Structure and mechanism of microtubule stabilization and motor regulation by MAP9

Structure and mechanism of microtubule stabilization and motor regulation by MAP9 | 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 Structure and mechanism of microtubule stabilization and motor regulation by MAP9 Burak Cetin , Aryan Taheri , Mert Golcuk , Brigette Y. Monroy , Jonathan Fernandes , Kassandra M. Ori-McKenney , Mert Gur , Eva Nogales , Ahmet Yildiz doi: https://doi.org/10.1101/2025.11.17.688911 Burak Cetin 1 Department of Molecular and Cell Biology, University of California , Berkeley, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Aryan Taheri 1 Department of Molecular and Cell Biology, University of California , Berkeley, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Mert Golcuk 2 Department of Computational and Systems Biology, University of Pittsburgh , Pittsburgh, PA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Brigette Y. Monroy 3 Department of Molecular and Cellular Biology, University of California , Davis, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jonathan Fernandes 4 Department of Chemistry, University of California , Berkeley, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Kassandra M. Ori-McKenney 3 Department of Molecular and Cellular Biology, University of California , Davis, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Mert Gur 2 Department of Computational and Systems Biology, University of Pittsburgh , Pittsburgh, PA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Eva Nogales 1 Department of Molecular and Cell Biology, University of California , Berkeley, CA, USA 5 Howard Hughes Medical Institute , Chevy Chase, MD, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site For correspondence: enogales{at}lbl.gov yildiz{at}berkeley.edu Ahmet Yildiz 1 Department of Molecular and Cell Biology, University of California , Berkeley, CA, USA 6 Physics Department, University of California , Berkeley, CA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site For correspondence: enogales{at}lbl.gov yildiz{at}berkeley.edu Abstract Full Text Info/History Metrics Supplementary material Preview PDF Abstract Microtubule-associated proteins (MAPs) regulate the organization of microtubules and control intracellular transport, but their individual contributions to microtubule dynamics and motor regulation remain poorly understood. Here, we identify MAP9 as a critical factor that stabilizes microtubules and facilitates neuronal morphogenesis. MAP9 knockdown abolishes the outgrowth of neurites, a phenotype not observed through the loss of other neuronal MAPs. Cryo-electron microscopy revealed that, unlike other MAPs that bind along protofilaments, MAP9 binds around the microtubule as a long alpha helix using five consecutive repeats. This unique binding mode enables MAP9 to staple adjacent protofilaments, thereby preventing microtubule depolymerization. We also showed that MAP9 selectively permits kinesin-3 motility while hindering kinesin-1 through interactions with a divergent loop-8 of their motor domains. Our results establish MAP9 as a key MAP required for neuronal growth and uncover how it differentially regulates intracellular transport driven by kinesin motors. Introduction The microtubule (MT) cytoskeleton is the architectural backbone of neurons and supports long-range intracellular transport driven by kinesin and dynein. Structural MAPs such as tau, MAP2, MAP7, and doublecortin exhibit differential localization across neuronal compartments, creating functional domains such as the axon initial segment 1 . Traditionally, MAPs have been viewed as essential stabilizers of neuronal MTs, providing structural support to axons and dendrites and protecting MTs from depolymerization 2 , 3 . However, recent studies have begun to challenge this long-standing view. Genetic depletion or knockout of individual MAPs, including tau, MAP1B, and even combinations of MAPs, often results in surprisingly mild or context-dependent phenotypes, without the widespread MT destabilization 4 – 7 . These observations led to the proposals that MAPs have overlapping functions in stabilizing MTs, such that the loss of one MAP can be compensated for by others 1 or MAPs serve regulatory roles in MT organization rather than acting as generic stabilizers 8 . These ideas have not been carefully tested, and it remains unclear if other MAPs play indispensable, non-redundant roles in driving early axonal extensions. MAP9, also known as Aster Associated Protein (ASAP), is a highly conserved MAP expressed in mitotic cells, cilia, and neurons 9 – 13 . MAP9 stabilizes and organizes the MT network in these cell lines. Depletion of MAP9 causes defects in mitotic spindle organization, ultrastructural defects in axonemes, and embryonic lethality or severe malformations during neural development 9 – 13 .These findings hinted that MAP9 may play an essential, non-redundant role in MT organization across diverse developmental contexts. However, MAP9 is understudied compared to other MAPs, and the molecular basis for its contributions to MT regulation remains poorly understood. MAP9 is known to differentially regulate kinesin motors involved in intracellular cargo transport. Unlike MAP7, which activates kinesin-1 motility 6 , 14 , 15 and inhibits kinesin-3 16 , MAP9 has the opposite effect, inhibiting kinesin-1 while enabling kinesin-3 motility 16 . The mechanism by which MAPs distinguish between kinesin motors that are closely related is still emerging. Functional studies demonstrated that MAP7 selectively activates kinesin-1 by interacting with the coiled-coil stalk of the motor 6 , 14 – 16 . However, MAP9 does not appear to strongly interact with either of these two kinesins, and it remains unclear how it can distinguish between different kinesin motors. Here, we identify MAP9 as a critical regulator of early neuronal growth and intracellular transport, acting through a distinct MT-binding mode and motor-selectivity mechanism. Using cryo-electron microscopy (cryo-EM), we show that MAP9 binds MTs through a repetitive motif that engages up to five adjacent protofilaments simultaneously, forming a lattice-stabilizing belt that suppresses MT catastrophe. In neurons, MAP9 knockdown abolishes neurite extension, highlighting its essential, non-redundant role in early neuronal morphogenesis. We further show that MAP9 selectively enables kinesin-3-driven transport by interacting with loop-8 of the kinesin motor domain. Insertion of loop-8 of kinesin-3 enabled kinesin-1 to walk uninhibited by MAP9, demonstrating that MAP9 differentially regulates transport kinesins through a divergent loop within the kinesin motor domain. Together, these findings reveal MAP9 as a unique neuronal MAP that couples MT stabilization with selective motor regulation to promote early neuronal development. Results MAP9 binds across five protofilaments on the MT We expressed full-length (FL) human MAP9, which contains a disordered projection domain (PD) and an MT binding domain predicted to form a 200-residue-long α-helix (MTBD; Fig. 1a ) 9 . Purified MAP9 is a monomer in solution and decorates MTs with a dissociation constant (K D ) of 61 ± 10 nM (±SE; Fig. 1b-d ). To determine the structure and MT binding mode of MAP9, we decorated MTs with excess FL MAP9 and obtained a cryo-EM reconstruction using the pseudo-helical symmetry of MTs ( Fig. 1e , Extended Data Figs. 1 - 2 , Table S1 ). The reconstruction revealed that MAP9 binds to MTs through a single extended α-helix, consistent with the structure prediction of its MTBD. Unlike MAP2, MAP4, MAP7, and tau, which bind along a single protofilament 15 , 17 – 19 , MAP9 binds across multiple protofilaments encircling the MT ( Fig. 1e ). Symmetry expansion of two laterally adjacent tubulins revealed how an 81-residue-long segment of α-helix within MAP9 MTBD binds to β-tubulin across the two neighboring tubulin dimers near the longitudinal interdimer interface ( Fig. 1f ). Download figure Open in new tab Fig. 1. MAP9 binds across up to five protofilaments around the MTs. a, Domain schematic showing that MAP9 contains an N-terminal PD and a C-terminal MTBD. b, MT binding of fluorescently labeled MAP9 in 100 mM salt. c, The fluorescence intensity (mean ± s.d.) of MAP9 was fit to the Langmuir isotherm (solid curve) to calculate K D (± SE; n = 30 MTs for all data points; two technical replicates). d, Mass photometry of purified MAP9 (mean ± s.d.). The red dashed line represents the expected monomer mass. e, Cryo-EM reconstruction of a MAP9-decorated MT obtained by imposing pseudo-helical symmetry. α-tubulin, β-tubulin, and MAP9 are shown in green, light blue, and royal blue, respectively. f, (Top) End-on view of the MAP9-MT cryo-EM reconstruction obtained by symmetry expansion using two adjacent protofilaments. (Bottom right) Alignment of the five MT binding repeats within MAP9 MTBD. (Bottom left) The fitted atomic model is shown as ribbons with all-atom representations of the side chains involved in hydrophobic (green) and ionic (magenta) interactions between the R1 pseudo-repeat of MAP9 (regular font) and β-tubulin (italic). g, Ribbon representation for the model of the MAP9 helix bound to five adjacent protofilaments through its pseudorepeats. h, MT decoration of 100 nM wild-type (WT) and mutant MAP9 constructs. i, The fluorescence intensity of WT and mutant MAP9 on MTs (From left to right, n = 23, 22, 22, 36, 22, 22 MTs). R indicates all three conserved residues were mutated in the specified pseudo-repeat (ND: not determined; see Extended Data Fig. 6a ). The interaction of MAP9 with β-tubulin involves a newly identified and well-conserved repeating sequence ФXXWXXXK, where Ф is mostly tyrosine or phenylalanine, W is tryptophan, K is lysine, and X is any amino acid ( Fig. 1f , Extended Data Figs. 3 - 4 ). Ф and W form hydrophobic interactions with F389, and K forms a salt bridge with E412 of β-tubulin ( Fig. 1f ). This pseudo-repeat sequence is present five times along MAP9 MTBD, with 35 or 36 residue spacers that allow the repeats along the helix to reach across neighboring protofilaments ( Fig. 1f-g ). Therefore, the 300 Å α-helix of MAP9 can bind up to five protofilaments, and three MAPs can cover the full circumference of the MT. All-atom molecular dynamics (MD) simulations (1 µs in total length , Extended Data Table 1 ) confirmed that all five repeats of MAP9 bind identical residues in β-tubulin as they bridge neighboring protofilaments ( Extended Data Fig. 5 , Video 1 ). In addition to the pairwise interactions that we identified from the cryo-EM reconstruction, simulations detected hydrogen bonding between conserved W and E412 of β-tubulin, hydrophobic interactions of MAP9 with M406 of β-tubulin, and transient electrostatic interactions with negatively charged residues in the α-tubulin’s C-terminal tail (CTT; Extended Data Fig. 5 ). To understand how the identified repeats and pairwise interactions affect the MT binding of MAP9, we generated mutations in the binding repeats of a MAP9 construct containing only the MTBD (MAP9-MTBD, Extended Data Fig. 6a ). Alanine substitutions or charge reversal of the conserved residues within the pseudo-repeats almost fully eliminated MAP9 binding to surface-immobilized MTs ( Fig. 1h-i ). Mutating the three conserved residues in only two pseudo-repeats in varying combinations also substantially (∼80%) reduced MT binding ( Fig. 1i , Extended Data Fig. 6a-b ), demonstrating that all repeats contribute to MT binding of MAP9. In comparison, mutating the residues that contact αCTT did not affect MT binding ( Extended Data Fig. 6c-e ), likely because these interactions are relatively infrequent (21 ± 5% of the simulation time, Extended Data Fig. 5 ) and thus contribute less to the overall MT affinity of MAP9. MAP9 is essential for neurite growth The unique binding of MAP9 across protofilaments suggests a strong effect of this MAP on MT dynamics. To test this possibility, we monitored how MAPs regulate MT polymerization and shrinkage in the presence of GTP and free tubulin in vitro. We observed that MAP9 substantially promotes MT polymerization by doubling the MT growth rate ( Fig. 2a-b , Video 2 ), similar to tau and MAP7 20 , 21 . Remarkably, MAP9 also nearly eliminated MT catastrophe ( Fig. 2b ), consistent with the bridging of neighboring protofilaments by this MAP. In comparison, MAP7 and tau had little to no effect on rescuing MTs against catastrophes. These results contrast with earlier reports of the MT stabilization role of tau and MAP7 2 , 22 , 23 , and are more consistent with recent studies showing that tau does not significantly affect MT stabilization in neurons 24 . Download figure Open in new tab Fig. 2. MAP9 promotes polymerization and suppresses the catastrophe of dynamic MTs. a, Kymographs of dynamic MTs in the presence or absence of 300 nM MAP7, tau, or MAP9. b, The catastrophe frequency and growth rate of dynamic MTs in the presence and absence of MAPs (mean ± s.d.; n = 30 MTs for all data points; two technical replicates). c, Representative kymographs of MT depolymerization in the presence and absence of MAPs. MTs were polymerized from GMP-CPP seeds (magenta). MT extensions and MAPs are shown in green and cyan, respectively. d, Depolymerization rates of MTs after removing free tubulin and GTP in the flow chamber in the presence and absence of MAPs (from left to right, n = 57, 98, 75, 55 MTs, two technical replicates). e, Immunostaining of primary cultured neurons with antibodies against MAP9 (magenta) and alpha-tubulin (green). f, Confocal image of DIV4 neurons transfected with MAP9 shRNA. Arrows show the MAP9 knockdown neuron (green) has only one neurite and is positive for tubulin (blue), but not for MAP9 (pink). g, The number of neurite processes extending from the cell body of DIV4 neurons in control and MAP9 shRNA-transfected neurons (n = 12 and 8 neurons, respectively, from 2 independent trials). In b and d, the centerline and whiskers represent mean and SD, respectively. In b and g, p-values were calculated from Welch’s two-tailed t-test. To directly observe whether these MAPs protect MTs against depolymerization, we removed free tubulin and GTP but kept the MAPs in the assay chamber after growing dynamic MTs ( Fig. 2c ). In the absence of MAPs, MTs depolymerized at 0.3 µm s - 1 , as previously reported 25 . MAP9 and tau slowed the depolymerization rate by 6-fold and 2.5-fold, respectively, whereas MAP7 failed to slow MT depolymerization ( Fig. 2d ). These results show that, among the MAPs we tested, MAP9 was most effective in stabilizing MTs and protecting them against depolymerization. We next tested whether MAP9 is essential for MT stability in neurons. Immunostaining in primary neuronal cultures revealed that MAP9 is distributed throughout neuronal extensions and is enriched in both axons and dendrites ( Fig. 2e , Extended Data Fig. 7a ), supporting its role in shaping polarized MT networks 16 . MAP9 staining appears weaker than that of tubulin in the soma, suggesting that MAP9 does not bind to soluble tubulin and instead decorates polymerized MTs ( Fig. 2e ). Remarkably, the shRNA knockdown of MAP9 nearly fully eliminated neurite outgrowth, as evidenced by a large reduction in the number of extended neurites in shRNA-positive neurons compared to non-transfected or control neurons ( Fig. 2f-g , Extended Data Fig. 7b ). These results indicate that MAP9 is a key regulator of neuronal growth through its distinct MT-binding mode that robustly stabilizes MTs. MAP9 differentially regulates kinesin motors We next investigated how MAP9 binding to MTs affects the motility of MT motors in vitro. We expressed a constitutively active human kinesin-1 KIF5B (K560, amino acids 1-560), human kinesin-2 KIF3A homodimer 26 , FL human kinesin-3 KIF1A, and mammalian dynein/dynactin/BicDR1 (DDR complex 27 ; Fig. 3a ). Motility assays revealed that MAP9 increases the frequency of kinesin-3 runs on MTs at low concentrations (<60 nM) up to 2-fold 16 , but higher concentrations of MAP9 slightly inhibited the motility ( Fig. 3b-c , Extended Data Fig. 8 , Video 3 ). Similar results were obtained for tail-truncated kinesin-3 ( Extended Data Fig. 9 ), indicating that MAP9 regulates kinesin-3 motility through its motor domain. Similarly, MAP9 did not affect dynein motility at low concentrations but reduced its run frequency at higher concentrations (Video 4) . In comparison, kinesin-1 and kinesin-2 motility were completely inhibited even at low concentrations of MAP9 (Video 5) , consistent with an increase in kinesin-2 velocity upon knocking out MAP9 in C. elegans 12 . To understand why higher concentrations of MAP9 inhibit all motors, we performed motility experiments using MAP9-MTBD, which lacks the majority of the PD ( Fig. 3d ) . MAP9-MTBD did not inhibit kinesin-3 or dynein motility at higher concentrations, while still inhibiting kinesin-1 motility as strongly as FL MAP9 ( Fig. 3d , Extended Data Fig. 10 , Video 6) . Therefore, the MTBD of MAP9 inhibits kinesin-1 while allowing kinesin-3 and dynein to walk on MTs. At high surface density, the negatively charged PD of MAP9 may screen the MT surface against motor binding when this MAP decorates the MT. Download figure Open in new tab Fig. 3. MAP9 differentially regulates kinesin and dynein motors. a, Schematic of single molecule motility assays of kinesin-1 (KIF5B), kinesin-2 (KIF3A), kinesin-3 (KIF1A), and dynein (DDR) motors on MAP9-decorated MTs. The tail of KIF5B was truncated to generate a constitutively active motor (K560). b, Kymographs of K560, KIF3A, KIF1A, and DDR motility in the presence and absence of MAP9. c, Run frequency of K560, KIF3A, KIF1A, and DDR motors under different MAP9 concentrations (from left to right, n = 18, 19, 18, 22, 9, 12, 12, 15, 24, 19, 20, 29, 18, 19, 16, 18, 12, 13, 13, 10, 11, 12, 11 MTs). d, Kymographs and run frequency of KIF1A, K560, and DDR under different MAP9-MTBD concentrations (From left to right, n = 24, 17, 12, 22, 21, 19, 16, 18, 17, 15, 15, 19, 19, 13, 23, 18, 17 MTs). e, Run frequency of KIF1A and K560 under different concentrations of chimeras MAP7-MTBD/MAP9-PD (from left to right, n = 18, 12, 9, 24, 22, 25, 21 MTs). f, Run frequency of FL KIF5B and KIF1A under different concentrations of MAP9-MTBD/MAP7-PD (from left to right, n = 15, 28, 15, 21, 24, 24, 21, 23, 24 MTs). In c-f, the centerline and whiskers represent mean and SD, respectively. In b and d, the scale bars are 5 µm and 5 s in the x and y axes, respectively. To gain insight into the differential regulation of kinesin-1 and kinesin-3 by MAP9, we generated chimeric MAPs by fusing the MTBDs and PDs of MAP9 and MAP7 16 . In contrast to MAP9, MAP7 inhibits kinesin-3 while allowing kinesin-1 to walk on MTs ( Extended Data Fig. 11 ) 6 , 14 , 15 . The MTBD of MAP7 overlaps with the kinesin binding site on tubulin, while its PD relieves kinesin-1 autoinhibition, acting as a tether to improve its landing on MTs 6 , 14 , 15 . A chimera of the MAP9 PD and MAP7 MTBD completely inhibited both motors ( Fig. 3e , Extended Data Fig. 12 ) , because this chimeric MAP overlaps with the motor binding site and lacks an activating tether for kinesin-1. In comparison, a chimera of MAP9 MTBD and MAP7 PD does not overlap with the kinesin binding site and contains an activating tether for kinesin-1 ( Fig. 3f ). A low concentration of this chimeric MAP activated FL kinesin-1 motility, similar to that observed for MAP7 15 , while it did not substantially inhibit kinesin-3 motility, even at high concentrations similar to MAP9 ( Fig. 3f , Extended Data Fig. 12 ) 16 . These results indicate that MAP9 differentially regulates kinesin-1 and kinesin-3 motility through its MTBD, not its PD. MAP9 distinguishes kinesins through their loop-8 It remained unclear how MAP9 could distinguish between kinesin-1 and −3, because these motors bind to MTs through a well-conserved motor domain within the kinesin family 28 . Docking the motor domains of these motors onto our map did not show an apparent steric clash with the MAP9 helix ( Extended Data Fig. 13 ). To gain insight into how MAP9 regulates kinesins without an overlap on MTs, we obtained the cryo-EM structure of MTs co-decorated with MAP9 MTBD and the kinesin-3 motor domain and performed 3D focused classification of the symmetry-expanded map of tubulin dimers ( Fig. 4a-b , Extended Data Figs. 2 and 14 ) . Unlike MAP7 MTBD, which precludes kinesin binding 15 , MAP9 binding allows kinesin-3 binding on the same tubulin dimer ( Fig. 4b ) . Download figure Open in new tab Fig. 4. MAP9 distinguishes between kinesin-1 and −3 through loop8 of the kinesin motor domain. a, Cryo-EM reconstruction with pseudo-helical symmetry imposed for an MT co-decorated with MAP9 and kinesin-3 motor domain. α-tubulin, β-tubulin, MAP9, and kinesin-3 are shown in green, light blue, royal blue, and purple, respectively. b, Surface representation and percentage of the two classes resulting from symmetry expansion and 3D classification on a single tubulin dimer and bound kinesin-3 and MAP9 segment. c, Atomic model from the cryo-EM reconstruction of MAP9 and kinesin-3 bound to the same tubulin. d, (Left) Interactions between R2 of MAP9 and loop-8 of kinesin-1 (top) and kinesin-3 (bottom) in MD simulations. Hydrophobic, negative, and positive side chains are highlighted in green, red, and blue, respectively. (Right) The frequencies of attractive and repulsive interactions between kinesin and MAP residues are color-coded based on interaction frequency. The red arrow highlights the loop-8b residues used to generate the chimeric K560 L8swap motor. e, Kymographs (top) and run frequency (bottom) of the chimeric K560 L8swap in the presence and absence of MAP9 (from left to right, n = 15, 21, 22 MTs). The centerline and whiskers represent the mean and SD, respectively. The blue dashed curve is the run frequency of wild-type K560 for comparison. The model built into the cryo-EM map showed that the MAP9 helix is positioned near loop-8 of the kinesin-3 motor domain ( Fig. 4c ) . Loop-8 of kinesin-3 consists of hydrophobic residues, whereas the equivalent region of kinesin-1 consists of positively charged residues ( Extended Data Fig. 15 ) . To test the possibility that differences in loop-8 between kinesin-1 and kinesin-3 are critical for their selective regulation by MAP9, we ran MD simulations of kinesin-1 and −3 motor domains interacting with each repeat of MAP9-MTBD on the MT (10 µs total simulation time). We identified attractive and repulsive interactions between MAP9 and the two kinesins by calculating positive and negative correlations of the interatomic distances of their amino acid C α atoms 29 . In kinesin-3, loop-8 formed favorable hydrophobic and ionic interactions with MAP9 ( Fig. 4d , Extended Data Fig. 16 , Video 7) . In comparison, kinesin-1 loop-8 did not form hydrophobic interactions and exhibited increased negative correlations due to electrostatic repulsion from similarly charged residues in MAP9 ( Fig. 4d , Extended Data Fig. 17 , Video 8) . To test these observations, we substituted five residues of loop-8 of kinesin-1 with those of kinesin-3 (K560 L8swap; Fig. 4e ). This modification did not affect kinesin-1 motility on undecorated MTs, but significantly reduced its inhibition by MAP9 ( Fig. 4e , Extended Data Fig. 18 ), in agreement with our model that MAP9 differentially regulates kinesin-1 and −3 via the differences between their divergent loop-8 regions. Discussion Our findings establish MAP9 as a structurally and functionally distinct MAP that not only stabilizes the MT cytoskeleton through a novel binding architecture but also encodes specificity in intracellular transport. While tau and MAP7 bind MTs along protofilaments 15 , 18 , MAP9 adopts a unique circumferential binding mode, spanning up to five adjacent protofilaments. This configuration effectively “staples” the protofilaments together, dramatically promoting MT growth and suppressing MT catastrophe compared to tau and MAP7. Previous studies suggested that lateral contacts between protofilaments play a more significant role than longitudinal contacts in resisting depolymerization 30 – 32 . Our results provide direct evidence supporting this model. The ability of MAP9 to staple adjacent protofilaments likely prevents protofilament curvature and peeling 30 , 32 , thereby protecting MTs against depolymerization more effectively than those MAPs that stabilize longitudinal contacts within single protofilaments 15 , 18 . Consistent with this mechanistic difference, we show that depletion of MAP9 results in complete arrest of neurite outgrowth. In comparison, knockdown of tau or MAP7 induces axonal branching and dendritic phenotypes in mice, but these neurons were able to develop and extend both dendrites and axons without major perturbations to the overall morphology 4 – 7 . The contrast suggests that not all MAPs contribute equally to lattice stability in vivo and highlights MAP9 as an essential and nonredundant factor required for sustaining the structural framework that supports neurite extension. More broadly, these results challenge the long-held view that MAPs act as functionally interchangeable stabilizers and instead point to a division of labor in which the specific binding geometry of individual MAPs dictates their ability to preserve axonal and dendritic MTs. These insights position MAP9 as a key regulator of neuronal architecture and a promising target for future studies into the mechanisms of neuronal development. Our results also demonstrate how MAP9 exerts selective control over kinesin-1 and kinesin-3 driven transport without overlapping with their MT footprints. Previous functional assays attributed this selectivity to the presence of a tandem repeat of lysine residues within loop-12 of the kinesin-3 motor domain 16 , but the underlying mechanism remained unclear. Our cryo-EM studies revealed that the MAP9 helix does not directly overlap with the binding sites of dynein and kinesin motors at the intradimer interface of α- and β-tubulin. Instead, MAP9 is positioned at the plus-end-directed side of kinesin, facing towards loop-8 of the kinesin motor domain. Loop-8 exhibits sequence divergence among kinesins and has been identified as a determinant of MT binding and processivity of kinesin-3 33 , 34 . Simulations and functional assays revealed that MAP9 forms attractive interactions with the loop-8 of kinesin-3, whereas its interactions with the loop-8 of kinesin-1 are repulsive, establishing the mechanistic basis of how MAP9 distinguishes between these motors despite their similarities. Substituting these residues in kinesin-1 with the corresponding kinesin-3 residues prevents inhibition by MAP9, showing that MAP9 inhibits kinesin-1 motility by recognizing its loop-8 sequence. Our work introduces a new regulatory principle for how MAPs recognize evolutionary divergent sequences of kinesin motors to gate their transport. Sequence divergence in these loops contributes to differences in autoregulation 35 and force generation 36 of kinesin motors, as well as to their class-specific functions 28 , such as MT depolymerization 37 , 38 . Our findings add a new layer to this regulatory code, showing that loop divergence can also serve as a recognition module for kinesin motors by MAPs. Author contributions B.C., A.T., E.N., and A.Y. conceived the project and analyzed the data. B.C. purified the proteins and performed single-molecule experiments. A.T. performed cryo-EM sample preparation, data collection, and analysis. B.Y.M. and K.M.O.M. performed studies in cultured neurons. M. Golcuk and M. Gur performed MD simulations and analysis. B.C., A.T., M. Golcuk, K.M.O.M., M. Gur, E.N., and A.Y. wrote the manuscript. Competing interests The authors declare no competing interests. Data and materials availability Materials are available from A.Y. under a material agreement with the University of California, Berkeley. The coordinates for MAP9 bound to tubulin, and MAP9 and the KIF1A motor domain bound to tubulin are available at the Protein Data Bank (PDB) with accession codes XXX and XXX, respectively. All cryo-EM maps are available at the EMDB with accession codes XXX (info will be filled before final acceptance). Code availability Scripts for fluorescence image processing have been deposited at GitHub ( https://github.com/Yildiz-Lab/YFIESTA ). Scripts for cryo-EM image processing are available from E.N. upon request. Methods Plasmids and Construct Design The plasmid contained the FL KIF1A sequence with a C-terminal mScarlet fusion in a pOmniBac backbone was received from R. McKenney, UC Davis 39 . This tag was replaced with a yBBr tag using Gibson assembly for organic dye labeling. The DNA sequence that expresses mouse kinesin-2 KIF3A motor domain and neck linker, dimerized through kinesin-1 coiled-coil was obtained from AddGene (Plasmid #129768). This sequence was cloned into a pET28(a) + vector with a C-terminal strepII and yBBr tag for purification and organic dye labeling, using Nde1 and XhoI cut sites. Human KIF1A (1-393) fused to an artificial dimerization leucine zipper tag, ybbR and Twin-Strep tag were cloned into a pET28(a)+ vector using Gibson assembly. E.coli expression vectors for WT and mutant MAP9-MTBD constructs with C-terminal GFP and StrepII tags, chimeric MAP7/MAP9 constructs with N-terminal GFP and C-terminal StrepII tags, and monomeric KIF1A (residues 1-393) with an N-terminal mTurquoise2 and C-terminal StrepII tag were ordered from Twist Biosciences. Briefly, the open reading frame was cloned into the pET28a(+) vector using Nde1 and XhoI cut sites, which introduces a cleavable His-tag into the N-terminus of the expressed protein. The K560 L8swap construct with a C-terminal mScarlet fusion was cloned into the pET29b vector using the same restriction sites. The constructs used in this study are listed in Extended Data Table 1 . Protein Expression and Purification The plasmid for expression of FL human KIF1A with the pOmniBac backbone was transformed into DH10Bac competent cells, followed by plating onto Bacmid plates with BluoGal at 37 °C for 2 days under antibiotic selection. A white colony was selected and grown in LB media overnight. Bacmid plasmids were isolated and transfected into adherent SF9 cells using Lipofectamine. The transfected SF9 cells were incubated at 27 °C for 3 days to grow the initial virus (P1). 2 mL of the P1 virus was added to 50 mL suspended SF9 cell culture and incubated at 27 °C in a shaking incubator for 3 days. The P2 virus was collected by centrifuging at 4,000 g for 10 min and stored at 4 °C in the dark for several months. The amplified P2 virus was then used to infect a 2 L suspension SF9 culture, and the protein was expressed for 72 h in serum-containing media. The cells were harvested via centrifugation at 4,000 g , and the resulting pellet was snap-frozen in liquid nitrogen. The cell pellet was stored at −80 °C. To purify the protein, the cell pellet was thawed and resuspended in SFP buffer (50 mM HEPES-KOH, pH 7.4, 300 mM KCl, 10 mM MgCl 2 , 10% glycerol). The resuspended cells were lysed by a dounce homogenizer. The lysate was cleared via centrifuging at 150,000 g for 45 min in a Ti70 rotor (Beckman Coulter). The clarified lysate was then applied to 1 mL of Strep-Tactin XT resin equilibrated in the lysis buffer using gravity flow with disposable columns. The resin containing the Strep-tag containing protein was then washed with 20 column volumes of SFP buffer. The FL KIF1A was then fluorescently labeled with LD655-CoA dye (Lumidyne) using SFP synthase and SFP buffer. Briefly, a 2 mL solution containing 50 nmol dye and 10 µM SFP synthase in SFP buffer was passed through the column, which was then capped at both ends and incubated overnight at 4 °C. The resin was then washed with 10 column volumes of SFP buffer to wash away the free dye. The protein was then eluted with SFP buffer containing 50 mM d-biotin. Labeled aliquots of FL KIF1A were flash-frozen and stored at −80 °C. Human KIF1A-LZ (residues 1-393) and mouse kinesin-2 KIF3A homodimer were expressed in E. coli BL21 (DE3) cells. Briefly, a 2L culture was grown in a TB medium to an OD 600 of 0.5, and protein expression was induced by adding 0.5 mM IPTG at 16 °C for 18 h. These proteins were purified and labeled identically as FL KIF1A, except the cells were lysed via tip sonication after harvesting. Human MAP9 fused to GFP and Twin-Strep tag (a gift from R. McKenney, UC Davis) and MAP9-MTBD were expressed in Rosetta (DE3) cells. Briefly, a 2 L culture was grown in TB medium to an OD 600 of 0.5, and protein expression was induced by adding 0.5 mM IPTG at 16 °C overnight. The protein was purified via the Strep-II tag without overnight labeling and proceeded directly to elution after washing the resin with the lysis buffer. For structural studies, the eluate was then diluted threefold using SFP buffer and applied to a heparin HP column (Cytiva). The column was washed with an SFP buffer containing 100 mM KCl, followed by stepwise elution with an SFP buffer under increasing KCl concentration from 100 to 500 mM. The eluate was concentrated and further purified via gel filtration to remove other contaminants on a Superdex 200 column. Pig-brain dynactin and the cargo adaptor mouse BicDR1 were purified as described previously 27 . The mutant of human dynein that cannot form the autoinhibited phi conformation (Phi mutant) was purified and fluorescently labeled as previously described 40 . The LD655-labeled single-use frozen aliquots were stored at −80 °C and used after thawing the same day. C-terminally His-tagged human KIF5B (K560) was purified as described previously 41 . Sample Preparation for Light Microscopy Plain glass coverslips were cleaned with water, acetone, and water by sonication for 10 min, followed by sonication in a bath sonicator for 40 min in 1 M KOH. The coverslips were then rinsed with ddH 2 O, incubated in 3-Aminopropyltriethoxysilane in acetate and methanol for 10 min while sonicating for 1 min between successive steps, further cleaned with methanol, and finally air-dried. 30 µl of 25% biotin-PEG-succinimidyl valerate in a NaHCO 3 buffer (pH 7.4) was sandwiched between two coverslips, followed by incubation at 4 °C overnight. The coverslips were then cleaned with water, air-dried, vacuum-sealed, and stored at −20 °C. Flow chambers were built by sandwiching a double-sided tape with a PEG-coated coverslip and a glass slide. Light Microscopy The fluorescent imaging was performed with a custom-built multicolor objective-type TIRF microscope equipped with a Nikon Ti-E microscope body, a 100X magnification 1.49 N.A. apochromatic oil-immersion objective (Nikon) with a Perfect Focus System. The fluorescence signal was detected using an electron-multiplying charge-coupled device camera (Andor, Ixon EM+, 512 × 512 pixels). The effective camera pixel size after magnification was 160 nm. Alexa488/GFP/mNeonGreen, LD555, and LD655 probes were excited using 488 nm, 561 nm, and 633 nm laser beams (Coherent) coupled to a single mode fiber (Oz Optics), and their emission was filtered using 525/40, 585/40, and 697/75 bandpass filters (Semrock), respectively. The microscope was controlled using Micromanager 1.4. The flow chambers were incubated with a 5 mg mL −1 streptavidin solution for 3 min and washed with BRB80 buffer (80 mM PIPES pH 6.8, 1 mM EGTA, 2 mM MgCl 2 ). The total salt concentration of the PIPES buffer after pH adjustment was approximately 100 mM. Assays were performed in the absence of additional salt. Biotinylated MTs were then added to the flow chamber and incubated for 3 min. The optimal concentration of biotinylated MTs was determined empirically. The chamber was then washed with BRB80 buffer supplemented with 1 mg mL - 1 casein, 0.5% pluronic acid. Kinesins, motors, and MAP proteins were diluted in BRB80 buffer supplemented with 1 mg mL - 1 casein, 0.5% pluronic acid. Motors were diluted in BRB80 supplemented with 1 mg mL - 1 casein, 0.5% pluronic acid, and 2 mM ATP. Compared to KIF1A-LZ, a higher motor concentration (40X) of FL KIF1A was flown into the chamber to dimerize and activate the motor. Movies were recorded for ∼2-4 min at 200 ms per frame. Motors and ATP were not added for experiments measuring the fluorescence intensity of MAPs on MTs. For DDR motility, 200 nM LD655-labeled phi-mutant dynein was incubated with 200 nM pig brain dynactin and 350 nM mouse BicDR1 in BRB80 buffer supplemented with 1 mg mL - 1 casein, 0.5% pluronic acid, and 2 mM ATP. The complex was incubated for 15 min at room temperature and diluted 25-fold. The diluted dynein was added to the flow chamber and imaged on the microscope for 10 mins in the presence or absence of MAP9. Dynamic MT imaging Biotinylated GMPCPP-labeled seeds were generated, as described previously 42 . Pig brain tubulin (30 mg mL - 1 ) was mixed with 2% Cy5-labeled tubulin and centrifuged at 4 °C at 100,000 rpm using a TLA 120.1 rotor (Beckman). The supernatant was kept on ice and used for MT polymerization in a flow chamber. Flow chambers with a biotin-PEG coverslip were preincubated with 1 mg mL - 1 streptavidin for 3 min and washed with BRB80 to remove unbound streptavidin. Biotin-labeled GMPCPP seeds were immobilized on the coverslip for 3 min, and the chamber was washed with BRB80 buffer containing 1 mg mL - 1 casein, 0.5% pluronic acid. After immobilizing the seeds, the pig brain tubulin with a mixture of labeled and unlabeled tubulin was diluted 10-fold in BRB80 buffer supplemented with 4 mM GTP, 0.2% methylcellulose, 1 mg mL - 1 casein, 0.5% pluronic acid, 0.1 mg mL −1 glucose oxidase, 0.02 mg mL −1 catalase, 0.8% D-glucose in the presence and absence of 300 nM MAP9. The mixture was flowed into the chamber and imaged immediately on the TIRF microscope at 22 °C. Movies were recorded for 100 frames with 5 s between frames for 8 min. Fluorescence Image Analysis Single-molecule motility of kinesins and dyneins was recorded for 250-1,000 frames per imaging area. A modified version of FIESTA (YFIESTA) was used to generate MT tracks and kymographs from record movies. The run frequency was calculated by observing the number of processive motors on each MT divided by the length of the MT and time. The run length and velocity were calculated by manually scoring the beginning and the ending of single trajectories in the kymographs. For the fluorescence intensity of MTs, pictures of MTs with MAP9 or MAP7 bound were imported into ImageJ, and the average fluorescent intensity of MTs was measured using the measure tool. The data was plotted in Prism. Neuronal Culture Isolation For neuronal cultures, hippocampi were dissected out from E16.5–18.5 mouse embryonic brains, and hippocampal neurons were isolated using the Worthington Papain Dissociation System (Worthington Biochemical Corporation) according to the manufacturer’s protocol. Primary neurons were plated at a density of 500,000 cells per glass coverslip coated with 0.05 mg/mL Poly-D-Lysine (molecular weight 70,000–150,000, Millipore-Sigma) and cultured in Neurobasal media supplemented with glucose, GlutaMAX (ThermoFisher), B27, and penicillin/streptomycin (P/S). The media was changed every two days until neurons were fixed at DIV4. A high plating density was used because MAP9 knockdown led to substantial neuronal loss, resulting in relatively few MAP9 shRNA-transfected neurons. RNA Interference To knock down MAP9, we used the GIPZ Lentiviral shRNA glycerol set (Horizon Discovery/Dharmacon), which targets the mouse MAP9 gene. The GIPZ shRNA sequences have been validated by the manufacturer for efficient and specific knockdown of MAP9 expression in mammalian cells. The set consisted of six different shRNA constructs (in pGIPZ vectors) with the following clone IDs and mature antisense sequences: RMM4431-200312303, Clone ID: V2LMM_124251, Antisense: TCCGTTCAATTCTTTCTTG RMM4431-200314601, Clone ID: V2LMM_124254, Antisense: TCTCTTTGAGGTATTCTAG RMM4431-200348107, Clone ID: V2LMM_240138, Antisense: TTCTTTGAAGGAAGATGAG RMM4431-200353088, Clone ID: V2LMM_211213, Antisense: ATGTGATCAGCCTTGTCTG RMM4431-200368801, Clone ID: V3LMM_418357, Antisense: CTAACTAGACTTTCATCTG RMM4431-200365512, Clone ID: V3LMM_418356, Antisense: TTGATAGCAAGGAATCCCA All six constructs were combined in equal ratios (1:1) to generate a pooled shRNA mixture. This mixture was transfected into primary hippocampal neurons using Lipofectamine 2000 (ThermoFisher) according to the manufacturer’s instructions. Transfection efficiency was assessed using the internal GFP reporter cassette included in the GIPZ vector. Neurons were fixed at DIV4 and analyzed for MAP9 expression and morphological changes. Immunocytochemistry The cultures were fixed in 4% paraformaldehyde for 20 min at room temperature, washed several times with phosphate buffer saline (PBS), permeabilized with 0.3% Triton X-100 in PBS (PBS-TX), and blocked with 5% BSA in PBS for 1 h at room temperature. Cultures were then incubated overnight at 4 °C with primary antibodies: rabbit anti-MAP9 (1:500, Invitrogen PA5-58145) and mouse anti-alpha tubulin (1:1000, Sigma Clone DM1A, T9026). The next day, cultures were incubated with secondary antibodies (1:1000) for 1 h at room temperature: Cy3 donkey anti-rabbit, Cy5 donkey anti-mouse, or Cy3 goat anti-chicken. Following several PBS washes, coverslips were mounted using VectaShield mounting medium (Vector Laboratories). Imaging was performed using a Leica SPE laser scanning confocal microscope with a 60x or 100x oil immersion objective. The total number of neuronal processes was quantified in GFP-positive MAP9 shRNA-transfected neurons and compared to control neurons. Cryo-EM Sample Preparation Lyophilized porcine brain tubulin (Cytoskeleton) was resuspended to 10 mg mL - 1 in BRB80 buffer supplemented with 10% (v/v) glycerol, 1 mM GTP, and 1 mM DTT. 10 μL of the tubulin solution was polymerized at 37 °C for 15 min. 1 μL of 2 mM taxol was added to the tubulin solution and incubated at 37 °C for 10 min. 1 μL of 2 mM taxol was added again, and the solution was incubated for 30 min. MTs were pelleted by centrifugation at 37 °C and 17,000 rcf for 20 min. The supernatant was discarded, and the pelleted MTs were resuspended in resuspension buffer (BRB80 buffer supplemented with 250 μM taxol). An aliquot of the resuspended MTs was depolymerized by CaCl 2 to measure the concentration of polymerized tubulin. The remaining MT solution was diluted to 1.5 μM in dilution buffer (BRB80 buffer supplemented with 100 μM taxol). Immediately before sample preparation, all MT-binding proteins were desalted to BRB80 using Zeba Spin columns (Pierce). To prepare MT-bound MAP9 samples, 2 μL of 1.5 μM taxol-MTs were incubated on a glow-discharged holey carbon cryo-EM grid (QuantiFoil, Cu 300 R 1.2/1.3) for 30 s, manually blotted with Whatman filter paper, and 2 μL of 7 μM MAP9 was added to the grid. The grid was transferred to a Vitrobot (ThermoFisher) set at 25 °C and 80% humidity, plunge-frozen in liquid ethane with a blot force of 5 pN and a blot time of 6 s, and transferred to liquid nitrogen. MTs were decorated with MAP9-MTBD plus KIF1A (1-393) following a similar procedure. 2 μL of 1.5 μM taxol-MTs were incubated on a glow-discharged holey carbon cryo-EM grid for 30 s, manually blotted as before, and incubated with 2 μL of 6 μM MAP9-MTBD, manually blotted, then incubated with 6 μM KIF1A (1-393) in 2 mM AMP-PNP. The grid was plunge-frozen in the Vitrobot under identical conditions as the previous sample and transferred to liquid nitrogen. Cryo-EM Data Collection Cryo-EM data for MT-bound MAP9 and MT-bound MAP9-MTBD and KIF1A (1-393) were collected on an Arctica microscope (ThermoFisher), operated at 200 kV with a K3 direct electron detector (Gatan). Images were acquired at 36,000x nominal magnification at 1.14 Å/pixel. All data was acquired in the super-resolution mode with a dose rate of ∼7.2 electrons/pixel/second and exposure time of ∼9 s dose-fractionated into 50 frames. All data were collected using the SerialEM software package 43 . Cryo-EM Image Processin g For the MT-MAP9 dataset, the movie stacks were imported to CryoSPARC 44 and motion-corrected ( Extended Data Table 2 ). The CTF parameters were estimated with the patch CTF job and manually curated to remove bad micrographs. We followed the pipeline used in our previous studies, which deals with the pseudo helical symmetry of MTs and uneven binding of multiple MAPs 15 , 42 . Briefly, particles were automatically picked with the filament tracer, initially without a reference, and later using 2D templates as input 45 . The segment separation was set to 82 Å, the length of an αβ tubulin dimer. Particle images were initially extracted with a box size of 512 pixels and Fourier-cropped to 256 pixels. Four rounds of 2D classification were performed and classes showing clear density for the MT were selected. MTs containing 13 and 14 PF were separated by using initial models of 13 and 14 PF MTs as references in heterogeneous refinement. In both cryo-EM datasets, the majority of particles were classified as 14 PF MTs, so this class was used for downstream processing. 14 PF MT particles were subjected to helical refinement with an initial rise estimate of 82.5 Å and a twist of 0°. Particles were then subjected to local refinement using a cylindrical mask enclosing the MT. To obtain a reconstruction that accounts for the MT seam, we used a seam search routine that accounts for the fact that helical symmetry is broken at a site of heterologous lateral contacts between tubulins using Frealign 46 , with custom scripts that determine the seam position on a per-particle basis 47 . For this purpose, we converted the CryoSPARC alignment file from the last local refinement, first to the STAR format using the csparc2star script from the PyEM suite (10.5281/zenodo.3576630), and then to PAR Frealign format using a custom Python script 42 . Upon completion of the seam search protocol, the particles were imported back to CryoSPARC with the seam-corrected alignments and re-extracted without Fourier cropping (box size: 512 pixels). The imported particles were subjected to local refinement, and CTF refinement was performed to estimate the per-particle CTF. Another local refinement was run on this particle set to produce the final C1 reconstruction. The symmetry search job was used to determine the rise and the twist of the map, and these parameters were input to exploit the pseudo-symmetry of the MT. A local refinement was performed on the newly expanded particle stack with a cylindrical mask around the MT to yield the final symmetrized reconstruction of the entire MT. A final local refinement was performed with a smaller mask around the good PF (PF opposite to the seam) and the PF adjacent to it, producing an approximately 3Å resolution map of MAP9-bound MTs. The MT-MAP9-MTBD-KIF1A (1-393) dataset was processed using the same protocol to produce a consensus reconstruction at 2.9Å resolution. To identify particle distributions associated with MAP9, KIF1A, or both MAP9 and KIF1A co-bound to MTs, 3D classification using four classes was performed in CryoSPARC with a more constrained mask only covering one tubulin dimer, one KIF1A monomer, and the MAP9 region covering a single tubulin dimer. The 3D classification of four classes yielded two classes with MAP9 bound to a tubulin dimer without KIF1A monomer and two classes with MAP9 and KIF1A monomer co-bound to a tubulin dimer. The particles in similar classes were merged into the two described states, each comprising approximately half of the data: MAP9 alone on MTs, and both the KIF1A monomer and MAP9 co-bound to MTs. Model Building and Refinement The KIF1A-MAP9-MT structure was modeled by first fitting the deposited model of the two-head-bound KIF1A AMPPNP-MT structure (PDB ID: 8UTO 36 ). Once the model was adjusted to fit the structure, it was morphed using ISOLDE 48 . The MAP9 R1-R2 segment was modeled into the density map by first identifying the MAP9 residues involved in MT binding across the first two pseudorepeats. The main identifying feature was the clear tryptophan residue protruding into the β-tubulin density. The AlphaFold2-predicted MAP9 model 49 was then tailored to include only the start of the MTBD and an extension matching the length of the helix identified in the density map. This tailored model was initially fitted into the map and subsequently morphed using ISOLDE. The identified MT-binding pseudorepeats fit well within the density map. Next, the model was combined with the adjusted KIF1A-MT model, and the combined structure was subjected to real-space refinement in Phenix 50 . The MAP9 R1-R2 (with a 36 a.a. spacing between pseudorepeats) and the MAP9 R2-R3 (with a 35 a.a. spacing between pseudorepeats)-MT model were generated by fitting tubulins (PDB: 6DPV) and tailoring the AlphaFold2 MAP9 model to extend the helix containing the residues identified as the first and second or the second and third MT-binding pseudorepeats, respectively. This extended helix was fitted and morphed into the density map using ISOLDE. Finally, both the MAP9 R1-R2-MT and MAP9 R2-R3-MT models underwent real-space refinement in Phenix 50 . MD Simulations An atomic model of the MAP9 MT-binding domain (MTBD; residues A421–H624) bound to five (α–β–α) protofilaments, including the tubulin CTTs, was built based on the cryo-EM reconstruction. Additional models were generated in which the MTs were co-decorated with MAP9-MTBD and KIF5B or KIF1A motor domains ( Extended Data Table 3a ). All systems were solvated in a TIP3P water box, ensuring at least 30 Å of water padding in each dimension, and ionized to final concentrations of 150 mM KCl and 1 mM MgCl 2 . Each system contained approximately 1.5 M atoms (Table S3A). MD simulations were performed using NAMD3 51 and the CHARMM36m 52 all-atom additive protein force field. The simulations used a 2-fs time step at 310 K and 1 atm. Long-range electrostatics were treated with the particle-mesh Ewald method, and a 12 Å cutoff was applied for van der Waals interactions. To equilibrate the system, the protein was first held fixed during 10,000 steps of minimization, followed by 1 ns of equilibration. The protein constraints were then released, and the system underwent an additional 10,000 steps of minimization. Subsequently, a 5-ns equilibration was performed while restraining the C α atoms with a harmonic potential of 1 kcal mol −1 Å −2 . Following these two minimization-equilibration phases, the production run was initiated. To account for the missing structural elements of MAP9 and the MT, C α atoms of select tubulin residues ( Extended Data Table 3b ) were restrained with a harmonic potential of 1 kcal mol −1 Å −2 . Eight simulation runs of 125 ns length each were performed for the MAP9-MT complex, eight runs of 700 ns each were conducted for the kinesin-3-bound MAP9-MT complex, and eight runs of 500 ns each were carried out for the kinesin-1-bound MAP9-MT complex. For analysis of MD simulations, all runs for each system were concatenated into a single trajectory. Correlations between the amino acids of MAP9 and kinesin were evaluated using a three-step method 29 , 53 . First, we constructed an inter-protein covariance matrix C , defined as C = 〈(R − 〈R〉)(R − 〈R〉) T 〉, where R represents the configuration vector containing only the C α coordinates for a given MD simulation conformation, and ⟨ R ⟩ denotes the trajectory-averaged configuration vector. Next, we filtered residue pairs based on their interatomic distances and finally focused on the correlated residues in close contact to pinpoint the specific residue interactions responsible for these observed correlations. Salt bridges were identified as interactions between basic nitrogen and acidic oxygen atoms within 6 Å 54 . Hydrogen bonds were defined as interactions where the donor–acceptor distance was ≤3.5 Å and the donor–hydrogen–acceptor angle was ≤30° 55 . Hydrophobic interactions were defined as contacts between side-chain carbon atoms within an 8 Å cutoff distance. Repulsive correlations were identified when distances between pairs of either basic nitrogen or acidic oxygen atoms were within 12 Å. Extended Data Extended Data Tables View this table: View inline View popup Extended Data Table 1. The list of constructs used in this study (EDF: Extended Data Figure). View this table: View inline View popup Download powerpoint Extended Data Table 2. Cryo-EM dataset and model building statistics. View this table: View inline View popup Download powerpoint Extended Data Table 3. Conformations and constraints used in MD simulations. a. The list of MD simulations performed in this study. The CTTs were modeled either as an α-helix or an unstructured peptide. Constraints were introduced at the amino acid residues of tubulin to minimize the fluctuations of tubulin subunits independent of one another. Additional constraints were introduced at MAP9 terminals to ensure MAP9 remains bound to the MT. Two sets of simulations were performed in the presence and absence of kinesin-1 and −3 motor domains using each combination of constraints and CTT models. b. The list of amino acids from MAP9 termini, α-tubulin, and β-tubulin, whose C α atoms were constrained during MD simulations. Extended Data Figures Download figure Open in new tab Extended Data Fig. 1. Cryo-EM data processing pipeline used to determine the structure of MAP9 bound to MTs. Download figure Open in new tab Extended Data Fig. 2. Resolution of the cryo-EM maps. a, Top left, Fourier shell correlation (FSC) plots of C1 (top) and symmetrized (bottom) reconstructions of MAP9 bound to MTs (left) and MAP9 and the KIF1A motor domain bound to MTs (right). b, Resolution maps of the symmetrized reconstruction of MAP9 bound to MTs (left) and MAP9 and the KIF1A motor domain bound to MTs (right). Download figure Open in new tab Extended Data Fig. 3. Sequence conservation of the MAP9-MT interface. a, The sequence alignment of MAP9 MTBD across species reveals the conservation of its MT-binding repeats. The alignment covers the MTBD of human MAP9 (residues 424 - 624). b, The sequence alignment of β-tubulin across species reveals the conservation of its residues interacting with the MAP9 helix (arrowheads). The alignment covers residues 380 – 420 of human β3 tubulin. The sequences were aligned and colored using the Clustal O algorithm in Jalview. Download figure Open in new tab Extended Data Fig. 4. Fit of the five MT-binding domains of MAP9 into the cryo-EM density. The view of repeats 1, 2, and 3 are built atomic models within the cryo-EM density map. Repeats 4 and 5 are fitted atomic models into the density map. Download figure Open in new tab Extended Data Fig. 5. Interactions between MAP9 and tubulin observed in all-atom MD simulations. (Left) Representative structural snapshots highlight the interactions between MAP9 (dark blue) and the tubulin dimer (α-tubulin in green and β-tubulin in blue). The protein backbones are depicted as new cartoons, while interacting side chains are shown in licorice representation. (Right) The fraction of specific intermolecular interactions formed between MAP9 and β-tubulin observed over the course of the MD simulations. The positively charged residues on MAP9 also transiently interacted with the glutamic acid side chains of the β-tubulin tail (light yellow; E441, E443, E445, E446, E447, E450). Conserved MAP9 residues in pseudo-repeats are highlighted in green. Download figure Open in new tab Extended Data Fig. 6. Mutations to the MAP9 residues that interact with αCTT do not affect the overall MT binding of MAP9. a, The schematic shows the mutations on MAP9 to disrupt the pairwise interactions between MAP9 and β-tubulin. The R index indicates that all three conserved residues were mutated to alanine in the specified pseudo-repeats. b, Example images show MT decoration of 100 nM mutant MAP9 constructs. c, Schematic of the αCTT mutant to disrupt potential interactions between MAP9 and the CTT of α-tubulin. d, Representative image of GFP-αCTT binding to MTs in 100 mM salt. e, Fluorescent intensity measurements reveal that αCTT and WT MAP9-MTBD decorate MTs with similar affinity (n = 20 MTs for each condition). Download figure Open in new tab Extended Data Fig. 7. MAP9 knockdown impairs neurite outgrowth in primary neuronal cultures. a, Immunostaining of primary cultured neurons with antibodies against MAP9 (Invitrogen PA5-58145, magenta) and alpha-tubulin (Sigma Clone DM1A T9026, green). MAP9 staining appears weaker than that of tubulin in the soma, suggesting that MAP9 does not bind to soluble tubulin and instead decorates polymerized MTs. b, Additional example image of DIV4 neurons transfected with MAP9 shRNA (green) at DIV0 and counterstained with antibodies against alpha-tubulin (Sigma Clone DM1A T9026, blue) and MAP9 (Invitrogen PA5-58145, magenta). Download figure Open in new tab Extended Data Fig. 8. The average velocity and run length of MT motors under different MAP9 concentrations. From left to right, n = 805, 580, 181, 105, 142 motors for K560, 508, 156, 177, 74 motors for KIF3A, 384, 316, 379, 383, 173, 254, 373, 400, 130 motors for KIF1A, and 547, 527, 334, 226, 297 motors for DDR. Download figure Open in new tab Extended Data Fig. 9. The regulation of tail-truncated KIF1A motility by MAP7 and MAP9. a, Schematic shows the domain organization of FL and tail-truncated artificial dimer of KIF1A (MD: motor domain, NC: neck-coil, CC: coiled-coil, FHA: fork-head association domain, PH: pleckstrin homology domain, LZ: leucine zipper). b, Run frequency of KIF1A (1-393)-LZ under different concentrations of MAP7 (from left to right, n = 14, 23, 28, 29, 16) and MAP9 (from left to right, n = 14, 17, 21, 16, 9, 19). The centerline and whiskers represent the mean and SD, respectively. Download figure Open in new tab Extended Data Fig. 10. The average velocity and run length of motors under different MAP9-MTBD concentrations. a, Mass photometry of purified MAP9-MTBD (mean ± s.d.). The red dashed line represents the expected monomer mass. b, (Left) MT decoration of fluorescently labeled MAP9-MTBD under different concentrations. (Right) The fluorescence intensity (mean ± SD) of MAP9-MTBD was fit to the Langmuir isotherm (solid curve) to estimate K D (±SE). From left to right, n = 28/29, 30/29, 36/32, 33/29, 30/29 MTs (two technical replicates). c, (Left) Example kymographs of KIF1A, K560, and DDR motility under different MAP9-MTBD concentrations. The scale bar is 5 µm on the x-axis and 5 s on the y-axis. (Right) The velocity and run length of motors under different MAP9-MTBD concentrations. From left to right, n = 384, 290, 319, 446, 639, 502, 443 motors for KIF1A, 805, 705, 645, 298, 390, 339 motors for KIF5B, and 547, 518, 502, 408 motors for DDR. Download figure Open in new tab Extended Data Fig. 11. MAP7 inhibits KIF1A. a, Kymograph of KIF1A motility in the presence of 100 nM MAP7. The scale bar is 5 µm on the x-axis and 5 s on the y-axis. b, Run frequency, velocity, and run length of KIF1A in the presence of various amounts of MAP7. From left to right, n = 24, 13, 18, and 11 MTs for run frequency and 384, 96, and 142 motors for run length and velocity measurements. Download figure Open in new tab Extended Data Fig. 12. Construction of chimeric MAPs and their regulatory role in KIF1A and KIF5B motility in vitro. a, Schematic of WT and chimeric MAP7 and MAP9 constructs. The black arrow points to the KIF5B interaction site of MAP7 PD. b, Kymographs of FL KIF5B in the presence and absence of MAP9-MTBD/MAP7-PD chimera. c, The run frequency, run length, and velocity of FL KIF5B under different MAP9-MTBD/MAP7-PD concentrations. (From left to right, n = 15, 28, 15, 21, 24 MTs for run frequency and 28, 393, 180, 76, and 76 motors for the run length and velocity measurements. d, Kymographs (left) and run frequency (right) of KIF1A and K560 under different concentrations of chimeras MAP7-MTBD/MAP9-PD and MAP9-MTBD/MAP7-PD. e, Run frequency of K560 motors under different concentrations of MAP9-MTBD/MAP7-PD (from left to right, n = 18, 25, 21, 21, 20 MTs). The centerline and whiskers represent mean and SD, respectively. f, Run length and velocity of KIF5B and KIF1A under different concentrations of chimeric MAPs. From left to right, n = 805, 473, 224 motors for KIF5B and MAP7-MTBD/MAP9-PD, 805, 812, 409, 254, 226 motors for KIF5B and MAP9-MTBD/MAP7-PD, 384, 178, 152, 130 motors for KIF1A and MAP7-MTBD/MAP9-PD, and 384, 541, 537, 326 motors for KIF1A and MAP9-MTBD/MAP7-PD. In b and d, the scale bar is 5 µm on the x-axis and 5 s on the y-axis. Download figure Open in new tab Extended Data Fig. 13. MAP9 does not sterically overlap with kinesin or dynein on the MT. The cryo-EM structure of MAP9 was superimposed onto those of the kinesin-3 motor domain (PDB ID: 8UTO 36 ) and dynein MTBD (PDB ID: 6RZB 58 ), all bound to the MT, demonstrating that MAP9 occupies a distinct location without an obvious overlap with the footprint of these motors on the MT. Download figure Open in new tab Extended Data Fig. 14. Cryo-EM data processing pipeline used to determine the MAP9-KIF1A-MT structure. The symmetry expansion was performed for two tubulin dimers laterally adjacent to each other. Download figure Open in new tab Extended Data Fig. 15. Alignment of the loop-8 region of transport kinesins from humans. a, The sequences are oriented from the N terminus to the C terminus, and members of the kinesin motor family were grouped based on their kinesin class. The sequences were aligned using the Clustal O algorithm in Jalview and colored based on the Clustal algorithm. b, Comparison of the atomic models of loop-8 from kinesin-3 (this work) and kinesin-1 (PDB ID:3J8Y) 59 . Hydrophobic, negative, and positive side chains are highlighted in green, red, and blue, respectively. The sequence highlighted with a red dashed rectangle in a is most closely positioned to and interacts with the pseudorepeats of the MAP9 helix, and this sequence in kinesin-1 was swapped with that of KIF1A to generate the K560 L8swap mutant. The sequence highlighted with a blue dashed rectangle in a forms electrostatic interactions with the MAP9 helix. Download figure Open in new tab Extended Data Fig. 16. Interactions between MAP9 and kinesin-3 loop 8 from MD simulations. a, A representative snapshot from the MD simulation system containing five kinesin-3 motor domains (purple, cartoon representation), five tubulin lattice segments (surface representation; α-tubulin and β-tubulin in light green and light blue, respectively), and MAP9 (royal blue, cartoon representation), simulated in the presence of explicit solvent. b, Close-up structural views of each MAP9 repeat (R1–R5) highlighting the hydrophobic cores formed between MAP9 amino acids (regular font) and kinesin loop-8 (italic), as well as the salt bridges formed with loop-8. c, (Top) Correlations between kinesin-3 and MAP9 amino acids throughout the MD simulations. (Bottom) Interaction frequency heat maps reveal the amino acid pairs responsible for these correlations. The magnitudes of attractive and repulsive correlations and interactions are color-coded according to the scale bar on the right. Download figure Open in new tab Extended Data Fig. 17. Interactions between MAP9 and kinesin-1 loop 8 from MD simulations. a, A representative snapshot from the MD simulation system containing five kinesin-1 motor domains (purple, cartoon representation), five tubulin lattice segments (surface representation; α-tubulin and β-tubulin in light green and light blue, respectively), and MAP9 (royal blue, cartoon representation), simulated in the presence of explicit solvent. b, Close-up structural views of each MAP9 repeat (R1–R5) highlighting the salt bridges formed between MAP9 amino acids (regular font) and kinesin loop-8b region, as well as the electrostatic repulsion with loop-8b and −8c regions (italic). c, (Top) Correlations between kinesin-3 and MAP9 amino acids throughout the MD simulations. (Bottom) Interaction frequency heat maps reveal the amino acid pairs responsible for these correlations. The magnitudes of attractive and repulsive correlations and interactions are color-coded according to the scale bar on the right. Download figure Open in new tab Extended Data Fig. 18. The run length and velocity of KIF5B L8swap under different concentrations of MAP9. From left to right, n = 631, 662,420 motors. Video Legends Video 1. Interactions between MAP9 and tubulin observed in all-atom MD simulations. All-atom MD simulations highlighting interactions between MAP9 (royal blue) and the tubulin dimer (α-tubulin in green, β-tubulin in blue). Simulations contain five tubulin lattice segments and MAP9 MTBD in the presence of explicit solvent (not shown). Protein backbones are depicted in cartoon representation. Interacting side chains are shown as licorice. Positively charged residues are colored blue, negatively charged residues are red, and hydrophobic residues are green. The MAP-tubulin interactions are magnified in separate panels for regions R1 through R5. The combined trajectory has a total duration of 1 µs and comprises eight separate MD simulation runs. The trajectory time (in ns) and the corresponding MD run number for each displayed conformation are indicated in the top-left and top-right corners, respectively. Video 2. MAP9 promotes MT polymerization and suppresses catastrophe. Single color imaging of MT polymerization (red) from GMP-CPP seeds (not shown) immobilized to the glass surface through biotin-streptavidin linkage. The fluorescence background is due to the presence of unpolymerized tubulin in the flow chamber. Images were acquired at 5 s per frame, with a pixel size of 160 nm. Video 3. Processive motility of FL KIF1A in the presence and absence of varying MAP9 concentrations . MAP9 inhibits KIF1A at high concentrations. One-color imaging of LD655-labeled FL KIF1A on surface-immobilized MTs in the presence of 0, 15, and 600 nM MAP9. Images were acquired at 200 ms per frame, with a pixel size of 160 nm. Video 4. Processive motility of DDR complexes in the presence of MAP9 and MAP9-MTBD . One-color imaging of LD655-labeled DDR on surface-immobilized MTs without a MAP, with 320 nM MAP9-MTBD and 500 nM MAP9. FL MAP9 inhibits DDR motility, however, no significant inhibition is observed with MAP9-MTBD. Images were acquired at 200 ms exposure per frame. Video 5. Processive motility of KIF5B (K560) complexes in the presence and absence of varying concentrations of MAP9. One-color imaging of LD655-labeled K560 on surface-immobilized MTs with no MAP9, and with 40 and 80 nM MAP9, respectively. MAP9 inhibits the MT binding and motility of K560. Images were acquired at 200 ms exposure per frame. Video 6. Processive motility of FL KIF1A in the presence and absence of varying MAP9-MTBD concentrations . MAP9-MTBD does not inhibit FL KIF1A. One-color imaging of LD655-labeled FL KIF1A without MAP9-MTBD, and with 280, 560 nM MAP9-MTBD. Images were acquired at 200 ms per frame. Video 7. Interactions between MAP9 and kinesin-3 loop-8 from MD simulations. All-atom MD simulations highlight interactions between kinesin-3 motor domains (purple, cartoon representation) and MAP9 (royal blue, cartoon representation). Simulations contain five kinesin-3 motor domains, MAP9, and five tubulin lattice segments (surface representation; α-tubulin and β-tubulin colored in light green and light blue, respectively), and explicit solvent (not shown). Close-up structural views of each MAP9 repeat (R1–R5) highlight the hydrophobic cores and the salt bridges formed between MAP9 residues and the kinesin-3 loop-8. Video 8. Interactions between MAP9 and kinesin-1 loop-8 from MD simulations. All-atom MD simulations highlight attractive as well as repulsive interactions between kinesin-1 motor domains (purple, cartoon representation) and MAP9 (royal blue, cartoon representation). Simulations comprise five kinesin-1 motor domains, MAP9, five tubulin lattice segments (surface representation; α-tubulin and β-tubulin in light green and light blue, respectively), and explicit solvent (not shown). Close-up structural views of each MAP9 repeat (R1–R5) highlight the salt bridges as well as electrostatic repulsions between MAP9 residues and loop-8 of kinesin-1. Acknowledgments We thank R. McKenney for constructs, M. Aslan, J. Slivka, Y. Zhao and J. Fernandes for assistance with protein purification, J. Weng for assistance with fluorescence image acquisition and processing, J. Peukes for microscopy and computational support, J. Al-Bassam for helpful discussions regarding MAP9 sequence motifs, the Cal-Cryo facility at UC Berkeley for EM imaging, and Indiana Jetstream2 and NCSA Delta for MD simulations. This work was supported by grants from the National Institute of General Medical Sciences (GM094522 (A.Y.), GM127018 (E.N.), GM133688 (K.M.O.M.)) and the National Science Foundation (MCB-1617028 and MCB-1055017, A.Y.; DGE-2146752, A.T.), Advanced Cyberinfrastructure Coordination Ecosystem (ACCESS, BIO240144, M.Gur). E.N. is a Howard Hughes Medical Institute Investigator. References 1. ↵ Bodakuntla , S. , Jijumon , A.S. , Villablanca , C. , Gonzalez-Billault , C. & Janke , C . Microtubule-Associated Proteins: Structuring the Cytoskeleton . Trends in cell biology 29 , 804 – 819 ( 2019 ). OpenUrl CrossRef PubMed 2. ↵ Weingarten , M.D. , Lockwood , A.H. , Hwo , S.Y. & Kirschner , M.W . A protein factor essential for microtubule assembly . Proceedings of the National Academy of Sciences of the United States of America 72 , 1858 – 1862 ( 1975 ). OpenUrl Abstract / FREE Full Text 3. ↵ Drubin , D.G. & Kirschner , M.W . Tau protein function in living cells . J Cell Biol 103 , 2739 – 2746 ( 1986 ). OpenUrl Abstract / FREE Full Text 4. ↵ Ke , Y.D. et al. Lessons from tau-deficient mice . Int J Alzheimers Dis 2012 , 873270 ( 2012 ). OpenUrl PubMed 5. Tymanskyj , S.R. , Yang , B. , Falnikar , A. , Lepore , A.C. & Ma , L . MAP7 Regulates Axon Collateral Branch Development in Dorsal Root Ganglion Neurons . The Journal of neuroscience 37 , 1648 – 1661 ( 2017 ). OpenUrl Abstract / FREE Full Text 6. ↵ Monroy , B.Y. et al. Competition between microtubule-associated proteins directs motor transport . Nature communications 9 , 1487 ( 2018 ). OpenUrl PubMed 7. ↵ Velazquez , R. et al. Acute tau knockdown in the hippocampus of adult mice causes learning and memory deficits . Aging cell 17 , e12775 ( 2018 ). OpenUrl CrossRef PubMed 8. ↵ Kapitein , L.C. & Hoogenraad , C.C . Building the Neuronal Microtubule Cytoskeleton . Neuron 87 , 492 – 506 ( 2015 ). OpenUrl CrossRef PubMed 9. ↵ Saffin , J.M. et al. ASAP, a human microtubule-associated protein required for bipolar spindle assembly and cytokinesis . Proceedings of the National Academy of Sciences of the United States of America 102 , 11302 – 11307 ( 2005 ). OpenUrl Abstract / FREE Full Text 10. Fontenille , L. , Rouquier , S. , Lutfalla , G. & Giorgi , D . Microtubule-associated protein 9 (Map9/Asap) is required for the early steps of zebrafish development . Cell cycle 13 , 1101 – 1114 ( 2014 ). OpenUrl CrossRef PubMed 11. Jensen , V.L. et al. Whole-Organism Developmental Expression Profiling Identifies RAB-28 as a Novel Ciliary GTPase Associated with the BBSome and Intraflagellar Transport . PLoS Genetics 12 , e1006469 ( 2016 ). OpenUrl 12. ↵ Tran , M.V. et al. MAP9/MAPH-9 supports axonemal microtubule doublets and modulates motor movement . Developmental Cell 59 , 199 – 210 e111 ( 2024 ). OpenUrl PubMed 13. ↵ Wang , S. et al. MAP9 Loss Triggers Chromosomal Instability, Initiates Colorectal Tumorigenesis, and Is Associated with Poor Survival of Patients with Colorectal Cancer . Clin Cancer Res 26 , 746 – 757 ( 2020 ). OpenUrl Abstract / FREE Full Text 14. ↵ Hooikaas , P.J. et al. MAP7 family proteins regulate kinesin-1 recruitment and activation . J Cell Biol 218 , 1298 – 1318 ( 2019 ). OpenUrl Abstract / FREE Full Text 15. ↵ Ferro , L.S. et al. Structural and functional insight into regulation of kinesin-1 by microtubule-associated protein MAP7 . Science 375 , 326 – 331 ( 2022 ). OpenUrl CrossRef PubMed 16. ↵ Monroy , B.Y. et al. A Combinatorial MAP Code Dictates Polarized Microtubule Transport . Developmental Cell 53 , 60 – 72 e64 ( 2020 ). OpenUrl CrossRef PubMed 17. ↵ Al-Bassam , J. , Ozer , R.S. , Safer , D. , Halpain , S. & Milligan , R.A . MAP2 and tau bind longitudinally along the outer ridges of microtubule protofilaments . J Cell Biol 157 , 1187 – 1196 ( 2002 ). OpenUrl Abstract / FREE Full Text 18. ↵ Kellogg , E.H. et al. Near-atomic model of microtubule-tau interactions . Science 360 , 1242 – 1246 ( 2018 ). OpenUrl Abstract / FREE Full Text 19. ↵ Shigematsu , H. et al. Structural insight into microtubule stabilization and kinesin inhibition by Tau family MAPs . J Cell Biol 217 , 4155 – 4163 ( 2018 ). OpenUrl Abstract / FREE Full Text 20. ↵ Kadavath , H. et al. Tau stabilizes microtubules by binding at the interface between tubulin heterodimers . Proceedings of the National Academy of Sciences of the United States of America 112 , 7501 – 7506 ( 2015 ). OpenUrl Abstract / FREE Full Text 21. ↵ Tymanskyj , S.R. & Ma , L . MAP7 Prevents Axonal Branch Retraction by Creating a Stable Microtubule Boundary to Rescue Polymerization . The Journal of Neuroscience 39 , 7118 – 7131 ( 2019 ). OpenUrl Abstract / FREE Full Text 22. ↵ Cleveland , D.W. , Hwo , S.Y. & Kirschner , M.W . Purification of tau, a microtubule-associated protein that induces assembly of microtubules from purified tubulin . J Mol Biol 116 , 207 – 225 ( 1977 ). OpenUrl CrossRef PubMed Web of Science 23. ↵ Masson , D. & Kreis , T.E . Binding of E-MAP-115 to microtubules is regulated by cell cycle-dependent phosphorylation . J Cell Biol 131 , 1015 – 1024 ( 1995 ). OpenUrl Abstract / FREE Full Text 24. ↵ Qiang , L. et al. Tau Does Not Stabilize Axonal Microtubules but Rather Enables Them to Have Long Labile Domains . Current Biology 28 , 2181 – 2189 e2184 ( 2018 ). OpenUrl CrossRef PubMed 25. ↵ Mitchison , T. & Kirschner , M . Dynamic instability of microtubule growth . Nature 312 , 237 – 242 ( 1984 ). OpenUrl CrossRef PubMed Web of Science 26. ↵ Zhang , Y. & Hancock , W.O . The two motor domains of KIF3A/B coordinate for processive motility and move at different speeds . Biophysical Journal 87 , 1795 – 1804 ( 2004 ). OpenUrl CrossRef PubMed Web of Science 27. ↵ Urnavicius , L. et al. Cryo-EM shows how dynactin recruits two dyneins for faster movement . Nature 554 , 202 – 206 ( 2018 ). OpenUrl CrossRef PubMed 28. ↵ Yildiz , A . Mechanism and regulation of kinesin motors . Nature Reviews. Molecular cell biology ( 2024 ). 29. ↵ Golcuk , M. & Gur , M. A Practical Covariance-Based Method for Efficient Detection of Protein-Protein Attractive and Repulsive Interactions in Molecular Dynamics Simulations . bioRxiv, 2025.2003.2024.644990 ( 2025 ). 30. ↵ Ravelli , R.B. et al. Insight into tubulin regulation from a complex with colchicine and a stathmin-like domain . Nature 428 , 198 – 202 ( 2004 ). OpenUrl CrossRef PubMed Web of Science 31. Ayaz , P. , Ye , X. , Huddleston , P. , Brautigam , C.A. & Rice , L.M . A TOG:alphabeta-tubulin complex structure reveals conformation-based mechanisms for a microtubule polymerase . Science 337 , 857 – 860 ( 2012 ). OpenUrl Abstract / FREE Full Text 32. ↵ Zhang , R. , Alushin , G.M. , Brown , A. & Nogales , E . Mechanistic Origin of Microtubule Dynamic Instability and Its Modulation by EB Proteins . Cell 162 , 849 – 859 ( 2015 ). OpenUrl CrossRef PubMed 33. ↵ Scarabelli , G. et al. Mapping the Processivity Determinants of the Kinesin-3 Motor Domain . Biophysical Journal 109 , 1537 – 1540 ( 2015 ). OpenUrl CrossRef PubMed 34. ↵ Atherton , J. et al. Conserved mechanisms of microtubule-stimulated ADP release, ATP binding, and force generation in transport kinesins . eLife 3 , e03680 ( 2014 ). OpenUrl CrossRef PubMed 35. ↵ Larson , A.G. , Naber , N. , Cooke , R. , Pate , E. & Rice , S.E . The conserved L5 loop establishes the pre-powerstroke conformation of the Kinesin-5 motor, eg5 . Biophysical Journal 98 , 2619 – 2627 ( 2010 ). OpenUrl CrossRef PubMed Web of Science 36. ↵ Benoit , M. , Rao , L. , Asenjo , A.B. , Gennerich , A. & Sosa , H . Cryo-EM unveils kinesin KIF1A’s processivity mechanism and the impact of its pathogenic variant P305L . Nature Communications 15 , 5530 ( 2024 ). OpenUrl PubMed 37. ↵ Ogawa , T. , Nitta , R. , Okada , Y. & Hirokawa , N . A common mechanism for microtubule destabilizers-M type kinesins stabilize curling of the protofilament using the class-specific neck and loops . Cell 116 , 591 – 602 ( 2004 ). OpenUrl CrossRef PubMed Web of Science 38. ↵ Benoit , M. , Asenjo , A.B. & Sosa , H . Cryo-EM reveals the structural basis of microtubule depolymerization by kinesin-13s . Nature Communications 9 , 1662 ( 2018 ). OpenUrl PubMed 39. ↵ Lam , A.J. et al. A highly conserved 3(10) helix within the kinesin motor domain is critical for kinesin function and human health . Sci Adv 7 ( 2021 ). 40. ↵ Elshenawy , M.M. et al. Cargo adaptors regulate stepping and force generation of mammalian dynein-dynactin . Nat Chem Biol 15 , 1093 – 1101 ( 2019 ). OpenUrl CrossRef PubMed 41. ↵ Yildiz , A. , Tomishige , M. , Gennerich , A. & Vale , R.D . Intramolecular strain coordinates kinesin stepping behavior along microtubules . Cell 134 , 1030 – 1041 ( 2008 ). OpenUrl CrossRef PubMed Web of Science 42. ↵ Perez-Bertoldi , J.M. , Zhao , Y. , Thawani , A. , Yildiz , A. & Nogales , E . HURP regulates Kif18A recruitment and activity to synergistically control microtubule dynamics . Nature Communications 15 , 9687 ( 2024 ). OpenUrl PubMed 43. ↵ Mastronarde , D.N . SerialEM: A Program for Automated Tilt Series Acquisition on Tecnai Microscopes Using Prediction of Specimen Position . Microscopy and Microanalysis 9 , 1182 – 1183 ( 2003 ). OpenUrl 44. ↵ Punjani , A. , Rubinstein , J.L. , Fleet , D.J. & Brubaker , M.A . cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination . Nature Methods 14 , 290 – 296 ( 2017 ). OpenUrl PubMed 45. ↵ Cook , A.D. , Manka , S.W. , Wang , S. , Moores , C.A. & Atherton , J . A microtubule RELION-based pipeline for cryo-EM image processing . J Struct Biol 209 , 107402 ( 2020 ). 46. ↵ Grigorieff , N . Frealign: An Exploratory Tool for Single-Particle Cryo-EM . Methods in Enzymology 579 , 191 – 226 ( 2016 ). OpenUrl CrossRef PubMed 47. ↵ Zhang , R. & Nogales , E . A new protocol to accurately determine microtubule lattice seam location . J Struct Biol 192 , 245 – 254 ( 2015 ). OpenUrl CrossRef PubMed 48. ↵ Croll , T.I . ISOLDE: a physically realistic environment for model building into low-resolution electron-density maps . Acta Crystallography D Struct Biol 74 , 519 – 530 ( 2018 ). OpenUrl 49. ↵ Jumper , J. et al. Highly accurate protein structure prediction with AlphaFold . Nature 596 , 583 – 589 ( 2021 ). OpenUrl CrossRef PubMed 50. ↵ Afonine , P.V. et al. Real-space refinement in PHENIX for cryo-EM and crystallography . Acta Crystallography D Struct Biol 74 , 531 – 544 ( 2018 ). OpenUrl 51. ↵ Phillips , J.C. et al. Scalable molecular dynamics on CPU and GPU architectures with NAMD . The Journal of Chemical Physics 153 , 044130 ( 2020 ). 52. ↵ Huang , J. et al. CHARMM36m: an improved force field for folded and intrinsically disordered proteins . Nature Methods 14 , 71 – 73 ( 2017 ). OpenUrl PubMed 53. ↵ Gur , M. & Erman , B . Quasi-harmonic analysis of mode coupling in fluctuating native proteins . Physical biology 7 , 046006 ( 2010 ). 54. ↵ Taka , E. et al. Critical Interactions Between the SARS-CoV-2 Spike Glycoprotein and the Human ACE2 Receptor . The journal of physical chemistry. B 125 , 5537 – 5548 ( 2021 ). OpenUrl PubMed 55. ↵ Durrant , J.D. & McCammon , J.A . HBonanza: a computer algorithm for molecular-dynamics-trajectory hydrogen-bond analysis . J Mol Graph Model 31 , 5 – 9 ( 2011 ). OpenUrl CrossRef PubMed 56. Shastry , S. & Hancock , W.O . Neck linker length determines the degree of processivity in kinesin-1 and kinesin-2 motors . Current biology 20 , 939 – 943 ( 2010 ). OpenUrl CrossRef PubMed 57. Schlager , M.A. , Hoang , H.T. , Urnavicius , L. , Bullock , S.L. & Carter , A.P . In vitro reconstitution of a highly processive recombinant human dynein complex . The EMBO Journal 33 , 1855 – 1868 ( 2014 ). OpenUrl Abstract / FREE Full Text 58. ↵ Lacey , S.E. , He , S. , Scheres , S.H. & Carter , A.P . Cryo-EM of dynein microtubule-binding domains shows how an axonemal dynein distorts the microtubule . eLife 8 ( 2019 ). 59. ↵ Shang , Z. et al. High-resolution structures of kinesin on microtubules provide a basis for nucleotide-gated force-generation . eLife 3 , e04686 ( 2014 ). OpenUrl CrossRef PubMed View the discussion thread. Back to top Previous Next Posted November 17, 2025. Download PDF Supplementary Material 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 Structure and mechanism of microtubule stabilization and motor regulation by MAP9 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 Structure and mechanism of microtubule stabilization and motor regulation by MAP9 Burak Cetin , Aryan Taheri , Mert Golcuk , Brigette Y. Monroy , Jonathan Fernandes , Kassandra M. Ori-McKenney , Mert Gur , Eva Nogales , Ahmet Yildiz bioRxiv 2025.11.17.688911; doi: https://doi.org/10.1101/2025.11.17.688911 Share This Article: Copy Citation Tools Structure and mechanism of microtubule stabilization and motor regulation by MAP9 Burak Cetin , Aryan Taheri , Mert Golcuk , Brigette Y. Monroy , Jonathan Fernandes , Kassandra M. Ori-McKenney , Mert Gur , Eva Nogales , Ahmet Yildiz bioRxiv 2025.11.17.688911; doi: https://doi.org/10.1101/2025.11.17.688911 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 Biophysics Subject Areas All Articles Animal Behavior and Cognition (7633) Biochemistry (17681) Bioengineering (13890) Bioinformatics (41930) Biophysics (21446) Cancer Biology (18586) Cell Biology (25493) Clinical Trials (138) Developmental Biology (13374) Ecology (19897) Epidemiology (2067) Evolutionary Biology (24308) Genetics (15607) Genomics (22498) Immunology (17736) Microbiology (40385) Molecular Biology (17175) Neuroscience (88584) Paleontology (666) Pathology (2831) Pharmacology and Toxicology (4823) Physiology (7641) Plant Biology (15149) Scientific Communication and Education (2045) Synthetic Biology (4293) Systems Biology (9823) Zoology (2271)