquitin ligase Rsp5 and inhibited by unsaturated lipids ( Hoppe et al , 2000 ; Shcherbik et al , 2003 ). Cleavage by the proteasome at an internal site( Piwko & Jentsch, 2006 ) and -according to our working model-degradation of the membrane-anchor liberates the active transcription factor, which then migrates to the nucleus and upregulates OLE1 expression( Hoppe et al , 2000 ; Rape et al , 2001 ). This creates a negative feedback loop in which UFA production restores membrane lipid packing. At the center of this sensory system is a biophysical mechanism provided by the transmembrane helix of Mga2( Covino et al , 2016 ). The transcription factor forms a dimer, with each monomer containing a bulky tryptophan residue that probes lipid packing in the hydrophobic core of the ER membrane at the level of Δ9 double bonds in lipid fatty acyl chains. The dimeric transmembrane helices dynamically sample different rotational states and the overall structural organization renders Mga2 sensitive to lipid saturation, but less sensitive to the lipid headgroup composition( Covino et al , 2016 ; Ballweg et al , 2020 ). When lipid packing is high (i.e., membranes are more saturated), the tryptophan residues are more likely to face each other in the dimer interface, thereby stabilizing a spectrum of conformations that promotes ubiquitylation. When lipid packing is low, the helices populate alternative conformations, and ubiquitylation efficacy drops( Covino et al , 2016 ; Ballweg et al , 2020 ). Notably, even modest shifts in the population of these conformations — as seen in coarse-grained molecular dynamics simulations and cwEPR experiments — are sufficient to drive changes in Mga2 activity both in vitro and in vivo ( Covino et al , 2016 ; Ballweg et al , 2020 ). This raises a central question: How do these rather subtle, membrane-driven conformational changes lead to decisive ubiquitylation events occurring ≈50 amino acids away, in the cytosol? The ubiquitylation machinery that activates Mga2 follows the canonical E1–E2–E3 cascade: ubiquitin is activated by an E1 enzyme, transferred to a conjugating enzyme (E2), and finally attached to substrate lysines by an E3 ubiquitin ligase( Shcherbik et al , 2003 ; Pohl & Dikic, 2019 ). In the case of Mga2, the E3 is Rsp5, a HECT-domain containing ligase from the Nedd4 family with a multitude of cellular functions( Shcherbik et al , 2004 ; Bhattacharya et al , 2009 , 2008 ). Rsp5 contains an N-terminal C2 domain, three WW domains for substrate recognition, and a catalytic HECT domain. While the third WW domain and the HECT domain are sufficient to complement RSP5 ’s essential functions( Hoppe et al , 2000 ), the roles of the N-terminal C2 and the other two WW domains can vary depending on the pathway and substrate( Wang et al , 1999 ; Chang et al , 2000 ; Dunn & Hicke, 2001 ; Bhattacharya et al , 2008 ). Like other HECT domain ligases, Rsp5 is thought to transfer ubiquitin sequentially, building polyubiquitin chains on its client one molecule at a time( Wang & Pickart, 2005 ; Pierce et al , 2009 ; French et al , 2017 ). Apart from its role in lipid regulation, Rsp5 is involved in endocytosis, the heat shock response, and other stress-related pathways — but how it allocates its activity among these competing demands remains unclear( Kaliszewski & Zoładek, 2008 ; Sardana & Emr, 2021 ). Interestingly, Rsp5 can also autoubiquitylate, a process that inhibits its own ligase activity( Attali et al , 2017 ). Ubiquitylation is a remarkably versatile regulatory tool. The anaphase-promoting complex (APC), for instance, establishes a defined temporal order of degradation for a variety of key cell-cycle regulators in G1 and mitosis by sequential ubiquitylation of its substrates( Peters, 2002 ; Rape et al , 2006 ). Pioneering work has demonstrated that the processivity of multiubiquitylation in conjunction with deubiquitylating enzymes (DUBs) determines this ordering of substrates( Peters, 2002 ). Similarly, the ER-associated degradation (ERAD) depends on the selective ubiquitylation of membrane and lumenal client proteins, which targets them for proteasomal clearance( Needham et al , 2019 ; Christianson & Carvalho, 2022 ). Reconstituting the decision-making process revealed that a deubiquitylating activity is critical for amplifying modest differences in E3 ligase-substrate interactions to facilitate robust decisions( Zhang et al , 2013 ). These systems showcased how E3 ligase processivity combined with deubiquitylating activities can translate weak or transient signals into all-or-none decisions by amplifying small molecular changes( Rape et al , 2006 ; Zhang et al , 2013 ). Inspired by this principle, we wondered whether biophysical changes in the lipid environment might also be amplified through coupling to a ubiquitylation reaction, even in the absence of the opposing activity of DUBs. This would set the system apart from many classical signaling pathways that rely on a reversible push-pull architecture (e.g. ubiquitylation/DUB or kinase/phosphatase cycles)( Ferrell & Xiong, 2001 ; Markevich et al , 2004 ; Stegmeier et al , 2007 ; Zhang et al , 2013 ). In this case, negative feedback could instead be supplied by Rsp5’s autoubiquitylation, which reduces its activity and limits signal propagation( Attali et al , 2017 ). Hence, we set out to test whether ubiquitylation alone — without any DUB activity — is sufficient to amplify small differences in membrane lipid packing. Using a reconstituted in vitro assay with purified components and fluorescent in-gel readouts, we quantified Mga2 ubiquitylation and modeled its kinetics. We find that the first five transfers of ubiquitin provide signal amplification, yielding up to 4.5-fold differences in the abundance of sufficiently ubiquitylated Mga2-Ub species for different membrane environments, thereby facilitating transcription factor mobilization. Strikingly, the lipid environment of Mga2 also provides negative feedback by diverting the flux of ubiquitin toward Rsp5 autoubiquitylation. Together, our findings support a signal amplification mechanism in which a combination of positive feedback and a diversion of ubiquitin flux serve as regulatory logic. This mechanism may represent a more general strategy by which cells translate subtle biophysical inputs into decisive molecular responses. Results Development of a fully defined in vitro assay to study Mga2 ubiquitylation Using the lipid saturation sensor Mga2 as a paradigm, we wanted to learn how fluctuating signals from the transmembrane region( Covino et al , 2016 ; Ballweg et al , 2020 ) can be amplified to robustly regulate transcription factor activation. Suspecting an important contribution of sensor ubiquitylation, we first optimized the reconstitution of the minimal sense-and-response construct ZIP-MBP Mga2 in liposomes( Ballweg et al , 2020 ). The construct contains an N-terminal leucine zipper (ZIP) for efficient dimerization, the maltose binding protein (MBP) from Escherichia coli as an affinity and solubility tag, and both the juxtamembrane (aa 951-1037) and C-terminal transmembrane region (aa 1038-1055) of Mga2 ( Figure 1A ). The juxtamembrane region includes the 967 LPKY 970 motif for recruiting the E3 ubiquitin ligase Rsp5, and three lysine residues K980, K983, K985 that are ubiquitylated by Rsp5 in vivo with comparable responsiveness to lipid saturation as the full length protein( Shcherbik et al , 2004 ; Bhattacharya et al , 2009 ). We installed a cysteine at position S1003C for fluorescence labeling using ATTO488-maleimide to facilitate in-gel quantification of ubiquitylated ZIP-MBP Mga2 with a broad dynamic range. After purifying labeled ZIP-MBP Mga2 in Octyl-β-D-glucopyranoside (OG)( Ballweg et al , 2020 ), we reconstituted the sensor protein in liposomes at a molar protein-to-lipid ratio of 1:8000. Quantitative detergent removal was accomplished by dialysis against a detergent-free buffer containing SM-2 BioBeads (BBs)( Ballweg et al , 2020 ). We used a 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) lipid matrix to mimic conditions that facilitate Mga2 ubiquitylation, and a loosely packed 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) matrix to mimic conditions that do not support transcription factor modification and activation( Covino et al , 2016 ; Ballweg et al , 2020 ). Labeled ZIP-MBP Mga2 was successfully reconstituted in both lipid environments with protein recoveries of >70% ( Figure 1B ) yielding proteoliposomes with an average diameter of ≈80 nm as studied by DLS ( Figure 1C ). In contrast to previous reconstitution protocols( Ballweg et al , 2020 ), our optimized procedures resulted in an almost unidirectional topology of the reconstituted protein as judged from proteinase K protection assays with >80% of the sequence elements relevant for ubiquitylation being accessible from the outside of the proteoliposomes ( Figure 1E ). The successful reconstitution was further validated by sucrose density gradient centrifugation ( Figure S1A ), and extraction assays using sodium carbonate, urea, and high salt ( Figure S1B ). Hence, the sensor protein was successfully and stably integrated into the lipid bilayer. Download figure Open in new tab Figure 1: Efficient reconstitution of ZIP-MBP Mga2 in distinct lipid environments. (A) Schematic of the Mga2 construct comprising an N-terminal leucine zipper (ZIP) for dimerization, the maltose binding protein (MBP) purification tag, and a sequence stretch derived from Mga2 (aa 951-1062). The juxtamembrane region of Mga2 (aa 951-1037) contains a binding site for the E3 ubiquitin ligase Rsp5 ( 967 LPKY 970 ) and three lysine residues used for ubiquitylation (K980, K983, K985). Serine 1003 found in wild-type Mga2 is replaced by cysteine (S1003C) for maleimide-based fluorescence labeling with ATTO 488. The single transmembrane helix (TMH, Mga2 aa 1038-1055) anchors the protein in the membrane. ( B) ZIP-MBP Mga2 recovery monitored by ATTO 488 fluorescence after dialysis. Prior to fluorescence spectroscopy, each sample was adjusted to 75 mM OG to solubilize the proteoliposomes and reduce scattering. The background signal from the buffer was subtracted, and the samples were normalized to the intensity of the input sample prior to dialysis. Shown are the mean and SD of n = 5 independent reconstitution experiments. (C) Size of proteoliposomes measured by dynamic light scattering, showing mean diameters for POPC and DOPC environments (n = 3 reconstitutions). (D) Proteinase K protection assay to assess ZIP-MBP Mga2 orientation in proteoliposomes. Samples were separated by SDS-PAGE and analyzed by in-gel fluorescence scanning. Graphs show the mean and SD of n = 5 independent experiments. (E) SDS-PAGE analysis of recombinant proteins (0.5 µg per lane) used for the in vitro ubiquitylation of ZIP-MBP Mga2, stained with InstantBlue ® . (F) Scheme of Mga2 in vitro ubiquitylation, detailing the roles of the E1, E2, and E3 enzymes, as well as highlighting Rsp5 autoubiquitylation. To reconstitute the ubiquitylation of ZIP-MBP Mga2, we mixed proteoliposomes with ubiquitin, 1 mM ATP, an ATP-regenerating system, and a set of enzymes for ubiquitin activation (E1), conjugation (E2), and ligation (E3). Due to their known functional equivalence, we could use 6xHistidine-tagged Uba1 from mouse (E1, 70 nM) and human GST-tagged UbcH5B (E2, 500 nM) to activate and deliver ubiquitin to the yeast ubiquitin ligase Rsp5 (E3, 200 nM) ( Figure 1E, F ). Notably, the E2 identity has no impact on the ubiquitin chains formed by Rsp5( Kim & Huibregtse, 2009 ) and this set of proteins was previously used to characterize the mechanism of ubiquitin ligation upon E2-to-E3-to-substrate transfer( Kamadurai et al , 2013 ). We purposely omitted the inclusion of DUBs, such as Ubp2 or Ubp15, which are thought to antagonize Rsp5 and remodel ubiquitin chains after their formation( Kee et al , 2005 ; Ho et al , 2017 ). Different Rsp5 variants ( Figure S2A , B) ubiquitylated ZIP-MBP Mga2 with distinct efficacies ( Figure S2C ), while also undergoing autoubiquitylation to a varying degree ( Figure S2D ). Compared to full-length Rsp5, a truncated variant (HECT, aa 383-809) lacking the N-terminal C2 domain and the first two of three WW domains ( Figure S2A ) provided only a minimal degree of ZIP-MBP Mga2 ubiquitylation ( Figure S2C ) in a tightly packed membrane environment (100 mol% POPC), yet it underwent significant autoubiquitylation ( Figure S2D ). Because C2 and WW domains can stabilize non-productive conformations of HECT E3 ligases in the absence of a client protein( Wiesner et al , 2007 ; Wang et al , 2010 ; Mari et al , 2014 ; Riling et al , 2015 ; Zhu et al , 2017 ; Wang et al , 2019 ), we speculate that the engagement of ZIP-MBP Mga2 with full-length Rsp5 can direct the ‘flux’ of ubiquitin towards the lipid saturation sensor. While ineffective towards their clients, HECT E3 ligases lacking C2 and WW domains are catalytically hyperactive, thereby causing increased E3 ligase autoubiquitylation( Wiesner et al , 2007 ; Wang et al , 2010 ; Mari et al , 2014 ; Riling et al , 2015 ; Zhu et al , 2017 ; Wang et al , 2019 ). This is consistent with our observations on Rsp5 autoubiquitylation (Fig S2D) and may serve auto-regulatory purposes, as the ubiquitylation of Rsp5’s lysine residues K411, K432, or K438 inhibits its E3 ligase activity( Attali et al , 2017 ). Consistently, the substitution of these three lysines with arginine (3KR) decreases Rsp5 autoubiquitylation ( Figure S2D ) and increases the ubiquitylation of the lipid saturation sensor ZIP-MBP Mga2 ( Figure S2C ). For all following experiments, we decided to use wild-type, full-length Rsp5 after removal of the GST-tag, as its presence limited ZIP-MBP Mga2 ubiquitylation ( Figure S2C ). Physical and kinetic modeling of Mga2 ubiquitylation A detailed kinetic analysis of ZIP-MBP Mga2 ubiquitylation requires quantitative data on the abundance of individual ZIP-MBP Mga2-Ub species. To this end, we performed in vitro ubiquitylation assays using fluorescently labeled ZIP-MBP Mga2 reconstituted in liposomes composed of tightly packing lipids (100 mol% POPC). We quantified both unmodified and ubiquitylated species by in-gel fluorescence over time ( Figure 2A ), which neither distinguishes between different linkage types nor between multiple mono- and polyubiquitylations. Because a minimum of 3-6 ubiquitin molecules is required for Cdc48 Npl4-Ufd1 -dependent handling of ubiquitylated substrates and the handover to the proteasome( Thrower et al , 2000 ; Bodnar & Rapoport, 2017 ; Williams et al , 2023 ; Tsuchiya et al , 2017 ; Kiss et al , 2025 ), we focused our kinetic analysis on ZIP-MBP Mga2-Ub species with up to six ubiquitin. Download figure Open in new tab Figure 2: ZIP-MBP Mga2 and fitting the ubiquitylation reaction kinetics. (A) SDS-PAGE analysis of an in vitro ubiquitylation assay with ATTO 488-labeled ZIP-MBP Mga2 reconstituted into a 100 mol% POPC membrane. Assay components: 2 µM ZIP-MBP Mga2, 200 nM Rsp5, 500 nM E2 (GST-UbcH5B), 70 nM E1 (6xHis-Uba1), 15 µM ubiquitin (8xHis-Ub), ATP. The reactions were incubated at 30°C. In total, 0.55 µg of ZIP-MBP Mga2 was loaded for SDS-PAGE (gel: 7.5% Mini-PROTEAN ® TGX™). Fluorescence emission was detected with a Typhoon laser scanner (488 nm laser, Cy2 filter, 25 µm resolution, 360 V PMT voltage). Different ZIP-MBP Mga2 species are color-coded. (B)-(H) Densitometric quantification of ZIP-MBP Mga2-Ub species during time course ubiquitylation assays with POPC-based membranes, normalized to unmodified Mga2 at t = 0 min. Presented are the mean and SD of n = 5 replicates, fitted to the model (solid lines). (B) Unmodified ZIP-MBP Mga2, (C) ZIP-MBP Mga2-Ub 1 , (D) ZIP-MBP Mga2-Ub 2 , (E) ZIP-MBP Mga2-Ub 3 , (F) ZIP-MBP Mga2-Ub 4 , (G) ZIP-MBP Mga2-Ub 5 , (H) ZIP-MBP Mga2-Ub 6 . (I) Schematic of the physical model for fitting ubiquitylation data, assuming sequential ubiquitylation with forward rate kx and inhibition rate kx’, where ‘x’ stands for the number of ubiquitin molecules with which Mga2 is modified. (J) Contribution of ubiquitylation (competent fraction) and inhibition (incompetent fraction) to the fit (total) for unmodified Mga2 and ubiquitylated Mga2-Ub x species. (K) Exploration of the parameter space for k1 and k1’ and determination of log 10 (ξ 2 ) as quality parameter for the goodness of the overall fit. ξ 2 was calculated as the sum of squares of the difference between the mean of the data and the fit, weighted for the standard deviation . (L) Enlarged view of the dashed section shown in (K) . White dashed line: line scan across the area containing the optimal solution. Expectedly, unmodified ZIP-MBP Mga2 was readily consumed over time ( Figure 2A, B ). However, less than 60% of the protein was modified after 60 min of reaction ( Figure 2A , B), even though >80% should be accessible to the E3 ligase ( Figure 1D ). This raised the question, why some ZIP-MBP Mga2 molecules were excluded from ubiquitylation. The abundance of ubiquitylated intermediates ( ZIP-MBP Mga2-Ub x ) followed a characteristic time course ( Figure 2C-H ) with an initial build-up towards a maximal abundance, which was followed by a phase of consumption towards a plateau. Different ZIP-MBP Mga2-Ub species reached their peak abundance consecutively: first, ZIP-MBP Mga2-Ub 1 , then ZIP-MBP Mga2-Ub 2 , then ZIP-MBP Mga2-Ub 3 , and so forth, which is characteristic for HECT domain E3 ligases and consistent with multiple transfers of single ubiquitin molecules( Pierce et al , 2009 ; Deol et al , 2019 ). Yet, none of the ubiquitylated intermediates ZIP-MBP Mga2-Ub x were fully consumed for the formation of ZIP-MBP Mga2-Ub x+1 ( Figure 2C-H ), thereby raising the question, why these conversions were incomplete. Because neither ATP nor free ubiquitin are limiting under our experimental conditions, these observations point toward one or several inhibitory mechanisms that counteract the complete ubiquitylation of the sensor protein ( Figure 2B ) and the full conversion of each ubiquitylated intermediate ( Figure 2C-H ). Among other possible mechanisms, this may include steric hindrances from growing ubiquitin chains, a competition of growing ubiquitin chains for Rsp5 binding, and an inhibitory autoubiquitylation of Rsp5( Attali et al , 2017 ). In any case, a quantitative description of the reaction kinetics should consider not only the sequential addition of ubiquitin molecules to the lipid saturation sensor, but also inhibitory mechanisms that counteract a complete conversion of reaction intermediates. To accommodate this requirement, we repurposed the sequential ubiquitylation model for single turnover reactions pioneered by Pierce et al .( Pierce et al , 2009 ) and applied it to our steady-state conditions in which the E3 ubiquitin ligase can encounter its client multiple times ( Figure 2I ). Hence, each reaction intermediate can either receive a new ubiquitin modification with the forward rate kx or undergo an irreversible inhibition with the rate kx ’ , which serves as a collective term for a variety of possible inhibitory mechanisms. It follows that each ubiquitylated ZIP-MBP Mga2 intermediate exists in two pools: one that is ubiquitylation-competent, and one that is inhibited and therefore excluded from further rounds of ubiquitylation (incompetent) ( Figure 2J ). The model robustly recapitulates the abundance of ubiquitylation intermediates over time with a characteristic initial accumulation of ZIP-MBP Mga2-Ub x towards a peak abundance, and a subsequent consumption towards a plateau ( Figure 2B-H ). Global fitting of our data using the average abundance of each intermediate at each time point ( Figure 2B-H ) was performed to estimate the individual rates for forward ubiquitylation and irreversible inhibition. We found that the first addition of ubiquitin is particularly slow (k0 = 0.041 min −1 ) and that the rate of ubiquitin transfer increases dramatically at later steps yielding 6.4-fold (k1 = 0.262 min −1 ) and even 120-fold higher rates (k2 = 4.85 min −1 ) for ubiquitin transfer. These data suggest that the addition of the first ubiquitin is rate-limiting and key to support subsequent rounds of ubiquitylation, e.g. by increasing the local density of ubiquitin attachment sites for the rather unselective E3 ubiquitin ligase Rsp5( Saeki et al , 2009 ; Kim & Huibregtse, 2009 ; Sardana et al , 2019 ). In contrast to the impressive acceleration of the forward reaction, the inhibitory rates k0’ to k6’ were identical within a factor of 2 for all reaction steps (≈0.03 min −1 to 0.05 min −1 ) and in the same order of magnitude as the rate-limiting step of ubiquitylation (0.041 min −1 ). Hence, the inhibitory mechanism seems to be insensitive to the number of ubiquitins attached to the sensor. The relevance of the inhibitory rates for the shape of curves in our time course experiments ( Figure 2B-H ) becomes more obvious when the ubiquitylation-competent pool of ZIP-MBP Mga2 and the inhibited pool are plotted individually ( Figure 2J , S3A-G): without any inhibitory mechanisms at work, each reaction intermediate of ZIP-MBP Mga2-Ub x would be fully consumed rather than approaching a steady-state concentration. Hence, the inclusion of an inhibitory mechanism is essential to reliably fit our experimental data. More importantly, the combination of an accelerating forward reaction with an inhibitory feedback mechanism fulfills an important prerequisite for signal amplification based on the ubiquitylation kinetics( Ferrell & Xiong, 2001 ). We wanted to test the validity of our 14-parameter fit and learn if the extracted rates indeed represent an optimal solution. To this end, we established a quality parameter log 10 (ξ 2 ), which is calculated as the sum of squared deviations of the fit to the average abundance at each time point, weighted by the standard deviation . While the rates k0 and k0’ can be analytically established, the rates of the following steps are affected by all preceding rates that determine how fast a certain reaction intermediate is formed and ‘replenished’. We decided to test the quality of the fit progressively by recapitulating the ubiquitylation reaction step by step. Initially, we fixed the analytically determined rates k0 and k0’ to systematically explore a wide parameter space for the forward and inhibition rates k1 and k1’, respectively, while plotting the resulting log 10 (ξ 2 ) after global fitting ( Figure 2E ). This analysis revealed a clearly defined minimum of the quality parameter log 10 (ξ 2 ) for both k1 and k1’ indicating optimal solutions ( Figure 2K ). A possible caveat of this analysis, however, is that an overestimation of the forward rate k1 may be compensated by an equivalent overestimation of the inhibitory rate k1’, such that good fits are achieved whenever a certain ratio of k1 to k1’ is preserved. Indeed, systematically plotting the quality parameter log 10 (ξ 2 ) for k1 and k1’ pairs identified a strong correlation between k1 and k1’ for yielding good fits ( Figure 2K ). Nonetheless, even along the diagonal of k1 and k1’ pairs yielding good fits, we found a defined minimum of log 10 (ξ 2 ) ( Figure 2L ), hence indicating that the estimated rates are indeed derived from an optimal solution. We then fixed all rates up until this step and systematically explored the parameter space for k2 and k2’, then for k3 and k3’ and so forth to consecutively identify their impact on the overall quality of the fit ( Figure S3H-L ). This analysis demonstrated that the rates derived from the best solution are particularly reliable for the first five rounds of ubiquitylation. In summary, our detailed kinetic analysis reveals a remarkable acceleration of forward ubiquitylation and uncovers an important contribution of an irreversible inhibition for describing the reaction. Ubiquitylation of ZIP-MBP Mga2 contributes to signal amplification We wanted to assess if lipid saturation affects the kinetics of ZIP-MBP Mga2 ubiquitylation. Previously, it was shown that the supplementation with UFAs to the culture medium abolishes Mga2 processing in vivo ( Jiang et al , 2002 ; Covino et al , 2016 ) and that the in vitro ubiquitylation of ZIP-MBP Mga2 is controlled by lipid packing( Covino et al , 2016 ; Ballweg et al , 2020 ) ( Figure 3A ). We confirmed a significant impact of lipid packing on ZIP-MBP Mga2 ubiquitylation in our in vitro system ( Figure 3B ). A direct comparison of ZIP-MBP Mga2 ubiquitylation in tightly packed (100 mol% POPC) versus loosely packed (100 mol% DOPC) membrane environments revealed that ZIP-MBP Mga2 species with four or more ubiquitin moieties are roughly 4.5-fold more abundant in the more tightly packed membrane environment ( Figure 3B ). Given this robust difference in Mga2 ubiquitylation, we wondered about the key differences in the ubiquitylation kinetics imposed by the lipid environments, which feature almost identical membrane viscosities, as judged from lipid diffusion assays( Ballweg et al , 2020 ; Ragaller et al , 2024 ). Hence, we reconstituted ZIP-MBP Mga2 into 100% DOPC PLs, performed in vitro time-course ubiquitylation experiments, and quantified ZIP-MBP Mga2-Ub species over time ( Figure 3C-J ). The consumption of unmodified ZIP-MBP Mga2 was roughly 2-fold slower in the loosely packed DOPC membrane environment (k0 DOPC = 0.020 min −1 ) compared to the POPC condition ( Figure 3D ). More dramatic were the differences observed for ubiquitylated reaction intermediates, both qualitatively and quantitatively ( Figure 3E-J ). In stark contrast to the ubiquitylation in POPC-based membranes, with a characteristic time course representing the buildup, peak abundance, consumption, and plateau phase of ubiquitylated intermediates, only a slow accumulation of the intermediates toward a plateau was found for the more loosely packed, DOPC-based membrane environment ( Figure 3E-J ). This suggests major differences in the ubiquitylation kinetics of ZIP-MBP Mga2 imposed by the different lipid environments. Remarkably, the maximal difference in the abundance of ubiquitylated intermediates between the two membrane environments increased with each ubiquitylation step up to the ZIP-MBP Mga2-Ub 5 intermediate ( Figure 3D-H ). This observation suggests a gradual signal amplification in the first rounds of ubiquitylation. Download figure Open in new tab Figure 3: Reconstituting and kinetic modeling of lipid packing-dependent Mga2 ubiquitylation. (A) Schematic of lipid packing sensing by Mga2 using a sensory tryptophan residue in the hydrophobic core of the membrane. Left : In densely packed, saturated membranes (100 mol% POPC), the sensory tryptophan is more likely to ‘hide’ from the lipid environment in the dimer interface, and Mga2 can be ubiquitylated. Right : In loosely packed, unsaturated membranes (100 mol% DOPC), the sensory tryptophan residues spend more time facing the loosely packed lipid environment, thereby lowering the efficacy of Mga2 ubiquitylation. (B) Comparison of ZIP-MBP Mga2 ubiquitylation in tightly packed POPC versus loosely packed DOPC membranes after 60 min of ubiquitylation. Assay components: 2 µM ZIP-MBP Mga2, 200 nM Rsp5, 500 nM E2 (GST-UbcH5B), 70 nM E1 (6xHis-Uba1), 15 µM ubiquitin (8xHis-Ub), ATP. Of each sample, 0.55 µg of ZIP-MBP Mga2 in 6.0 µL were analyzed by SDS-PAGE and detected using a Typhoon laser scanner (488 nm laser, Cy2 filter, 25 µm resolution, 360 V PMT voltage). (C) In vitro ubiquitylation of ZIP-MBP Mga2 in 100 mol% DOPC membranes. Assay components as described in (B). Analysis of 0.55 µg of ZIP-MBP Mga2 by SDS-PAGE (gel: 7.5% Criterion™ TGX™) followed by in-gel fluorescence detection using a Typhoon laser scanner (488 nm laser, Cy2 filter, 25 µm resolution, 360 V PMT voltage). Different ZIP-MBP Mga2 species are color-coded. (D)-(J) Densitometric measurement of ZIP-MBP Mga2 species fluorescence, normalized to unmodified ZIP-MBP Mga2 at t = 0 min. Mean and SD of n = 5 experiments. ZIP-MBP Mga2 species are color-coded as in (B). Comparison with ubiquitylation in 100 mol% POPC membranes (in gray, from Figure 2 ). Data fitted to the theoretical model (solid lines) with optimal rates indicated. (D) Unmodified ZIP-MBP Mga2, (E) ZIP-MBP Mga2-Ub 1 , (F) ZIP-MBP Mga2-Ub 2 , (G) ZIP-MBP Mga2-Ub 3 , (H) ZIP-MBP Mga2-Ub 4 , (I) ZIP-MBP Mga2-Ub 5 , (J) ZIP-MBP Mga2-Ub 6 . (K) Calculation of the kx/kx’ ratio for POPC- and DOPC-based assays. Ratios that are higher with POPC-based membranes are indicated in green. A detailed analysis of the reaction kinetics further underscored this conclusion. Global fitting of the time course data in the loosely packed membrane environment yielded rates for the forward ubiquitylation and the irreversible inhibition at each step of the reaction ( Figure 3D-J ). We determined the rate k0 DOPC = 0.020 min −1 for the addition of the first ubiquitin and the equivalent rate of irreversible inhibition k0’ DOPC = 0.024 min −1 . We also found robust evidence for increasing rates of forward ubiquitylation ( Figure 3D, E ). However, properties inherent to the kinetics of the system made it impossible to determine the forward rate kx and the irreversible inhibition kx’ for subsequent reaction steps with certainty. This is because an overestimation of the forward reaction rate kx can be compensated by an equivalent overestimation of the rate of inhibition kx’ ( Figure S4A-F ). To address the challenges arising from this uncertainty, we introduced a dimensionless ubiquitylation efficacy factor kx/kx’, which should cancel out misestimations of the two interdependent rates ( Figure 3K , Figure S4G-M ). Comparing the ubiquitylation efficacy factors at each step of the reaction determined for the tightly packed POPC and the loosely packed DOPC membrane environments revealed consistently higher values for the first five ubiquitin additions in POPC. Hence, each of these steps contributes to signal amplification and the robust distinction between saturated and unsaturated membranes by Mga2 ( Figure 3K ). Of particular interest in this context is the first, rate-limiting ubiquitin addition. The ubiquitylation efficacy factor for this step is ≈1.25 for the saturated membrane environment, thus favoring the forward reaction, but only ≈0.75 for the unsaturated membrane, thereby favoring irreversible inhibition. Hence, forward ubiquitylation dominates in the saturated membrane environments, while irreversible inhibition dominates in more loosely packed membranes. Ultimately, these differences result in a 3- to 4.5-fold higher abundance of signaling-active ZIP-MBP Mga2 species in tightly packed membrane environments. Ubiquitin at the crossroads – Rerouting ubiquitin flux from the lipid saturation sensor to Rsp5 Given the central importance of the irreversible inhibition for signal amplification, we wanted to further characterize the underlying molecular basis, and focused our attention on the autoubiquitylation of Rsp5, which can inhibit its E3 ligase activity( Attali et al , 2017 ). Hence, we assayed the autoubiquitylation of Rsp5 in the presence of ZIP-MBP Mga2-containing proteoliposomes featuring either a tightly packed (POPC) or loosely packed (DOPC) lipid environment by immunoblotting ( Figure 4A ). In stark contrast to the ubiquitylation of the lipid saturation sensor ZIP-MBP Mga2, we found that the Rsp5 autoubiquitylation proceeded faster and to a greater extent in loosely packed DOPC-based membranes ( Figure 4B ). As the autoubiquitylation of Rsp5 inhibits its E3 ligase activity( Attali et al , 2017 ), the increased ubiquitylation in loosely packed DOPC-based membranes suggests an elegant mechanism of autoregulation, which may contribute to signal amplification in lipid saturation sensing. Naturally, this does not exclude additional inhibitory mechanisms or alternative scenarios. In fact, Rsp5 contains a C2 domain, which has been implicated in lipid binding and regulating receptor-mediated and fluid-phase endocytosis by directing client protein ubiquitylation( Dunn & Hicke, 2001 ; Dunn et al , 2004 ). Considering the differences in Rsp5 autoubiquitylation observed for proteoliposomes with different lipid formulations ( Figure 4B ), there was a formal possibility that Rsp5, rather than ZIP-MBP Mga2, might serve as the lipid saturation sensor. To address this possible caveat, we performed Rsp5 autoubiquitylation assays in the presence of extruded, protein-free liposomes composed of either POPC or DOPC ( Figure 4C ). In this case, no difference in Rsp5 autoubiquitylation was detected between the two membrane environments, therefore validating the crucial role of ZIP-MBP Mga2 as the lipid saturation sensor and its important role in modulating the autoubiquitylation of Rsp5. Download figure Open in new tab Figure 4: Membrane packing redirects the ubiquitin flux (A) Comparing the ubiquitylation of ZIP-MBP Mga2 in two different lipid environments. ZIP-MBP Mga2 was reconstituted into tightly packing 100% POPC or loosely packed 100% DOPC membranes. Components of the in vitro ubiquitylation assay: 2 µM ZIP-MBP Mga2 (in 100% POPC or DOPC), 200 nM Rsp5 (aa 1-809), 500 nM E2 (GST-UbcH5B), 70 nM E1 (6xHis-Uba1), 15 µM ubiquitin (8xHis-Ub), ATP regenerating system. Per time point, 0.55 µg of ZIP-MBP Mga2 were loaded for SDS-PAGE (gels: 7.5% Mini-PROTEAN ® TGX™). Fluorescent ZIP-MBP Mga2 species were detected using a Typhoon laser scanner (488 nm laser, Cy2 filter, 25 µm resolution, PMT voltage of 360 V). (B) For detection of Rsp5, the samples of the in vitro ubiquitylation assay were diluted 1:10 and 8.3 ng of Rsp5 were used for SDS-PAGE and anti-Rsp5 immunoblotting. Primary antibody: anti-Rsp5 (rabbit, polyclonal serum 1:1000). Secondary antibody: IRDye ® 800 CW goat anti-rabbit (1:15000). The blots were scanned using a LI-COR Odyssey ® scanner at 700 and 800 nm (84 µm resolution). (C) Rsp5 autoubiquitylation assays were performed in the presence of empty large unilamellar vesicles (LUVs) consisting of 100% POPC or 100% DOPC. Assay components: 200 nM Rsp5, 500 nM E2 (GST-UbcH5B), 70 nM E1 (6xHis-Uba1), 15 µM ubiquitin (8xHis-Ub), ATP regenerating system, 16 mM lipids. Lipids were prepared from multilamellar vesicles by extrusion (100 nm filter). 83 ng of Rsp5 were used for SDS-PAGE (gels: 7.5% Mini-PROTEAN ® TGX™). Rsp5 was detected by immunoblotting as described above. (D)-(F) Mass spectrometry analysis of membrane-dependent ZIP-MBP Mga2 ubiquitylation. ZIP-MBP Mga2 was reconstituted into POPC and DOPC membranes. In vitro ubiquitylation assays were performed as described. The signal was normalized to the total intensity. Shown is the log 2 of the normalized intensities of n = 3 replicates. (D) Analysis of the ubiquitin linkage types. ( E) Detection of ubiquitylation sites in ZIP-MBP Mga2. (F) Detection of ubiquitylation sites in Rsp5. (G)-(I) Mass spectrometry analysis of assays containing ZIP-MBP Mga2 mutants in 100 mol% POPC membranes. In vitro ubiquitylation assays were performed as described, and the reactions were stopped after 20 min. ΔLPKY: ZIP-MBP Mga2 lacking the Rsp5 binding motif. 3KR: ZIP-MBP Mga2 lacking all three native lysines in the juxtamembrane region. Shown is the log 2 of the normalized intensities of n = 3 replicates. (G) Detection of ubiquitin linkage types. (H) Analysis of ZIP-MBP Mga2 mutant ubiquitylation. (I) Detection of ubiquitylation sites in Rsp5. (J) In vitro ubiquitylation of ATTO 590-labeled ZIP-MBP Mga2 ΔLPKY after reconstitution into 100 mol% POPC membranes. Assay components as detailed in (A) . 0.55 µg of ZIP-MBP Mga2 ΔLPKY per time point were analyzed by SDS-PAGE. In-gel fluorescence was detected using a Typhoon laser scanner (635 nm laser, Cy5 filter, 25 µm resolution, PMT voltage of 500 V). (K) Samples from assays shown in (J) were diluted 1:10 to analyze Rsp5 autoubiquitylation by immunoblotting. Per lane, 8.3 ng of Rsp5 were loaded. Primary antibody: anti-Rsp5 (rabbit, polyclonal serum 1:1000). Secondary antibody: IRDye ® 800 CW goat anti-rabbit (1:15000). The blots were scanned using a LI-COR Odyssey ® scanner at 700 and 800 nm (84 µm resolution). Together, our findings suggest that the ‘flux’ of ubiquitin towards either ZIP-MBP Mga2 or Rsp5 is regulated by the lipid environment by modulating the structural dynamics of the lipid saturation sensor. According to this model, the autoubiquitylation of Rsp5 would be controlled by how effectively Mga2 ‘accepts’ ubiquitin modifications. The interplay of the E3 ligase and the lipid saturation sensor is sufficient to generate the negative feedback required for signal amplification, as Rsp5 can undergo an autoinhibitory self-ubiquitylation. Detailed Characterization of Mga2 and Rsp5 Ubiquitylation To further dissect the ubiquitylation of ZIP-MBP Mga2 and Rsp5, and to evaluate the influence of membrane lipid composition, we performed in vitro ubiquitylation assays for 20 min with ZIP-MBP Mga2 in either POPC- or DOPC-based membrane environments. Subsequent mass spectrometry analysis ( Figure 4D ) revealed the formation of ubiquitin chains containing K11-, K48-, and K63-linkages, and to a lesser extent, K27-linkages. These observations are indicative of an Rsp5-dependent ubiquitin chain formation with mixed linkages, consistent with the previously reported linkage preferences of Rsp5( Kim & Huibregtse, 2009 ; Fang et al , 2016 ; French et al , 2009 ). As expected for an in vitro system utilizing a client protein with 50 lysine residues, we detected ubiquitylation across the entire ZIP-MBP Mga2 fusion protein, including the N-terminal leucine zipper, the MBP tag, and Mga2-derived sequence elements ( Figure S5A ). Among these, K980 and K983 were particularly interesting, as they were previously identified as primary sites of Mga2 ubiquitylation in vivo ( Bhattacharya et al , 2009 ). Quantitative analysis revealed increased ubiquitylation in the POPC condition relative to DOPC ( Figure 3B-J ), with K980 and K983 modifications elevated approximately 1.5-fold in more saturated membranes ( Figure 4E , S5A). Similar increases were also observed for non-native lysines such as K339 in the MBP tag ( Figure 4E , S5A). Overall, these results underscore that membrane saturation promotes ZIP-MBP Mga2 ubiquitylation. The autoubiquitylation of Rsp5, on the other hand, followed an opposite trend. When the assay was performed with ZIP-MBP Mga2 in loosely packed, DOPC-based membranes, Rsp5 modification was detected on five lysine residues (K45, K78, K79, K258, and K411), but only at two (K79 and K258) and with lower intensities under POPC conditions ( Figure 4F ). Most strikingly, the modification of K411 — known to attenuate E3 ligase activity — was exclusive to the DOPC condition ( Figure 4F ). These findings suggest that the recruitment of Rsp5 to ZIP-MBP Mga2 in loosely packed membranes promotes Rsp5 auto-inhibition. To assess the substrate specificity of the ubiquitylation reaction, we compared the wild-type ZIP-MBP Mga2 reconstituted in POPC-based liposomes with a variant lacking the LPKY motif, which is crucial for Rsp5 recruitment in vivo ( Figure 4H , S5B)( Shcherbik et al , 2004 ). Consistent with our previous observations( Ballweg et al , 2020 ), deletion of this motif led to a near-complete loss of sensor ubiquitylation ( Figure 4J ), including all detectable lysine modifications ( Figure 4H , Figure S5B ). In contrast, Rsp5 autoubiquitylation, including modification of the inhibitory K411 site, was markedly enhanced ( Figure 4I, K ). These data underscore the high specificity of the reaction and establish a strict dependence on the LPKY motif for Rsp5 engagement and substrate ubiquitylation. In the absence of a suitable substrate, the ubiquitin ‘flux’ is redirected towards an inhibitory Rsp5 self-modification. In vivo , K980, K983, and K985 are the principal sites of Mga2 ubiquitylation( Bhattacharya et al , 2009 ). To evaluate their contribution in vitro , we replaced these residues with arginine (3KR). In vitro ubiquitylation and analysis by mass spectrometry revealed a broad redistribution of ubiquitins modifications across alternative lysine residues in the ZIP-MBP Mga2 construct ( Figure 4H , S5B) and reduced levels of Ub-Ub linkages ( Figure 4G ). The autoubiquitylation of Rsp5 remained largely unchanged in this condition ( Figure 4I ). These observations suggest a preferential formation of mixed polyubiquitin chains on those lysine residues, which are targeted by Rsp5 in vivo . Overall, the detailed characterization of ubiquitylation highlights a dynamic ‘rewiring’ of the ubiquitin attachment depending on the conformational dynamics of ZIP-MBP Mga2 controlled by the lipid environment, an appropriate recruitment of Rsp5, and the availability of preferred ubiquitylation sites. Download figure Open in new tab Figure 5: Model for the regulation of Mga2 ubiquitylation by a rewired ubiquitin flux. Tight lipid packing in a saturated membrane promotes Rsp5-mediated ubiquitylation of Mga2. Loose lipid packing due to unsaturated lipids lowers Mga2 ubiquitylation and supports inhibitory Rsp5 autoubiquitylation. Discussion Signal amplification in biology is often achieved through multilayer processes that integrate weak and noisy input signals for robust signaling outcomes. Here, we dissect a signal amplification mechanism for decisive ubiquitylation of the lipid saturation sensor Mga2. We propose that the remarkable sensitivity of Mga2 is based on membrane-triggered conformational changes in the transmembrane region( Covino et al , 2016 ) that are transmitted to the site of ubiquitylation( Ballweg et al , 2020 ), and amplified by the kinetics of sensor ubiquitylation. The described mechanism provides an explanation for amplification even in the absence of deubiquitylating activities and relies solely on the functional interplay of Mga2 and its E3 ligase Rsp5. At the heart of the mechanism lies a kinetic barrier for the first ubiquitin transfer and a divergent ‘flux’ of ubiquitin. When lipid packing is tight, the ubiquitin ‘flux’ is directed toward the lipid saturation sensor Mga2. In loosely packed membranes, the flux of ubiquitin is directed toward Rsp5, thereby limiting its E3 ligase activity and titrating it away from the lipid saturation sensor. A kinetic barrier, positive feedback, and irreversible inhibition for signal amplification Sensitive quantification of reaction intermediates over a broad dynamic range provided the basis for a detailed characterization of the reaction kinetics. Even though we used steady-state conditions that allow for multiple encounters between the E3 ligase and its client, we uncovered strikingly similar kinetic features as observed in single-encounter experiments, yet on a different time scale( Pierce et al , 2009 ; Kim & Huibregtse, 2009 ). Firstly, the initial ubiquitin transfer to Mga2 is markedly slower than subsequent additions, thereby suggesting a kinetic barrier that must be overcome to initiate productive multiubiquitylation. Secondly, subsequent ubiquitylations occur at many-fold higher rates. This dramatic acceleration hints at positive feedback, presumably because ubiquitin modifications increase the local density of lysine residues as attachment sites for subsequent rounds of ubiquitylation and stabilize the interaction with Rsp5( French et al , 2009 ; Kim et al , 2011 ; Zhu et al , 2022 ). Thirdly, an irreversible inhibition at every step of the reaction is crucial to reliably fit the data. The underlying molecular basis is likely to be more complex under steady-state conditions than in single-encounter experiments, where the inhibition can be attributed to the dissociation of the processive E3 ligase from its client( Pierce et al , 2009 ). As dissociation is likely to occur multiple times during our experiments, it is likely that the E3 ligase is gradually titrated away from unmodified and poorly ubiquitylated clients under steady-state conditions and attracted towards more abundantly ubiquitylated species. This would explain the incomplete consumption of reaction intermediates ( Figure 2A-H ). Furthermore, autoubiquitylation of Rsp5, known to limit E3 ligase activity ( Figure 4F )( Attali et al , 2017 ), would limit its activity towards the lipid saturation sensor in this scenario. How does the ubiquitylation kinetics contribute to decision making by a lipid saturation sensor? Casual inspection of our model suggests that even modest changes in the ubiquitylation efficiency, which we define as the forward ubiquitylation rate over the rate of inhibition in each round of ubiquitylation, can have a decisive impact on the overall ubiquitylation of the sensor. This is particularly relevant in the first round of ubiquitylation, which is rate-limiting and acts as the ‘gatekeeper’ for multiubiquitylation ( Figure 3D,K ). In tightly packed membranes, the ubiquitylation efficiency is 1.24, thereby favoring sensor ubiquitylation over inhibition, while it is only 0.83 in loosely packed membrane environments, hence favoring inhibition. Subsequent rounds of ubiquitylation further amplify the differences imposed by the lipid environment up until the fifth ubiquitin transfer ( Figure 4D-K ). Hence, in contrast to the carefully characterized decision-making processes based on ubiquitylation/deubiquitylation cycles, already the first ubiquitin modification contributes to the decision( Rape et al , 2006 ; Zhang et al , 2013 ). The mechanism of signal amplification described here is reminiscent of the process of kinetic proofreading introduced to explain the accuracy of tRNA loading, protein translation, and DNA replication( Hopfield, 1974 ). According to this concept, a series of modifications must occur until the system commits to proceed via an irreversible step. Each modification is required to amplify small differences in binding affinities or rate constants. In the case of protein translation, a 100-fold difference in the binding affinity of charged tRNAs to their respective ribosome binding site is enhanced at the expense of GTP hydrolysis to provide higher precision and a 10,000-fold higher probability for incorporating the correct amino acid into the growing polypeptide chain( Hopfield, 1974 ; Gromadski & Rodnina, 2004 ). In analogy, the ATP-dependent formation of polyubiquitylated Mga2 species is a multistep process that bears striking conceptual similarities to kinetic proofreading: initially modest differences in the rates of ubiquitylation are amplified with each step, resulting in several-fold differences in sufficiently ubiquitylated species of Mga2 that are irreversibly mobilized from the ER membrane for transcription factor activation. Several observations support the notion that the flux of ubiquitin is diverted when Mga2 encounters different membrane environments: Rsp5 undergoes pronounced autoubiquitylation when it interacts with Mga2 in a loosely packed membrane environment, but much less so when Mga2 is in a tightly packed environment ( Figure 4B,F ). The removal of the ‘natural’ ubiquitin attachment sites from Mga2 redirects the ubiquitin flux, yielding a vastly distorted pattern of ubiquitylation on the Mga2-based sensor construct ( Figure 4H ). Likewise, removing Mga2’s LPKY motif, which recruits Rsp5, abrogates ubiquitylation of Mga2 altogether, and again causes pronounced Rsp5 autoubiquitylation ( Figure 4F,I,K ). As the ubiquitylation of Rsp5 at residue K411 is inhibitory( Attali et al , 2017 ), these observations point toward a built-in feedback mechanism that limits the E3 ligase activity in the absence of permissive clients. It is tempting to speculate that this auto-regulatory feedback serves to establish an inactive, yet readily activatable pool of Rsp5. This makes sense considering the many functions of Rsp5, which is involved in endocytosis, vacuolar degradation, and during cytosolic heat stress( Kaliszewski & Zoładek, 2008 ; Fang et al , 2014 ; Ballweg & Ernst, 2017 ; Sardana & Emr, 2021 ). While Rsp5 can undergo inhibitory autoinhibition in the absence of abundant substrates, it can be readily activated from the inactive pool when the demand increases, e.g. during heat stress( Fang et al , 2014 ). If this auto-regulatory cycle indeed supports the orchestration of the many Rsp5 functions, and if this mechanism has wider relevance also for other multi-functional ubiquitin ligases remains to be tested( Kaliszewski & Zoładek, 2008 ). In summary, signal amplification by Rsp5 is facilitated by three features even in the absence of deubiquitylating activities: 1) distinct ubiquitylation efficiencies – particularly at the first, rate-limiting step, 2) acceleration of ubiquitylation, which establishes a positive feed-forward loop( Pierce et al , 2009 ; French et al , 2017 ; Markevich et al , 2004 ; French et al , 2009 ), and 3) negative feedback provided by a diverted flux of ubiquitin, which can limit the E3 ligase activity. Naturally, this does not exclude an additional role for DUBs in this process, which bear the potential to further enhance the fidelity of the decision-making process ( Zhang et al , 2013 ). A comparison to the Anaphase-promoting complex (APC) and ERAD pathway A fine balance between ubiquitylation and deubiquitylation has been identified to mediate key decisions in cell cycle progression via the anaphase-promoting complex (APC) and in substrate discrimination via the ERAD pathway( Rape et al , 2006 ; Zhang et al , 2013 ). Central to these mechanisms is the processivity of the ubiquitin ligase and a substantial, rather unspecific deubiquitylation activity that removes ubiquitin handles from insufficiently modified substrates. This means that substrates with a high affinity to the E3 ligase are more likely to become sufficiently modified in a single encounter with a processive E3 ligase for downstream signaling, while poorer, less processive substrates that feature a more distributive ubiquitylation are less likely to become activated. Instead, they are converted back to their unmodified ‘ground state’ by DUBs. In both cases, it was proposed that the main difference between a ‘good’ substrate and a ‘poor’ one is the processivity of the ubiquitylation reaction, and that the discrimination between substrates is sharpened by deubiquitylating enzymes( Rape et al , 2006 ; Zhang et al , 2013 ). While this elegant mechanism ensures a correct order of substrate ubiquitylation by the anaphase-promoting complex (APC) and sensitive substrate discrimination by the ERAD machinery( Rape et al , 2006 ; Zhang et al , 2013 ), it is distinct from the mode of signaling amplification in lipid saturation sensing by Mga2: the rate for the first ubiquitylation supports discrimination of different lipid environments, while different substrates of the ERAD machinery feature an identical rate for the first ubiquitin transfer( Zhang et al , 2013 ). Furthermore, an auto-regulatory diversion of ubiquitin flux to the E3 ligase has - to our knowledge – not been implicated in the fidelity of ubiquitylation-dependent decisions. In summary, we have dissected the mechanism of signal amplification in lipid saturation sensing, which works in the absence of deubiquitylating activities. Future work will address the role of DUBs in this process and explore how autoregulatory self-inhibition of Rsp5 contributes to orchestrating its many functions in vivo . SUPPLEMENTARY FIGURES Download figure Open in new tab Figure S1: Validation of correct membrane insertion of ZIP-MBP Mga2 (A ) Incorporation of the protein into the lipid bilayer during reconstitution was tested by sucrose density gradient centrifugation. 25 µg of protein were set to 40% (w/v) sucrose and overlaid with successively decreasing sucrose concentrations (20%, 10%, 5%, 0%). The gradients were centrifuged at 100,000 x g for 18 hours and samples were collected from top (low sucrose) to bottom (high sucrose). The protein content was determined via SDS-PAGE (10 µL of sample was loaded, gels: 7.5% Mini-PROTEAN ® TGX™) and in-gel fluorescence scanning (Typhoon laser scanner, 488 nm laser, Cy2 filter, 25 µm resolution, 360 V PMT voltage). The lipids were tracked by staining 100 µL of each fraction with 7 µM of Hoechst 33342 ( Jumpertz et al , 2011 ; Cordeiro et al , 2023 ; Halbleib et al , 2017 ). The signal was detected using a TECAN plate reader: ex. = 355 nm, em. = 459 nm, bandwidth = 20 nm. (B) Membrane integration and association or aggregation of ZIP-MBP Mga2 were tested by alkaline carbonate extraction, urea extraction, and high salt extraction. Each condition was performed with 5 µg of protein to which buffer containing 200 mM Na 2 CO 3 , 2 M urea, or 500 mM NaCl was added. The mix was incubated for 30 min, after which proteoliposomes were separated from soluble components by ultracentrifugation (350,000 x g, 2 h, 4°C). The supernatant (SN) containing soluble components was recovered. The proteoliposome-containing pellet (P) was resuspended in a volume equal to the recovered SN (V SN = V P ). The samples of the SN and P were analyzed by SDS-PAGE (10 µL of each fraction was loaded, gels: 7.5% Mini-PROTEAN ® TGX™) and in-gel fluorescence scanning (Typhoon laser scanner, 488 nm laser, Cy2 filter, 25 µm resolution, 360 V PMT voltage). Download figure Open in new tab Figure S2: Full-length Rsp5 supports ZIP-MBP Mga2 ubiquitylation in vitro (A) Domain architecture of the full-length Rsp5 (aa 1-809) and truncated construct (383-809). The FL construct contains the N-terminal C2 domain (dark gray) as well as the three WW domains (gray) and the catalytically active HECT domain (orange). The short construct encompasses the third WW and the HECT domain. Both constructs are purified as N-terminal GST fusion constructs. The GST-tag is removable by protease cleavage using a TEV (short construct) or HRV 3C protease (FL construct). The three lysine residues that are responsible for autoubiquitylation induced inhibition, are highlighted in red. (B) SDS-PAGE of recombinant Rsp5 construct after isolation from E. coli . GST-HECT: short construct (aa 383-809) containing the affinity purification tag. HECT: short construct (aa 383-809) after protease cleavage. GST-HECT 3KR: GST-tagged short construct with K-to-R substitution of all three inhibitory lysine residues (aa 383-809; K411R, K432R, K438R). (C) In vitro ubiquitylation of ZIP-MBP Mga2 with different Rsp5 constructs after reconstitution into 100% POPC membranes. Assay components: 70 nM of E1, 500 nM of E2, 200 nM of the indicated Rsp5 construct, 2 µM of ZIP-MBP Mga2, 15 µM of ubiquitin, 10x ATP regenerating system. The reactions were incubated at 30°C and samples were taken by mixing 7.5 µL of the reaction with 4x MSB followed by boiling at 95°C for 5 min. Of each sample, 6.0 µL was loaded for SDS-PAGE (gels: 7.5% Mini-PROTEAN ® TGX™). In-gel fluorescence was detected with a Typhoon laser scanner (488 nm laser, Cy2 filter, 25 µm resolution, PMT voltage of 360 V). (D) Analysis of construct-dependent Rsp5 autoubiquitylation after ubiquitylation of ZIP-MBP Mga2. Samples from C were diluted 1:10 and used for anti-Rsp5 immunoblotting. Primary antibody: polyclonal anti-Rsp5 serum (rabbit, 1:1000). Secondary antibody: IRDye ® 800 CW goat anti-rabbit (1:15000). The blots were scanned on a LI-COR Odyssey ® scanner at 700 and 800 nm. Red arrow: unmodified construct at time point t = 0 min. Asterisk: unspecific recognition of ZIP-MBP Mga2 by the anti-Rsp5 polyclonal serum. Download figure Open in new tab Figure S3: Quality assessment of fits for ZIP-MBP Mga2 ubiquitylation in 100 mol% POPC membranes (A)-(G) Contribution of ubiquitylation and inhibition to the fit. The sum of the ubiquitylated portion (dotted lines) and the inhibited pool (dashed lines) gives the overall shape of the fit (solid line). The fits are re-plotted from figure 2 and were generated by fitting the in vitro ubiquitylation data to the theoretical model. The optimal values for the parameters kx and kx’ are indicated on the upper right. (H)-(L) Heatmap analysis of the parameter space for each kx/kx’ parameter pair. The log 10 (ξ 2 ) for different kx/kx’ pairs is shown. All rates were successively increased in 0.01 min −1 steps. (H) Screened area k2/k2’: k2 from 0.01 min −1 to 5.0 min −1 and k2’ from 0.01 min −1 to 1.0 min −1 . (I) Screened area for k3/k3’: both parameters from 0.01 min −1 to 1.0 min −1 . (J) Screened area for k4/k4’: k4 from 0.01 min −1 to 1.5 min −1 and k4’ from 0.01 min −1 to 1.0 min −1 . (K) Screened area for k5/k5’: k5 from 0.01 min −1 to 3.0 min −1 and k5’ from 0.01 min −1 to 1.0 min −1 . (L) Screened area for k6/k6’: 0.4 min −1 to 2.0 min −1 and k6’ from 0.01 min −1 to 1.0 min −1 . Download figure Open in new tab Figure S4: Quality assessment of fits for ZIP-MBP Mga2 ubiquitylation in 100% DOPC membranes (A)-(F) Heatmap representation of log 10 (ξ 2 ). The parameters kx and kx’were successively increased, and the quality of the fit was evaluated. (A) Screened area for k1/k1’: k1 from 0.01 min −1 to 2.0 min −1 and k1’ from 0.01 min −1 to 1.0 min −1 . The line scans show the effect of different complementary rates on ξ 2 for a fixed rate k1 or k1’. Red curve: quality values for a fixed k1 = 0.35 min −1 with various k1’ values. Blue curve: quality values for a fixed k1’ = 0.11 min −1 with increasing values for k1. Dashed box: focus on a zoomed in area of the heatmap with increased resolution. Screened area: k1 from 0.002 min − 1 to 1.0 min −1 and k1’ from 0.002 min −2 to 1.0 min −1 . The rates were successively increased by 0.002 min −1 . White dashed line: scan along the optimal values for analysis of the optimum. (B) Selected area for k2/k2’: k2 from 40 min −1 to 59.96 min −1 in 0.04 min −1 increments and k2’ from 0.01 min −1 to 5.0 min −1 in 0.01 min −1 steps. (C) Screened area for k3/k3’: k3 from 0.5 min −1 to 2.49 min −1 and k3’ from 0.01 min −1 to 2.0 min −1 , both in steps of 0.01 min −1 . (D) Screened area for k4/k4’: k4 from 12.0 min −1 to 19.96 min −1 in steps of 0.04 min −1 and k4’ from 0.01 min −1 to 2.0 min −1 with 0.01 min −1 steps. (E) Screened area for k5/k5’: k5 from 18.0 min −1 to 21.98 min −1 in steps of 0.02 min −1 and k5’ from 0.01 min −1 to 2.0 min −1 in 0.01 min −1 steps. (F) Screened area for k6/k6’: k6 from 2.5 min −1 to 4.49 min −1 and k6’ from 0.01 min −1 to 2.0 min −1 , both in steps of 0.01 min −1 . (G)-(M) Bar graph representation of the kx/kx’ ratio of the optimal rate constants determined for assays with 100 mol% POPC (black) and 100 mol% DOPC (gray) proteoliposomes plotted in Figure 3 . Download figure Open in new tab Figure S5: Identified ubiquitylation sites in ZIP-MBP Mga2 by mass spectrometry Detailed overview of all ubiquitylated lysine residues identified by mass spectrometry. ZIP: leucine zipper. MBP: maltose binding protein. Native: native lysines of Mga2, which are ubiquitylated in vivo . The intensity was normalized to the total intensity. Shown is the log 2 of the normalized intensities of n = 3 replicates. (A) Identification of ubiquitylated lysines in ZIP-MBP Mga2 after reconstitution into tightly packed (POPC) and loosely packed (DOPC) membranes and ubiquitylation for 20 min. (B) Analysis of ubiquitylation sites in ZIP-MBP Mga2 and mutant constructs after reconstitution into 100 mol% POPC membranes and ubiquitylation for 20 min. ΔLPKY: removal of the Rsp5 binding motif. 3KR: lysine-to-arginine substitution of ZIP-MBP Mga2’s native lysines. Materials and Methods Plasmids View this table: View inline View popup Download powerpoint Table 1: List of plasmids used in this study Oligonucleotides View this table: View inline View popup Download powerpoint Table 2: List of oligonucleotides used in this sutdy Molecular cloning The oligonucleotides listed in Table 2 were used to generate plasmids with the indicated amino acid substitutions. The site-directed mutagenesis PCRs were performed using the CloneAmp ™ HiFi PCR mix (Takara Bio). Heterologous expression of recombinant proteins For heterologous expression of all recombinant proteins mentioned in this work, chemically competent BL21 Star (DE3) pLysS or CodonPlus-RIL (DE3) E. coli were transformed with the corresponding expression vectors as indicated in Table 3 . View this table: View inline View popup Download powerpoint Table 3: plasmids and bacterial strains used for expression and synthesis of recombinant proteins The GST-tagged Rsp5 constructs, the 6xHis-tagged Uba1 (E1 enzyme), and the GST-tagged UbcH5B (E2 enzyme) were expressed in E. coli cultivated in ZYM-5052-autoinduction medium (1% (w/v) tryptone, 0.5% (w/v) yeast extract, 25 mM Na 2 HPO 4 , 25 mM KH 2 PO 4 , 50 mM NH 4 Cl, 5 mM Na 2 SO 4 , 2 mM MgSO 4 , 0.5% (w/v) glycerol, 0.05% (w/v) glucose, 0.2% (w/v) α-lactose, 0.02% (v/v) 1000x trace metals) ( Studier, 2005 ). An overnight pre-culture of bacteria cultivated in LB medium was used to inoculate a main culture (autoinduction medium) to an optical density of 0.05 OD 600 units (ODU). The bacteria were cultivated at 37°C until an OD 600 of 1.0 ODU was reached. The cultures were switched to incubation at 18°C for 18-20 hours to allow optimal protein synthesis. The bacteria were harvested by centrifugation (5000 x g, 30 min, 4°C) and washed with cold PBS (137 mM NaCl, 2.7 mM KCl, 10 mM Na 2 HPO 4 , 1.8 mM KH 2 PO 4 , pH 7.4). The cell pellets were stored at −20°C. For the expression of ubiquitin constructs, LB medium was used for main cultures. The ZIP-MBP Mga2 constructs were cultivated in a main culture of LB medium with 0.2% (w/v) glucose. In both cases, main cultures were inoculated to an optical density of 0.05 ODU using an overnight pre-culture. The cultures were incubated at 37°C (220 rpm shaking) and gene expression was induced by addition of 0.3 mM IPTG as the cultures reached an optical density of 0.6 ODU. After three hours of expression, the bacteria were harvested by centrifugation (5000 x g, 30 min, 4°C), washed with cold PBS (pH 7.4), and stored at −20°C. Affinity purification of GST fusion constructs Constructs of yeast Rsp5 (E3 enzyme) and human UbcH5B (E2 enzyme) contain an N-terminal GST-tag. Affinity purification was performed at 4°C or on ice. Frozen cell pellets were thawed and resuspended in Buffer 1 (50 mM Tris-HCl pH 8.0, 300 mM NaCl, 10% (w/v) glycerol, 0.2% (v/v) Tergitol™ Type NP-40 (Sigma-Aldrich), 1 mM DTT, 0.01% (v/v) Benzonase ® nuclease (Sigma-Aldrich), 0.1% (w/v) chymostatin, 0.1% (w/v) antipain, 0.1% (w/v) pepstatin A). The resuspended cells were lysed by sonication. Insoluble cell debris was removed by ultracentrifugation (100,000 x g, 30 min, 4°C). The supernatant was mixed with 3.0 mL of glutathione Sepharose™ 4B resin (Cytiva) per liter of bacterial culture. The matrix was equilibrated with water and Buffer 2 (50 mM Tris-HCl pH 8.0, 300 mM NaCl, 10% (w/v) glycerol, 1 mM DTT). The supernatant-resin suspension was incubated at 4°C for 30 min, followed by transfer to gravity columns. The resin was washed with 3x 10 column volumes (CVs) of Buffer 2. The GST-tagged protein was eluted in 3x 3.0 mL of Buffer 3 (50 mM Tris-HCl pH 8.0, 300 mM NaCl, 10% (w/v) glycerol, 1 mM DTT, 10 mM reduced glutathione) after 5 min of incubation. The eluate was collected in 1 mL fractions. The protein concentration was estimated using the absorption at 280 nm, the theoretical molar extinction coefficient, and the molecular weight of the protein. For GST-E2 preparations, the affinity-purified protein was used for size exclusion chromatography. If indicated, the GST-tag was removed from Rsp5 constructs by protease cleavage. To this end, 1 µg of TEV (short Rsp5 constructs) or PreScission ® protease (full-length Rsp5 constructs) was used per 100 µg of recombinant protein. Cleavage was allowed for 20 hours at 4°C with constant agitation. Then, the protease-protein mix was incubated with 500 µL of washed glutathione Sepharose™ 4B and 500 µL Ni-NTA resin for 60 min at 4°C to remove any uncleaved proteins and the proteases, respectively. The resin material was removed from the solution using gravity columns. The protein concentration was determined as described. Affinity-purified proteins were further purified by size exclusion chromatography. Purification and fluorescence labeling of ZIP-MBP Mga2 constructs Cell pellets were resuspended in detergent-containing Buffer 4 (25 mM HEPES pH 7.4, 150 mM NaCl, 50 mM n-octyl-D-glucopyranoside (OG), 10 mM TCEP, 1 mM EDTA, 0.01% (v/v) Benzonase ® nuclease (Sigma-Aldrich), 0.1% (w/v) chymostatin, 0.1% (w/v) antipain, 0.1% (w/v) pepstatin A) and lysed with ultrasound. The lysate was incubated at 4°C for 60 min with moderate agitation to allow solubilization of the membrane proteins. The insoluble components were removed by ultracentrifugation (100,000 x g, 30 min, 4°C), and the supernatant was recovered. For affinity purification, 3.0 mL of amylose resin (NEB) per liter of bacterial cultures were washed with water and Buffer 5 (25 mM HEPES pH 7.4, 150 mM NaCl, 50 mM OG, 1 mM EDTA). The recovered supernatant of the lysate was mixed with the washed resin. The supernatant-resin suspension was incubated at 4°C for 60 min with slight agitation. The suspension was transferred to gravity columns and washed with 3x 10 CVs of Buffer 5 to remove the reducing agent. The resin-bound protein was used for covalent labeling of the cysteine with a maleimide-coupled fluorophore. To this end, 3.0 mL of ATTO 488- or ATTO 590-maleimide solution (250 µM of dye in Buffer 5) were added to the gravity column and mixed with the resin-bound protein. The column was sealed and incubated at 4°C with constant, mild agitation for 18 - 20 hours. After labeling, the resin was washed with 30 CVs of Buffer 5. The labeled protein was eluted from the resin by incubating 3x 3.0 mL of Buffer 6 (25 mM HEPES pH 7.4, 150 mM NaCl, 50 mM OG, 1 mM EDTA, 10 mM maltose) for 5 min, followed by collection of 1.0 mL eluate fractions. The protein concentration and the labeling efficiency were determined for each fraction. To this end, the absorbance at 280 nm and the maximal absorption of ATTO 488 (λ abs = 500 nm, λ ems = 520 nm) or ATTO 590 ((λ abs = 593 nm, λ ems = 622 nm) were measured. The corrected protein concentration and the labeling efficiency were calculated using the protein-specific molecular weight and molar extinction coefficient and the dye-specific parameters according to the equations and specifications given by ATTO-TEC in the dye’s manual. The affinity fractions with the highest protein concentration and best labeling efficiency (>75%) were pooled and further purified by size exclusion chromatography. Purification of His-tagged ubiquitin constructs Cell pellets were resuspended in Buffer 7 (50 mM HEPES pH 7.4, 250 mM NaCl, 20 mM imidazole, 0.01% (v/v) Benzonase ® nuclease (Sigma-Aldrich), 0.1% (w/v) chymostatin, 0.1% (w/v) antipain, 0.1% (w/v) pepstatin A) and lysed with ultrasound. The crude lysate was centrifuged (100,000 x g, 30 min, 4°C) to remove the cell debris, and the supernatant was recovered. For each liter of bacterial culture, 3.0 mL of Ni-NTA agarose (Qiagen) were equilibrated with water and Buffer 8 (50 mM HEPES pH 7.4, 250 mM NaCl, 20 mM imidazole). The equilibrated resin was mixed with the recovered supernatant. The supernatant-resin mix was incubated at 4°C for 30 min with slight agitation. The suspension was transferred to gravity columns, followed by washing with 30 CVs of Buffer 8. The resin-bound, His-tagged proteins were eluted by incubating 3x 3.0 mL of Buffer 9 (50 mM HEPES pH 7.4, 250 mM NaCl, 400 mM imidazole) for 5 min. The protein concentration was estimated using the absorbance at 280 nm and the protein-specific molar extinction coefficient (χ = 1490 M −1 *cm −1 ) and molecular weight (MW = 9863.2 g/mol). The affinity-purified protein was further purified by size exclusion chromatography. Purification of His-tagged mouse E1 The protocol for affinity purification of the mouse E1 enzyme was adapted from published protocols ( Carvalho et al , 2012 ). In brief, cells were resuspended in Buffer 10 (50 mM Tris-HCl pH 8.0, 150 mM NaCl, 0.15 (w/v) Triton X-100, 20 mM imidazole pH 8.0, 1 mM EDTA, 0.01% (v/v) Benzonase ® nuclease (Sigma-Aldrich), 0.1% (w/v) chymostatin, 0.1% (w/v) antipain, 0.1% (w/v) pepstatin A, 0.1% (v/v) PMSF) and lysed using ultrasound. The insoluble cell debris was removed by ultracentrifugation (100,000 x g, 30 min, 4°C), and the supernatant was collected. Per liter of bacterial culture, 3.0 mL of Ni-NTA agarose (Qiagen) were equilibrated with water and Buffer 11 (50 mM Na 2 HPO 4 , 150 mM NaCl, 20 mM imidazole pH 8.0, 1 mM DTT). The supernatant and resin were mixed and incubated at 4°C for 60 min with slight agitation. The resin-supernatant suspension was transferred to gravity columns, and the resin was washed with 30 CVs of Buffer 11. The bound proteins were eluted after 5 min of incubation with 3x 3.0 mL of Buffer 12 (50 mM Na 2 HPO 4 , 150 mM NaCl, 400 mM imidazole pH 8.0, 1 mM DTT). The protein concentration was determined as described. Size exclusion chromatography was performed for further purification. Size exclusion chromatography (SEC) SEC was performed on ÄKTA pure 25 L chromatography systems (GE Healthcare) using a Superdex ® 200 Increase 10/300 GL column. For SEC with the E1, E2, and Rsp5 constructs, the column was equilibrated with detergent-free Buffer 13 (25 mM HEPES pH 7.4, 150 mM NaCl, 1 mM TCEP). For runs with the Mga2 construct, detergent-containing Buffer 14 (25 mM HEPES pH 7.4, 150 mM NaCl, 1 mM EDTA, 50 mM OG) was used. Prior to loading, the affinity-purified protein was concentrated to 600 µL using Vivaspin ® 20 spin concentrators (Sartorius) with the appropriate molecular weight cut-off. The concentrated protein solution was applied using a 500 µL loop. Each run was performed with a flow rate of 0.5 mL/min. Eluate fractions were collected after the void volume (8.0 mL). The protein content of each fraction was estimated using the A 280 and the molar extinction coefficient. The labeling efficiency and protein concentration for fluorescently labeled ZIP-MBP Mga2 were calculated as described. The protein was concentrated, and the glycerol content was adjusted to 20% (w/v) using glycerol stocks (80% (w/v)), prepared in the respective SEC buffers (Buffers 13 or 14). The protein concentration was set to the desired final concentration using 20% (w/v) glycerol (in Buffers 13 or 14). The protein solutions were snap-frozen in liquid nitrogen for storage at −80°C. Preparation of multilamellar vesicles Multilamellar vesicles were prepared as 10 mM stocks from 18:1 (Δ9- cis ) phosphatidylcholine (DOPC) and 16:0-18:1(Δ9- cis ) phosphatidylcholine (POPC) (Avanti Polar Lipids). POPC and DOPC stocks (25 mg/mL) in chloroform were mixed in solvent-resistant 2.0 mL Eppendorf tubes to yield 10 µmol of lipids with a desired molar ratio of saturated and unsaturated lipid acyl chains. Chloroform was evaporated at 60°C under a constant stream of nitrogen until the formation of a lipid film. Remaining chloroform was removed using vacuum (2 - 4 mbar) for 1 hour at room temperature. Lipids were rehydrated with 1.0 mL Buffer 15 (25 mM HEPES pH 7.4, 150 mM NaCl, 5% (w/v) glycerol), followed by an incubation at 60°C with constant agitation (1200 rpm). The resulting multilamellar liposomes were sonicated for 20 min at 60°C in a water bath (VWR ® ultrasonic cleaner THD, power setting 9). The multilamellar vesicles were rapidly frozen with liquid nitrogen in aliquots at a lipid concentration of 10 mM. Reconstitution of ZIP-MBP Mga2 constructs into liposomes ZIP-MBP Mga2 was reconstituted into defined lipid environments with a protein:lipid ratio of 1:8000. Reconstitution mixes were prepared in the following order to limit aggregation of ZIP-MBP Mga2. First, 10 mM stocks of multilamellar vesicles were solubilized with 40 mM of n-octyl-D-glucopyranoside (OG) by adding Buffer 16 (25 mM HEPES pH 7.4, 150 mM NaCl, 5% (w/v) glycerol, 20% (w/v) OG). The mix was incubated at 4°C for 10 min with mild rotation. Second, Buffer 16 and Buffer 17 (25 mM HEPES pH 7.4, 150 mM NaCl, 50 mM OG, 1 mM EDTA, 20% (w/v) glycerol) were added to reach an OG concentration of 23.5 mM. The mix was incubated at 4°C with slight agitation. Third, fluorescently labeled ZIP-MBP Mga2 (0.1 mg/mL) was added last to reach a final OG concentration of 25.5 mM and a protein:lipid ratio of 1:8000. The mix was incubated at 4°C for 10 min with constant rotation. For detergent removal, 9.0 mL of reconstitution mix were dialyzed against 1.0 L of Buffer 18 (25 mM HEPES pH 7.4, 150 mM NaCl, 5% (w/v) glycerol, 1 mM EDTA). A total of 3.0 mL of the prepared reconstitution mix were transferred to Slide-A-Lyzer™ G2 dialysis cassettes with a molecular weight cut-off of 10 kDa (Thermo Scientific). The dialysis was performed at 4°C with constant mixing of the dialysis buffer in a total of four steps. First, the reconstitution mixes were dialyzed for 1 hour against 1 L of Buffer 18 containing 400 mg of Bio-Beads ® SM-2 resin (Bio-Rad) to provide a sink for detergent molecules. Then, two steps of dialysis against 1 L of fresh Buffer 18 for 1 hour followed. Fourth, the cassettes were placed in 1 L of Buffer 18 containing 800 mg of Bio-Beads ® SM-2 resin and incubated for 16-18 hours. The reconstitution reactions containing the proteoliposomes were recovered from the cassettes. The proteoliposomes were diluted 1:5 with Buffer 19 (20 mM HEPES pH 7.4, 75 mM NaCl) and harvested by centrifugation (257,000 x g at 4°C, 20 h). The pelleted proteoliposomes were resuspended in Buffer 15 to reach a final concentration of ZIP-MBP Mga2 of 4.0 µM. The proteoliposomes were snap-frozen in liquid nitrogen and stored at −80°C. Determination of protein recovery after dialysis Samples of the reconstitution mixes were taken before and after the dialysis procedure (see section above). OG was added to 50 µL of these samples to reach a final OG concentration of 50 mM. The final volume was set to 150 µL by adding Buffer 15. The fluorescence intensity of the samples was determined by scanning in a TECAN plate reader (ex. = 485 nm, em. = 535 nm, bandwidth = 20 nm). To this end, 100 µL of each sample were transferred to a 96-well plate (black, flat bottom, chimney well, non-binding, Greiner Bio-One). The fluorescence intensity of each sample was corrected to a background sample (50 mM OG in Buffer 16). Dilution or concentration of the fluorescence signal was accounted for by considering the total volume of the reconstitution mix before and after dialysis. Protein recovery after dialysis was normalized to the fluorescence signal of the reconstitution mix prior to dialysis. Preparation of large unilamellar vesicles by extrusion Large unilamellar vesicles (LUVs) were generated from MLVs (1 mM final lipid concentration) by extrusion as previously described ( MacDonald et al , 1991 ). In brief, MLVs were passed through the extruder with a 100 nm filter 21 times to create LUVs. The LUVs were diluted 1:6 with Buffer 19 and harvested by centrifugation (500,000 x g, 16 h, 4°C). The pelleted LUVs were resuspended in a small volume to increase the lipid concentration to 10 mM. The LUVs were briefly stored at 4°C and quickly used after preparation. Urea, carbonate, and high salt extraction of membrane associated proteins 20 µL of the proteoliposomes (4 µM protein concentration) were mixed with an equal volume of carbonate buffer (Buffer 20: 20 mM HEPES pH 7.4, 75 mM NaCl, 200 mM Na 2 CO 3 pH 11), urea buffer (Buffer 21: 20 mM HEPES pH 7.4, 75 mM NaCl, 2 M urea), high salt buffer (Buffer 22: 20 mM HEPES pH 7.4, 500 mM NaCl) or Buffer 19 (20 mM HEPES pH 7.4, 75 mM NaCl). The mixes were then incubated at room temperature for 30 min. The mixes were diluted 25-fold with Buffer 19 followed by a centrifugation step with 350,000 x g at 4°C for 2 hours. The supernatant was recovered. The pellet was resuspended in a volume equal to the supernatant. Samples of each fraction were then mixed with 5x membrane sample buffer, boiled at 95°C for 5 min and analyzed via SDS-PAGE (10 µL of samples were loaded) followed by in-gel fluorescence scanning (Typhoon Laser Scanner, Cy2 Laser, 25 µm resolution, 360 V PMT voltage). Proteinase K protection assay The reaction mix contained 12.5 µg of reconstituted ZIP-MBP Mga2 (final protein concentration 0.25 µg/µL), which was diluted from the concentrated proteoliposome stock with Buffer 15 (25 mM HEPES pH 7.4, 150 mM NaCl, 5% (w/v) glycerol). To reach complete degradation of the protein, 1% (w/v) SDS was added to one condition. Then, 1 µL of proteinase K (New England Biolabs) was added to the SDS-free and SDS-containing reactions. An untreated reaction (proteoliposomes without proteinase K or SDS) served as a control. The reactions were incubated at room temperature for 60 min. The proteinase K was inactivated by boiling at 90°C for 10 min and addition of 0.1 mM PMSF. The inactivation procedure was also performed for the control condition. Then, 5x membrane sample buffer was added to the reactions, followed by an incubation at 95°C for 5 min. The samples were analyzed by SDS-PAGE and in-gel fluorescence detection to trace the degradation of the fluorescently labeled ZIP-MBP Mga2 construct (Typhoon Laser Scanner, Cy2 Laser, 25 µm resolution, 360 V PMT voltage). Dynamic light scattering measurements Dynamic light scattering (DLS) was performed on a Zetasizer Nano-S (Malvern Panalytical). For measurements, 50 µL of proteoliposomes (12.5 µg of ZIP-MBP Mga2) were used in a ZEN2112 quartz cuvette. The temperature was allowed to equilibrate to 30°C for 2 min. The material properties were set to phospholipid-based proteoliposomes (refractive index: 1.450, absorbance: 0.001) and the buffer composition was adjusted accordingly (viscosity: 0.9298 cP, refractive index: 1.338). Sucrose density gradient centrifugation The proteoliposomes with a protein content of 25 µg were adjusted to 40% (w/v) sucrose (prepared in Buffer 15) in a volume of 600 µL. The protein solution was transferred to 13.2 mL open-top thin-wall ultracentrifugation tubes. This suspension was overlaid with 2.0 mL layers of successively decreasing sucrose concentrations: 20%-10%-5%-0% (w/v) sucrose. The gradients were centrifuged in a SW41-Ti swing-out rotor at 100,000 x g for 18 hours (temperature: 4°C). The centrifuge was set to accelerate slowly and decelerate without brakes. Then, 1 mL fractions were collected from top to bottom. The protein content of each fraction was assessed by SDS-PAGE and in-gel fluorescence scanning (Typhoon Laser Scanner, Cy2 Laser, 25 µm resolution, 360 V PMT voltage). In vitro ubiquitylation assays under constant turnover conditions In vitro ubiquitylation assays consisted of 70 nM E1 (6xHis-Uba1, mouse), 500 nM E2 (GST-UbcH5B, human), 200 nM E3 (Rsp5, yeast), 2 µM reconstituted fluorescently labeled ZIP-MBP Mga2, and 15 µM 8xHis-ubiquitin (human) in Buffer 15 (25 mM HEPES pH 7.4, 150 mM NaCl, 5% (w/v) glycerol). Autoubiquitylation assays of Rsp5 were performed under the same conditions, but without reconstituted ZIP-MBP Mga2. All components except for the 10x ATP regenerating system (10 mM ATP, 500 mM creatine phosphate, 2 mg/mL creatine phosphokinase in 25 mM HEPES pH 7.4, 150 mM NaCl, 5 mM MgCl 2 ) were mixed. The ubiquitylation was started by the addition of the 10x ATP regenerating system. A master mix of the reaction was incubated at 30°C under constant shaking (300 rpm). Samples were drawn at indicated time points by mixing 15 µL of the reactions with 5 µL of 4x MSB (100 mM Tris-HCl pH 6.8, 8 M urea, 3.2% (w/v) SDS, 0.15% (v/v) bromophenol blue, 4% (v/v) glycerol, 5 mM EDTA) and incubation at 95°C for 5 min. Antibodies The anti-Rsp5 antibody (rabbit, polyclonal) was kindly provided by Jeffrey Brodsky. The polyclonal anti-Rsp5 serum was generated by the antibody production facility of the Department of Medical Biochemistry and Molecular Biology of Saarland University headed by Dr. Martin Jung. Rabbits were immunized with the WW-HECT domain construct of Rsp5 (aa 383-809). The IRDye ® 800CW antibody was purchased from LI-COR Biosciences. SDS-PAGE and sample analysis The samples collected in the in vitro ubiquitylation assays were analyzed by SDS-PAGE using 7.5% Mini-PROTEAN ® TGX™ or 7.5% Criterion™ TGX™ precast gels (Bio-Rad). Per well, 6 µL of sample containing 0.55 µg of ZIP-MBP Mga2 was loaded. The gels were run at 180 V (Mini-PROTEAN ® TGX™ gels) or 200 V (Criterion™ TGX™ gels). Fluorescently labeled ZIP-MBP Mga2 species were subsequently detected by in-gel fluorescence detection (ATTO 488-labeled: Cy2, laser, 25 µm resolution, photomultiplier voltage of 360 V; ATTO 590-labeled: Cy5 laser, 25 µm resolution, photomultiplier voltage of 500 V). Unlabeled Rsp5 species and the MBP were detected via immunoblotting. The proteins were blotted onto nitrocellulose membranes using the Trans-Blot ® Turbo™ transfer system (Bio-Rad). The membranes were blocked in Buffer 23 (5% (w/v) BSA in TBS-T (20 mM Tris pH 7.5, 137 mM NaCl, 0.2% (v/v) Tween-20)) for 30 min. The primary anti-Rsp5 serum was diluted 1:2000 in Buffer 23 and incubated with the membranes for 1 hour at room temperature or at 4°C for overnight incubation. The membranes were washed with TBS-T (20 mM Tris-HCl pH 7.5, 137 mM NaCl, 0.2% (v/v) Tween-20). The IRDye ® 800CW goat anti-rabbit IgG secondary antibody (LI-COR Biosciences) was prepared in Buffer 23, applied in a 1:15000 dilution, and incubated with the membranes for 1 hour at room temperature. The membranes were washed with TBS-T prior to imaging on an Odyssey ® CLx imaging system (LI-COR Biosciences) at both 700 and 800 nm. Preparation and processing of samples for mass spectrometry In vitro ubiquitylation assays were performed as described. Ubiquitylation was stopped with 1 mM of N-ethyl-maleimide at indicated time points. Purified proteins from the ubiquitylation reactions were solved in SDC buffer (2% SDC, 1 mM TCEP, 4 mM CAA, 50 mM Tris pH 8.5) and heated for 10 min at 95°C. 100 ng of trypsin and LysC were added in 50 mM Tris pH 8.5 and incubated overnight at 37°C. The digestion was stopped upon addition of 150 µL of 1% TFA in isopropanol. Peptide clean-up was performed using SDB-RPS stage tips (Sigma-Aldrich). Peptides were added to stage tips and washed first with 1% TFA in isopropanol and then with 0.2% TFA in water. Lastly, peptides were eluted in 80% acetonitrile plus 1.25% ammonia and dried in a vacuum concentrator. Mass spectrometry data acquisition Samples were analyzed on a Q Exactive HF coupled to an EASY-nLC 1200 (ThermoFisher Scientific) using a 35 cm long, 75 µm ID fused-silica column packed in-house with 1.9 µm C18 particles (ReproSil-Pur, Dr. Maisch) and kept at 50°C using an integrated column oven (Sonation). HPLC solvents consisted of 0.1% formic acid in water (Buffer A) and 0.1% formic acid, 80% acetonitrile in water (Buffer B). Peptides were eluted by a linear gradient from 5% to 30% B over 30 minutes, followed by a stepwise increase to 95% B in 6 minutes which was held for another 9 minutes. Full scan MS spectra (350-1650 m/z) were acquired in Profile mode at a resolution of 60,000 at m/z 200, a maximum injection time of 20 ms and an AGC target value of 3 x 10 6 . Up to 15 most intense peptides per full scan were isolated using a 1.4 m/z window and fragmented using higher energy collisional dissociation (normalized collision energy of 27). MS/MS spectra were acquired in profile mode with a resolution of 30,000, a maximum injection time of 54 ms and an AGC target value of 1 x 10 5 . Singley charged ions, ions with a charge state above 5 and ions with unassigned charge states were not considered for fragmentation. Mass spectrometry data analysis MS raw data were analyzed using MaxQuant (v1.6.7.0) ( Tyanova et al , 2016 ). Acquired spectra were searched against a custom database containing Mga2, Rsp5 (both NCBI txid: 1247190), Uba1 (NCBI txid: 10090), UbcH5B and ubiquitin sequences (both NCBI txid: 9606) and a collection of common contaminants using the Andromeda search engine integrated in MaxQuant ( Cox et al , 2011 ). Identifications were filtered to obtain false discovery rates (FDR) below 1% for both peptide spectrum matches (PSM; minimum length of 7 amino acids). Spectra were searched with a mass tolerance of 6 ppm in MS mode, 20 ppm in HCD MS2 mode, strict trypsin specificity, and allowing up to 2 miscleavages. Carbamidomethylated cysteine was set as a fixed modification and oxidation of methionine, N-terminal protein acetylation and GlyGly on Lysine as variable modifications allowing up to 5 modifications per peptide. Data Fitting The model used for fitting was adapted from Pierce et al . ( Pierce et al , 2009 ) (analytical solutions to differential equations, see below). Kinetic rate constants were estimated by ξ 2 minimization of mean normalized data. Parameter pairs were optimized through an iterative refinement procedure analogous to simulated annealing: initial search intervals covered four orders of magnitude (starting at 0.001 min⁻¹), with 1000 samples per iteration. The best-performing values defined the subsequent interval boundaries. This process was repeated 1000 times per parameter pair. For unmodified Mga2, rates were derived assuming that abundance after 60 min is solely determined by the inhibitory rate k0’, allowing exact determination of k0 and k0’. Fit quality was quantified as weighted ξ 2 , defined as the sum of squared deviations between calculated curves (y fit ) and mean data (y mean ), normalized by the standard deviation (SD) at each time point: Global parameter landscapes were assessed by heat map analysis, in which parameters were varied pairwise in fixed increments and evaluated by ξ 2 . All computations were implemented in C++. Disclosure and competing interest statement The authors declare that they have no conflict of interest. Acknowledgements This work was funded by the Deutsche Forschungsgemeinschaft in the framework of the SFB1027 to HR and RE, and with an LC-MS system (easy nLC 1200, QExactive HF) used in this study (Project-ID: 259130777, SFB1177 – Selective Autophagy). Furthermore, the project was funded by the European Research Council under the European Union’s Horizon 2020 research and innovation program (grant agreement no. 866011) to RE. References 1. ↵ Alsayyah C , Rodrigues E , Hach J , Renne MF & Ernst R ( 2025 ) Reversible tuning of membrane sterol levels by cyclodextrin in a dialysis setting . Biophys J 124 : 1433 – 1445 OpenUrl 2. ↵ Antonny B ( 2011 ) Mechanisms of membrane curvature sensing . Annu Rev Biochem 80 : 101 – 123 OpenUrl CrossRef PubMed Web of Science 3. ↵ Attali I , Tobelaim WS , Persaud A , Motamedchaboki K , Simpson-Lavy KJ , Mashahreh B , Levin-Kravets O , Keren-Kaplan T , Pilzer I , Kupiec M , et al. ( 2017 ) Ubiquitylation-dependent oligomerization regulates activity of Nedd4 ligases . EMBO J 36 : 425 – 440 OpenUrl Abstract / FREE Full Text 4. ↵ Ballweg S & Ernst R ( 2017 ) Control of membrane fluidity: the OLE pathway in focus . Biol Chem 398 : 215 – 228 OpenUrl CrossRef PubMed 5. ↵ Ballweg S , Sezgin E , Doktorova M , Covino R , Reinhard J , Wunnicke D , Hänelt I , Levental I , Hummer G & Ernst R ( 2020 ) Regulation of lipid saturation without sensing membrane fluidity . Nat Commun 11 : 756 OpenUrl CrossRef PubMed 6. ↵ Bhattacharya S , Shcherbik N , Vasilescu J , Smith JC , Figeys D & Haines DS ( 2009 ) Identification of lysines within membrane-anchored Mga2p120 that are targets of Rsp5p ubiquitination and mediate mobilization of tethered Mga2p90 . J Mol Biol 385 : 718 – 725 OpenUrl CrossRef PubMed 7. ↵ Bhattacharya S , Zoladek T & Haines DS ( 2008 ) WW domains 2 and 3 of Rsp5p play overlapping roles in binding to the LPKY motif of Spt23p and Mga2p . Int J Biochem Cell Biol 40 : 147 – 157 OpenUrl CrossRef PubMed Web of Science 8. ↵ Bigay J & Antonny B ( 2012 ) Curvature, lipid packing, and electrostatics of membrane organelles: defining cellular territories in determining specificity . Dev Cell 23 : 886 – 895 OpenUrl CrossRef PubMed Web of Science 9. ↵ Bodnar NO & Rapoport TA ( 2017 ) Molecular mechanism of substrate processing by the Cdc48 ATPase complex . Cell 169 : 722 – 735 .e9 OpenUrl CrossRef PubMed 10. ↵ Boumann HA , Gubbens J , Koorengevel MC , Oh C-S , Martin CE , Heck AJR , Patton-Vogt J , Henry SA , de Kruijff B & de Kroon AIPM ( 2006 ) Depletion of phosphatidylcholine in yeast induces shortening and increased saturation of the lipid acyl chains: evidence for regulation of intrinsic membrane curvature in a eukaryote . Mol Biol Cell 17 : 1006 – 1017 OpenUrl Abstract / FREE Full Text 11. ↵ Carvalho AF , Pinto MP , Grou CP , Vitorino R , Domingues P , Yamao F , Sá-Miranda C & Azevedo JE ( 2012 ) High-yield expression in Escherichia coli and purification of mouse ubiquitin-activating enzyme E1 . Mol Biotechnol 51 : 254 – 261 OpenUrl CrossRef PubMed 12. ↵ Chang A , Cheang S , Espanel X & Sudol M ( 2000 ) Rsp5 WW domains interact directly with the carboxyl-terminal domain of RNA polymerase II . J Biol Chem 275 : 20562 – 20571 OpenUrl Abstract / FREE Full Text 13. ↵ Chellappa R , Kandasamy P , Oh CS , Jiang Y , Vemula M & Martin CE ( 2001 ) The membrane proteins, Spt23p and Mga2p, play distinct roles in the activation of Saccharomyces cerevisiae OLE1 gene expression. Fatty acid-mediated regulation of Mga2p activity is independent of its proteolytic processing into a soluble transcription activator . J Biol Chem 276 : 43548 – 43556 OpenUrl Abstract / FREE Full Text 14. ↵ Christianson JC & Carvalho P ( 2022 ) Order through destruction: how ER-associated protein degradation contributes to organelle homeostasis . EMBO J 41 : e109845 OpenUrl CrossRef PubMed 15. ↵ Cordeiro MM , Filipe HAL , Santos PD , Samelo J , Ramalho JPP , Loura LMS & Moreno MJ ( 2023 ) Interaction of Hoechst 33342 with POPC membranes at different pH values . Molecules 28 16. ↵ Covino R , Ballweg S , Stordeur C , Michaelis JB , Puth K , Wernig F , Bahrami A , Ernst AM , Hummer G & Ernst R ( 2016 ) A eukaryotic sensor for membrane lipid saturation . Mol Cell 63 : 49 – 59 OpenUrl CrossRef PubMed 17. ↵ Covino R , Hummer G & Ernst R ( 2018 ) Integrated functions of membrane property sensors and a hidden side of the unfolded protein response . Mol Cell 71 : 458 – 467 OpenUrl CrossRef PubMed 18. ↵ Cox J , Neuhauser N , Michalski A , Scheltema RA , Olsen JV & Mann M ( 2011 ) Andromeda: a peptide search engine integrated into the MaxQuant environment . J Proteome Res 10 : 1794 – 1805 OpenUrl CrossRef PubMed Web of Science 19. ↵ Cybulski LE , Martín M , Mansilla MC , Fernández A & de Mendoza D ( 2010 ) Membrane thickness cue for cold sensing in a bacterium . Curr Biol 20 : 1539 – 1544 OpenUrl CrossRef PubMed 20. ↵ Deol KK , Lorenz S & Strieter ER ( 2019 ) Enzymatic logic of ubiquitin chain assembly . Front Physiol 10 : 835 OpenUrl CrossRef 21. ↵ Dunn R & Hicke L ( 2001 ) Domains of the Rsp5 ubiquitin-protein ligase required for receptor-mediated and fluid-phase endocytosis . Mol Biol Cell 12 : 421 – 435 OpenUrl Abstract / FREE Full Text 22. ↵ Dunn R , Klos DA , Adler AS & Hicke L ( 2004 ) The C2 domain of the Rsp5 ubiquitin ligase binds membrane phosphoinositides and directs ubiquitination of endosomal cargo . J Cell Biol 165 : 135 – 144 OpenUrl Abstract / FREE Full Text 23. ↵ Ernst R , Ballweg S & Levental I ( 2018 ) Cellular mechanisms of physicochemical membrane homeostasis . Curr Opin Cell Biol 53 : 44 – 51 OpenUrl CrossRef PubMed 24. ↵ Ernst R , Renne MF , Jain A & von der Malsburg A ( 2024 ) Endoplasmic reticulum membrane homeostasis and the unfolded protein response . Cold Spring Harb Perspect Biol 16 : a041400 OpenUrl Abstract / FREE Full Text 25. ↵ Fang NN , Chan GT , Zhu M , Comyn SA , Persaud A , Deshaies RJ , Rotin D , Gsponer J & Mayor T ( 2014 ) Rsp5/Nedd4 is the main ubiquitin ligase that targets cytosolic misfolded proteins following heat stress . Nat Cell Biol 16 : 1227 – 1237 OpenUrl CrossRef PubMed 26. ↵ Fang NN , Zhu M , Rose A , Wu K-P & Mayor T ( 2016 ) Deubiquitinase activity is required for the proteasomal degradation of misfolded cytosolic proteins upon heat-stress . Nat Commun 7 : 12907 OpenUrl CrossRef PubMed 27. ↵ Ferrell JE & Xiong W ( 2001 ) Bistability in cell signaling: How to make continuous processes discontinuous, and reversible processes irreversible . Chaos 11 : 227 – 236 OpenUrl CrossRef PubMed Web of Science 28. ↵ French ME , Klosowiak JL , Aslanian A , Reed SI , Yates JR 3rd . & Hunter T ( 2017 ) Mechanism of ubiquitin chain synthesis employed by a HECT domain ubiquitin ligase . J Biol Chem 292 : 10398 – 10413 OpenUrl Abstract / FREE Full Text 29. ↵ French ME , Kretzmann BR & Hicke L ( 2009 ) Regulation of the RSP5 ubiquitin ligase by an intrinsic ubiquitin-binding site . J Biol Chem 284 : 12071 – 12079 OpenUrl Abstract / FREE Full Text 30. ↵ Gromadski KB & Rodnina MV ( 2004 ) Kinetic determinants of high-fidelity tRNA discrimination on the ribosome . Mol Cell 13 : 191 – 200 OpenUrl CrossRef PubMed Web of Science 31. ↵ Haider A , Wei Y-C , Lim K , Barbosa AD , Liu C-H , Weber U , Mlodzik M , Oras K , Collier S , Hussain MM , et al. ( 2018 ) PCYT1A regulates phosphatidylcholine homeostasis from the inner nuclear membrane in response to membrane stored curvature elastic stress . Dev Cell 45 : 481 – 495 .e8 OpenUrl CrossRef PubMed 32. ↵ Halbleib K , Pesek K , Covino R , Hofbauer HF , Wunnicke D , Hänelt I , Hummer G & Ernst R ( 2017 ) Activation of the unfolded protein response by lipid bilayer stress . Mol Cell 67 : 673 – 684 .e8 OpenUrl CrossRef PubMed 33. ↵ Harayama T & Riezman H ( 2018 ) Understanding the diversity of membrane lipid composition . Nat Rev Mol Cell Biol 19 : 281 – 296 OpenUrl CrossRef PubMed 34. ↵ Ho H-C , MacGurn JA & Emr SD ( 2017 ) Deubiquitinating enzymes Ubp2 and Ubp15 regulate endocytosis by limiting ubiquitination and degradation of ARTs . Mol Biol Cell 28 : 1271 – 1283 OpenUrl Abstract / FREE Full Text 35. ↵ Holthuis JCM & Menon AK ( 2014 ) Lipid landscapes and pipelines in membrane homeostasis . Nature 510 : 48 – 57 OpenUrl CrossRef PubMed Web of Science 36. ↵ Hopfield JJ ( 1974 ) Kinetic proofreading: a new mechanism for reducing errors in biosynthetic processes requiring high specificity . Proc Natl Acad Sci U S A 71 : 4135 – 4139 OpenUrl Abstract / FREE Full Text 37. ↵ Hoppe T , Matuschewski K , Rape M , Schlenker S , Ulrich HD & Jentsch S ( 2000 ) Activation of a membrane-bound transcription factor by regulated ubiquitin/proteasome-dependent processing . Cell 102 : 577 – 586 OpenUrl CrossRef PubMed Web of Science 38. ↵ Jiang Y , Vasconcelles MJ , Wretzel S , Light A , Gilooly L , McDaid K , Oh C-S , Martin CE & Goldberg MA ( 2002 ) Mga2p processing by hypoxia and unsaturated fatty acids in Saccharomyces cerevisiae: impact on LORE-dependent gene expression . Eukaryot Cell 1 : 481 – 490 OpenUrl Abstract / FREE Full Text 39. ↵ Jiang Y , Vasconcelles MJ , Wretzel S , Light A , Martin CE & Goldberg MA ( 2001 ) MGA2 is involved in the low-oxygen response element-dependent hypoxic induction of genes in Saccharomyces cerevisiae . Mol Cell Biol 21 : 6161 – 6169 OpenUrl Abstract / FREE Full Text 40. ↵ Jumpertz T , Tschapek B , Infed N , Smits SHJ , Ernst R & Schmitt L ( 2011 ) High-throughput evaluation of the critical micelle concentration of detergents . Anal Biochem 408 : 64 – 70 OpenUrl PubMed 41. ↵ Kaliszewski P & Zoładek T ( 2008 ) The role of Rsp5 ubiquitin ligase in regulation of diverse processes in yeast cells . Acta Biochim Pol 55 : 649 – 662 OpenUrl CrossRef PubMed 42. ↵ Kamadurai HB , Qiu Y , Deng A , Harrison JS , Macdonald C , Actis M , Rodrigues P , Miller DJ , Souphron J , Lewis SM , et al. ( 2013 ) Mechanism of ubiquitin ligation and lysine prioritization by a HECT E3 . Elife 2 : e00828 OpenUrl CrossRef PubMed 43. Kamadurai HB , Souphron J , Scott DC , Duda DM , Miller DJ , Stringer D , Piper RC & Schulman BA ( 2009 ) Insights into ubiquitin transfer cascades from a structure of a UbcH5B approximately ubiquitin-HECT(NEDD4L) complex . Mol Cell 36 : 1095 – 1102 OpenUrl CrossRef PubMed Web of Science 44. ↵ Kee Y , Lyon N & Huibregtse JM ( 2005 ) The Rsp5 ubiquitin ligase is coupled to and antagonized by the Ubp2 deubiquitinating enzyme . EMBO J 24 : 2414 – 2424 OpenUrl Abstract / FREE Full Text 45. ↵ Kim HC & Huibregtse JM ( 2009 ) Polyubiquitination by HECT E3s and the determinants of chain type specificity . Mol Cell Biol 29 : 3307 – 3318 OpenUrl Abstract / FREE Full Text 46. ↵ Kim HC , Steffen AM , Oldham ML , Chen J & Huibregtse JM ( 2011 ) Structure and function of a HECT domain ubiquitin-binding site . EMBO Rep 12 : 334 – 341 OpenUrl Abstract / FREE Full Text 47. ↵ Kiss L , James LC & Schulman BA ( 2025 ) UbiREAD deciphers proteasomal degradation code of homotypic and branched K48 and K63 ubiquitin chains . Mol Cell 85 : 1467 – 1476 .e6 OpenUrl PubMed 48. ↵ MacDonald RC , MacDonald RI , Menco BP , Takeshita K , Subbarao NK & Hu LR ( 1991 ) Small-volume extrusion apparatus for preparation of large, unilamellar vesicles . Biochim Biophys Acta 1061 : 297 – 303 OpenUrl CrossRef PubMed 49. ↵ Mari S , Ruetalo N , Maspero E , Stoffregen MC , Pasqualato S , Polo S & Wiesner S ( 2014 ) Structural and functional framework for the autoinhibition of Nedd4-family ubiquitin ligases . Structure 22 : 1639 – 1649 OpenUrl CrossRef PubMed 50. ↵ Markevich NI , Hoek JB & Kholodenko BN ( 2004 ) Signaling switches and bistability arising from multisite phosphorylation in protein kinase cascades . J Cell Biol 164 : 353 – 359 OpenUrl Abstract / FREE Full Text 51. ↵ Nakagawa Y , Sakumoto N , Kaneko Y & Harashima S ( 2002 ) Mga2p is a putative sensor for low temperature and oxygen to induce OLE1 transcription in Saccharomyces cerevisiae . Biochem Biophys Res Commun 291 : 707 – 713 OpenUrl CrossRef PubMed 52. ↵ Needham PG , Guerriero CJ & Brodsky JL ( 2019 ) Chaperoning endoplasmic reticulum-associated degradation (ERAD) and protein conformational diseases . Cold Spring Harb Perspect Biol 11 : a033928 OpenUrl Abstract / FREE Full Text 53. ↵ Peters J-M ( 2002 ) The anaphase-promoting complex: proteolysis in mitosis and beyond . Mol Cell 9 : 931 – 943 OpenUrl CrossRef PubMed Web of Science 54. ↵ Pierce NW , Kleiger G , Shan S-O & Deshaies RJ ( 2009 ) Detection of sequential polyubiquitylation on a millisecond timescale . Nature 462 : 615 – 619 OpenUrl CrossRef PubMed Web of Science 55. ↵ Piwko W & Jentsch S ( 2006 ) Proteasome-mediated protein processing by bidirectional degradation initiated from an internal site . Nat Struct Mol Biol 13 : 691 – 697 OpenUrl CrossRef PubMed Web of Science 56. ↵ Pohl C & Dikic I ( 2019 ) Cellular quality control by the ubiquitin-proteasome system and autophagy . Science 366 : 818 – 822 OpenUrl Abstract / FREE Full Text 57. ↵ Ragaller F , Sjule E , Urem YB , Schlegel J , El R , Urbancic D , Urbancic I , Blom H & Sezgin E ( 2024 ) Quantifying fluorescence lifetime responsiveness of environment-sensitive probes for membrane fluidity measurements . J Phys Chem B 128 : 2154 – 2167 OpenUrl CrossRef 58. ↵ Rape M , Hoppe T , Gorr I , Kalocay M , Richly H & Jentsch S ( 2001 ) Mobilization of processed, membrane-tethered SPT23 transcription factor by CDC48(UFD1/NPL4), a ubiquitin-selective chaperone . Cell 107 : 667 – 677 OpenUrl CrossRef PubMed Web of Science 59. ↵ Rape M , Reddy SK & Kirschner MW ( 2006 ) The processivity of multiubiquitination by the APC determines the order of substrate degradation . Cell 124 : 89 – 103 OpenUrl CrossRef PubMed Web of Science 60. ↵ Renne MF & Ernst R ( 2023 ) Membrane homeostasis beyond fluidity: control of membrane compressibility . Trends Biochem Sci 48 : 963 – 977 OpenUrl CrossRef PubMed 61. ↵ Riling C , Kamadurai H , Kumar S , O’Leary CE , Wu K-P , Manion EE , Ying M , Schulman BA & Oliver PM ( 2015 ) Itch WW domains inhibit its E3 ubiquitin ligase activity by blocking E2-E3 ligase trans-thiolation . J Biol Chem 290 : 23875 – 23887 OpenUrl Abstract / FREE Full Text 62. Saeki Y , Isono E & Toh-E A ( 2005 ) Preparation of ubiquitinated substrates by the PY motif-insertion method for monitoring 26S proteasome activity . Methods Enzymol 399 : 215 – 227 OpenUrl CrossRef PubMed Web of Science 63. ↵ Saeki Y , Kudo T , Sone T , Kikuchi Y , Yokosawa H , Toh-e A & Tanaka K ( 2009 ) Lysine 63-linked polyubiquitin chain may serve as a targeting signal for the 26S proteasome . EMBO J 28 : 359 – 371 OpenUrl Abstract / FREE Full Text 64. ↵ Sardana R & Emr SD ( 2021 ) Membrane protein quality control mechanisms in the endo-lysosome system . Trends Cell Biol 31 : 269 – 283 OpenUrl CrossRef PubMed 65. ↵ Sardana R , Zhu L & Emr SD ( 2019 ) Rsp5 Ubiquitin ligase-mediated quality control system clears membrane proteins mistargeted to the vacuole membrane . J Cell Biol 218 : 234 – 250 OpenUrl Abstract / FREE Full Text 66. ↵ Shcherbik N , Kee Y , Lyon N , Huibregtse JM & Haines DS ( 2004 ) A single PXY motif located within the carboxyl terminus of Spt23p and Mga2p mediates a physical and functional interaction with ubiquitin ligase Rsp5p . J Biol Chem 279 : 53892 – 53898 OpenUrl Abstract / FREE Full Text 67. ↵ Shcherbik N , Zoladek T , Nickels JT & Haines DS ( 2003 ) Rsp5p is required for ER bound Mga2p120 polyubiquitination and release of the processed/tethered transactivator Mga2p90 . Curr Biol 13 : 1227 – 1233 OpenUrl CrossRef PubMed 68. ↵ Stegmeier F , Rape M , Draviam VM , Nalepa G , Sowa ME , Ang XL , McDonald ER 3rd . , Li MZ , Hannon GJ , Sorger PK , et al. ( 2007 ) Anaphase initiation is regulated by antagonistic ubiquitination and deubiquitination activities . Nature 446 : 876 – 881 OpenUrl CrossRef PubMed Web of Science 69. ↵ Studier FW ( 2005 ) Protein production by auto-induction in high density shaking cultures . Protein Expr Purif 41 : 207 – 234 OpenUrl CrossRef PubMed Web of Science 70. ↵ Stukey JE , McDonough VM & Martin CE ( 1989 ) Isolation and characterization of OLE1, a gene affecting fatty acid desaturation from Saccharomyces cerevisiae . J Biol Chem 264 : 16537 – 16544 OpenUrl Abstract / FREE Full Text 71. ↵ Surma MA , Klose C , Peng D , Shales M , Mrejen C , Stefanko A , Braberg H , Gordon DE , Vorkel D , Ejsing CS , et al. ( 2013 ) A lipid E-MAP identifies Ubx2 as a critical regulator of lipid saturation and lipid bilayer stress . Mol Cell 51 : 519 – 530 OpenUrl CrossRef PubMed Web of Science 72. ↵ Thrower JS , Hoffman L , Rechsteiner M & Pickart CM ( 2000 ) Recognition of the polyubiquitin proteolytic signal . EMBO J 19 : 94 – 102 OpenUrl Abstract / FREE Full Text 73. ↵ Tsuchiya H , Ohtake F , Arai N , Kaiho A , Yasuda S , Tanaka K & Saeki Y ( 2017 ) In vivo ubiquitin linkage-type analysis reveals that the Cdc48-Rad23/Dsk2 axis contributes to K48-linked chain specificity of the proteasome . Mol Cell 66 : 488 – 502 .e7 OpenUrl CrossRef PubMed 74. ↵ Tyanova S , Temu T & Cox J ( 2016 ) The MaxQuant computational platform for mass spectrometry-based shotgun proteomics . Nat Protoc 11 : 2301 – 2319 OpenUrl CrossRef PubMed 75. ↵ Wang G , Yang J & Huibregtse JM ( 1999 ) Functional domains of the Rsp5 ubiquitin-protein ligase . Mol Cell Biol 19 : 342 – 352 OpenUrl Abstract / FREE Full Text 76. ↵ Wang J , Peng Q , Lin Q , Childress C , Carey D & Yang W ( 2010 ) Calcium activates Nedd4 E3 ubiquitin ligases by releasing the C2 domain-mediated auto-inhibition . J Biol Chem 285 : 12279 – 12288 OpenUrl Abstract / FREE Full Text 77. ↵ Wang M & Pickart CM ( 2005 ) Different HECT domain ubiquitin ligases employ distinct mechanisms of polyubiquitin chain synthesis . EMBO J 24 : 4324 – 4333 OpenUrl Abstract / FREE Full Text 78. ↵ Wang Z , Liu Z , Chen X , Li J , Yao W , Huang S , Gu A , Lei Q-Y , Mao Y & Wen W ( 2019 ) A multi-lock inhibitory mechanism for fine-tuning enzyme activities of the HECT family E3 ligases . Nat Commun 10 : 3162 OpenUrl CrossRef PubMed 79. ↵ Wiesner S , Ogunjimi AA , Wang H-R , Rotin D , Sicheri F , Wrana JL & Forman-Kay JD ( 2007 ) Autoinhibition of the HECT-type ubiquitin ligase Smurf2 through its C2 domain . Cell 130 : 651 – 662 OpenUrl CrossRef PubMed Web of Science 80. ↵ Williams C , Dong KC , Arkinson C & Martin A ( 2023 ) The Ufd1 cofactor determines the linkage specificity of polyubiquitin chain engagement by the AAA+ ATPase Cdc48 . Mol Cell 83 : 759 – 769 .e7 OpenUrl CrossRef PubMed 81. ↵ Zhang S , Skalsky Y & Garfinkel DJ ( 1999 ) MGA2 or SPT23 is required for transcription of the delta9 fatty acid desaturase gene, OLE1, and nuclear membrane integrity in Saccharomyces cerevisiae . Genetics 151 : 473 – 483 OpenUrl Abstract / FREE Full Text 82. ↵ Zhang Z-R , Bonifacino JS & Hegde RS ( 2013 ) Deubiquitinases sharpen substrate discrimination during membrane protein degradation from the ER . Cell 154 : 609 – 622 OpenUrl CrossRef PubMed 83. ↵ Zhu K , Shan Z , Chen X , Cai Y , Cui L , Yao W , Wang Z , Shi P , Tian C , Lou J , et al. ( 2017 ) Allosteric auto-inhibition and activation of the Nedd4 family E3 ligase Itch . EMBO Rep 18 : 1618 – 1630 OpenUrl Abstract / FREE Full Text 84. ↵ Zhu L , Zhang Q , Cordeiro CD , Banjade S , Sardana R , Mao Y & Emr SD ( 2022 ) Adaptor linked K63 di-ubiquitin activates Nedd4/Rsp5 E3 ligase . Elife 11 View the discussion thread. Back to top Previous Next Posted November 06, 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 Ubiquitin-dependent signal amplification in lipid saturation sensing 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 Ubiquitin-dependent signal amplification in lipid saturation sensing Jona Causemann , Barbara Schmidt , Daniel Granz , Thorsten Mosler , Martin Jung , Ivan Dikic , Heiko Rieger , Robert Ernst bioRxiv 2025.11.05.686737; doi: https://doi.org/10.1101/2025.11.05.686737 Share This Article: Copy Citation Tools Ubiquitin-dependent signal amplification in lipid saturation sensing Jona Causemann , Barbara Schmidt , Daniel Granz , Thorsten Mosler , Martin Jung , Ivan Dikic , Heiko Rieger , Robert Ernst bioRxiv 2025.11.05.686737; doi: https://doi.org/10.1101/2025.11.05.686737 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 (7618) Biochemistry (17636) Bioengineering (13860) Bioinformatics (41847) Biophysics (21401) Cancer Biology (18536) Cell Biology (25424) Clinical Trials (138) Developmental Biology (13353) Ecology (19860) Epidemiology (2067) Evolutionary Biology (24287) Genetics (15583) Genomics (22463) Immunology (17701) Microbiology (40300) Molecular Biology (17141) Neuroscience (88434) Paleontology (666) Pathology (2825) Pharmacology and Toxicology (4813) Physiology (7633) Plant Biology (15107) Scientific Communication and Education (2042) Synthetic Biology (4285) Systems Biology (9808) Zoology (2268)