To uncover their contrasting mechanisms, antimitotic drugs that inhibit Eg5 (kinesin-5) were analyzed in mixed-motor gliding assays of kinesin-1 and Eg5 motors in which Eg5 “braking” dominates motility. Loop-5 inhibitors (monastrol, STLC, ispinesib and filanesib) increased gliding speeds, consistent with inducing a weak-binding state in Eg5, whereas BRD9876 slowed gliding, consistent with locking Eg5 in a rigor state. Biochemical and single-molecule assays demonstrated that BRD9876 acts as an ATP- and ADP-competitive inhibitor with 4 nM KI. Consistent with its microtubule polymerase activity, Eg5 was shown to stabilize microtubules against depolymerization. This stabilization activity was eliminated in monastrol, but was enhanced by BRD9876. Finally, in metaphase-arrested RPE-1 cells, STLC promoted spindle collapse, whereas BRD9876 did not. Thus, different Eg5 inhibitors impact spindle assembly and architecture through contrasting mechanisms, and rigor inhibitors may paradoxically have the capacity to stabilize microtubule arrays in cells.

Highlights d FLA8/10/KAP is necessary and sufficient for the anterograde IFT in C. reinhardtii d Heterotrimerization with KAP fully activates the autoinhibited FLA8/10 motor d Coupled kinesin-2 motors work mostly independently during co-transport

Despite its fundamental importance to understanding the molecular bases of these cellular processes, a quantitative and testable mechanochemical model that describes how structural differences between kinesins alter kinetic steps in the ATPase cycle to produce functional changes in processivity is lacking. Here we use high resolution single-molecule microscopy to directly observe the stepping behavior of kinesin-1 and À2 motors with different neck linker lengths. We identify a one-head bound vulnerable state where a kinetic race between attachment of the tethered head and detachment of the bound head occurs, and find that its duration is negatively correlated with processivity. Using a cross-family comparative approach, we map functional differences back to structural differences to build an understanding of the design principles underlying motor processivity. Overall, our results provide a quantitative framework for understanding kinesin processivity both biophysically and as it pertains to the physiologically relevant emergent behaviors that result from its tuning.

20The construction and function of virtually all cilia require the universally conserved 21 process of Intraflagellar Transport (IFT) [1, 2]. During the atypically fast IFT in the green 22 alga C. reinhardtii, up to ten kinesin-2 motors 'line up' in a tight assembly on the trains [3], 23 provoking the question of how these motors coordinate their action to ensure smooth and 24 fast transport along the flagellum without standing in each other's way. Here, we show 25 that the heterodimeric FLA8/10 kinesin-2 alone is responsible for the atypically fast IFT in 26 C. reinhardtii. Notably, in single-molecule studies, FLA8/10 moved at speeds matching 27 those of in vivo IFT [4], but additionally displayed a slow velocity distribution, indicative of 28 auto-inhibition. Addition of the KAP subunit to generate the heterotrimeric FLA8/10/KAP 29 relieved this inhibition, thus providing a mechanistic rationale for heterotrimerization with 30 the KAP subunit in fully activating FLA8/10 for IFT in vivo. Finally, we link fast FLA8/10 31 and slow KLP11/20 kinesin-2 from C. reinhardtii and C. elegans through a DNA tether to 32 understand the molecular underpinnings of motor coordination during IFT in vivo. For 33 motor pairs from both species, the co-transport velocities very nearly matched the single-34 molecule velocities, and the complexes both spent roughly 80% of the time with only one 35 of the two motors attached to the microtubule. Thus, irrespective of phylogeny and kinetic 36properties, kinesin-2 motors prefer to work alone without sacrificing efficiency. Our 37 findings thus offer a simple mechanism for how efficient IFT is achieved across diverse 38 organisms despite being carried out by motors with different properties. 39Cilia or flagella (used interchangeably) are microtubule-based organelles that project from 40 the surface of most eukaryotic cells. As evidenced from their remarkably diverse 41 deployment across eukaryotes, they are among the most evolutionarily adaptive of all 42 organelles. This diversity is manifested as the large number of seemingly unrelated 43 human disorders linked to impaired ciliary biogenesis, ranging from infertility to vision 44 degeneration [5][6][7][8][9][10][11][12]. Such functional diversity starkly contrasts with the highly conserved 45 nature of IFT, the process that builds and maintains virtually all cilia from unicellular 46 organisms up to humans. IFT was first observed in the unicellular green alga C. reinhardtii 47 as movement of large, non-membrane-bound 'trains' between the ciliary tip and base 48 ( Figure 1) [13]. Given that these trains moved on axonemesan elaborate microtubule-49 based scaffoldit soon became clear that IFT is powered by cilia-specific kinesin-2 and 50 dynein-2 motors [1,[14][15][16][17][18][19][20][21]. The kinesin-2 motors that carry out IFT appear to have co-51 evolved with the axonemal structure, suggesting that their mechanochemical properties 52 are adapted to the highly specialized ciliary environment [22, 23]. Unlike most kinesins 53 that form ho...

maintain the motor in a tightly bound state, making each an attractive explanation for superprocessivity. In contrast, a hydrolysis trigger, which has been recently established for kinesin-1, involves a post-hydrolysis weakly-bound state susceptible to detachment, a feature that hinders superprocessivity. To test these models, we used stopped-flow fluorescence spectroscopy, steady-state kinetics, and singlemolecule motility assays to characterize the chemomechanical cycle of the kinesin-3, KIF1A. The KIF1A on-rate, k on Mt , is 17.3 mM À1 s À1 , 20-fold faster than kinesin-1. The k cat from solution ATPase assays is 278 s À1 , in agreement with the fast motility. KIF1A releases one ADP when combined with nucleotidefree microtubules, ruling out the no-trigger model. When this 1HB complex was combined with ATPgS, the tethered-head ADP release rate was 16.9 s À1 whereas ATP triggered half-site release was 162 s À1 . This result indicates ATP hydrolysis occurs before tethered head binding and undermines the ATP-binding trigger model. The similarity of the half-site release rate to the k cat suggests the ratelimiting step is before or during the release of ADP. We conclude that KIF1A exercises the same general stepping framework as other neuronal transport families but differs in the kinetics of specific transitions within the chemomechanical cycle.