Hsp70s use ATP hydrolysis to disrupt protein-protein associations and to move macromolecules. One example is the Hsc70- mediated disassembly of the clathrin coats that form on vesicles during endocytosis. Here, we exploited the exceptional features of these coats to test three models—Brownian ratchet, power-stroke and entropic pulling—proposed to explain how Hsp70s transform their substrates. Our data rule out the ratchet and power-stroke models and instead support a collision-pressure mechanism whereby collisions between clathrin-coat walls and Hsc70s drive coats apart. Collision pressure is the complement to the pulling force described in the entropic pulling model. We also found that self-association augments collision pressure, thereby allowing disassembly of clathrin lattices that have been predicted to be resistant to disassembly. These results illuminate how Hsp70s generate the forces that transform their substrates.
Subscribe to Journal
Get full journal access for 1 year
only $18.75 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
Kim, Y.E., Hipp, M.S., Bracher, A., Hayer-Hartl, M. & Hartl, F.U. Molecular chaperone functions in protein folding and proteostasis. Annu. Rev. Biochem. 82, 323–355 (2013).
Sousa, R. & Lafer, E.M. The role of molecular chaperones in clathrin mediated vesicular trafficking. Front. Mol. Biosci. 2, 26 (2015).
Kityk, R., Kopp, J., Sinning, I. & Mayer, M.P. Structure and dynamics of the ATP-bound open conformation of Hsp70 chaperones. Mol. Cell 48, 863–874 (2012).
Qi, R. et al. Allosteric opening of the polypeptide-binding site when an Hsp70 binds ATP. Nat. Struct. Mol. Biol. 20, 900–907 (2013).
Zhuravleva, A., Clerico, E.M. & Gierasch, L.M. An interdomain energetic tug-of-war creates the allosterically active state in Hsp70 molecular chaperones. Cell 151, 1296–1307 (2012).
Misselwitz, B., Staeck, O. & Rapoport, T.A. J proteins catalytically activate Hsp70 molecules to trap a wide range of peptide sequences. Mol. Cell 2, 593–603 (1998).
Zuiderweg, E.R. et al. Allostery in the Hsp70 chaperone proteins. Top. Curr. Chem. 328, 99–153 (2013).
Sousa, R.J. Structural mechanisms of chaperone mediated protein disaggregation. Front. Mol. Biosci. 1, 12 (2014).
Rampelt, H. et al. Metazoan Hsp70 machines use Hsp110 to power protein disaggregation. EMBO J. 31, 4221–4235 (2012).
Iosefson, O., Sharon, S., Goloubinoff, P. & Azem, A. Reactivation of protein aggregates by mortalin and Tid1: the human mitochondrial Hsp70 chaperone system. Cell Stress Chaperones 17, 57–66 (2012).
Winkler, J., Tyedmers, J., Bukau, B. & Mogk, A. Chaperone networks in protein disaggregation and prion propagation. J. Struct. Biol. 179, 152–160 (2012).
Voisine, C. et al. The protein import motor of mitochondria: unfolding and trapping of preproteins are distinct and separable functions of matrix Hsp70. Cell 97, 565–574 (1999).
Matlack, K.E., Misselwitz, B., Plath, K. & Rapoport, T.A. BiP acts as a molecular ratchet during posttranslational transport of prepro-alpha factor across the ER membrane. Cell 97, 553–564 (1999).
Misselwitz, B., Staeck, O., Matlack, K.E. & Rapoport, T.A. Interaction of BiP with the J-domain of the Sec63p component of the endoplasmic reticulum protein translocation complex. J. Biol. Chem. 274, 20110–20115 (1999).
Sousa, R. & Lafer, E.M. Keep the traffic moving: mechanism of the Hsp70 motor. Traffic 7, 1596–1603 (2006).
De Los Rios, P., Ben-Zvi, A., Slutsky, O., Azem, A. & Goloubinoff, P. Hsp70 chaperones accelerate protein translocation and the unfolding of stable protein aggregates by entropic pulling. Proc. Natl. Acad. Sci. USA 103, 6166–6171 (2006).
Goloubinoff, P. & De Los Rios, P. The mechanism of Hsp70 chaperones: (entropic) pulling the models together. Trends Biochem. Sci. 32, 372–380 (2007).
Ungewickell, E. et al. Role of auxilin in uncoating clathrin-coated vesicles. Nature 378, 632–635 (1995).
Böcking, T., Aguet, F., Harrison, S.C. & Kirchhausen, T. Single-molecule analysis of a molecular disassemblase reveals the mechanism of Hsc70-driven clathrin uncoating. Nat. Struct. Mol. Biol. 18, 295–301 (2011).
Böcking, T. et al. Key interactions for clathrin coat stability. Structure 22, 819–829 (2014).
Rothnie, A., Clarke, A.R., Kuzmic, P., Cameron, A. & Smith, C.J. A sequential mechanism for clathrin cage disassembly by 70-kDa heat-shock cognate protein (Hsc70) and auxilin. Proc. Natl. Acad. Sci. USA 108, 6927–6932 (2011).
Van Jaarsveld, P.P., Nandi, P.K., Lippoldt, R.E., Saroff, H. & Edelhoch, H. Polymerization of clathrin protomers into basket structures. Biochemistry 20, 4129–4135 (1981).
Xing, Y. et al. Structure of clathrin coat with bound Hsc70 and auxilin: mechanism of Hsc70-facilitated disassembly. EMBO J. 29, 655–665 (2010).
Jiang, J., Prasad, K., Lafer, E.M. & Sousa, R. Structural basis of interdomain communication in the Hsc70 chaperone. Mol. Cell 20, 513–524 (2005).
Jiang, J. et al. Structure-function analysis of the auxilin J-domain reveals an extended Hsc70 interaction interface. Biochemistry 42, 5748–5753 (2003).
Jiang, J. et al. Structural basis of J cochaperone binding and regulation of Hsp70. Mol. Cell 28, 422–433 (2007).
Rapoport, I., Boll, W., Yu, A., Böcking, T. & Kirchhausen, T. A motif in the clathrin heavy chain required for the Hsc70/auxilin uncoating reaction. Mol. Biol. Cell 19, 405–413 (2008).
Benaroudj, N., Batelier, G., Triniolles, F. & Ladjimi, M.M. Self-association of the molecular chaperone HSC70. Biochemistry 34, 15282–15290 (1995).
Aprile, F.A. et al. Hsp70 oligomerization is mediated by an interaction between the interdomain linker and the substrate-binding domain. PLoS One 8, e67961 (2013).
Preissler, S. et al. Physiological modulation of BiP activity by trans-protomer engagement of the interdomain linker. eLife 4, e08961 (2015).
Kramer, E.M. & Myers, D.R. Osmosis is not driven by water dilution. Trends Plant Sci. 18, 195–197 (2013).
Kramer, E.M. & Myers, D.R. Five popular misconceptions about osmosis. Am. J. Phys. 80, 694–699 (2012).
Wilbanks, S.M., Chen, L., Tsuruta, H., Hodgson, K.O. & McKay, D.B. Solution small-angle X-ray scattering study of the molecular chaperone Hsc70 and its subfragments. Biochemistry 34, 12095–12106 (1995).
Ha, J.H. & McKay, D.B. ATPase kinetics of recombinant bovine 70 kDa heat shock cognate protein and its amino-terminal ATPase domain. Biochemistry 33, 14625–14635 (1994).
Northrup, S.H. & Erickson, H.P. Kinetics of protein-protein association explained by Brownian dynamics computer simulation. Proc. Natl. Acad. Sci. USA 89, 3338–3342 (1992).
Zanten, J.H.V. & Monbouquette, H.G. Characterization of vesicles by classical light scattering. J. Colloid Interface Sci. 146, 330–336 (1991).
Andréasson, C., Fiaux, J., Rampelt, H., Mayer, M.P. & Bukau, B. Hsp110 is a nucleotide-activated exchange factor for Hsp70. J. Biol. Chem. 283, 8877–8884 (2008).
Schuermann, J.P. et al. Structure of the Hsp110:Hsc70 nucleotide exchange machine. Mol. Cell 31, 232–243 (2008).
Liu, Q., D'Silva, P., Walter, W., Marszalek, J. & Craig, E.A. Regulated cycling of mitochondrial Hsp70 at the protein import channel. Science 300, 139–141 (2003).
Vilker, V.L., Colton, C.K. & Smith, K.A. The osmotic-pressure of concentrated protein solutions: effect of concentration and pH in saline solutions of bovine serum-albumin. J. Colloid Interface Sci. 79, 548–566 (1981).
Svoboda, K. & Block, S.M. Force and velocity measured for single kinesin molecules. Cell 77, 773–784 (1994).
Coy, D.L., Wagenbach, M. & Howard, J. Kinesin takes one 8-nm step for each ATP that it hydrolyzes. J. Biol. Chem. 274, 3667–3671 (1999).
Scheele, U., Kalthoff, C. & Ungewickell, E. Multiple interactions of auxilin 1 with clathrin and the AP-2 adaptor complex. J. Biol. Chem. 276, 36131–36138 (2001).
Fotin, A. et al. Structure of an auxilin-bound clathrin coat and its implications for the mechanism of uncoating. Nature 432, 649–653 (2004).
Ben-Zvi, A., De Los Rios, P., Dietler, G. & Goloubinoff, P. Active solubilization and refolding of stable protein aggregates by cooperative unfolding action of individual hsp70 chaperones. J. Biol. Chem. 279, 37298–37303 (2004).
Gao, X. et al. Human Hsp70 disaggregase reverses Parkinson's-linked α-synuclein amyloid fibrils. Mol. Cell 59, 781–793 (2015).
Mattoo, R.U., Sharma, S.K., Priya, S., Finka, A. & Goloubinoff, P. Hsp110 is a bona fide chaperone using ATP to unfold stable misfolded polypeptides and reciprocally collaborate with Hsp70 to solubilize protein aggregates. J. Biol. Chem. 288, 21399–21411 (2013).
Zhu, X. et al. Structural analysis of substrate binding by the molecular chaperone DnaK. Science 272, 1606–1614 (1996).
Brizzard, B.L., Chubet, R.G. & Vizard, D.L. Immunoaffinity purification of FLAG epitope-tagged bacterial alkaline phosphatase using a novel monoclonal antibody and peptide elution. Biotechniques 16, 730–735 (1994).
Ye, W. & Lafer, E.M. Clathrin binding and assembly activities of expressed domains of the synapse-specific clathrin assembly protein AP-3. J. Biol. Chem. 270, 10933–10939 (1995).
Mindell, J.A. & Grigorieff, N. Accurate determination of local defocus and specimen tilt in electron microscopy. J. Struct. Biol. 142, 334–347 (2003).
Abrishami, V. et al. A pattern matching approach to the automatic selection of particles from low-contrast electron micrographs. Bioinformatics 29, 2460–2468 (2013).
Scheres, S.H. RELION: implementation of a Bayesian approach to cryo-EM structure determination. J. Struct. Biol. 180, 519–530 (2012).
Scheres, S.H. & Chen, S. Prevention of overfitting in cryo-EM structure determination. Nat. Methods 9, 853–854 (2012).
de la Rosa-Trevín, J.M. et al. Xmipp 3.0: an improved software suite for image processing in electron microscopy. J. Struct. Biol. 184, 321–328 (2013).
Penczek, P.A. Three-dimensional spectral signal-to-noise ratio for a class of reconstruction algorithms. J. Struct. Biol. 138, 34–46 (2002).
Pettersen, E.F. et al. UCSF Chimera: a visualization system for exploratory research and analysis. J. Comput. Chem. 25, 1605–1612 (2004).
We thank L. Wang of UTHSCSA for technical assistance; T. Kirchhausen of HMS for CHC cDNA; C. Brantner and D. Sackett of NICHD, the NHLBI EM Core Facility for help with EM at an early stage of the work; and D. Sackett and R. Nossal at NICHD, D. Luque from ISCIII and J. Stachowiak from UT-Austin for helpful discussions. This work was supported by NS029051 (to E.M.L.) and GM118933 (to E.M.L. and R.S.); the NIH Intramural Research Program (NIBIB); and by grants BFU2013-44202 from the Spanish Ministry of Economy and Innovation and S2013/MIT-2807 from the Madrid Regional Government to J.M.V. H.-S.L. was supported in part by a scholarship from the Taiwan National Science Council (NSC103-2917-I-564-072).
The authors declare no competing financial interests.
Integrated supplementary information
Supplementary Figure 1 Standardization, specificity and controls for experiments measuring disassembly by light-scattering.
a: Standard curves for scattering by various concentrations of either clathrin cages (black squares) or triskelia (red circles) expressed as concentration of CHC. Error bars show +/- SD for 6 (cages) or 13 (triskelia) repetitions of one experiment each. Across the concentration ranges used in our experiments the amount of light scattering by either cages or triskelia is linear (R2 for both fits >0.99). Panels b-e: Removal of auxilin, ATP, Hsc70 ATP binding or hydrolysis activity, or deletion of the Hsc70 NBD greatly slows or abrogates Hsc70 mediated cage disassembly. b: Blue trace shows a complete reaction with 2 μM Hsc70 as in fig. 2A. Black trace shows an identical reaction with ATP omitted. Red trace shows a reaction in which the Hsc70 was replaced with 2 μM PBD (missing the NBD). c. Blue trace as in b; black=auxilin omitted; red=ATP and auxilin omitted; c: Blue as in a; black=ATPγS replaces ATP; red= ADP replaces ATP. Experimental: All reactions in b-d were carried out on the same day with the same set of cages, with each reaction condition repeated 4-11 times (trace thickness corresponds to +S.E.). The blue trace is identical in all plots but data for different reaction conditions was plotted separately for clarity as the reaction profiles otherwise overlap. e: Blue trace is complete reaction with WT Hsc70; red trace is with Hsc70 mutant that cannot bind ATP (G201E/G202D; Morgan, J.R. et al. A role for an Hsp70 nucleotide exchange factor in the regulation of synaptic vesicle endocytosis. J Neurosci 33, 8009 (2013)); black trace is for ATP hydrolysis deficient mutant (E175S; Jiang, J. et al. Structural basis of J cochaperone binding and regulation of Hsp70. Mol Cell 28, 422 (2007)). Trace thickness shows +/- s.e. for 8-15 repetitions of 1-3 independent experiments. Panels f, g: Rapid disassembly of cages in which the canonical Hsc70 binding site is moved is not due to Hsc70 binding at non-canonical sites. f: Light scattering vs. time plotted for disassembly reactions with 2 μM Hsc70 and either cages in which the Hsc70 binding site was moved by 25AA (red; average of 11 replicates with trace thickness giving +/- s.e.) or in which it was eliminated (black; average of 9 replicates with +25AA FLAG tag cages). g: As in f, but using 2 μM Hsc70ΔC (averages of 7 and 10 replicates, respectively). The residual Hsc70-mediated disassembly seen with cages in which the Hsc70 site is replaced with a FLAG-tag appears to be primarily due to Hsc70 binding weakly to other sites on the tails, as we found that complete deletion of the tails markedly slowed disassembly even further than simply removing the Hsc70 binding site by mutation (data not shown, available on request). h: Anti-FLAG Fabs can’t dissociate WT cages. WT cages reacted with 2 μM Hsc70 (black trace; 5 replicates averaged) or 2 μM Fab (red trace; 7 replicates averaged). i: Auxilin stabilizes clathrin cage against disassembly. +10AA FLAG-tag Clathrin cages with auxilin in 1.3x excess of CHC were reacted with anti-FLAG Fab as described in fig. 2K (black trace), or in the absence of auxilin (red trace). Effects on auxilin on cage stability and contributions to disassembly are difficult to parse as it also makes an essential contribution to disassembly by recruiting the Hsc70 to the cage. By using FLAG-tag cages and anti-FLAG Fab we can remove the recruitment contribution and measure just auxilin’s contribution to cage stability. It is seen that the Fab disassemble the FLAG-tag cages more slowly in the presence of auxilin than without it, indicating that auxilin binding stabilizes cages and makes them more resistant to disassembly.
Supplementary Figure 2 Tight coupling limits fitting of microscopic reaction parameters from ensemble experiments that measure disassembly by light-scattering.
A: Fit (magenta) of a reaction profile (blue) for one experiment using the ‘Simplefit’ script (Supplementary Information). The obtained values of the variables (ka=Hsc70 binding rate; kd=cage disassembly rate; A1=scattering amplitude of the transient Hsc70:cage complex) were used to calculate the overall rate of disassembly (kov) and the time to 50% reaction complete (t1/2) according to the equation: kov=1/(1/([Hsc70]*ka)+1/kd). B: As in A, but kd was fixed at 2x the value in A. Note the large differences in ka and A1 between panels A and B, but the very similar value for kov and t1/2 for both fits. While the fit in A is better, both fits are quite good. This is because a reaction characterized by “fast Hsc70 binding (ka), slow disassembly (kd) and a small scattering amplitude for the transient cage:Hsc70 complex (A1)” will generate a profile very similar to a “slow binding, fast disassembly, large transient scattering amplitude increase” reaction. In practice, we found that experimental variation was greater than the differences in the quality of fits shown above so that microscopic rate parameters like ka, kd or A1 could not be accurately determined from such data. However, the kov values were relatively insensitive to this coupling.
Supplementary Figure 3 Bivalent anti-FLAG antibodies are ineffective at disassembling FLAG-tag cages.
A. Light scattering vs. time in disassembly reactions with +0AA FLAG cages run as in fig. 2J but with the indicated concentrations of anti-FLAG IgG rather than Fab. Traces are averages from 7-9 repetitions. B: The same data as in A, but plotted with error bars giving +/- s.e. range for each experiment (error bars removed in A for clarity due to data overlap). C. Plot of the 1st 15 seconds of the reactions to better resolve the initial antibody binding step. Note that the amplitude of the scattering increase (~40%) is similar to that seen with Fab (fig. 2J-K), and is therefore consistent with bivalent binding of the antibody since monovalent binding would result in binding of similar numbers of Ab and Fab molecules and with a larger scattering increase for the former due to its larger MW. D: Gel analysis confirms that anti-FLAG IgG binds cages bivalently. Plot of IgG/CHC ratios vs. [IgG] (at 1, 3 and 9 μM; errors are +/- s.e. for n=3) shows that 0.45-0.65 IgG molecules per CHC bind to cages, while 0.8-1.2 Fab molecules per CHC are seen to bind in identical experiments shown in figure 5.
Supplementary Figure 4 The large Hsc70-induced increase in scattering at pH 6.0 is not due to cage aggregation, because the amplitude and rate of the scattering increase is insensitive to cage concentration.
Hsc70 could induce large increases in scattering by causing cages to aggregate. Since aggregation would be at least a 2nd-order reaction, its rate should be sensitive to cage concentration. A. Scattering vs. time in pH 6.0 disassembly reactions with [Hsc70] at 2 μM and clathrin cages at the indicated concentrations (expressed in molar concentration of CHC with data plotted representing averages +s.e. of 10, 19, 15 and 14 replicates at 38, 75, 150 and 300 nM, respectively) and with scattering normalized so that the starting level in the reaction with 150 nM cages is 1.0. B. Data from A replotted so that the starting level of scattering in all the reactions is normalized to 1.0.
Supplementary Figure 5 Binding of large numbers of Hsc70s per CHC in pH 6–stabilized cages requires ATP, auxilin and an Hsc70-binding site in the CHC tails.
Plot of the # of Hsc70s per CHC pelleting with cages in absence of ATP, auxilin or an Hsc70 binding site (FLAG-tag cages; error is +se for 2 analyses of one experiment). Residual Hsc70 binding to cages in which the Hsc70 binding site is replaced with a FLAG-tag appears to be primarily due to Hsc70 binding weakly to other sites on the tails, as we found that complete deletion of the tails reduced binding (as assessed by the amount of 70 that pellets with cages) even more (to a level not meaningfully different than seen in the absence of cages; data not shown, available on request) than does altering the Hsc70 binding site by mutation. The nature of the 'non-specific' sites at which 70's bind to such cages is unknown, but such binding is consistent with the known promiscuous binding specificity of Hsc70 and the conclusion (Bocking, T., Aguet, F., Harrison, S.C. & Kirchhausen, T. Single-molecule analysis of a molecular disassemblase reveals the mechanism of Hsc70-driven clathrin uncoating. Nat Struct Mol Biol 18, 295-301 (2011)), that, even with WT cages, approximately half of the Hsc70 molecules that associate with cages bind at 'non-specific' sites that are ineffective for disassembly.
Supplementary Figure 6 Hsc70ΔC competes for Hsc70 binding to cages and inhibits cage disassembly by Hsc70 at low pH.
Reduced disassembly activity and scattering increases with Hsc70ΔC vs. Hsc70 could be due to weaker binding of Hsc70ΔC to cages, rather than reduced self-association. If so then adding excess Hsc70ΔC to reactions with Hsc70 shouldn't inhibit disassembly by the latter since Hsc70ΔC shouldn't effectively compete with Hsc70 for cage binding. Conversely, if Hsc70ΔC and Hsc70 bind cages with similar affinity, but Hsc70ΔC doesn't disassemble as effectively once bound, then Hsc70ΔC should compete with and inhibit Hsc70 disassembly. To test this we carried out disassembly at pH 6.3 and 1 μM Hsc70, conditions under which disassembly is slow and depends on Hsc70 self-association. A. Cage disassembly reactions carried out at pH 6.3 with 1 μM Hsc70 and with 0 (black; average of 5 replicates), 3 μM (red; average of 8 replicates with trace thickness showing +/- se), or 9 μM (blue; average of 8 replicates) Hsc70ΔCterm. Hsc70ΔCterm at 3- and 9x excess over Hsc70 is observed to reduce disassembly rates by ~2x and ~4x, respectively. B. SDS PAGE analysis of pellet fractions of 1mM Hsc70, 0.23 μM Auxilin, 0.15 μM (lanes 1-6) or 0 μM (lanes 7-12) cages (concentrations of cages expressed as [CHC]) and with Hsc70ΔC at 1 (lanes 2, 8), 2 (lanes 3, 9), 4 (lanes 4, 10), 8 (lanes 5, 11) or 16 (lanes 6, 12) μM. Hsc70ΔC reduces the amount of Hsc70 that pellets with the cages. This experiment was independently performed 2x with similar results.
Supplementary Figure 7 Hsc70, but not Hsc70ΔC, disassembles cages and forms filamentous oligomers at pH 6.0.
Cages were mixed with ATP and either Hsc70ΔC or Hsc70 as described for cryo-EM, but then incubated for 30 minutes to allow disassembly to occur and then negatively stained for EM. In the Hsc70ΔC reaction, cages were still present against a background of excess Hsc70ΔC (panel A). In reactions with Hsc70 almost no cages could be seen (panel B), though clearly obvious triskelia were not visible because the spindly triskelia are likely obscured by Hsc70 molecules in 10-fold molar excess of CHC. Long Hsc70 oligomers were abundant (a higher magnification view of these is shown in panel C), with each unit of the oligomer estimated as large enough to accommodate 1-2 Hsc70 molecules. In panels A and B a model of a reconstructed cage is inset to provide a scale marker.
A. Left panel: QNM AFM at 4nm pixel resolution and 100 pN tip force of clathrin-auxilin cages and under buffer (pH 6.0) reveals characteristic polygonal surface structure. Top right panel: 3D height view of a field of clathrin-auxilin cages. Lower right panel: Nanoscale intra-cage compressibility profiles of the same field. The shared X-Y scale bar of 100 nm and Z-color bar representation are for the height (100 nm) and the compression (30 nm). The white threshold lines in the right panels demarcate 35 nm height level around each observed cage to identify cage tops for the computation of the observed maximal and mean compression for each cage. B. QNM AFM of fields of cages with the indicated associated proteins. The top panel plots cage height while the lower panel plots compression (shared Z scale of 100 and 50 nm for height & compression, respectively). Note the wide distribution of compression profiles in the +Hsc70 panel where cages display both very large and small compressions, while the distribution of compression sizes in the other panels is more uniform. C. Height distributions of cages with the indicated associated proteins (200 pN tip force). D. Left panel: Distribution of mean and maximum compression values for cages with the indicated proteins (200 pN tip force). Mean compression corresponds the average of the compression observed when each cage was sampled with either a 3x3 or 4x4 grid (9 or 16 measurements respectively, depending on cage size). Maximum compression corresponds to the largest compression measured during AFM probing of each cage.
About this article
Cite this article
Sousa, R., Liao, H., Cuéllar, J. et al. Clathrin-coat disassembly illuminates the mechanisms of Hsp70 force generation. Nat Struct Mol Biol 23, 821–829 (2016). https://doi.org/10.1038/nsmb.3272
Journal of Molecular Biology (2020)
International Journal of Molecular Sciences (2019)
Nature Reviews Molecular Cell Biology (2019)
Trends in Cell Biology (2019)
Nature Communications (2019)